paperID stringlengths 36 36 | pwc_id stringlengths 8 47 | arxiv_id stringlengths 6 16 ⌀ | nips_id float64 | url_abs stringlengths 18 329 | url_pdf stringlengths 18 742 | title stringlengths 8 325 | abstract stringlengths 1 7.27k ⌀ | authors stringlengths 2 7.06k | published stringlengths 10 10 ⌀ | conference stringlengths 12 47 ⌀ | conference_url_abs stringlengths 16 198 ⌀ | conference_url_pdf stringlengths 27 199 ⌀ | proceeding stringlengths 6 47 ⌀ | taskID stringlengths 7 1.44k | areaID stringclasses 688
values | embedding stringlengths 9.26k 12.5k | umap_embedding stringlengths 29 44 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
c20d8ecc-3469-4ba2-a06e-7d9762c7a9a5 | does-structure-matter-leveraging-data-to-text | 2112.04344 | null | https://arxiv.org/abs/2112.04344v1 | https://arxiv.org/pdf/2112.04344v1.pdf | Does Structure Matter? Leveraging Data-to-Text Generation for Answering Complex Information Needs | In this work, our aim is to provide a structured answer in natural language to a complex information need. Particularly, we envision using generative models from the perspective of data-to-text generation. We propose the use of a content selection and planning pipeline which aims at structuring the answer by generating intermediate plans. The experimental evaluation is performed using the TREC Complex Answer Retrieval (CAR) dataset. We evaluate both the generated answer and its corresponding structure and show the effectiveness of planning-based models in comparison to a text-to-text model. | ['Lynda Tamine', 'Karen Pinel-Sauvagnat', 'Laure Soulier', 'Thomas Gerald', 'Hanane Djeddal'] | 2021-12-08 | null | null | null | null | ['data-to-text-generation'] | ['natural-language-processing'] | [ 3.44652325e-01 8.60465288e-01 2.64304489e-01 -4.27442312e-01
-1.26777756e+00 -6.07177138e-01 1.47411418e+00 4.23219085e-01
-3.23872119e-01 8.09112847e-01 9.06891286e-01 -4.22091961e-01
-1.75815478e-01 -1.02905357e+00 -3.46581101e-01 6.44945800e-02
2.02077940e-01 1.38917875e+00 5.32104909e-01 -5.96461654e-01
7.52017081e-01 6.11870512e-02 -1.80074322e+00 8.70681345e-01
9.65459108e-01 7.60282278e-01 3.91341239e-01 7.99062252e-01
-8.39718163e-01 1.12443352e+00 -7.96689272e-01 -4.30170685e-01
-1.23425521e-01 -8.80389571e-01 -1.43946004e+00 1.15086250e-02
7.95436874e-02 -1.80314764e-01 2.90831029e-01 5.12594700e-01
4.48711872e-01 2.74136335e-01 8.27877522e-01 -9.50692356e-01
-1.45411894e-01 6.97259188e-01 4.90512460e-01 -2.18145922e-01
1.42689621e+00 -6.00102805e-02 8.75339985e-01 -9.31110024e-01
1.32325971e+00 1.38351166e+00 -1.71269804e-01 7.31992006e-01
-9.48363483e-01 2.39513844e-01 -2.62512062e-02 1.80956081e-01
-1.10360146e+00 -3.89309615e-01 6.11088395e-01 -3.55978400e-01
1.18745089e+00 4.14259404e-01 5.37169218e-01 6.63191378e-01
3.53811264e-01 9.70593095e-01 9.28726733e-01 -7.94137001e-01
5.66414833e-01 1.17533788e-01 1.40469089e-01 5.76667249e-01
-2.51138717e-01 -6.61276281e-02 -6.76390648e-01 -1.65875450e-01
6.60066083e-02 -5.65841079e-01 -1.21683806e-01 -9.62865055e-02
-9.09270406e-01 1.05096459e+00 1.35377690e-01 1.99617207e-01
-6.61050022e-01 7.74746481e-03 3.34530562e-01 3.37287068e-01
4.86314744e-01 7.23886251e-01 3.92135326e-03 -1.49298623e-01
-7.75325477e-01 1.11719525e+00 1.51949465e+00 1.43421316e+00
5.38322985e-01 -4.32872772e-01 -8.64059269e-01 1.98175639e-01
7.04260886e-01 4.88554001e-01 4.85514134e-01 -9.04717624e-01
7.40360975e-01 9.11787748e-01 5.35527766e-01 -8.29097807e-01
-3.82111996e-01 -1.72294229e-01 -2.14784235e-01 -4.88745719e-02
-1.37700811e-01 -1.88245744e-01 -8.86689365e-01 1.04684317e+00
4.02605236e-01 -7.67237365e-01 5.68378508e-01 4.28121209e-01
1.51344323e+00 8.82592916e-01 -7.32484832e-02 -3.94283444e-01
1.12180138e+00 -1.00448823e+00 -1.09694469e+00 -1.20073862e-01
7.36963332e-01 -8.67520154e-01 1.03222680e+00 8.69348198e-02
-1.29802001e+00 -7.17981815e-01 -6.07437015e-01 -5.51274270e-02
-5.03115773e-01 -1.40086198e-02 7.92542025e-02 3.16283047e-01
-1.17558086e+00 -1.09665934e-02 -2.05878764e-01 -5.42277396e-01
-1.88485444e-01 -5.68551831e-02 1.56763464e-01 -2.86039948e-01
-1.10382795e+00 9.71063912e-01 8.68765354e-01 -3.57716709e-01
-8.36476266e-01 -3.52251410e-01 -8.80489707e-01 -8.06588866e-03
3.46734822e-01 -1.07016492e+00 1.59194875e+00 -1.38639256e-01
-1.61463153e+00 3.56119812e-01 -3.40739995e-01 -6.38616800e-01
4.74953473e-01 -1.75927699e-01 -6.13964349e-02 4.01154220e-01
3.19602132e-01 1.04719877e+00 5.52556634e-01 -1.53007329e+00
-6.75654590e-01 -9.01384726e-02 3.07726920e-01 4.99578178e-01
2.49651462e-01 2.41588637e-01 -6.20699227e-01 -3.16836417e-01
-3.92621942e-02 -7.09294319e-01 -4.20149326e-01 -8.91987205e-01
-3.93215209e-01 -6.79847360e-01 6.18224978e-01 -5.69530368e-01
1.64692628e+00 -1.37616765e+00 1.82652608e-01 2.61135966e-01
-3.83198895e-02 9.78702530e-02 -2.22952232e-01 1.24018955e+00
2.67816901e-01 3.63796502e-01 7.06682950e-02 -4.79603380e-01
3.04479152e-01 8.32463503e-02 -6.64028525e-01 -5.51214993e-01
3.10756922e-01 1.04289424e+00 -7.97784269e-01 -9.92217481e-01
-1.07698016e-01 -4.69262637e-02 -7.44556069e-01 6.64103031e-01
-1.29298985e+00 3.19965363e-01 -8.64987016e-01 4.44822013e-01
2.03753874e-01 6.69839755e-02 5.70875630e-02 2.05579564e-01
-1.63719907e-01 4.30777669e-01 -8.86450589e-01 1.80709946e+00
-3.78169954e-01 2.54271999e-02 -4.87414956e-01 -4.35774140e-02
1.17843235e+00 5.49387336e-01 4.08228971e-02 -1.00515354e+00
9.75658968e-02 3.03541690e-01 -2.15136230e-01 -7.70678461e-01
1.21101415e+00 -1.22790873e-01 -4.79593098e-01 8.91258240e-01
3.17473598e-02 -1.03903234e+00 8.51487994e-01 6.51434064e-01
1.10225344e+00 5.93850613e-01 2.18362257e-01 -1.02777429e-01
6.95220053e-01 6.45978808e-01 -3.36063236e-01 9.76410329e-01
4.82950658e-01 6.05908871e-01 5.78740895e-01 -3.31250221e-01
-7.38714159e-01 -6.35340452e-01 2.59133190e-01 5.56179523e-01
-1.62678361e-01 -9.07622695e-01 -1.09352016e+00 -9.98982906e-01
-6.15359724e-01 1.46180809e+00 -4.86118883e-01 2.44100079e-01
-5.22195578e-01 -7.01462850e-02 1.59910381e-01 7.98070058e-02
4.04853016e-01 -1.50587070e+00 -1.06964302e+00 3.52850169e-01
-6.25607312e-01 -8.61937702e-01 -2.63494462e-01 -2.08159640e-01
-6.73257053e-01 -9.02719557e-01 -6.08973742e-01 -3.97421062e-01
5.25758028e-01 -1.33302122e-01 1.52278900e+00 1.63923055e-01
7.42930770e-02 8.41047764e-01 -1.08620930e+00 -8.36024880e-01
-1.10007250e+00 4.08719003e-01 -7.49447465e-01 -2.19978228e-01
-1.64306611e-02 5.09327324e-03 -4.40538138e-01 2.36286297e-01
-1.52287817e+00 3.40101689e-01 3.94644022e-01 2.71653861e-01
5.23976505e-01 -1.15572117e-01 7.01611996e-01 -9.49822843e-01
1.52992237e+00 -5.01444221e-01 -4.51816708e-01 7.26887882e-01
-7.26786196e-01 5.28972447e-01 1.50077403e-01 1.85616523e-01
-1.65522707e+00 1.38113141e-01 -4.98568892e-01 3.55866790e-01
-2.78775573e-01 9.45598245e-01 -6.74151490e-03 4.38587129e-01
8.53207529e-01 2.43687674e-01 -1.91255376e-01 -2.60509908e-01
6.94435000e-01 3.47789168e-01 1.01954892e-01 -7.73918211e-01
5.79387784e-01 1.49442172e-02 -2.40282621e-02 -4.23731863e-01
-6.19174182e-01 -5.62595844e-01 -4.95471507e-01 -8.23987782e-01
1.02773714e+00 -2.75769114e-01 -2.51471668e-01 -8.15884992e-02
-1.75767481e+00 -2.52906740e-01 -5.97092152e-01 1.95207726e-02
-1.05570567e+00 1.31597698e-01 -4.53329273e-02 -9.50092196e-01
-7.14748859e-01 -1.06776321e+00 1.27049363e+00 9.36932489e-02
-5.78078926e-01 -9.54462826e-01 5.66696346e-01 4.48385328e-01
5.93769491e-01 2.57668018e-01 1.20798433e+00 -8.83421123e-01
-8.53492141e-01 -4.92498875e-01 3.06035597e-02 -4.08772588e-01
-2.96566069e-01 -1.14301115e-01 -5.99219918e-01 2.80132592e-01
3.70023027e-02 -6.65875733e-01 4.70411122e-01 -1.78913236e-01
5.48757732e-01 -5.26039720e-01 -3.73533294e-02 -4.06313747e-01
1.43616498e+00 3.83322179e-01 1.09212852e+00 1.00762516e-01
3.65569815e-02 1.25087726e+00 1.05265820e+00 4.38925028e-01
7.32393563e-01 8.51654172e-01 4.47361052e-01 2.27869138e-01
-3.62572908e-01 -6.18142724e-01 -8.82859975e-02 6.44097328e-01
3.61079574e-02 -7.18767762e-01 -1.13670707e+00 4.42656994e-01
-2.04912162e+00 -1.14992857e+00 -4.29160088e-01 1.92357779e+00
7.32804537e-01 1.65731639e-01 -1.13986313e-01 -1.96379364e-01
-1.22531923e-03 4.49617878e-02 1.26053110e-01 -5.19382060e-01
2.12450773e-01 3.51285368e-01 -2.97582239e-01 9.00605440e-01
-6.07662559e-01 1.02593219e+00 6.64727020e+00 9.09618199e-01
-2.93332398e-01 -9.37965605e-03 2.99976766e-01 2.63204932e-01
-8.96896303e-01 3.16855103e-01 -1.03842103e+00 -7.70231336e-02
1.14497864e+00 -4.56447423e-01 5.26329465e-02 5.55059075e-01
2.08755895e-01 -5.77717185e-01 -1.04066360e+00 3.56945664e-01
5.30913234e-01 -1.67431879e+00 6.56035841e-01 -9.78060141e-02
7.73229957e-01 -4.70398635e-01 -4.08109576e-01 6.01022959e-01
4.07562852e-01 -8.61822486e-01 1.10250378e+00 1.16670704e+00
2.54967302e-01 -6.21578991e-01 9.55151975e-01 8.05728018e-01
-1.01121938e+00 2.81645060e-01 8.71843770e-02 7.33873621e-02
6.47536039e-01 2.20947608e-01 -1.46002340e+00 9.10309792e-01
1.81155026e-01 -1.34417266e-01 -8.71497631e-01 9.74748433e-01
-4.23739791e-01 1.10639416e-01 -7.51483887e-02 -6.81071401e-01
2.84586519e-01 6.43775538e-02 4.58148926e-01 1.18313527e+00
4.04861867e-01 2.67238587e-01 5.46044290e-01 9.57993567e-01
1.65295959e-01 6.03236616e-01 -9.04936969e-01 -2.34238654e-02
1.31986663e-01 9.66280878e-01 -7.37196982e-01 -5.57628751e-01
-2.94240303e-02 7.75106132e-01 -2.04227194e-02 3.40521663e-01
-3.29842806e-01 -4.24202532e-01 -5.55930078e-01 2.19067782e-01
1.62129384e-02 -2.39347834e-02 8.93742312e-03 -7.25883365e-01
7.56637305e-02 -1.13055301e+00 3.37548345e-01 -1.15117085e+00
-6.43742919e-01 1.09870315e+00 6.58431828e-01 -9.89998996e-01
-1.28507841e+00 1.78281986e-03 -7.37210453e-01 1.04642749e+00
-1.30527210e+00 -1.22419345e+00 -3.07939798e-01 3.95080507e-01
9.36472118e-01 -1.17430054e-01 1.13717330e+00 -7.16677234e-02
1.07911371e-01 -1.47045836e-01 -5.78751266e-01 -4.97217536e-01
3.60494375e-01 -1.19643831e+00 3.96804541e-01 7.34782159e-01
9.57034230e-02 6.03674412e-01 1.03592157e+00 -1.07984841e+00
-1.25306511e+00 -9.36369896e-01 1.49540234e+00 -6.91954374e-01
3.94165725e-01 -1.53694868e-01 -4.72816646e-01 2.20306635e-01
8.17198157e-01 -8.97348225e-01 6.27582908e-01 -2.22512260e-01
1.83690459e-01 3.65945399e-01 -1.16618061e+00 5.00930190e-01
5.91274559e-01 -3.56672108e-01 -8.61677885e-01 5.84446490e-01
9.96690631e-01 -4.79628831e-01 -5.09218931e-01 4.16152269e-01
2.03323320e-01 -7.87474275e-01 5.68845510e-01 -6.40531600e-01
8.32019746e-01 -4.15738612e-01 -3.66741568e-01 -1.20738876e+00
3.32562417e-01 -6.22449279e-01 -3.81951756e-03 1.19522369e+00
9.51658845e-01 -4.85364087e-02 7.14693666e-01 9.10038531e-01
-2.70495594e-01 -7.07040668e-01 -7.52360284e-01 -3.50439996e-01
-1.57372087e-01 -5.47078252e-01 6.75587058e-01 6.24118745e-02
1.10400140e-01 7.65120447e-01 -2.72911415e-02 -4.14933711e-01
1.10872246e-01 1.49839714e-01 1.13416779e+00 -1.06384027e+00
-1.73267603e-01 -2.22413495e-01 9.74584073e-02 -9.75809813e-01
-6.76154345e-02 -7.42616117e-01 3.79142612e-01 -2.31741691e+00
2.22653169e-02 -1.79772288e-01 6.58022821e-01 -1.56254217e-01
5.15609942e-02 -3.07048827e-01 4.36938107e-01 1.12568319e-01
-1.07111585e+00 7.24243641e-01 1.28773558e+00 -1.03818700e-02
-3.46732318e-01 2.50864327e-01 -5.50883234e-01 4.88355815e-01
8.07417691e-01 -5.37479401e-01 -8.25256348e-01 -1.98245332e-01
8.30089569e-01 5.53515673e-01 5.84714971e-02 -9.13805068e-01
4.78327155e-01 -2.01396823e-01 -1.98304608e-01 -1.25322568e+00
2.22193912e-01 -5.10919452e-01 9.68542099e-02 5.30562580e-01
-7.74625301e-01 2.96369791e-01 -7.76541932e-03 3.50475162e-01
-5.20004094e-01 -1.07700384e+00 -8.08150694e-02 -3.35577101e-01
-4.90612209e-01 -1.22042619e-01 -6.07964277e-01 2.85226226e-01
8.45939517e-01 2.54869014e-01 -5.04879177e-01 -1.05196500e+00
-4.28223342e-01 5.29367208e-01 -2.70857802e-03 2.79936403e-01
1.09115338e+00 -1.22107708e+00 -9.18525755e-01 -2.08312362e-01
6.26828253e-01 -1.21912040e-01 3.37552163e-03 2.03519940e-01
-7.87185848e-01 1.09500313e+00 2.45477393e-01 -2.64448225e-01
-1.10226941e+00 5.82454562e-01 8.05841088e-02 -8.66621971e-01
-2.29173630e-01 4.16783899e-01 -1.11802824e-01 -7.13084698e-01
2.95780838e-01 -1.98186845e-01 -8.44543338e-01 2.62137771e-01
7.76149929e-01 1.20358221e-01 1.66425183e-01 -2.69552886e-01
2.36332074e-01 1.25006482e-01 7.63270259e-02 -9.22022998e-01
1.00168097e+00 -2.37068255e-02 -9.58867967e-02 1.06831081e-01
7.17800260e-01 -1.28411576e-02 -4.95179921e-01 -2.72631377e-01
5.46679318e-01 -2.71632254e-01 -3.19930106e-01 -9.77100253e-01
-1.98401421e-01 6.58913672e-01 3.33485276e-01 6.98379099e-01
9.03785408e-01 4.57219869e-01 4.66016859e-01 9.13552344e-01
5.68546712e-01 -1.10557032e+00 5.09674788e-01 8.84797454e-01
1.69396281e+00 -9.46015656e-01 -1.22786991e-01 -3.50050896e-01
-7.70996213e-01 1.24600947e+00 4.26649272e-01 4.97158825e-01
5.25599897e-01 7.62763098e-02 4.03384604e-02 -7.71803856e-01
-1.25204754e+00 -7.23682702e-01 5.72308481e-01 7.05733240e-01
5.19314647e-01 -2.74473935e-01 -8.01832974e-01 9.87437367e-02
-3.30541670e-01 2.14571521e-01 5.59395194e-01 1.22844195e+00
-7.86397457e-01 -1.54243803e+00 -3.23904902e-01 3.42836320e-01
-1.06702283e-01 -1.56916738e-01 -1.07822824e+00 6.87465608e-01
-2.15012804e-01 1.57111931e+00 -2.94363111e-01 -1.78020835e-01
5.99518538e-01 3.27248961e-01 2.09071487e-01 -1.13709223e+00
-9.47309196e-01 1.00608334e-01 8.64388645e-01 -3.71461004e-01
-6.02325082e-01 -6.11496389e-01 -1.42843032e+00 2.28172228e-01
-3.09023727e-02 5.22639930e-01 1.01875007e+00 1.07065845e+00
3.46864969e-01 6.28224313e-01 3.48986506e-01 -4.83411968e-01
-4.17041093e-01 -1.17616868e+00 1.33397028e-01 2.61956334e-01
-1.50127724e-01 -1.90504603e-02 1.84083715e-01 1.26632676e-01] | [11.553937911987305, 8.32755184173584] |
f0ecaf4f-7360-480f-9842-dcbaa40339f8 | customizing-knowledge-graph-embedding-to | 2212.14102 | null | https://arxiv.org/abs/2212.14102v1 | https://arxiv.org/pdf/2212.14102v1.pdf | Customizing Knowledge Graph Embedding to Improve Clinical Study Recommendation | Inferring knowledge from clinical trials using knowledge graph embedding is an emerging area. However, customizing graph embeddings for different use cases remains a significant challenge. We propose custom2vec, an algorithmic framework to customize graph embeddings by incorporating user preferences in training the embeddings. It captures user preferences by adding custom nodes and links derived from manually vetted results of a separate information retrieval method. We propose a joint learning objective to preserve the original network structure while incorporating the user's custom annotations. We hypothesize that the custom training improves user-expected predictions, for example, in link prediction tasks. We demonstrate the effectiveness of custom2vec for clinical trials related to non-small cell lung cancer (NSCLC) with two customization scenarios: recommending immuno-oncology trials evaluating PD-1 inhibitors and exploring similar trials that compare new therapies with a standard of care. The results show that custom2vec training achieves better performance than the conventional training methods. Our approach is a novel way to customize knowledge graph embeddings and enable more accurate recommendations and predictions. | ['Murthy Devarakonda', 'Iya Khalil', 'Xiong Liu'] | 2022-12-28 | null | null | null | null | ['knowledge-graph-embedding', 'knowledge-graph-embeddings', 'knowledge-graph-embeddings'] | ['graphs', 'graphs', 'methodology'] | [-6.79140016e-02 4.88045692e-01 -9.37494278e-01 -3.21456701e-01
-2.89910495e-01 -7.13697433e-01 3.55505288e-01 8.76995206e-01
-5.19079804e-01 6.46670222e-01 6.56159639e-01 -7.70880044e-01
-6.09022200e-01 -8.28960359e-01 -4.78549749e-01 -3.79985005e-01
-1.45837322e-01 8.08605194e-01 2.04380035e-01 -3.16262782e-01
-2.32381761e-01 5.63794911e-01 -8.41248155e-01 3.01870704e-01
6.21913493e-01 6.21094346e-01 -1.38449490e-01 6.56272650e-01
-6.92930967e-02 3.41156214e-01 -2.63344496e-01 -7.14384437e-01
-1.33693218e-04 1.26016483e-01 -6.93283975e-01 -4.80385989e-01
2.38557726e-01 -2.65869386e-02 -6.96785510e-01 7.27931261e-01
8.80965531e-01 6.31907657e-02 7.62803614e-01 -1.28746164e+00
-1.23717117e+00 7.67681777e-01 7.19733611e-02 1.30101815e-01
3.66790980e-01 -9.95880440e-02 1.58910668e+00 -5.23092985e-01
1.02962863e+00 8.50753903e-01 7.74675369e-01 5.19451082e-01
-1.28031325e+00 -4.78548646e-01 -1.42887235e-04 4.21390325e-01
-1.40068805e+00 1.77749306e-01 4.90354002e-01 -6.35207832e-01
1.23214340e+00 4.66740310e-01 7.08163798e-01 1.23566496e+00
5.55335104e-01 3.21416914e-01 3.16052109e-01 -6.59640208e-02
1.17705785e-01 6.73881173e-01 4.13389832e-01 8.52013528e-01
6.75143838e-01 6.59354404e-02 -2.08850011e-01 -7.49285996e-01
4.65775281e-01 4.23050374e-01 -7.94161499e-01 -7.44529009e-01
-1.06904840e+00 1.16694593e+00 6.66658819e-01 3.07788074e-01
-3.87033910e-01 1.59709528e-01 5.85103452e-01 3.71309757e-01
2.79562682e-01 7.57355332e-01 -7.44558752e-01 3.86777937e-01
-5.65705061e-01 -2.13704407e-01 9.85392928e-01 8.37997437e-01
4.49944347e-01 -3.63257349e-01 -5.24661899e-01 6.55681491e-01
5.16450584e-01 6.96726236e-03 9.50210631e-01 -1.94041803e-01
-5.06605080e-04 7.31633306e-01 -1.33235037e-01 -1.03350782e+00
-4.77917224e-01 -5.37398040e-01 -4.46094364e-01 -1.46315634e-01
-2.73690850e-01 -1.79805681e-01 -8.71070981e-01 1.43435049e+00
3.41320664e-01 4.42555487e-01 2.84735203e-01 6.69928849e-01
1.11407864e+00 1.94721475e-01 4.69684005e-01 -1.64810508e-01
1.28961766e+00 -9.86468375e-01 -8.17306101e-01 3.85426670e-01
1.09882081e+00 -5.91718376e-01 7.98265994e-01 -1.25555977e-01
-3.13185006e-01 -2.14474857e-01 -1.10968912e+00 4.97225113e-02
-9.51825738e-01 -1.46891445e-01 1.01891911e+00 7.63082802e-01
-1.12963808e+00 9.64524269e-01 -5.66452742e-01 -6.46220267e-01
4.85181838e-01 7.33837247e-01 -6.61887825e-01 -1.74092680e-01
-1.46973360e+00 9.66929197e-01 4.23985332e-01 -4.36971068e-01
-5.40279806e-01 -1.22805643e+00 -9.40833747e-01 3.03854316e-01
2.19126701e-01 -1.17156339e+00 4.93231714e-01 -3.15746635e-01
-1.29021096e+00 4.67140198e-01 4.19564664e-01 -4.21629220e-01
1.69519082e-01 1.78313047e-01 -8.44884455e-01 -2.46364772e-01
-2.33962670e-01 3.40035766e-01 3.79739732e-01 -6.77312970e-01
-6.11461960e-02 -2.07824603e-01 6.25168085e-02 9.46623161e-02
-9.14287210e-01 -5.35899997e-01 -6.38934195e-01 -5.27494848e-01
-6.85710609e-01 -9.56414282e-01 -6.68828070e-01 1.60280406e-01
-5.25721610e-01 -4.90654737e-01 7.73303092e-01 -4.16561544e-01
1.43009007e+00 -1.89188266e+00 3.16514641e-01 5.35837054e-01
5.52104115e-01 4.78612453e-01 -4.87452090e-01 5.68690419e-01
-2.51969010e-01 5.01531005e-01 2.63089478e-01 1.79808587e-01
1.28305137e-01 4.81205791e-01 2.03118399e-01 2.00101912e-01
8.56734961e-02 1.19526947e+00 -1.23816693e+00 -4.68177587e-01
1.16671890e-01 8.18151891e-01 -8.36911142e-01 1.97901309e-01
-3.14729631e-01 -2.23456711e-01 -7.75709152e-01 3.90296429e-01
2.75659055e-01 -6.00411355e-01 9.30176377e-01 -6.77470744e-01
5.34911215e-01 -1.11845739e-01 -8.06059897e-01 1.52587652e+00
-3.18630964e-01 4.24710333e-01 -5.59883773e-01 -7.59404719e-01
5.20151496e-01 5.05196393e-01 7.02771842e-01 -1.13777786e-01
2.95323789e-01 -6.97418600e-02 8.24586302e-02 -6.80680573e-01
2.56026149e-01 6.00958914e-02 1.10116728e-01 3.46503228e-01
4.23294753e-01 1.50852591e-01 -1.41592681e-01 3.44603688e-01
1.53547168e+00 -2.69885391e-01 7.36180365e-01 -5.31884879e-02
2.68005073e-01 -6.70153357e-04 4.40437734e-01 4.24056977e-01
-1.67205840e-01 2.31418028e-01 5.43103576e-01 -4.41478938e-01
-6.34002686e-01 -1.04184031e+00 -1.12955794e-01 9.41646993e-01
-3.34793031e-01 -7.40551591e-01 5.30767068e-02 -1.45725310e+00
6.47349238e-01 6.87821925e-01 -1.10119736e+00 -3.10925752e-01
6.33289665e-02 -6.29574537e-01 2.98423409e-01 6.49295688e-01
-5.45979381e-01 -6.40765429e-01 1.34840414e-01 3.19885135e-01
5.84091365e-01 -8.46593976e-01 -1.06956530e+00 2.94945598e-01
-7.57856786e-01 -1.58942401e+00 -6.47551477e-01 -8.75330091e-01
7.65881479e-01 6.42739236e-03 1.00788677e+00 1.99761093e-01
-5.02430916e-01 8.38380218e-01 -5.74236214e-01 -3.01532090e-01
-1.84861362e-01 -7.61492923e-02 2.06934482e-01 -1.64981335e-01
3.79318357e-01 -4.39469516e-01 -7.64374375e-01 1.37587398e-01
-1.09803081e+00 -5.12804568e-01 6.16950095e-01 1.09509015e+00
8.04776609e-01 -2.15933025e-01 5.53716719e-01 -1.53299057e+00
1.21591401e+00 -8.76318097e-01 -1.91247642e-01 6.23778105e-01
-1.15255857e+00 3.41737777e-01 4.96421307e-01 -5.96200109e-01
-3.08043212e-01 9.46133882e-02 -4.80530635e-02 -6.91160500e-01
3.06821316e-01 9.69887197e-01 1.57679707e-01 -4.47331160e-01
8.68545830e-01 -2.78546572e-01 -3.35675366e-02 -2.57057250e-01
8.73760343e-01 6.28681064e-01 -7.62126073e-02 -3.23764235e-02
5.54960251e-01 -3.39826420e-02 -1.07146315e-01 -5.69407821e-01
-4.73621696e-01 -5.87863088e-01 -2.86996126e-01 2.39045247e-01
9.51557994e-01 -6.73718214e-01 -8.14182699e-01 -7.91810989e-01
-1.00997603e+00 -2.93578178e-01 -3.86007339e-01 8.32431376e-01
-7.10396618e-02 4.01902646e-01 -3.46599132e-01 1.48283452e-01
-6.63183630e-01 -1.12510812e+00 4.60005462e-01 1.99793391e-02
-4.45134342e-01 -1.76729417e+00 6.92470253e-01 1.19308233e-01
3.68809044e-01 2.26567373e-01 1.29913282e+00 -1.34638798e+00
-1.97430938e-01 -4.62045044e-01 -2.78730810e-01 1.23032458e-01
5.01622558e-01 3.48977782e-02 -6.36875033e-01 -4.08933640e-01
-9.26474035e-01 -4.09245305e-02 7.01204002e-01 1.80650026e-01
1.32955849e+00 -3.41082901e-01 -8.89775634e-01 5.44960618e-01
1.69348884e+00 3.68381925e-02 4.70975578e-01 3.96173075e-02
8.67518187e-01 1.76611602e-01 1.30527005e-01 3.88441056e-01
3.13088566e-01 6.71152234e-01 2.94964194e-01 -2.00644834e-03
-1.53183356e-01 -3.14363271e-01 1.24582186e-01 7.69475400e-01
2.62599848e-02 -5.82837641e-01 -7.24352002e-01 6.45865500e-01
-1.78881574e+00 -7.24588394e-01 1.18299931e-01 2.20214844e+00
1.06896806e+00 -1.18455119e-01 -2.91732728e-01 -3.02303642e-01
4.64929223e-01 -1.31845891e-01 -6.78539395e-01 -6.98554158e-01
2.34591335e-01 8.13803375e-01 7.82375455e-01 6.27215743e-01
-8.03272605e-01 8.15904677e-01 6.45629597e+00 5.89850307e-01
-1.06433380e+00 3.11792731e-01 1.04946382e-01 1.35621935e-01
-8.81297648e-01 -2.76219249e-01 -4.93012607e-01 2.41664425e-01
9.29701447e-01 -6.32727087e-01 -9.82531384e-02 7.47114301e-01
-1.26870587e-01 9.08207059e-01 -1.41622210e+00 8.14388633e-01
2.71524161e-01 -1.99680305e+00 4.18575138e-01 1.37804106e-01
7.69197345e-01 6.40240386e-02 6.27432242e-02 5.45818746e-01
7.02227712e-01 -1.01983893e+00 -4.65143085e-01 5.62436461e-01
8.83080781e-01 -3.17640841e-01 1.14121354e+00 -4.37286168e-01
-9.84095812e-01 1.11332014e-01 -4.38241363e-01 6.24016821e-01
9.53122391e-04 6.23345613e-01 -1.61565661e+00 9.08734620e-01
2.48316914e-01 9.68860984e-01 -7.08345115e-01 1.21269143e+00
-1.59499183e-01 4.91913617e-01 3.14405896e-02 -3.95981729e-01
9.80699956e-02 1.51081696e-01 2.54969567e-01 1.32004368e+00
4.00284022e-01 -1.43233165e-01 2.11970717e-01 3.38794976e-01
-4.95274037e-01 6.17540896e-01 -9.96811211e-01 -5.71763337e-01
3.85651916e-01 1.32762325e+00 -2.24684864e-01 -1.71311930e-01
-4.96351242e-01 1.06264281e+00 3.86534274e-01 3.51382583e-01
-7.74505496e-01 -5.80315709e-01 9.28948343e-01 1.30648911e-01
5.11911094e-01 1.88166380e-01 2.49182299e-01 -8.44161332e-01
-5.47075152e-01 -4.97664243e-01 9.76944089e-01 -6.43590808e-01
-1.55285954e+00 7.28593767e-01 -4.26978260e-01 -1.19648051e+00
1.52863845e-01 -9.51003313e-01 -7.59207368e-01 5.87555647e-01
-1.68068290e+00 -1.23233116e+00 -1.15275398e-01 5.67754030e-01
-2.83993259e-02 -4.58108544e-01 1.37567878e+00 4.78139877e-01
-5.58723330e-01 1.31778574e+00 2.53807813e-01 -9.42630693e-03
1.12024498e+00 -1.17810214e+00 -3.23447883e-01 -1.24027535e-01
1.57308757e-01 6.81938052e-01 4.77761924e-01 -8.09208274e-01
-1.46609128e+00 -1.57093763e+00 9.41806495e-01 -5.44847906e-01
1.12286365e+00 2.61423081e-01 -7.43336618e-01 9.08026636e-01
4.26931709e-01 1.69036791e-01 1.76701558e+00 4.00613040e-01
-5.73251605e-01 1.00873224e-01 -1.18453002e+00 7.00716019e-01
7.14388549e-01 -4.36342388e-01 -5.28648019e-01 7.69784212e-01
1.17028797e+00 -5.85660115e-02 -1.67123175e+00 2.92796969e-01
4.94157284e-01 -7.15368316e-02 1.19869471e+00 -1.46190119e+00
1.43587336e-01 -1.29335910e-01 -1.53350487e-01 -1.79372716e+00
-7.88033545e-01 -4.61456597e-01 -2.56016552e-01 6.92453384e-01
1.06064153e+00 -9.21118975e-01 9.23175931e-01 4.73416537e-01
-1.47166342e-01 -1.07430255e+00 -5.79527438e-01 -5.71028650e-01
-6.46852702e-02 -7.36828148e-02 5.62048495e-01 1.52015245e+00
4.71386552e-01 4.42479014e-01 -1.91523731e-01 3.27158719e-01
3.24460685e-01 -2.13250905e-01 3.71177554e-01 -1.40495372e+00
-4.31704700e-01 -4.03687835e-01 -1.12996399e+00 -3.44216883e-01
2.13826567e-01 -1.72671008e+00 -6.05648100e-01 -1.92657173e+00
2.68688321e-01 -4.80257362e-01 -1.10979664e+00 7.03750670e-01
-1.48254901e-01 4.41880561e-02 -2.81300366e-01 -2.99536288e-01
-4.74788308e-01 5.17624319e-01 1.21561491e+00 -6.73933923e-01
-2.31605709e-01 -1.74361885e-01 -1.03188717e+00 2.43250161e-01
7.06859708e-01 -5.65347433e-01 -8.43005836e-01 -2.67395347e-01
3.85430336e-01 -1.08420532e-02 5.27217016e-02 -5.39414346e-01
4.18977588e-01 -2.67989963e-01 2.22206324e-01 -4.80203442e-02
7.91021250e-03 -9.63523149e-01 3.17326903e-01 4.22132045e-01
-4.99813825e-01 -5.01947217e-02 4.26073581e-01 1.19903755e+00
4.38653305e-02 -1.00548714e-01 2.09382981e-01 3.51000637e-01
-7.07443178e-01 9.18022096e-01 -1.16230965e-01 -1.15694046e-01
1.16595459e+00 1.49335086e-01 -3.88909638e-01 -1.55837029e-01
-1.25620151e+00 5.71217775e-01 1.11710861e-01 5.66383958e-01
9.44689989e-01 -1.68406236e+00 -4.49289203e-01 -8.95098746e-02
6.64589643e-01 -8.39123011e-01 9.36838686e-02 7.11031795e-01
-4.37423199e-01 4.37506527e-01 -1.44222245e-01 2.18815985e-03
-1.45255208e+00 7.69361794e-01 8.67524296e-02 -5.92458308e-01
-3.02619278e-01 9.21348453e-01 -2.11765766e-01 -7.11251795e-01
2.39784867e-01 -1.73141450e-01 -6.36116982e-01 2.02082708e-01
1.34879053e-01 6.40924051e-02 1.16146967e-01 -1.87934712e-02
-4.63410825e-01 3.85939538e-01 -4.39475566e-01 4.54129666e-01
1.50689137e+00 5.75992823e-01 9.10451487e-02 -8.56136531e-02
1.59935558e+00 8.72865170e-02 -2.81090558e-01 -3.26553971e-01
-2.54527330e-01 -2.38556698e-01 4.47099864e-01 -9.54976976e-01
-1.26450515e+00 4.94726032e-01 8.76816511e-01 -9.73552018e-02
5.22386253e-01 4.69253324e-02 6.39227629e-01 6.92372262e-01
6.60726875e-02 -8.60333860e-01 -1.92136578e-02 1.97950467e-01
8.36189866e-01 -1.26026452e+00 3.42145860e-01 -3.75030696e-01
-6.75149083e-01 1.26538324e+00 3.11446726e-01 1.35054857e-01
1.42645407e+00 -1.94742173e-01 5.22426814e-02 -5.72552264e-01
-7.95724571e-01 -2.29281574e-01 6.90800309e-01 8.37325752e-01
7.49724686e-01 5.58881342e-01 -4.97132480e-01 8.85188639e-01
2.26291627e-01 9.36470404e-02 4.30271357e-01 7.42404640e-01
-3.64249349e-02 -1.60997093e+00 4.15023476e-01 1.03037977e+00
-1.12396277e-01 -2.69964904e-01 -5.59000671e-01 6.45468116e-01
2.31151022e-02 6.01935029e-01 -3.36610734e-01 -9.83310997e-01
6.26618028e-01 2.43498147e-01 2.42798150e-01 -1.03041458e+00
-5.83190799e-01 -4.19704080e-01 2.66013443e-01 -5.44603467e-01
5.57469688e-02 -1.21543460e-01 -9.33818460e-01 -1.74762860e-01
-5.21706164e-01 5.38382590e-01 4.42961216e-01 2.93508232e-01
9.07410264e-01 1.07662916e+00 5.04046500e-01 -3.20483744e-01
-4.87781197e-01 -4.98234928e-01 -5.68427444e-01 5.13664126e-01
-7.67448395e-02 -7.80443072e-01 -2.17012554e-01 -1.60257667e-01] | [8.052042961120605, 7.37106466293335] |
8d597861-9376-42d8-80d4-c4fe0904217f | low-dose-ct-image-reconstruction-using-vector | null | null | https://rdcu.be/dajRa | https://rdcu.be/dajRa | Low-Dose CT Image Reconstruction using Vector Quantized Convolutional Autoencoder with Perceptual Loss | Computed Tomography (CT) has become a useful screening procedure to identify disease or injury within various regions of the human body. The human beings’ health issues caused by CT radiation have attracted the interest of the researchers and academic community. Reducing the radiation dose is the solution, but the CT image generated with low-dose radiation results in excessive noise due to lower intensity and fewer angle measurements. Low-dose CT scan images reduce image quality and thus affect a doctor’s diagnosis. Deep learning methods have become increasingly popular in recent years, many models have been proposed for Low-Dose CT image reconstruction. Low-Dose CT Image Reconstruction is an active area of modern medical imaging research. Deep learning-based medical image reconstruction methods will be helpful to reduce noise without compromising image quality. Therefore, this paper introduces a novel CT image reconstruction method based on the vector quantization technique utilized in the convolutional autoencoder network. The quality of the results is evaluated based on the perceptual loss function. Experimental evaluations are conducted on the LoDoPaB-CT benchmark dataset. Its result showed that the proposed network obtained better performance metric values and better noise elimination results, in terms of quantitative and visual evaluation, respectively. | ['Mohan Ramasundaram', 'Shalini Ramanathan'] | 2023-03-28 | null | null | null | sadhana-2023-3 | ['image-reconstruction'] | ['computer-vision'] | [ 7.17448592e-02 -2.97142833e-01 -3.01859993e-02 -3.21232021e-01
-5.58811486e-01 2.09710807e-01 8.44277516e-02 3.66895080e-01
-6.09765351e-01 4.04215753e-01 3.59214813e-01 1.56577870e-01
-3.05313766e-01 -1.23661673e+00 -1.67779267e-01 -9.06402230e-01
-1.96560044e-02 2.35175073e-01 3.64295900e-01 -3.07957660e-02
7.62243494e-02 4.49158400e-01 -1.21643424e+00 3.40224147e-01
5.61241627e-01 1.07513952e+00 5.17131567e-01 3.63779217e-01
-9.39334258e-02 9.69919860e-01 -4.61623758e-01 -2.03156680e-01
3.18167895e-01 -5.08407354e-01 -7.13344276e-01 -2.71900301e-03
-2.51316756e-01 -5.03927767e-01 -5.87998152e-01 1.32869971e+00
9.62971866e-01 1.95546269e-01 6.14804506e-01 -6.32218003e-01
-6.84448361e-01 3.02541614e-01 -5.87499321e-01 7.03937829e-01
3.95104960e-02 8.47960711e-02 2.10843280e-01 -6.56712651e-01
1.27673388e-01 1.31041765e+00 5.77957034e-01 4.81641948e-01
-7.82742441e-01 -4.74189401e-01 -7.94308662e-01 5.71666002e-01
-1.29196227e+00 2.92484730e-01 7.35141933e-01 -3.18782777e-01
5.57638586e-01 3.81502122e-01 7.91893840e-01 5.01902819e-01
7.56033361e-01 4.48846757e-01 1.18255913e+00 -4.17922795e-01
8.11595097e-02 -8.17301869e-02 -3.11604500e-01 8.99581373e-01
3.53981763e-01 3.03972214e-01 1.01254784e-01 1.07303202e-01
9.89623368e-01 4.43934739e-01 -5.90704560e-01 -6.98289508e-03
-8.67333710e-01 1.10159767e+00 9.97409165e-01 7.60093331e-01
-5.56432664e-01 1.61216319e-01 6.64335310e-01 -1.45690545e-01
1.99767902e-01 6.40443340e-02 -3.58707295e-03 5.33103980e-02
-6.21757686e-01 6.60737930e-03 2.16546625e-01 1.76175117e-01
2.16588363e-01 1.23514809e-01 -1.56200781e-01 9.77194309e-01
3.48801821e-01 4.12486970e-01 1.01112497e+00 -8.59934986e-01
1.97011530e-01 5.70235133e-01 -4.58531052e-01 -1.42945004e+00
-5.37318468e-01 -5.73995948e-01 -1.35675108e+00 3.71709138e-01
-1.39270246e-01 1.68073505e-01 -1.11280692e+00 1.13879669e+00
5.14298201e-01 -8.97182822e-02 -1.12760380e-01 1.30542815e+00
1.14763927e+00 9.56858575e-01 2.79984951e-01 -3.26250672e-01
1.45621431e+00 -6.03579402e-01 -1.11329365e+00 1.80046394e-01
2.93064356e-01 -9.41974819e-01 9.61314917e-01 5.86927414e-01
-1.08403146e+00 -7.29953229e-01 -1.23580635e+00 -6.11693561e-02
5.20221442e-02 -6.89730654e-03 5.33829212e-01 8.41736674e-01
-6.43501341e-01 7.22948134e-01 -1.05840361e+00 -2.23002434e-01
7.70515859e-01 2.56008029e-01 -1.88182443e-01 -3.67543221e-01
-1.34160733e+00 1.00395656e+00 3.97720218e-01 9.24446136e-02
-7.22461760e-01 -6.17529631e-01 -5.96527100e-01 3.55610177e-02
1.55579969e-01 -6.81160569e-01 1.19013870e+00 -6.42919898e-01
-1.20835030e+00 6.26761496e-01 4.38250333e-01 -3.15267235e-01
3.20584178e-01 3.61956581e-02 -5.47944009e-01 6.27612352e-01
1.16800025e-01 4.80439425e-01 4.18975502e-01 -1.04929423e+00
-5.54443419e-01 -4.52239126e-01 -2.37911344e-01 3.55153292e-01
-1.62842646e-01 -4.69745919e-02 -2.90803373e-01 -8.02917242e-01
4.07460481e-01 -5.80658078e-01 -4.24854517e-01 1.75719485e-01
-9.37219709e-02 -6.83569238e-02 9.32355940e-01 -8.21126819e-01
1.17469859e+00 -1.77992201e+00 -1.39741153e-01 1.46409303e-01
1.46652311e-01 2.61650026e-01 5.08653998e-01 6.28664345e-02
-2.60325760e-01 2.73105890e-01 -4.57406282e-01 3.47461611e-01
-6.90945208e-01 2.26937950e-01 4.96664077e-01 6.58187568e-01
-2.15364352e-01 4.60931420e-01 -4.26170856e-01 -8.70216608e-01
5.85075080e-01 7.20394850e-01 -4.59328562e-01 1.67693481e-01
2.86587477e-01 5.74767530e-01 -5.98184943e-01 5.27462900e-01
1.00712955e+00 -2.44294405e-01 -3.57611150e-01 -7.19974816e-01
8.13058317e-02 -2.34132916e-01 -9.47949946e-01 1.56142056e+00
-4.80066150e-01 4.79725391e-01 -1.02634132e-01 -9.35789108e-01
5.64740539e-01 5.40200412e-01 8.45133901e-01 -1.30950272e+00
6.22267187e-01 6.53718561e-02 1.52793720e-01 -1.19209158e+00
-1.27060220e-01 -7.90839672e-01 2.23279133e-01 2.24586815e-01
-5.17409921e-01 -2.80626535e-01 -1.49450764e-01 -5.20620644e-02
8.25864434e-01 -4.70348626e-01 3.72554541e-01 -2.47617196e-02
5.96712410e-01 -2.43742205e-02 4.74012852e-01 1.55224770e-01
-4.31367815e-01 6.41354859e-01 -1.92486182e-01 -7.25469530e-01
-1.15155220e+00 -9.63707030e-01 -5.48166275e-01 2.93744057e-01
3.80020022e-01 1.79257840e-01 -7.61382282e-01 -2.16960639e-01
-4.27743584e-01 4.41759557e-01 -4.57884759e-01 -3.78129780e-01
-7.98029542e-01 -1.04528272e+00 4.78765309e-01 4.77570146e-01
1.10236394e+00 -1.17578292e+00 -9.25809562e-01 2.05171138e-01
-4.54613656e-01 -7.84649432e-01 -1.74236670e-01 -8.88507068e-03
-1.26362634e+00 -9.93555784e-01 -1.02202523e+00 -9.69649017e-01
4.88365233e-01 2.09440157e-01 8.44140291e-01 4.33876514e-01
-8.52023542e-01 1.00954093e-01 -4.04180974e-01 -3.25518817e-01
-5.29361129e-01 -4.81505781e-01 -2.88334101e-01 -3.83943379e-01
1.64472029e-01 -4.00660962e-01 -1.22830141e+00 8.70279297e-02
-1.36445975e+00 -1.90411538e-01 9.14501727e-01 9.83295798e-01
8.73964012e-01 8.08016598e-01 1.78174630e-01 -7.99334824e-01
6.05079770e-01 -3.89827132e-01 -3.28926057e-01 -7.38093853e-02
-5.15174210e-01 -4.75892611e-03 5.41865468e-01 5.85827790e-02
-1.30974090e+00 -3.91360819e-01 -6.30945563e-01 -2.41336569e-01
-1.17454015e-01 5.68073452e-01 -5.44881448e-02 -1.56752810e-01
7.43475139e-01 3.01344782e-01 -1.27256513e-01 -3.44928026e-01
-3.22831929e-01 6.82034492e-01 3.52773428e-01 8.19604471e-02
3.43731523e-01 5.41933417e-01 3.47003251e-01 -8.29288423e-01
-4.74743992e-01 -4.56915200e-01 -1.00147106e-01 -4.58499849e-01
1.30507016e+00 -7.90060461e-01 -8.26998711e-01 1.76156670e-01
-1.12456882e+00 3.73585343e-01 6.47206753e-02 1.04453492e+00
-1.71900824e-01 6.33130074e-01 -8.57864082e-01 -3.82528991e-01
-6.49524927e-01 -1.61037743e+00 5.85788012e-01 3.98496896e-01
2.36977041e-01 -7.76228189e-01 3.73784490e-02 4.44391876e-01
4.40315187e-01 5.34518361e-01 1.35138810e+00 2.41347980e-02
-6.02944672e-01 -2.09179968e-01 -2.82472253e-01 6.53069973e-01
3.14257890e-01 -5.70614696e-01 -7.96606719e-01 -2.14082241e-01
7.44154930e-01 -1.95048794e-01 7.65542924e-01 8.23936760e-01
1.80786979e+00 -1.70234740e-01 -5.45468293e-02 7.72708654e-01
2.01503587e+00 6.58386230e-01 9.18277323e-01 4.01872903e-01
6.48828506e-01 3.07687223e-01 5.50367892e-01 4.23209101e-01
-1.44619405e-01 3.47862661e-01 7.40019619e-01 -4.10990566e-01
-5.52024484e-01 -2.18168879e-03 -3.39623451e-01 1.04979670e+00
-1.48933798e-01 -2.54722834e-01 -8.18737924e-01 4.24618214e-01
-1.12535143e+00 -1.05922997e+00 -5.33961356e-01 1.80719948e+00
7.02191532e-01 7.76786404e-03 -2.71480858e-01 7.18615413e-01
7.11795330e-01 -1.47694066e-01 -3.56369376e-01 -3.05859268e-01
1.46362782e-01 6.67338550e-01 6.44174814e-01 1.84382334e-01
-1.02279031e+00 9.42769051e-02 5.47365332e+00 1.07894933e+00
-1.45224106e+00 4.35997933e-01 7.86904752e-01 1.21019967e-02
-1.28421396e-01 -4.90148813e-01 -1.22607380e-01 7.21775711e-01
5.48376024e-01 1.00920834e-01 -1.15915872e-01 8.00460517e-01
5.69710195e-01 -4.93757010e-01 -5.38631022e-01 1.36579788e+00
-1.61840409e-01 -1.18554938e+00 2.95928389e-01 -5.34488335e-02
5.48359632e-01 -2.25867301e-01 3.27424586e-01 -1.27803665e-02
-2.32597604e-01 -1.26008856e+00 2.14716867e-01 3.96166831e-01
7.61564851e-01 -1.20427001e+00 1.17463589e+00 2.90415585e-01
-9.93043303e-01 -5.70366066e-03 -6.42423093e-01 1.83287904e-01
1.41138747e-01 8.25471282e-01 -7.40875185e-01 5.14674366e-01
1.07382333e+00 2.18351081e-01 -3.43198180e-01 1.36000264e+00
3.61685529e-02 6.39365613e-01 -1.34432778e-01 -7.81177059e-02
2.15139449e-01 -1.04176238e-01 3.21637958e-01 9.44383979e-01
4.47614014e-01 5.79986572e-01 1.16006546e-01 5.47023058e-01
-4.52095158e-02 4.41974372e-01 -5.18282712e-01 5.26038468e-01
9.74148437e-02 9.27059472e-01 -9.81113970e-01 -3.02890509e-01
-2.82850891e-01 9.02322173e-01 -3.83746028e-01 -1.24780498e-01
-9.32740152e-01 -2.85379499e-01 -1.32353246e-01 4.82455224e-01
-6.52890429e-02 2.10755527e-01 -4.41482544e-01 -5.70186734e-01
-1.40634745e-01 -8.16976309e-01 5.15831292e-01 -6.54088914e-01
-9.96691465e-01 6.69854462e-01 -1.26212277e-02 -1.38073087e+00
2.21182197e-01 -3.75772715e-01 -4.33464020e-01 7.80768037e-01
-1.29988933e+00 -6.67882144e-01 -5.82596898e-01 6.60762727e-01
7.14887857e-01 8.60646740e-02 7.43127882e-01 8.43953311e-01
-2.99610853e-01 2.84905612e-01 8.23554844e-02 7.81366825e-02
3.37554395e-01 -1.09040928e+00 -5.08840799e-01 7.06171334e-01
-4.06992793e-01 2.78696388e-01 7.06238925e-01 -5.86998880e-01
-9.17658150e-01 -1.02859592e+00 2.59890616e-01 3.75465900e-01
-2.39928424e-01 4.80264515e-01 -8.89274120e-01 1.46869794e-01
3.34708929e-01 2.71140963e-01 7.42614985e-01 -7.91285455e-01
5.48407793e-01 -2.18777165e-01 -1.80227470e+00 2.22974643e-01
3.68490100e-01 -1.30570471e-01 -4.95017022e-01 3.24865878e-01
4.99855936e-01 -3.50937039e-01 -1.06487584e+00 5.29256105e-01
3.32521915e-01 -1.20495987e+00 1.15608907e+00 -2.77880263e-02
6.32341087e-01 -3.19690108e-01 -7.81052187e-02 -1.26169574e+00
-6.48912311e-01 5.91655672e-01 5.80816865e-01 6.35041416e-01
-9.69363302e-02 -7.90962204e-02 9.50722158e-01 2.30982751e-01
-2.01671392e-01 -9.40338016e-01 -9.15163040e-01 -3.45033824e-01
4.83576655e-02 -2.67615616e-01 2.64948010e-01 7.90844321e-01
-4.34264958e-01 6.14685044e-02 -1.50761127e-01 9.32212844e-02
7.58548141e-01 -3.21591049e-01 7.75032490e-02 -9.73109484e-01
-2.98899621e-01 -2.59480655e-01 -7.15277672e-01 -6.69296324e-01
-7.01897740e-01 -8.59073222e-01 -1.23024076e-01 -1.92146909e+00
4.15089399e-01 -1.84659138e-01 -1.97562292e-01 8.70652199e-02
-1.46267369e-01 4.29556251e-01 -3.84649448e-02 2.86140561e-01
-1.96998734e-02 7.28885233e-01 1.94249499e+00 -3.64910692e-01
1.62379175e-01 -5.52799329e-02 -3.11560839e-01 9.55000579e-01
8.66854846e-01 -7.92863548e-01 -4.53665704e-01 -6.35222137e-01
5.81794605e-02 3.77768666e-01 3.31237406e-01 -1.45590270e+00
2.19670027e-01 7.00942576e-02 9.95826066e-01 -7.54957736e-01
2.48713911e-01 -1.15135384e+00 1.79936662e-01 1.15129924e+00
-1.08358651e-01 7.87319988e-02 1.44595861e-01 5.29869676e-01
-5.76943934e-01 -4.76144075e-01 1.50242746e+00 -6.39159501e-01
-6.61712885e-01 3.10716987e-01 -3.70138854e-01 -3.03227246e-01
1.19580340e+00 -2.74676710e-01 3.77376139e-01 -2.60865152e-01
-6.37859881e-01 -2.56645888e-01 -4.93721552e-02 -1.32036164e-01
1.14677882e+00 -1.52024484e+00 -5.41438639e-01 -3.60330865e-02
-7.24965483e-02 2.53881067e-01 7.42233098e-01 7.75333345e-01
-1.41141176e+00 2.67788261e-01 -5.21457970e-01 -7.29281783e-01
-1.27370787e+00 4.01489347e-01 5.52085042e-01 -5.00362337e-01
-6.22693956e-01 9.42076385e-01 1.36274636e-01 2.65436191e-02
1.21470001e-02 -5.82514346e-01 -4.33267534e-01 -4.56284076e-01
4.54392761e-01 6.10418379e-01 4.08397168e-01 -5.84260941e-01
-1.88980252e-01 7.44288743e-01 -1.18206464e-01 -4.54402231e-02
1.40708745e+00 -2.31636852e-01 1.97637989e-03 5.54016605e-02
1.52765918e+00 -2.87507176e-01 -5.24795353e-01 -2.66682226e-02
-2.98935026e-01 -6.51444733e-01 8.34423363e-01 -9.56190228e-01
-1.49273241e+00 1.12289333e+00 1.59496498e+00 1.39669165e-01
1.43753791e+00 -2.37573907e-01 1.35305798e+00 6.47751018e-02
3.26782167e-01 -1.04678261e+00 4.92680341e-01 -1.94847152e-01
8.29287350e-01 -1.61141598e+00 2.94583291e-01 -4.58734244e-01
-4.28148359e-01 1.21072721e+00 4.89515215e-01 -2.50734657e-01
8.49685550e-01 2.67836601e-01 2.19022781e-01 -4.42351818e-01
-6.08031303e-02 1.63544059e-01 -5.25131449e-02 6.00312948e-01
6.50276303e-01 -3.49307358e-02 -7.52811730e-01 5.18255830e-01
-3.15375179e-02 1.87402353e-01 4.17576849e-01 1.03222477e+00
-8.26571703e-01 -8.85166705e-01 -9.30415332e-01 6.63007319e-01
-1.03732979e+00 1.79783121e-01 2.85671681e-01 7.34477341e-01
5.43620408e-01 8.90067160e-01 -2.34911352e-01 -1.75235897e-01
3.52735251e-01 -6.44515276e-01 4.97099876e-01 -4.43239689e-01
-4.82032627e-01 1.39288113e-01 -4.89954770e-01 -3.73092264e-01
-6.37941122e-01 -1.19103044e-01 -1.66411638e+00 -3.51479024e-01
-2.25760728e-01 6.70780987e-02 8.94997358e-01 6.62993729e-01
-3.10871154e-01 1.04173422e+00 6.37190521e-01 -3.89456332e-01
-3.00987601e-01 -9.43130612e-01 -6.01525366e-01 5.62424660e-01
1.02407657e-01 -6.71581030e-01 -8.91952068e-02 9.78589505e-02] | [13.467547416687012, -2.5304830074310303] |
8aecf466-eb6f-4c1c-ae84-f6dd45ebead6 | sf-fsda-source-free-few-shot-domain-adaptive | 2306.04385 | null | https://arxiv.org/abs/2306.04385v1 | https://arxiv.org/pdf/2306.04385v1.pdf | SF-FSDA: Source-Free Few-Shot Domain Adaptive Object Detection with Efficient Labeled Data Factory | Domain adaptive object detection aims to leverage the knowledge learned from a labeled source domain to improve the performance on an unlabeled target domain. Prior works typically require the access to the source domain data for adaptation, and the availability of sufficient data on the target domain. However, these assumptions may not hold due to data privacy and rare data collection. In this paper, we propose and investigate a more practical and challenging domain adaptive object detection problem under both source-free and few-shot conditions, named as SF-FSDA. To overcome this problem, we develop an efficient labeled data factory based approach. Without accessing the source domain, the data factory renders i) infinite amount of synthesized target-domain like images, under the guidance of the few-shot image samples and text description from the target domain; ii) corresponding bounding box and category annotations, only demanding minimum human effort, i.e., a few manually labeled examples. On the one hand, the synthesized images mitigate the knowledge insufficiency brought by the few-shot condition. On the other hand, compared to the popular pseudo-label technique, the generated annotations from data factory not only get rid of the reliance on the source pretrained object detection model, but also alleviate the unavoidably pseudo-label noise due to domain shift and source-free condition. The generated dataset is further utilized to adapt the source pretrained object detection model, realizing the robust object detection under SF-FSDA. The experiments on different settings showcase that our proposed approach outperforms other state-of-the-art methods on SF-FSDA problem. Our codes and models will be made publicly available. | ['Luc van Gool', 'Konrad Schindler', 'Rui Gong', 'Han Sun'] | 2023-06-07 | null | null | null | null | ['robust-object-detection', 'pseudo-label'] | ['computer-vision', 'miscellaneous'] | [ 4.49208081e-01 6.02343865e-02 -2.12985903e-01 -3.53969306e-01
-7.70700157e-01 -5.47245026e-01 4.89824951e-01 -9.45781544e-02
-3.65051299e-01 7.34742820e-01 -3.53078306e-01 1.92610443e-01
1.14587910e-01 -6.09055042e-01 -7.18798459e-01 -7.94881523e-01
4.95892018e-01 5.24458230e-01 5.81796587e-01 8.28967839e-02
6.46247640e-02 2.19766483e-01 -1.62894917e+00 1.12965867e-01
9.26472485e-01 1.25614583e+00 4.85151052e-01 2.46987909e-01
-3.05728137e-01 4.21812385e-01 -5.77613294e-01 -2.54239917e-01
6.24080956e-01 -3.43217224e-01 -2.99850792e-01 7.03299880e-01
4.05113161e-01 -8.04159880e-01 -9.84363332e-02 1.43546259e+00
4.59791631e-01 1.45310611e-01 5.79790294e-01 -1.52372158e+00
-7.83128440e-01 2.64480144e-01 -7.01042295e-01 9.68738124e-02
1.43348724e-01 3.53904337e-01 5.27597487e-01 -1.24766183e+00
8.00267696e-01 1.10205841e+00 3.53652865e-01 8.07229936e-01
-1.02028799e+00 -9.59627032e-01 3.94462436e-01 6.58332184e-03
-1.59832966e+00 -6.39314175e-01 8.36439073e-01 -3.35260630e-01
1.89478830e-01 -5.81253543e-02 2.60980248e-01 1.21981549e+00
-5.50826192e-01 8.47090006e-01 1.01319551e+00 -4.64782238e-01
4.38480228e-01 7.14187086e-01 1.85734004e-01 6.00680888e-01
5.18286765e-01 1.49506643e-01 -3.35571855e-01 -2.18932733e-01
6.37631297e-01 2.20129937e-01 -2.39309996e-01 -7.77488768e-01
-1.10536134e+00 4.93424118e-01 2.32663959e-01 4.17336356e-03
-2.65544116e-01 -4.89820331e-01 4.55235064e-01 1.14083111e-01
4.30238962e-01 -6.56546801e-02 -5.26021957e-01 3.79804492e-01
-7.94998825e-01 1.71054706e-01 5.91655910e-01 1.69036269e+00
9.84467506e-01 1.13330908e-01 -1.77039579e-01 7.09136844e-01
3.45517993e-01 7.59804308e-01 4.25338328e-01 -6.39057398e-01
5.66097856e-01 8.19032848e-01 4.64442819e-01 -7.56993830e-01
2.61841174e-02 -4.98689294e-01 -5.60166240e-01 2.41095319e-01
6.91792309e-01 -4.96099144e-02 -1.15144193e+00 1.69608164e+00
7.63129473e-01 2.50014365e-01 1.71246096e-01 1.08133566e+00
6.90457642e-01 4.76851046e-01 1.45654887e-01 -4.83307153e-01
1.34192753e+00 -8.52480173e-01 -7.84695566e-01 -3.70916754e-01
5.28961062e-01 -5.34268498e-01 1.32996130e+00 2.52310574e-01
-6.03331625e-01 -6.62837923e-01 -1.19022954e+00 1.52815759e-01
-3.89740556e-01 4.11585003e-01 2.09139019e-01 6.92802191e-01
-3.49283218e-01 3.77944633e-02 -4.75199401e-01 -5.09480119e-01
7.32585669e-01 1.49721965e-01 -3.27257484e-01 -4.94081825e-01
-9.36576247e-01 4.62022007e-01 8.29705060e-01 -1.55735239e-01
-1.31570113e+00 -6.14114344e-01 -5.63005924e-01 -6.99071214e-02
1.02396607e+00 -4.02343661e-01 1.33869922e+00 -1.29242051e+00
-1.18515801e+00 9.35718954e-01 2.09320094e-02 -3.03975761e-01
7.89295018e-01 -2.48150080e-01 -4.80369717e-01 1.71816334e-01
3.06268424e-01 5.13331294e-01 1.22271037e+00 -1.45339417e+00
-9.88892913e-01 -4.12248105e-01 -5.70228621e-02 1.58361271e-01
-5.47864735e-01 -8.64626914e-02 -5.63246846e-01 -5.80756724e-01
-4.86894548e-02 -7.03275681e-01 -9.06745791e-02 4.57399338e-01
-3.95078003e-01 -1.74907267e-01 1.27827513e+00 -4.04598922e-01
9.48950171e-01 -2.47867990e+00 -3.29582274e-01 -7.38513321e-02
4.76675034e-02 5.10184407e-01 -1.04007088e-01 -1.60937328e-02
4.76397313e-02 -2.00514346e-01 -4.83629107e-01 -1.61033839e-01
-1.57246456e-01 9.42519456e-02 -3.60345721e-01 4.28662479e-01
3.04748088e-01 4.34699863e-01 -1.10997856e+00 -8.21435273e-01
1.80957794e-01 9.42371413e-03 -3.08731496e-01 3.16890776e-01
-5.61071813e-01 4.49576080e-01 -6.63029492e-01 8.78647685e-01
9.81243193e-01 -1.93183959e-01 -8.49855617e-02 -2.32994407e-01
7.50064924e-02 -2.90126413e-01 -1.70780563e+00 1.69897354e+00
-4.66431156e-02 7.19339699e-02 1.98302656e-01 -9.39072609e-01
1.15709960e+00 2.84825414e-01 2.68532097e-01 -5.39036453e-01
1.54061913e-01 3.97574574e-01 -1.74721122e-01 -5.35546720e-01
2.74159670e-01 -1.35507032e-01 -4.88503575e-02 3.72096032e-01
1.51416138e-01 2.57374913e-01 1.77669585e-01 2.71611959e-01
8.43049884e-01 3.82987857e-01 6.01485372e-01 -1.49844764e-02
6.34036362e-01 3.24660838e-01 8.96407962e-01 8.08350921e-01
-6.07220531e-01 5.86183012e-01 -1.18242139e-02 -2.40055442e-01
-1.08055091e+00 -1.03053606e+00 -1.12264246e-01 1.24317110e+00
5.16860723e-01 1.14498280e-01 -8.47947359e-01 -1.18445837e+00
-8.95918831e-02 7.12918222e-01 -3.67690742e-01 -1.60553128e-01
-3.18960965e-01 -4.74958450e-01 4.23219472e-01 4.23402041e-01
7.82744408e-01 -8.89562488e-01 -4.85675097e-01 1.80198953e-01
-9.93177295e-02 -1.21091795e+00 -6.16265655e-01 8.50041434e-02
-7.65710473e-01 -1.22442842e+00 -8.46924782e-01 -7.57211387e-01
8.92413199e-01 6.06656194e-01 5.93286335e-01 -1.86301321e-01
-3.33173245e-01 3.26201349e-01 -5.32736778e-01 -5.52914500e-01
-4.23845530e-01 -2.02143922e-01 2.28304341e-01 3.82747918e-01
7.28183985e-01 -2.41712093e-01 -4.67337072e-01 4.92270291e-01
-1.00036287e+00 -9.50050503e-02 6.96994603e-01 7.68686891e-01
7.96837151e-01 8.09556022e-02 9.27641809e-01 -1.13153529e+00
2.03251660e-01 -6.16018653e-01 -7.50723124e-01 2.87818491e-01
-6.22654736e-01 -2.53303051e-01 6.30557477e-01 -9.27400768e-01
-1.51762962e+00 6.48832142e-01 4.81172800e-01 -7.68949151e-01
-4.43369895e-01 -5.50450981e-02 -7.29534686e-01 6.01380169e-02
1.05172992e+00 2.93946356e-01 -1.55930907e-01 -5.74763477e-01
4.21493262e-01 1.01111257e+00 6.93078160e-01 -5.18839061e-01
1.09343326e+00 6.98360145e-01 -3.10677826e-01 -5.50161541e-01
-9.18026805e-01 -6.44700110e-01 -7.03575909e-01 -1.15147710e-01
5.82139730e-01 -1.02474296e+00 2.61168648e-02 4.37477440e-01
-1.05404997e+00 7.94792026e-02 -4.55990195e-01 3.59772950e-01
-4.05571371e-01 4.26391423e-01 -7.41633773e-02 -1.06233239e+00
-2.87903070e-01 -8.77527177e-01 9.78977025e-01 3.51689994e-01
3.04867595e-01 -4.30834025e-01 -4.22873646e-01 2.07151264e-01
3.06232013e-02 1.80613577e-01 6.80852950e-01 -1.12473726e+00
-8.59211564e-01 -3.55912745e-01 -5.96035421e-01 5.06521940e-01
3.09169799e-01 -3.08089882e-01 -1.25269926e+00 -2.98373699e-01
2.53901362e-01 -3.04536641e-01 5.22394061e-01 1.89969502e-02
9.35864151e-01 -3.02963823e-01 -4.61162269e-01 2.93067127e-01
1.39427209e+00 2.46365279e-01 1.79183662e-01 2.71908253e-01
5.90122640e-01 6.50350511e-01 1.34182680e+00 6.53800070e-01
1.11935169e-01 5.42331278e-01 4.03172165e-01 -2.86431313e-02
-3.50435823e-01 -3.81182998e-01 2.42747799e-01 2.55700052e-01
3.47631872e-01 -2.87602544e-01 -8.46886098e-01 6.68254852e-01
-1.92343569e+00 -6.61961854e-01 -1.01912200e-01 2.36408520e+00
8.65159273e-01 1.49714798e-01 2.57886976e-01 2.23640259e-02
1.21505153e+00 -2.24254414e-01 -9.84026015e-01 3.14530015e-01
-8.37306604e-02 -3.29717189e-01 5.81516862e-01 3.93524356e-02
-1.21878052e+00 9.28876638e-01 4.93182802e+00 9.46909606e-01
-8.49759459e-01 4.41628307e-01 2.13110149e-01 -2.76046172e-02
1.93092808e-01 4.20424379e-02 -1.05382490e+00 5.58042347e-01
4.73091692e-01 -3.96855831e-01 1.99956536e-01 1.47470284e+00
-5.55082224e-02 -4.20657545e-02 -1.26254535e+00 9.87529933e-01
1.03418201e-01 -7.96619236e-01 -1.29771950e-02 -1.10691870e-02
6.22068226e-01 -3.35272282e-01 -1.09056637e-01 4.80298460e-01
2.33428985e-01 -3.44155699e-01 8.29489708e-01 2.76566803e-01
1.21583617e+00 -4.39673573e-01 6.03205860e-01 8.50554645e-01
-1.15226233e+00 -3.36215436e-01 -6.15785301e-01 2.08669528e-01
-2.31910981e-02 6.08695626e-01 -1.02840757e+00 5.39494693e-01
6.18755937e-01 5.00066400e-01 -4.67708558e-01 1.01824796e+00
-9.55232456e-02 5.50403357e-01 -3.54570866e-01 2.37311751e-01
-3.55801247e-02 1.42812416e-01 7.34712780e-01 1.08222055e+00
3.24776500e-01 1.97294012e-01 5.26127398e-01 9.04529333e-01
-1.22048408e-01 1.35504156e-01 -7.28167295e-01 2.05220729e-02
7.94334948e-01 1.16242671e+00 -6.17240787e-01 -5.63852727e-01
-5.61256528e-01 9.40621316e-01 1.26204848e-01 4.55522984e-01
-7.71219730e-01 -3.60281616e-01 3.61543298e-01 2.28965640e-01
3.31208080e-01 1.39121518e-01 -2.55063951e-01 -1.19796455e+00
2.12236330e-01 -8.82355273e-01 5.64110816e-01 -5.86422324e-01
-1.52538168e+00 3.02429259e-01 1.48520902e-01 -1.66014898e+00
-6.48585334e-02 -4.71651137e-01 -1.67162210e-01 6.87393010e-01
-1.54929221e+00 -1.20746899e+00 -6.09921575e-01 7.96728432e-01
8.13067317e-01 -2.90003061e-01 6.80379629e-01 5.30653000e-01
-6.56230032e-01 7.11276770e-01 1.36269955e-02 1.11491956e-01
1.07081449e+00 -8.77211869e-01 -4.26963484e-03 1.04809594e+00
-1.34401754e-01 4.24550712e-01 6.31748021e-01 -8.08857083e-01
-1.13964868e+00 -1.47305381e+00 2.99334019e-01 -3.39807063e-01
3.92306358e-01 -4.98156816e-01 -1.22149253e+00 6.53238893e-01
-3.97787750e-01 4.20840681e-01 4.22221869e-01 -4.48046446e-01
-4.71822172e-01 -3.30718398e-01 -1.63795710e+00 3.30331534e-01
1.19871545e+00 -3.05106819e-01 -7.38231480e-01 3.48931521e-01
8.24121773e-01 -2.97763258e-01 -4.04913455e-01 3.93116117e-01
2.55095512e-01 -6.69822156e-01 8.06310713e-01 -3.94991547e-01
3.42781693e-02 -7.52996206e-01 -2.51776218e-01 -8.23214293e-01
-1.42068535e-01 -2.16697946e-01 -2.45381474e-01 1.64889884e+00
8.48783478e-02 -4.93766040e-01 7.46548831e-01 7.23090649e-01
-9.32266265e-02 -1.34442568e-01 -9.17991221e-01 -1.11266935e+00
-3.66880596e-01 -2.26542413e-01 5.85577488e-01 9.39048529e-01
-4.62235630e-01 2.29803607e-01 -4.21264082e-01 5.04587889e-01
8.91595006e-01 8.45687836e-02 1.03479910e+00 -1.20504069e+00
-1.59209847e-01 1.23387814e-01 -1.70397595e-01 -1.08582449e+00
-8.96190777e-02 -7.05633879e-01 4.02734637e-01 -1.11940539e+00
3.84139299e-01 -6.32326007e-01 -4.16959375e-01 5.35981417e-01
-3.26406986e-01 1.20650880e-01 2.83212095e-01 4.83765453e-01
-8.34480226e-01 4.32375818e-01 1.11863792e+00 -2.49133781e-02
-2.18950242e-01 -6.34329244e-02 -6.80988729e-01 6.99470401e-01
5.72692990e-01 -8.24399412e-01 -5.82739353e-01 -2.85921007e-01
-3.73209357e-01 -2.32671022e-01 4.74998474e-01 -1.00299132e+00
2.28718907e-01 -3.58598113e-01 3.82495165e-01 -4.91768688e-01
2.24923730e-01 -1.23683751e+00 -1.13780797e-01 2.72551924e-01
-5.41988090e-02 -8.11788201e-01 9.35742911e-03 1.13189900e+00
-7.64973089e-03 -3.89179260e-01 1.01830256e+00 -2.72291213e-01
-1.15728915e+00 3.72286320e-01 1.15969181e-01 8.04969594e-02
1.50672424e+00 -5.44300735e-01 -3.58842522e-01 4.72411998e-02
-5.12205362e-01 2.09572002e-01 6.38895988e-01 4.71796185e-01
4.78842139e-01 -1.27441132e+00 -5.62929928e-01 3.75244945e-01
5.94329357e-01 3.69615704e-01 2.51632422e-01 4.73395079e-01
1.82295199e-02 7.89906085e-02 -1.18635468e-01 -7.09823072e-01
-1.02412343e+00 1.16015780e+00 1.47224829e-01 1.35743305e-01
-5.97398639e-01 6.66269898e-01 7.29118049e-01 -3.37867290e-01
3.80132079e-01 4.15110663e-02 1.23247743e-01 2.19565779e-02
7.49173284e-01 4.34838146e-01 -9.80677903e-02 -4.39077377e-01
-3.57379943e-01 2.76097089e-01 -2.79374212e-01 8.94761011e-02
9.31238174e-01 -5.06769121e-01 4.50601369e-01 4.32477623e-01
7.37292528e-01 -1.50561675e-01 -1.71832538e+00 -8.45262051e-01
7.09651187e-02 -7.25699604e-01 -2.19256014e-01 -9.58355844e-01
-7.28173494e-01 7.79438436e-01 9.60705876e-01 -4.13710177e-02
1.27483821e+00 -3.13254558e-02 7.35861361e-01 3.83862317e-01
5.92982292e-01 -1.26083112e+00 2.31329232e-01 9.92636979e-02
5.72943449e-01 -1.53260779e+00 -1.80611238e-01 -6.20264709e-01
-8.56622875e-01 9.15900886e-01 1.05218709e+00 -2.99220290e-02
3.56558740e-01 1.63619831e-01 1.03995420e-01 1.49186581e-01
-3.73763591e-01 -3.96472305e-01 1.73348002e-02 9.45781112e-01
-2.60799497e-01 -6.43908828e-02 3.79737903e-04 9.91636872e-01
4.93590593e-01 3.28540385e-01 3.23629379e-01 8.96669567e-01
-8.35047126e-01 -9.28039074e-01 -6.73001528e-01 3.57874632e-01
-2.34762788e-01 1.43390685e-01 -3.84636164e-01 8.43914032e-01
5.02821028e-01 9.29017782e-01 -1.42415971e-01 1.82358529e-02
4.35324162e-01 3.39121252e-01 1.34330899e-01 -1.06929171e+00
-3.48503999e-02 3.05166766e-02 -1.43132061e-01 -1.77264065e-01
-2.53235132e-01 -5.81124306e-01 -1.33730090e+00 1.77007571e-01
-7.15296268e-01 -2.49331892e-01 5.07641256e-01 8.46136808e-01
4.01720822e-01 2.17228130e-01 5.80060899e-01 -6.11563683e-01
-8.23781788e-01 -1.01597595e+00 -6.86161458e-01 6.43517077e-01
3.85599136e-01 -8.51723671e-01 -3.19147527e-01 3.52154315e-01] | [9.432158470153809, 1.5762228965759277] |
dbad2daf-8461-495b-9542-df0922c91874 | divide-and-conquer-based-large-scale-spectral-1 | 2104.15042 | null | https://arxiv.org/abs/2104.15042v2 | https://arxiv.org/pdf/2104.15042v2.pdf | Divide-and-conquer based Large-Scale Spectral Clustering | Spectral clustering is one of the most popular clustering methods. However, how to balance the efficiency and effectiveness of the large-scale spectral clustering with limited computing resources has not been properly solved for a long time. In this paper, we propose a divide-and-conquer based large-scale spectral clustering method to strike a good balance between efficiency and effectiveness. In the proposed method, a divide-and-conquer based landmark selection algorithm and a novel approximate similarity matrix approach are designed to construct a sparse similarity matrix within low computational complexities. Then clustering results can be computed quickly through a bipartite graph partition process. The proposed method achieves a lower computational complexity than most existing large-scale spectral clustering methods. Experimental results on ten large-scale datasets have demonstrated the efficiency and effectiveness of the proposed method. The MATLAB code of the proposed method and experimental datasets are available at https://github.com/Li-Hongmin/MyPaperWithCode. | ['Tetsuya Sakurai', 'Akira Imakura', 'Xiucai Ye', 'Hongmin Li'] | 2021-04-30 | null | null | null | null | ['imagedocument-clustering'] | ['computer-vision'] | [-3.09017539e-01 -4.61265475e-01 -1.69513240e-01 -2.30744034e-01
-8.34110260e-01 -5.13991416e-01 8.08422714e-02 9.83225405e-02
-2.67859727e-01 4.62432176e-01 -2.37775147e-02 -1.87954664e-01
-5.47698677e-01 -8.09919775e-01 -2.01865807e-01 -9.15582299e-01
9.86373201e-02 4.94033486e-01 5.99415362e-01 3.68825734e-01
2.64783919e-01 1.40009820e-01 -1.11115646e+00 -1.32834896e-01
1.12681615e+00 6.87224925e-01 4.72045988e-01 2.81246841e-01
-1.23145640e-01 2.12398067e-01 -2.45532572e-01 1.13919027e-01
4.57308441e-01 -6.40921235e-01 -6.40430272e-01 3.38121682e-01
-1.04393631e-01 1.62693009e-01 -2.23288596e-01 1.22438133e+00
7.81042516e-01 2.71822572e-01 3.51524383e-01 -1.39581096e+00
-2.98786342e-01 6.73504293e-01 -1.27147913e+00 1.30831987e-01
2.51677185e-01 -1.71789154e-01 7.64452994e-01 -7.72990108e-01
4.60583776e-01 1.05556488e+00 6.36016905e-01 -9.27530825e-02
-1.29427183e+00 -1.07597804e+00 3.12733091e-02 3.13057333e-01
-2.23039103e+00 -2.17703462e-01 8.67038071e-01 -1.97598636e-01
3.83972019e-01 2.98994362e-01 7.29832947e-01 1.15078695e-01
-4.49167013e-01 6.53560221e-01 1.16871667e+00 -1.80549011e-01
3.56516004e-01 -5.74121289e-02 3.31043676e-02 7.30157733e-01
4.30752963e-01 -3.75429899e-01 -8.60779583e-02 -4.77366537e-01
5.64862192e-01 3.96334171e-01 -2.70471573e-01 -6.25209570e-01
-1.15997231e+00 7.81625390e-01 5.28732240e-01 4.31162775e-01
-1.76813111e-01 1.00854188e-01 3.40968937e-01 6.35185763e-02
2.28830367e-01 -2.07859546e-01 8.53875768e-04 4.83939936e-03
-1.28721869e+00 1.66908316e-02 5.33295870e-01 9.84658062e-01
1.07141149e+00 -2.11715654e-01 3.36219758e-01 1.03218579e+00
2.16435567e-01 6.21467650e-01 4.71814096e-01 -7.58154511e-01
2.55575299e-01 8.14277112e-01 -1.74193427e-01 -1.54307878e+00
-4.17907953e-01 -2.24037483e-01 -1.15429938e+00 -3.58051747e-01
2.91222960e-01 -1.29791141e-01 -4.31241512e-01 1.34816921e+00
8.22400212e-01 6.62542045e-01 -1.74457893e-01 1.05130744e+00
5.06927133e-01 8.98860633e-01 -1.78218707e-01 -6.35984659e-01
1.24071991e+00 -9.23320889e-01 -5.82205117e-01 1.17496833e-01
4.91718680e-01 -9.46953714e-01 7.99225569e-01 1.71160370e-01
-9.46580827e-01 -3.21381956e-01 -9.54624176e-01 3.25364888e-01
2.98524089e-03 1.73178896e-01 5.58938682e-01 5.49396873e-01
-8.85393858e-01 2.29317486e-01 -9.99092996e-01 -7.35894501e-01
2.77239352e-01 4.72373128e-01 -1.85693517e-01 -9.08378065e-02
-8.30827177e-01 1.24570429e-01 6.70694172e-01 8.13013539e-02
-3.62756252e-01 -3.59969765e-01 -3.94966334e-01 8.15070421e-02
8.52049351e-01 -3.88080657e-01 9.17831838e-01 -7.09545851e-01
-1.09501481e+00 5.11482716e-01 -2.40249977e-01 3.71710286e-02
2.20878959e-01 1.88871786e-01 -2.21152395e-01 3.85278136e-01
3.09653133e-01 2.09064424e-01 2.60347664e-01 -1.28001785e+00
-7.87724137e-01 -4.48551953e-01 -5.07158279e-01 3.13853383e-01
-4.41449702e-01 3.37995999e-02 -8.31077933e-01 -5.24844289e-01
5.44516742e-01 -1.11669552e+00 -5.45723617e-01 -3.66200358e-01
-4.05660361e-01 -2.33046919e-01 6.95991218e-01 -2.26209298e-01
1.71580803e+00 -2.24080563e+00 3.77797969e-02 7.82346010e-01
1.87878907e-01 2.29147404e-01 1.22219443e-01 7.70170987e-01
-4.92316559e-02 6.70502931e-02 -3.26151997e-01 2.27793053e-01
-1.45803794e-01 -3.10059041e-01 2.21023962e-01 7.67902613e-01
-5.84472835e-01 4.90490764e-01 -9.56686676e-01 -8.45893502e-01
1.47156477e-01 1.99643582e-01 -2.76458949e-01 1.16865389e-01
2.72873282e-01 -1.04538910e-02 -6.98946297e-01 6.30559444e-01
9.75766540e-01 -4.53326643e-01 6.17529631e-01 -3.76835883e-01
-8.22239518e-02 -3.63667607e-01 -1.93866849e+00 1.63019729e+00
-2.74675488e-02 1.16471484e-01 3.19032341e-01 -1.25056839e+00
9.44883406e-01 2.39772111e-01 8.43124866e-01 -3.84070665e-01
2.98241645e-01 5.18946648e-01 2.08834466e-02 -2.44774297e-01
2.06546441e-01 -1.53481722e-01 -2.14411244e-02 7.63438642e-01
-5.34969032e-01 6.29101023e-02 6.12984717e-01 4.78006631e-01
1.10356343e+00 -2.30223149e-01 5.51176071e-01 -5.39590895e-01
7.01814771e-01 3.23520213e-01 8.75401497e-01 3.10727894e-01
-3.25011998e-01 5.02085388e-01 1.84484497e-01 -7.45709687e-02
-9.08164859e-01 -9.12014484e-01 1.73758000e-01 6.82919025e-01
5.69837630e-01 -6.88193202e-01 -8.88877451e-01 -3.18576306e-01
-4.38036211e-02 7.17285881e-03 -2.13168353e-01 7.37332851e-02
-4.05335814e-01 -7.87176013e-01 4.08136487e-01 2.21595824e-01
6.96839035e-01 -6.86298013e-01 -3.63079935e-01 1.47590429e-01
-3.53929758e-01 -8.11297953e-01 -7.70434737e-01 -2.70691931e-01
-8.63532484e-01 -1.33380282e+00 -6.11718714e-01 -1.21615946e+00
9.86172378e-01 1.01373351e+00 4.71889049e-01 4.02358025e-01
-3.99814546e-01 -5.17095439e-02 -5.73802710e-01 1.40565231e-01
1.50132000e-01 2.41539478e-01 2.61684030e-01 2.18368635e-01
5.11862576e-01 -8.67175937e-01 -9.12543237e-01 7.53187656e-01
-1.00461805e+00 -1.66727118e-02 7.36778855e-01 6.07941329e-01
8.42949569e-01 6.57543719e-01 5.13323009e-01 -8.40834379e-01
8.06745768e-01 -5.77597320e-01 -7.67180681e-01 1.72944978e-01
-8.16662252e-01 -3.20724607e-01 7.76064694e-01 -3.63185793e-01
-7.45814562e-01 4.51952994e-01 2.38967672e-01 -3.24927539e-01
-3.27800885e-02 6.62402630e-01 -1.92500353e-01 -4.32936624e-02
2.39728585e-01 4.35704321e-01 -5.62720820e-02 -5.54158211e-01
4.04719532e-01 9.27477658e-01 2.86958963e-01 -6.11699581e-01
1.08166790e+00 6.52637362e-01 -1.74452402e-02 -6.41341388e-01
-3.87734681e-01 -9.79364753e-01 -6.51689649e-01 -8.48122239e-02
5.30839384e-01 -9.49787021e-01 -8.01305056e-01 3.83115947e-01
-4.85228866e-01 -3.68880220e-02 2.96804398e-01 5.81294596e-01
-3.48512560e-01 7.46519029e-01 -3.54728341e-01 -7.54385769e-01
-3.68060261e-01 -9.51939225e-01 6.31542563e-01 4.33218986e-01
1.10125160e-02 -6.69341266e-01 1.72269389e-01 3.95976484e-01
1.62247866e-01 2.69821256e-01 5.24798095e-01 -5.75500786e-01
-6.58963263e-01 -9.21445787e-02 -5.20296156e-01 -3.48354757e-01
3.30364436e-01 4.17425483e-02 -2.08957136e-01 -6.37015998e-01
-2.60839343e-01 -8.26429129e-02 4.99158472e-01 2.20859870e-01
1.05158293e+00 -2.68000782e-01 -6.83612108e-01 6.46370113e-01
1.71164489e+00 4.00901943e-01 3.65678132e-01 2.90861726e-01
7.16053307e-01 3.41974258e-01 9.01581764e-01 6.26174927e-01
3.94971818e-01 4.50525671e-01 -1.07418425e-01 -6.54740036e-02
1.65846631e-01 -2.85060138e-01 -2.20887698e-02 1.26422107e+00
9.89641547e-02 2.00919770e-02 -1.14289939e+00 7.80250669e-01
-2.25060821e+00 -1.04600978e+00 -4.55019146e-01 2.07249570e+00
8.21960926e-01 -2.05412358e-01 4.98440832e-01 3.34117264e-01
1.13635015e+00 2.43147314e-02 -3.92349482e-01 7.39233121e-02
1.34928063e-01 -2.60637701e-01 5.09036779e-01 4.39596474e-01
-8.95633280e-01 9.46369648e-01 4.97657299e+00 1.41341710e+00
-9.05924976e-01 1.44771963e-01 4.56854224e-01 -1.78138062e-01
5.25014987e-03 3.27182740e-01 -4.97005612e-01 6.36177897e-01
7.54541755e-01 -5.10545015e-01 5.74414968e-01 9.30178642e-01
4.96634007e-01 -3.72004151e-01 -4.82413143e-01 1.32543230e+00
-1.71642765e-01 -1.11020327e+00 -2.54238009e-01 9.75658074e-02
7.46610165e-01 -1.37813374e-01 -3.61030221e-01 -2.92513639e-01
3.41995597e-01 -5.48667371e-01 3.12814295e-01 6.16482012e-02
5.97778141e-01 -1.07750142e+00 5.64843535e-01 3.99889290e-01
-1.88187647e+00 -3.73792984e-02 -5.63509405e-01 1.58056706e-01
2.05591470e-01 8.51954937e-01 -4.15610135e-01 6.89973533e-01
8.75129819e-01 5.06885290e-01 -4.08182442e-01 1.38702297e+00
-1.68840250e-03 6.75846100e-01 -5.52261353e-01 -2.48875231e-01
3.09174061e-01 -8.60944033e-01 3.70044172e-01 1.15661001e+00
4.60371494e-01 5.78669071e-01 7.27373123e-01 5.17278969e-01
5.73003590e-02 6.81599677e-01 -3.97969842e-01 -9.91636813e-02
1.03123295e+00 1.57660449e+00 -1.41648686e+00 -3.08823824e-01
-4.82685298e-01 6.54308975e-01 2.88436264e-01 2.04287380e-01
-8.37510109e-01 -8.79061997e-01 1.49019688e-01 3.72158349e-01
4.19515491e-01 -3.54032815e-01 -1.35293320e-01 -1.01868856e+00
-2.44757496e-02 -9.04096603e-01 7.27499545e-01 -6.14312947e-01
-1.13016975e+00 3.72644663e-01 1.05020098e-01 -1.23851514e+00
4.44481641e-01 4.07579318e-02 -6.08707726e-01 5.32096028e-01
-9.72497165e-01 -1.10937309e+00 -4.96313691e-01 9.65461552e-01
1.71158746e-01 -1.58080533e-02 3.77128989e-01 4.89191473e-01
-6.75413072e-01 3.05195153e-01 6.96572244e-01 2.10979357e-01
7.17713356e-01 -9.83790100e-01 -2.04335004e-01 1.04444945e+00
-7.11830482e-02 9.52089429e-01 5.45069575e-01 -7.73152411e-01
-1.46206081e+00 -8.28612149e-01 4.46490824e-01 3.12084734e-01
7.94459581e-01 -2.57667035e-01 -8.40444863e-01 3.41814101e-01
2.81412721e-01 -9.98957902e-02 9.29950237e-01 5.52851819e-02
-1.61982536e-01 -5.30585527e-01 -1.10578036e+00 4.80183959e-01
1.00394213e+00 -2.88332820e-01 -2.61339277e-01 3.57652903e-01
2.73341954e-01 -1.34866431e-01 -8.56595039e-01 2.73986220e-01
2.76156545e-01 -9.83924568e-01 8.29694390e-01 1.68809921e-01
-1.24354310e-01 -1.00620127e+00 -4.91634347e-02 -1.13268268e+00
-5.07383585e-01 -7.19917238e-01 3.31472307e-01 1.38698566e+00
6.75657317e-02 -7.47205496e-01 8.15366268e-01 2.42626876e-01
3.47887337e-01 -8.71938884e-01 -6.13124192e-01 -8.67527008e-01
-2.23291934e-01 -6.00960571e-03 6.12678707e-01 1.10599732e+00
2.77954340e-01 3.91272515e-01 -2.01439455e-01 3.07013899e-01
7.73091197e-01 7.60562479e-01 8.28147948e-01 -1.05872881e+00
-1.23684980e-01 -3.39855522e-01 -3.82972866e-01 -8.05230200e-01
-1.00133725e-01 -9.56496477e-01 -7.29320720e-02 -1.76934969e+00
8.40765119e-01 -6.64254487e-01 -3.33463013e-01 4.46651071e-01
-3.94306391e-01 2.03781173e-01 1.63940310e-01 6.22439027e-01
-1.03694248e+00 4.18900639e-01 8.36460412e-01 1.35486647e-01
-2.77133733e-01 -2.54012853e-01 -7.28176594e-01 5.08961320e-01
9.46142852e-01 -7.15813339e-01 -7.55780339e-01 -4.81531695e-02
-1.78403780e-02 2.29613587e-01 -1.85673386e-01 -1.06592190e+00
6.82537556e-01 -4.44296241e-01 6.32571802e-02 -6.71482205e-01
-3.90855819e-02 -1.08479774e+00 6.61318779e-01 6.38973653e-01
7.68963993e-02 1.91813141e-01 -5.85290156e-02 7.68165886e-01
-3.33236963e-01 1.28365755e-01 1.02450907e+00 -3.72789428e-02
-5.89355409e-01 5.80064237e-01 -1.19893722e-01 8.32998529e-02
1.48138332e+00 -2.41802320e-01 -1.46038070e-01 -3.46514463e-01
-5.14184415e-01 6.14551067e-01 6.40786827e-01 1.85263917e-01
5.80221176e-01 -1.57552707e+00 -5.21713912e-01 -1.68164670e-01
-5.90053853e-03 -4.62609902e-02 3.55438858e-01 1.08997798e+00
-7.55897462e-01 3.66221517e-01 -1.05228364e-01 -4.76807982e-01
-1.68364787e+00 8.47549379e-01 5.98869137e-02 -9.20727775e-02
-4.52074558e-01 5.21222770e-01 -2.47055385e-02 -4.97828126e-01
-3.34292389e-02 3.33063602e-01 2.01562077e-01 3.37317437e-02
3.36876631e-01 6.85784996e-01 -3.98227155e-01 -7.96634495e-01
-6.50391102e-01 9.38741863e-01 2.68177658e-01 -7.39129931e-02
1.39472759e+00 -6.34280324e-01 -5.35743177e-01 1.79661095e-01
1.14656138e+00 3.08041573e-01 -6.79946721e-01 -3.34777594e-01
-9.25261759e-07 -8.66572320e-01 -4.64561693e-02 -2.36658186e-01
-1.32340264e+00 6.15320563e-01 4.56949264e-01 4.02008861e-01
1.38095200e+00 -1.46322906e-01 1.06555200e+00 1.92554697e-01
4.48075563e-01 -1.26989973e+00 -1.11449309e-01 -9.94653776e-02
2.72028357e-01 -1.03968060e+00 3.25956285e-01 -7.42666364e-01
-5.14618218e-01 7.73475289e-01 7.11853981e-01 -3.00913215e-01
8.62846076e-01 1.01590358e-01 4.57971916e-02 -2.03808457e-01
-3.45953912e-01 -3.14474702e-01 -1.36748284e-01 2.49869674e-01
3.33418459e-01 2.00126097e-01 -8.38698566e-01 5.99038601e-01
3.73564549e-02 -1.18368067e-01 4.29466695e-01 9.72357869e-01
-7.11521745e-01 -1.18881953e+00 -5.62984705e-01 3.48788202e-01
-3.53057384e-01 3.96417230e-02 -3.61128688e-01 5.19292712e-01
-3.97666879e-02 1.19235027e+00 -2.46591762e-01 -5.11848330e-01
8.62707123e-02 -2.21375644e-01 -3.77428788e-03 -3.59590769e-01
-2.18322515e-01 4.53249753e-01 -2.78090984e-01 -6.36384666e-01
-6.95875168e-01 -5.39535820e-01 -1.67834067e+00 -6.66857779e-01
-4.08727050e-01 7.85203516e-01 4.54830587e-01 4.25688624e-01
5.74544668e-01 1.25135064e-01 8.96233141e-01 -4.43868220e-01
-1.94410533e-01 -5.99743187e-01 -8.88019145e-01 4.67005640e-01
-2.18859032e-01 -4.56517786e-01 -3.15812349e-01 -3.38582471e-02] | [7.5334625244140625, 4.747193813323975] |
c1f489c8-e2f6-4efa-bb08-980708ead90c | zero-shot-blind-audio-bandwidth-extension | 2306.01433 | null | https://arxiv.org/abs/2306.01433v1 | https://arxiv.org/pdf/2306.01433v1.pdf | Zero-Shot Blind Audio Bandwidth Extension | Audio bandwidth extension involves the realistic reconstruction of high-frequency spectra from bandlimited observations. In cases where the lowpass degradation is unknown, such as in restoring historical audio recordings, this becomes a blind problem. This paper introduces a novel method called BABE (Blind Audio Bandwidth Extension) that addresses the blind problem in a zero-shot setting, leveraging the generative priors of a pre-trained unconditional diffusion model. During the inference process, BABE utilizes a generalized version of diffusion posterior sampling, where the degradation operator is unknown but parametrized and inferred iteratively. The performance of the proposed method is evaluated using objective and subjective metrics, and the results show that BABE surpasses state-of-the-art blind bandwidth extension baselines and achieves competitive performance compared to non-blind filter-informed methods when tested with synthetic data. Moreover, BABE exhibits robust generalization capabilities when enhancing real historical recordings, effectively reconstructing the missing high-frequency content while maintaining coherence with the original recording. Subjective preference tests confirm that BABE significantly improves the audio quality of historical music recordings. Examples of historical recordings restored with the proposed method are available on the companion webpage: (http://research.spa.aalto.fi/publications/papers/ieee-taslp-babe/) | ['Vesa Välimäki', 'Filip Elvander', 'Eloi Moliner'] | 2023-06-02 | null | null | null | null | ['bandwidth-extension', 'bandwidth-extension'] | ['audio', 'speech'] | [ 1.72596037e-01 -2.88373113e-01 1.69339404e-01 1.92642882e-01
-1.34176922e+00 -5.87859571e-01 3.61815751e-01 -2.23881751e-01
-1.53871641e-01 7.80089259e-01 7.67246008e-01 3.26165743e-02
-4.63932991e-01 -2.61073977e-01 -6.34784937e-01 -9.43295896e-01
-1.21508598e-01 1.13633908e-01 1.29480273e-01 -1.01232067e-01
2.15496838e-01 1.62329361e-01 -1.66585636e+00 2.03763619e-01
8.55202973e-01 8.38203311e-01 4.08894032e-01 9.30420995e-01
5.25148928e-01 4.28801805e-01 -8.21698487e-01 -2.81038970e-01
1.91024199e-01 -6.07764959e-01 -3.92753184e-01 1.97858319e-01
2.61390418e-01 -4.62680608e-01 -4.74485666e-01 1.29286587e+00
9.40469980e-01 5.15577197e-01 4.96586293e-01 -6.17632389e-01
-6.81853831e-01 6.11763179e-01 -1.91317067e-01 3.81516665e-01
7.26632059e-01 -8.72197077e-02 1.10362160e+00 -1.03019238e+00
1.10724919e-01 1.04862726e+00 7.83323467e-01 2.79165328e-01
-1.49155045e+00 -4.76028055e-01 -1.23856291e-01 4.46654916e-01
-1.43659854e+00 -9.41327453e-01 9.77592826e-01 -5.53988993e-01
5.22807181e-01 4.44040447e-01 4.94982958e-01 1.28589594e+00
-1.93592757e-01 8.70100558e-01 1.06308150e+00 -5.89659095e-01
4.31921244e-01 -1.72849223e-02 8.43037367e-02 7.75833055e-02
-4.75812070e-02 5.56510091e-01 -1.05851901e+00 -3.89376193e-01
6.65177524e-01 -5.74616551e-01 -1.15683866e+00 -9.04556364e-02
-1.11091638e+00 5.64275384e-01 1.84669793e-01 3.31346929e-01
-6.32742465e-01 -1.30757347e-01 1.18807465e-01 4.36865449e-01
7.34537065e-01 4.12842393e-01 2.23371647e-02 -2.21861690e-01
-1.41901004e+00 2.81777054e-01 7.85586059e-01 6.33393705e-01
1.34053394e-01 5.39190412e-01 -1.95416898e-01 1.17673469e+00
2.15743542e-01 5.21596253e-01 5.28671086e-01 -1.11286032e+00
1.46016061e-01 -5.00226974e-01 6.85982049e-01 -5.34034550e-01
-9.07277688e-02 -8.90666008e-01 -9.05370712e-01 5.70497103e-02
3.16284537e-01 -1.19122297e-01 -8.25974524e-01 1.63954413e+00
1.61000565e-01 4.05312210e-01 1.08862415e-01 1.29618680e+00
3.85641396e-01 9.82210577e-01 -3.79507422e-01 -5.72584152e-01
9.77080464e-01 -8.25018883e-01 -1.22890317e+00 -2.82662988e-01
-4.30881888e-01 -1.19161820e+00 1.15159643e+00 1.08831644e+00
-1.25486672e+00 -8.01839769e-01 -1.21176803e+00 3.59797031e-01
3.31434667e-01 2.76818991e-01 1.46336108e-01 9.76299107e-01
-1.06620300e+00 6.70157194e-01 -5.34026980e-01 -1.33681282e-01
8.83840546e-02 -5.30595146e-02 1.94877503e-04 -2.25038111e-01
-1.32997274e+00 4.88383144e-01 1.56759292e-01 4.63315919e-02
-1.47215974e+00 -8.89168382e-01 -4.92585629e-01 8.75787809e-03
2.46243939e-01 -5.72192788e-01 1.53990853e+00 -7.50911772e-01
-1.78681839e+00 1.44519880e-01 -5.92782274e-02 -6.40336931e-01
6.39158487e-01 -7.65227795e-01 -9.26394105e-01 4.06283379e-01
-1.46853805e-01 8.80575553e-02 1.63246155e+00 -1.31235826e+00
-2.08060160e-01 -9.28547457e-02 -1.70281097e-01 1.73140511e-01
-5.39168179e-01 -2.22248450e-01 -3.48713815e-01 -1.42332685e+00
1.16960846e-01 -7.00377345e-01 1.83075309e-01 -1.86278149e-01
-3.07987034e-01 4.30596828e-01 5.02418101e-01 -1.31606936e+00
1.46382904e+00 -2.61916828e+00 3.37359846e-01 7.13800862e-02
-2.49713302e-01 2.93874592e-01 -8.63831025e-03 8.16323102e-01
-9.45648551e-02 -4.37824368e-01 -5.93284488e-01 -3.25724155e-01
1.11883886e-01 -1.47926241e-01 -6.40363276e-01 7.28069484e-01
-3.52264583e-01 3.92480846e-03 -7.83272147e-01 -7.63128176e-02
1.84633091e-01 8.12631428e-01 -5.15182853e-01 3.02504569e-01
1.91473197e-02 7.02709794e-01 1.35357350e-01 5.53521335e-01
5.84355772e-01 7.96471983e-02 -7.65752941e-02 -2.80274868e-01
1.76780187e-02 1.11115873e-01 -1.44761646e+00 1.96231723e+00
-5.38907468e-01 6.46125019e-01 4.81232077e-01 -7.07411170e-01
8.60991776e-01 8.59541297e-01 -3.71718183e-02 -2.58823663e-01
-1.32889077e-01 4.16549861e-01 -5.89719750e-02 -4.41199273e-01
4.21618044e-01 -4.70545322e-01 4.33312237e-01 2.23984987e-01
2.85795003e-01 -3.43042314e-01 7.31192976e-02 5.52949831e-02
7.58141160e-01 9.28330496e-02 1.41262010e-01 -3.03449720e-01
5.73200107e-01 -5.77546775e-01 3.14112931e-01 7.95303702e-01
-1.36301115e-01 7.95611143e-01 -1.33924887e-01 3.59292209e-01
-9.26965475e-01 -1.37706745e+00 -3.92196655e-01 9.22465146e-01
-1.54891657e-02 -2.08244652e-01 -9.13776636e-01 1.89128056e-01
-1.04398064e-01 1.10788870e+00 -1.89974502e-01 -1.59048617e-01
-8.97465199e-02 -7.23297894e-01 3.81470025e-01 3.05787116e-01
3.84534866e-01 -5.39943397e-01 -5.11021279e-02 4.64196920e-01
-5.56676984e-01 -8.77136946e-01 -6.48114264e-01 -1.76577538e-01
-8.57312202e-01 -5.07572353e-01 -1.32132995e+00 -5.31461596e-01
2.01003999e-01 2.27861270e-01 6.85877502e-01 -5.20155013e-01
-1.64755717e-01 5.75079083e-01 -4.44840759e-01 -1.71927512e-01
-4.32535261e-01 -4.79377300e-01 4.29444879e-01 5.76612473e-01
-2.25845426e-01 -9.92246687e-01 -7.62733817e-01 4.48189139e-01
-8.55215907e-01 -3.17954689e-01 4.88080174e-01 1.03927994e+00
4.92610067e-01 4.51686770e-01 9.56762731e-01 -2.84180135e-01
9.30586159e-01 -2.15114713e-01 -5.97428501e-01 -4.77635972e-02
-4.01446521e-01 -3.34991425e-01 4.89404589e-01 -8.23570251e-01
-1.36822665e+00 -1.99248433e-01 -1.66136667e-01 -5.59249222e-01
5.19510843e-02 4.02223259e-01 -3.30592662e-01 7.13499486e-02
9.46689785e-01 3.55612695e-01 -3.06597501e-01 -1.13435113e+00
3.36427152e-01 1.04067230e+00 1.12374938e+00 -6.38282955e-01
6.10086501e-01 3.57160538e-01 -4.98380005e-01 -1.08371544e+00
-6.94597244e-01 -6.89925849e-01 -7.61348158e-02 -4.32142287e-01
3.64954650e-01 -1.08404851e+00 -2.33338907e-01 7.52217710e-01
-8.60787451e-01 -2.02785209e-01 -5.27572632e-01 1.01141822e+00
-8.50618780e-01 4.92906839e-01 -7.34537303e-01 -1.28379285e+00
-1.94434181e-01 -8.98035884e-01 7.88192809e-01 -1.40113793e-02
-3.07800740e-01 -8.16521227e-01 1.68324798e-01 4.26468372e-01
3.63246262e-01 -1.57850891e-01 6.87504888e-01 -3.80792081e-01
-3.51863593e-01 -4.47965711e-01 3.39099884e-01 7.85295248e-01
9.74490568e-02 -4.01168525e-01 -1.47193718e+00 -6.56522453e-01
3.76835793e-01 7.01903552e-03 8.90432477e-01 6.69758439e-01
7.97043979e-01 -4.16947216e-01 1.58270702e-01 3.54193956e-01
1.29029024e+00 4.39813845e-02 5.24099469e-01 1.62647124e-02
-2.97062322e-02 4.26861942e-01 4.78645712e-01 7.84128964e-01
-4.04855907e-01 6.49951816e-01 3.18858474e-01 1.87105939e-01
-5.77241182e-01 -3.27021509e-01 5.08893371e-01 1.13516927e+00
-1.16636135e-01 -4.73396748e-01 -5.61293721e-01 9.28553641e-01
-1.48518324e+00 -8.90654325e-01 2.29062624e-02 2.63177991e+00
9.88227487e-01 6.10484555e-02 1.79164559e-01 8.15626562e-01
9.96008754e-01 1.81574166e-01 -5.77333510e-01 3.76750976e-02
-2.59942472e-01 2.11975992e-01 2.82922834e-01 9.32467759e-01
-9.47329283e-01 4.02620703e-01 6.14745569e+00 1.01319981e+00
-5.39409399e-01 4.57787663e-01 -2.94440649e-02 -2.21838236e-01
-1.20811947e-01 -6.73998073e-02 -3.44794393e-01 2.89131939e-01
1.07323170e+00 -3.08750898e-01 8.96260798e-01 2.52039969e-01
4.59931880e-01 -1.60463452e-02 -9.01512623e-01 1.05082810e+00
8.85285363e-02 -8.94150674e-01 -1.86055496e-01 -4.18331176e-02
7.81589448e-01 -1.64108634e-01 3.71541053e-01 -1.30092531e-01
5.82507662e-02 -5.74779510e-01 1.15741789e+00 9.57280159e-01
8.56702685e-01 -6.50959730e-01 5.26491582e-01 3.85113657e-01
-8.17202032e-01 -3.83447945e-01 -1.51472449e-01 -1.69367455e-02
4.89081830e-01 1.06016982e+00 -5.89959383e-01 7.10292637e-01
8.10170114e-01 6.55508220e-01 -2.27370691e-02 1.71872568e+00
-4.90183860e-01 1.15421128e+00 -1.21314749e-01 5.82715511e-01
-1.88901201e-01 -5.98794110e-02 1.32966208e+00 1.04565215e+00
7.36550629e-01 -7.61865377e-02 -1.71629250e-01 7.16783047e-01
3.65219116e-02 -2.55266521e-02 -1.38227686e-01 -1.01165593e-01
4.68263298e-01 6.43114567e-01 -1.95104599e-01 -1.69100598e-01
-1.10367477e-01 1.01717126e+00 -5.57661593e-01 8.49484921e-01
-6.59848034e-01 -4.84610140e-01 4.15344834e-01 2.95237210e-02
4.99986976e-01 -3.87098156e-02 1.25398636e-01 -8.88941944e-01
1.03213303e-01 -1.07232392e+00 2.13737100e-01 -1.18765044e+00
-1.36903548e+00 5.87713480e-01 -2.37862151e-02 -1.58325505e+00
-2.33984053e-01 -2.43485153e-01 -1.78977847e-01 1.07251728e+00
-1.30694580e+00 -6.34678841e-01 6.81257322e-02 7.32077301e-01
8.25092256e-01 -2.60436237e-01 9.50474501e-01 5.28622925e-01
-2.01323450e-01 4.59192216e-01 6.62703395e-01 -3.98666501e-01
9.33215857e-01 -1.07436037e+00 1.84191808e-01 1.07818782e+00
3.38235438e-01 4.23356563e-01 1.29244375e+00 -5.50208628e-01
-1.11191297e+00 -7.63578713e-01 4.95598465e-01 4.89765033e-02
8.42113018e-01 -2.06469923e-01 -1.16003227e+00 2.34556228e-01
3.06960881e-01 -2.11400792e-01 8.35358024e-01 -2.42071152e-01
-2.87468702e-01 -1.77923769e-01 -9.82410192e-01 2.63697475e-01
7.26643085e-01 -8.73048425e-01 -8.16164076e-01 4.32457745e-01
7.01376855e-01 -2.36228362e-01 -8.90925348e-01 1.40776902e-01
5.33024490e-01 -1.02619851e+00 1.19944155e+00 -7.04105804e-03
-8.48935246e-02 -4.09465820e-01 -6.20926261e-01 -1.60947239e+00
-1.91213712e-01 -1.33148837e+00 -3.93205911e-01 1.38455987e+00
2.48816803e-01 -5.88338852e-01 1.35935798e-01 -2.02740818e-01
-2.51120478e-01 1.29328249e-02 -9.34161246e-01 -1.27158725e+00
-3.39122623e-01 -5.26435733e-01 4.33934927e-01 6.24923885e-01
-3.45277973e-02 1.54850483e-01 -8.06472778e-01 5.71581900e-01
9.79904771e-01 -1.66736737e-01 3.49785119e-01 -1.13957596e+00
-9.87197816e-01 -1.25498801e-01 -1.78589955e-01 -1.14100683e+00
-1.02919087e-01 -4.27124798e-01 2.43452057e-01 -1.37888265e+00
-4.55680281e-01 1.84737608e-01 -3.26997787e-01 -4.09495384e-02
8.37628916e-03 3.67037535e-01 9.99077111e-02 4.69691247e-01
1.19998291e-01 7.52562284e-01 1.26674771e+00 -2.44552329e-01
-2.74639964e-01 5.61949790e-01 -4.76694733e-01 6.83903575e-01
6.89115226e-01 -4.28755641e-01 -6.53705776e-01 -2.56420225e-01
-4.33943532e-02 6.07156754e-01 4.87505943e-01 -1.38921416e+00
3.14215839e-01 3.71592015e-01 4.55136709e-02 -3.00891757e-01
9.27688420e-01 -5.65115929e-01 4.14781600e-01 2.20471472e-01
-3.14846039e-01 -5.44708550e-01 3.38975519e-01 1.11533189e+00
-5.89037061e-01 -3.80285382e-01 7.90100276e-01 1.58420578e-01
-4.33047384e-01 -1.39769450e-01 -5.11698961e-01 -7.16347471e-02
2.79809415e-01 -3.69392410e-02 -1.56212240e-01 -8.53530169e-01
-1.26420259e+00 -4.91329789e-01 1.15577057e-01 1.47456974e-01
6.29850924e-01 -1.13689697e+00 -1.07400107e+00 1.98967010e-01
-1.56681277e-02 -6.64124250e-01 6.70621216e-01 8.21547925e-01
-1.61338776e-01 1.56190693e-01 1.29773647e-01 -4.62226421e-01
-1.03307343e+00 5.51623106e-01 2.78222412e-01 2.93219477e-01
-8.64549637e-01 1.05667686e+00 1.95841670e-01 1.92644894e-01
6.31967425e-01 1.16629034e-01 -8.38667601e-02 8.90655816e-03
8.76711965e-01 5.87708592e-01 4.11252558e-01 -5.54240704e-01
-5.79045601e-02 3.44206631e-01 2.69054711e-01 -8.54772329e-01
1.17803895e+00 -3.11410040e-01 6.73074871e-02 6.78168297e-01
8.50527406e-01 4.64403540e-01 -1.39025486e+00 -3.69205981e-01
-2.74301857e-01 -7.50942469e-01 6.21848762e-01 -1.09096503e+00
-8.03694904e-01 9.54641700e-01 9.16651368e-01 2.62518704e-01
1.55561495e+00 -3.39737266e-01 5.46842098e-01 -3.19176540e-02
2.11474076e-01 -9.00786161e-01 7.18116909e-02 1.07149547e-02
1.37198365e+00 -5.88216484e-01 1.27357719e-02 -2.68934727e-01
-3.81823182e-01 8.19957495e-01 -1.55427784e-01 -7.84992352e-02
6.62750959e-01 6.41774461e-02 1.16898865e-01 3.24676454e-01
-3.70238036e-01 -1.15203880e-01 4.65162843e-01 8.46561730e-01
1.39224112e-01 1.05567053e-01 -9.65293795e-02 7.45181620e-01
-4.73943323e-01 -1.23977132e-01 5.16341388e-01 5.56104898e-01
-5.44456482e-01 -8.85812342e-01 -1.09572840e+00 2.42187511e-02
-5.20551085e-01 -3.27612877e-01 -1.25867039e-01 3.12435776e-01
-4.40982848e-01 1.43422878e+00 -3.40247154e-01 -2.13729888e-01
3.41174096e-01 2.62229145e-01 3.41459602e-01 -1.86260268e-01
-2.84954727e-01 9.29322243e-01 1.59286857e-01 -2.28814989e-01
-5.49968183e-01 -7.58795917e-01 -6.34510159e-01 3.11281672e-03
-5.74705601e-01 5.07385373e-01 7.10763872e-01 4.34313089e-01
3.61362211e-02 8.90276074e-01 5.92972696e-01 -1.07297790e+00
-9.13001537e-01 -1.17898023e+00 -1.05196285e+00 2.91490071e-02
9.36737537e-01 -5.12400985e-01 -8.29764724e-01 3.76568943e-01] | [15.310530662536621, 5.80245304107666] |
ebf617f3-839c-4e37-8a92-104951350b98 | covxnet-a-multi-dilation-convolutional-neural | null | null | https://www.sciencedirect.com/science/article/pii/S0010482520302250?via%3Dihub | https://www.sciencedirect.com/science/article/pii/S0010482520302250?via%3Dihub | CovXNet: A multi-dilation convolutional neural network for automatic COVID-19 and other pneumonia detection from chest X-ray images with transferable multi-receptive feature optimization | With the recent outbreak of COVID-19, fast diagnostic testing has become one of the major challenges due to the critical shortage of test kit. Pneumonia, a major effect of COVID-19, needs to be urgently diagnosed along with its underlying reasons. In this paper, deep learning aided automated COVID-19 and other pneumonia detection schemes are proposed utilizing a small amount of COVID-19 chest X-rays. A deep convolutional neural network (CNN) based architecture, named as CovXNet, is proposed that utilizes depthwise convolution with varying dilation rates for efficiently extracting diversified features from chest X-rays. Since the chest X-ray images corresponding to COVID-19 caused pneumonia and other traditional pneumonias have significant similarities, at first, a large number of chest X-rays corresponding to normal and (viral/bacterial) pneumonia patients are used to train the proposed CovXNet. Learning of this initial training phase is transferred with some additional fine-tuning layers that are further trained with a smaller number of chest X-rays corresponding to COVID-19 and other pneumonia patients. In the proposed method, different forms of CovXNets are designed and trained with X-ray images of various resolutions and for further optimization of their predictions, a stacking algorithm is employed. Finally, a gradient-based discriminative localization is integrated to distinguish the abnormal regions of X-ray images referring to different types of pneumonia. Extensive experimentations using two different datasets provide very satisfactory detection performance with accuracy of 97.4% for COVID/Normal, 96.9% for COVID/Viral pneumonia, 94.7% for COVID/Bacterial pneumonia, and 90.2% for multiclass COVID/normal/Viral/Bacterial pneumonias. Hence, the proposed schemes can serve as an efficient tool in the current state of COVID-19 pandemic. | ['Tanvir Mahmud', 'Shaikh Anowarul Fattah', 'Md Awsafur Rahman'] | 2020-07-01 | null | null | null | computers-in-biology-and-medicine-2020-7 | ['pneumonia-detection'] | ['medical'] | [ 1.35857090e-01 -6.83561802e-01 9.99331474e-02 -2.40209594e-01
-3.16886753e-01 -4.45677966e-01 8.95353854e-02 -3.17559391e-02
-4.60437238e-01 6.30697966e-01 -1.74436212e-01 -5.37425458e-01
-4.65030521e-01 -8.10868561e-01 -5.07941604e-01 -8.51756215e-01
-1.18260451e-01 8.94809723e-01 1.71297655e-01 2.37541705e-01
-1.53191701e-01 8.06344569e-01 -1.16967714e+00 4.07109827e-01
6.91796601e-01 8.85482609e-01 9.24256682e-01 1.29618621e+00
2.00581253e-02 4.98446584e-01 -6.32273316e-01 1.51895061e-01
2.17187151e-01 -1.52065784e-01 -5.10649323e-01 -1.67546436e-01
4.05063443e-02 -1.07872665e+00 -1.38323218e-01 4.11543131e-01
6.11893117e-01 -2.34296918e-01 1.04503644e+00 -1.01271605e+00
-2.57051557e-01 -1.83223680e-01 -5.79570830e-01 7.29141474e-01
-1.42970175e-01 2.70788759e-01 5.70917487e-01 -9.03274536e-01
3.81321877e-01 9.49948370e-01 1.04491258e+00 6.76585734e-01
-4.57279921e-01 -6.92050159e-01 -4.73832071e-01 2.27002233e-01
-1.13385797e+00 4.57930773e-01 2.22152159e-01 -9.32843566e-01
1.14604199e+00 4.53620493e-01 7.18888700e-01 1.16453207e+00
4.16964412e-01 4.78932619e-01 7.46774495e-01 1.26913980e-01
-4.55961414e-02 6.01469576e-02 2.89042473e-01 8.02557468e-01
7.28803754e-01 -1.53867574e-02 3.74129355e-01 -4.52494651e-01
9.07430530e-01 1.08319664e+00 -4.88305986e-01 -3.43513303e-02
-9.57938790e-01 9.31225538e-01 5.11554956e-01 2.84987926e-01
-6.42670751e-01 -2.73465067e-01 5.80696166e-01 -2.71699697e-01
-1.22149229e-01 3.21630746e-01 -7.25247860e-01 3.80801886e-01
-7.10269213e-01 3.72641087e-01 2.21149623e-01 5.67228615e-01
4.91547048e-01 -3.80940214e-02 -2.42221653e-01 1.03328621e+00
2.75157690e-01 9.66815472e-01 6.38692737e-01 -3.60796094e-01
6.58828437e-01 7.79036105e-01 7.88365304e-02 -6.50442839e-01
-8.54812145e-01 -4.68062282e-01 -1.34398174e+00 -1.87891796e-01
-2.62949546e-03 -4.24377441e-01 -1.17657864e+00 1.34937024e+00
2.65653223e-01 2.66155392e-01 1.13529727e-01 8.76118481e-01
8.29733729e-01 8.97987068e-01 -4.07487713e-02 -2.61873573e-01
1.85426044e+00 -7.13470161e-01 -3.28308076e-01 1.38579950e-01
4.62186188e-01 -5.76789260e-01 8.01256061e-01 1.15457498e-01
-6.79479837e-01 -6.08935535e-01 -9.06991899e-01 4.07050639e-01
-3.14517587e-01 3.43637109e-01 1.09287508e-01 4.86703396e-01
-8.39209974e-01 7.54775703e-02 -8.29493403e-01 -7.01774240e-01
5.05039692e-01 5.06495655e-01 3.47531885e-02 -2.79304594e-01
-1.03969884e+00 6.30532086e-01 3.42357695e-01 4.05439176e-02
-9.45682943e-01 -7.08640933e-01 -3.74067783e-01 8.01206380e-02
1.96901515e-01 -1.13127458e+00 9.47133124e-01 -1.70943081e-01
-5.85226834e-01 6.87401593e-01 6.85650855e-02 -1.30046174e-01
3.87594581e-01 -3.70925009e-01 -3.57453376e-01 5.29579818e-01
1.76197886e-01 3.79974067e-01 7.97957778e-01 -1.16195428e+00
-8.47209394e-01 -3.83627653e-01 -2.47022897e-01 -5.06598130e-02
-2.85039783e-01 3.92487198e-01 -3.30335319e-01 -7.49574125e-01
-2.45182633e-01 -9.78805780e-01 -1.39549240e-01 -3.41250181e-01
-4.43475991e-01 -3.15291941e-01 1.40068567e+00 -6.89348280e-01
1.02017212e+00 -1.85627139e+00 -4.31185544e-01 1.67840198e-01
5.34738660e-01 8.66603434e-01 1.08879823e-02 2.42278978e-01
1.20805129e-02 1.29693180e-01 -4.92264867e-01 1.20145597e-01
-3.65337312e-01 3.87789547e-01 -1.08823583e-01 2.58728832e-01
6.45069122e-01 7.86313951e-01 -7.30101824e-01 -6.48416698e-01
3.50918531e-01 8.16298723e-01 -4.95804906e-01 7.85268068e-01
-4.52464037e-02 4.59425241e-01 -7.60550380e-01 8.64012718e-01
8.41450930e-01 -8.22668552e-01 -1.45893022e-01 -6.23596758e-02
7.15941936e-02 -2.81223446e-01 -6.89516842e-01 5.10784805e-01
-2.54870355e-01 3.19505632e-01 1.16868965e-01 -9.61262286e-01
5.42983890e-01 7.15825200e-01 6.92102909e-01 -7.49984086e-02
2.95933604e-01 2.11993501e-01 1.05123371e-02 -1.17514610e+00
9.63874757e-02 -6.29862100e-02 4.31257457e-01 7.34671295e-01
-3.65519941e-01 1.03860274e-01 1.00050755e-01 -1.77768871e-01
1.39971888e+00 -4.65313584e-01 2.27224946e-01 6.62470609e-02
6.81459248e-01 1.70847431e-01 5.52032948e-01 8.32154155e-01
-2.00545117e-01 1.00041926e+00 -1.47941504e-02 -6.50254667e-01
-1.13218212e+00 -1.19992685e+00 -4.47965503e-01 7.25151181e-01
-2.08807468e-01 2.39739895e-01 -8.01455677e-01 -6.97208762e-01
-2.46384099e-01 1.42992660e-01 -4.91744697e-01 3.22427630e-01
-1.18018913e+00 -1.18312395e+00 6.01084173e-01 8.45903337e-01
6.23217404e-01 -1.35444319e+00 -1.20259619e+00 2.80335277e-01
-6.99680746e-02 -8.40892255e-01 -2.02145159e-01 3.40251505e-01
-8.35700512e-01 -1.49777138e+00 -1.09082079e+00 -9.73928213e-01
6.62972748e-01 4.21955198e-01 8.39189768e-01 4.45688874e-01
-8.44959617e-01 1.05345905e-01 -2.89075583e-01 -5.22956550e-01
-3.89517814e-01 -5.39321788e-02 7.60142878e-02 -2.85126179e-01
2.67343819e-01 -1.69855922e-01 -9.51751530e-01 3.20803463e-01
-1.04990256e+00 -1.76920921e-01 7.96223104e-01 9.91065204e-01
7.33397543e-01 1.36874065e-01 5.95515907e-01 -6.73938394e-01
4.75050807e-01 -8.75237226e-01 -3.52192193e-01 2.10552439e-01
-3.30901861e-01 -5.57819068e-01 9.82333481e-01 -3.02738816e-01
-9.82899129e-01 -1.90336093e-01 -3.27542484e-01 -7.58979619e-01
-3.32311749e-01 1.61789566e-01 3.52768421e-01 4.63047713e-01
3.61637980e-01 2.12289989e-01 -2.26793259e-01 -4.51776266e-01
-3.84334475e-01 9.45440531e-01 4.74748015e-01 8.29347894e-02
6.91745043e-01 4.24768060e-01 -4.07489017e-02 -7.31274426e-01
-6.45794034e-01 -8.36434186e-01 -3.40738624e-01 -8.45854580e-02
1.66869402e+00 -8.25219274e-01 -5.36068976e-01 8.49056125e-01
-1.25780880e+00 5.30091301e-02 1.48623288e-01 8.36824775e-01
-2.48417497e-01 1.96233481e-01 -9.33971703e-01 -3.96839887e-01
-9.20743167e-01 -1.45755982e+00 9.83062804e-01 1.76593840e-01
-1.59923866e-01 -7.94240534e-01 3.52510661e-01 4.98689622e-01
4.61121798e-01 8.13584998e-02 1.37777078e+00 -1.03784263e+00
-6.23926878e-01 -2.08524942e-01 -7.10412085e-01 6.33117020e-01
5.87190747e-01 1.99279577e-01 -6.84196949e-01 -5.09068847e-01
4.11317706e-01 -3.47510427e-01 7.08267808e-01 6.87140167e-01
1.22326159e+00 -2.75891989e-01 -6.68066740e-01 8.38846385e-01
1.58418024e+00 8.69150877e-01 6.03434406e-02 1.86954007e-01
9.35910463e-01 2.09802762e-01 3.71476352e-01 6.71590745e-01
2.11929470e-01 1.05612516e-01 4.95919049e-01 -3.86627942e-01
1.43733829e-01 5.01071274e-01 -2.40266502e-01 8.24573040e-01
-2.47862130e-01 -4.61922258e-01 -1.26226795e+00 4.62507248e-01
-1.27971840e+00 -9.97616470e-01 -3.11953455e-01 1.55563629e+00
5.25101542e-01 -2.04414770e-01 -1.37932584e-01 1.56523064e-02
9.56118643e-01 -1.27957657e-01 -7.32325077e-01 -3.63294214e-01
1.04763620e-01 4.94718254e-01 1.48041114e-01 -6.44554123e-02
-1.36900353e+00 -5.46053313e-02 6.06730986e+00 2.43811131e-01
-1.51112306e+00 9.74253714e-02 5.36448598e-01 -3.12028211e-02
1.34139732e-01 -8.57645452e-01 -6.95011377e-01 5.30533373e-01
4.63976175e-01 4.81297284e-01 1.50304819e-02 9.37369108e-01
2.13952541e-01 3.14787745e-01 -7.91365623e-01 9.39166367e-01
-1.26240537e-01 -1.41205943e+00 3.29642184e-02 -1.68659002e-01
8.47166419e-01 5.37955701e-01 -4.88646962e-02 1.70918971e-01
-1.90795567e-02 -9.64450181e-01 -6.10974580e-02 3.22597176e-01
1.07792294e+00 -6.78896666e-01 1.28033805e+00 2.16019616e-01
-1.30011547e+00 -2.17077464e-01 -3.58709723e-01 5.02517402e-01
5.49459569e-02 3.87333214e-01 -1.44873142e+00 3.38946879e-01
1.18586433e+00 2.59731710e-01 -2.92627424e-01 8.34184527e-01
7.64391646e-02 7.48816192e-01 -3.41491401e-01 -1.34182006e-01
3.54845762e-01 3.42438221e-02 4.20019060e-01 1.66601920e+00
5.18790364e-01 3.61916155e-01 -2.15741508e-02 7.13344991e-01
3.93874794e-02 -4.53252392e-03 -6.09128535e-01 1.22650370e-01
3.71800542e-01 1.14128137e+00 -6.59783125e-01 -8.06232631e-01
-4.74479318e-01 5.48111677e-01 -1.29714116e-01 4.16953176e-01
-9.52858984e-01 -3.27805340e-01 6.00311100e-01 1.39736876e-01
6.95793629e-01 2.25111440e-01 -1.62797898e-01 -7.63369799e-01
-1.78275406e-01 -7.41716862e-01 5.18691361e-01 -8.50494623e-01
-1.23642755e+00 9.22533095e-01 -1.37979083e-03 -1.09551466e+00
-5.04838645e-01 -8.02006602e-01 -1.05397534e+00 9.05042589e-01
-1.34563327e+00 -7.05122352e-01 -6.57853246e-01 7.17503488e-01
6.05447292e-01 -3.88513118e-01 8.54004204e-01 1.80349693e-01
-7.15989232e-01 2.83266515e-01 4.33088332e-01 2.25996092e-01
2.70167798e-01 -1.00942624e+00 1.27294287e-01 7.02800870e-01
-7.47823238e-01 6.62046194e-01 1.61696598e-01 -7.21204817e-01
-1.08184159e+00 -1.83363497e+00 7.78150439e-01 -1.72979593e-01
1.98022053e-01 1.07236855e-01 -1.07862723e+00 5.90199172e-01
1.12588607e-01 -4.81182337e-02 7.23385870e-01 -7.87603676e-01
-9.26249325e-02 1.20592147e-01 -1.42450666e+00 2.46587455e-01
6.72060728e-01 -2.67981380e-01 -7.41837442e-01 5.01364887e-01
7.86282361e-01 -1.14075504e-01 -6.69450104e-01 9.18211579e-01
3.92394811e-01 -1.00523567e+00 1.17082608e+00 -4.00495768e-01
4.60614383e-01 -1.18355349e-01 -1.47451460e-01 -8.13005805e-01
-4.22979414e-01 1.98243186e-01 1.26819506e-01 5.83349049e-01
-1.01099409e-01 -7.89088190e-01 7.90614307e-01 -2.23711967e-01
-3.22408587e-01 -1.20008564e+00 -6.58931315e-01 -4.88402575e-01
-2.96178132e-01 -1.84595898e-01 6.25634670e-01 8.59273791e-01
-9.33228791e-01 1.15767457e-01 -1.00008756e-01 3.98797929e-01
3.40173006e-01 1.02999642e-01 5.13535976e-01 -1.11377490e+00
-3.58212769e-01 -3.05601329e-01 -9.45427939e-02 -5.63719332e-01
-5.21283388e-01 -6.37493670e-01 -6.44885823e-02 -1.88455939e+00
5.60349762e-01 -6.90828204e-01 -6.64538920e-01 3.84258658e-01
-4.53049511e-01 3.24436843e-01 -3.42010558e-02 4.69317198e-01
3.01509742e-02 -3.07959281e-02 1.36575520e+00 -1.68929458e-01
-1.86863840e-01 3.32929879e-01 -8.34199339e-02 8.47910345e-01
1.07673919e+00 -5.19647777e-01 -6.27489030e-01 -4.54548150e-01
-8.04439336e-02 3.23875844e-01 3.63279343e-01 -9.91403162e-01
-3.25966418e-01 -3.22101742e-01 5.80016732e-01 -1.49310625e+00
3.57519388e-01 -7.71184266e-01 2.58361518e-01 1.12518513e+00
2.20460355e-01 3.85102659e-01 1.00062571e-01 2.96977639e-01
-6.44853041e-02 -2.56339759e-01 8.38639557e-01 -1.72020614e-01
-3.16838235e-01 5.82297206e-01 -7.85895050e-01 4.68567647e-02
1.18166089e+00 -3.07896465e-01 -2.78360426e-01 2.91585684e-01
-3.40191782e-01 9.56991538e-02 2.41885960e-01 1.37902901e-01
1.05853832e+00 -9.39639509e-01 -8.75125647e-01 4.96463388e-01
4.71673086e-02 4.64262694e-01 5.23815811e-01 8.46713543e-01
-1.08776605e+00 7.31792092e-01 -2.75135517e-01 -1.13374531e+00
-1.56332839e+00 7.35810637e-01 4.13160682e-01 -5.11424422e-01
-6.24473095e-01 9.18754756e-01 8.33445668e-01 -3.22027653e-01
1.47186652e-01 -5.92608929e-01 -2.85509586e-01 -1.84705973e-01
6.02049768e-01 4.09855604e-01 8.35321620e-02 -4.96012807e-01
-6.41477048e-01 7.94091761e-01 -2.65806735e-01 7.45857537e-01
1.42253077e+00 2.69152582e-01 -1.74270064e-01 -7.62347654e-02
1.41926515e+00 -3.62566382e-01 -9.57486391e-01 -8.60455260e-02
-4.75939453e-01 -3.82812284e-02 -4.88224715e-01 -8.11839044e-01
-1.05695009e+00 9.20148611e-01 1.00947464e+00 -3.88283143e-03
1.36462319e+00 2.08814800e-01 1.12013161e+00 4.69317377e-01
-1.26056135e-01 -5.29971480e-01 1.94741070e-01 5.13963282e-01
7.80559540e-01 -1.15122497e+00 -1.92454621e-01 -3.55780460e-02
-2.37049893e-01 1.02313685e+00 6.88879669e-01 -1.29681617e-01
5.46459198e-01 3.51947635e-01 2.90965199e-01 -5.45008600e-01
-7.16211021e-01 2.00413138e-01 -1.54668707e-02 7.28387654e-01
6.64757937e-03 4.26364928e-01 -6.82428107e-02 5.72372258e-01
5.17858677e-02 -7.88019523e-02 1.16030283e-01 1.10535896e+00
-6.21894479e-01 -5.07815123e-01 -8.45971584e-01 1.13298202e+00
-6.71936691e-01 -2.09445000e-01 -2.72913976e-03 9.73259926e-01
6.05192602e-01 6.34959102e-01 2.67934740e-01 -3.56300890e-01
5.38401976e-02 -6.52558953e-02 1.95783973e-01 -6.52921140e-01
-6.95154369e-01 7.41288960e-02 -3.75603586e-01 -1.88203052e-01
-2.60531217e-01 -5.33512712e-01 -1.77338958e+00 -9.14150178e-02
-1.44280300e-01 -2.12499976e-01 5.61932266e-01 7.96601951e-01
5.94576560e-02 6.86688364e-01 6.85657918e-01 -3.85353178e-01
-5.46570480e-01 -8.29090297e-01 -4.82817680e-01 3.94301951e-01
7.92323828e-01 -4.58450466e-01 -3.17972809e-01 1.23832397e-01] | [15.567960739135742, -1.741341233253479] |
a3c35239-afdd-4541-b020-377b05106846 | application-of-data-engineering-approaches-to | 2307.00033 | null | https://arxiv.org/abs/2307.00033v2 | https://arxiv.org/pdf/2307.00033v2.pdf | Application of data engineering approaches to address challenges in microbiome data for optimal medical decision-making | The human gut microbiota is known to contribute to numerous physiological functions of the body and also implicated in a myriad of pathological conditions. Prolific research work in the past few decades have yielded valuable information regarding the relative taxonomic distribution of gut microbiota. Unfortunately, the microbiome data suffers from class imbalance and high dimensionality issues that must be addressed. In this study, we have implemented data engineering algorithms to address the above-mentioned issues inherent to microbiome data. Four standard machine learning classifiers (logistic regression (LR), support vector machines (SVM), random forests (RF), and extreme gradient boosting (XGB) decision trees) were implemented on a previously published dataset. The issue of class imbalance and high dimensionality of the data was addressed through synthetic minority oversampling technique (SMOTE) and principal component analysis (PCA). Our results indicate that ensemble classifiers (RF and XGB decision trees) exhibit superior classification accuracy in predicting the host phenotype. The application of PCA significantly reduced testing time while maintaining high classification accuracy. The highest classification accuracy was obtained at the levels of species for most classifiers. The prototype employed in the study addresses the issues inherent to microbiome datasets and could be highly beneficial for providing personalized medicine. | ['Shyam Kumar Sudhakar', 'Pavan Kumar Perepu', 'Isha Thombre'] | 2023-06-30 | null | null | null | null | ['classification-1', 'decision-making'] | ['methodology', 'reasoning'] | [ 4.70002532e-01 -6.46919191e-01 -2.29311399e-02 -1.02370977e-01
2.60217190e-01 -3.54242295e-01 3.96270812e-01 5.42173505e-01
-1.15901884e-02 1.12048817e+00 1.28560327e-02 -6.23433590e-01
-2.21255064e-01 -8.15175951e-01 -4.43294704e-01 -1.01781571e+00
-3.49929571e-01 1.17777482e-01 -2.11757809e-01 -3.61142039e-01
5.01175523e-01 3.47598821e-01 -2.13906145e+00 5.88600457e-01
1.14924741e+00 7.85323143e-01 3.16824853e-01 6.74794495e-01
-3.56698990e-01 3.38370383e-01 -4.03800339e-01 -2.74496991e-02
2.25238323e-01 -4.17083114e-01 -2.99865663e-01 -3.77801299e-01
2.97612268e-02 2.92077184e-01 7.86919296e-01 6.36354566e-01
6.61504567e-01 -2.80394971e-01 8.51221979e-01 -1.17067552e+00
-4.06060874e-01 9.24552046e-03 -5.80252111e-01 1.04415096e-01
5.51628232e-01 1.15458317e-01 2.90457070e-01 -6.25673234e-01
4.48094577e-01 1.23059225e+00 9.94636655e-01 3.30033571e-01
-1.52634621e+00 -7.62190282e-01 -1.75204530e-01 2.64971763e-01
-9.48582053e-01 -2.18019843e-01 3.57398957e-01 -1.12802243e+00
1.10496485e+00 5.77695489e-01 1.16595054e+00 1.07557714e+00
6.12066031e-01 4.42169383e-02 1.74122894e+00 -6.73394918e-01
5.57046473e-01 4.85195309e-01 7.12533891e-02 2.33932421e-01
7.54705012e-01 3.68961036e-01 -5.87198853e-01 -5.75267136e-01
2.41851747e-01 8.16150308e-02 -2.44657680e-01 -3.38092148e-01
-1.11598301e+00 8.57542753e-01 2.50668973e-01 2.25920588e-01
-6.89329922e-01 -3.25978756e-01 7.28246510e-01 2.63260335e-01
4.02912676e-01 5.88661075e-01 -8.13807905e-01 1.29662797e-01
-3.54352832e-01 -1.05621852e-01 9.64578390e-01 3.64546031e-01
5.15996158e-01 -1.22103430e-02 3.11827183e-01 1.10086918e+00
3.20636809e-01 3.23407054e-01 9.43496883e-01 -4.74462181e-01
-1.96876954e-02 7.09582388e-01 1.65223703e-01 -1.18778133e+00
-2.30220586e-01 -4.51288134e-01 -7.82322407e-01 2.33691260e-01
2.84364045e-01 -1.14275701e-03 -5.98936737e-01 1.30004668e+00
8.58289182e-01 -1.58563852e-01 5.24759851e-02 5.77304125e-01
5.30083179e-01 4.04015690e-01 3.86763841e-01 -2.20713660e-01
1.60619152e+00 -3.89159411e-01 -5.55231631e-01 -4.77058031e-02
4.37046885e-01 -1.02501583e+00 7.77862430e-01 7.76003659e-01
-1.53046399e-01 -3.36943984e-01 -1.18144917e+00 6.33938968e-01
-5.97498655e-01 -2.00297758e-01 7.69543409e-01 1.23112607e+00
-7.19592929e-01 5.97539246e-01 -7.80771792e-01 -8.38213921e-01
1.66329533e-01 4.05600548e-01 -4.45711762e-01 -1.78785294e-01
-7.75587559e-01 9.49827611e-01 1.61635935e-01 -4.84990254e-02
-2.89232105e-01 -5.78846812e-01 -5.95188737e-01 -4.70236927e-01
-4.15796757e-01 -9.90339875e-01 5.78660011e-01 -8.49654317e-01
-1.46209085e+00 7.27349818e-01 -1.63321659e-01 -1.75765648e-01
3.43639731e-01 -2.49428704e-01 -3.50489348e-01 -2.63911635e-01
8.56006518e-02 2.22545415e-01 3.20888489e-01 -1.13004470e+00
-5.02708554e-01 -8.87934923e-01 -6.96882963e-01 -3.14036578e-01
-3.01954985e-01 -1.72342323e-02 9.81085837e-01 -6.59907639e-01
4.30673629e-01 -9.43473697e-01 -1.71819285e-01 -3.63641500e-01
-2.47901440e-01 1.34356692e-01 6.85693383e-01 -9.39402342e-01
9.37019706e-01 -1.92718351e+00 2.72262171e-02 -1.23022832e-01
-4.93793905e-01 1.04825445e-01 1.70507923e-01 7.90204167e-01
-2.83405423e-01 2.00088561e-01 -1.15147322e-01 4.29443747e-01
-5.41104794e-01 6.69578761e-02 -1.42628714e-01 7.94272482e-01
1.41661853e-01 9.30468366e-02 -9.10360694e-01 -3.02135378e-01
4.78535175e-01 7.35048771e-01 -7.38357604e-01 1.42155558e-01
-2.44003072e-01 5.63784659e-01 -1.70435145e-01 9.96956825e-01
7.44796455e-01 1.49012148e-01 4.85703528e-01 -2.44810149e-01
-3.78111392e-01 5.47854491e-02 -1.06049347e+00 1.14696479e+00
-4.56753880e-01 4.20891136e-01 9.76154804e-02 -8.18847179e-01
1.06944144e+00 3.12448442e-01 3.73436540e-01 -3.55686963e-01
3.44876796e-02 6.39110506e-01 2.73671657e-01 -9.22399640e-01
5.20330183e-02 -5.24176657e-01 5.50704360e-01 1.44577026e-01
-2.65677691e-01 4.11492258e-01 -9.30814892e-02 -6.57057881e-01
5.56063950e-01 3.74221355e-01 8.48476946e-01 -7.28492558e-01
6.75917089e-01 5.66030025e-01 5.08000612e-01 1.29581898e-01
-2.60959059e-01 1.63604319e-01 2.55089402e-01 -6.42637312e-01
-8.38351071e-01 -4.60239083e-01 -5.84348619e-01 9.47149754e-01
-2.33757362e-01 1.00061700e-01 -3.63885731e-01 1.89163551e-01
3.24473411e-01 6.75495625e-01 -6.55255854e-01 1.16445728e-01
-2.88845956e-01 -1.52481318e+00 2.83564031e-01 3.56036089e-02
1.41986102e-01 -6.76712692e-01 -8.02914381e-01 4.75325286e-01
-1.90529320e-02 -4.37216610e-01 5.18491387e-01 4.68207389e-01
-1.20132887e+00 -1.49329042e+00 -3.41615528e-01 -7.51350343e-01
5.24767995e-01 2.68321514e-01 7.57116139e-01 -1.19833447e-01
-8.81522894e-01 -1.66394591e-01 -5.79033077e-01 -7.98673153e-01
-7.50359952e-01 -3.16455752e-01 2.75281310e-01 -2.97213852e-01
5.84320962e-01 -6.90930188e-01 -9.66872454e-01 4.08675224e-01
-5.07332742e-01 -1.25892758e-01 4.31175888e-01 1.17061627e+00
4.80131716e-01 1.55817124e-03 9.00917411e-01 -8.11790168e-01
5.32449365e-01 -9.34941947e-01 -1.35411277e-01 1.78713743e-02
-6.43263876e-01 -2.50948817e-01 5.09040296e-01 -4.97818828e-01
-7.56166935e-01 -1.17828496e-01 -1.34252459e-01 4.45644766e-01
-3.92608970e-01 3.04817349e-01 1.37508824e-01 -3.11907139e-02
1.09674668e+00 9.86802578e-02 5.90055227e-01 -7.11764991e-01
-2.34976798e-01 1.16936171e+00 -1.61155730e-01 -6.39514387e-01
-6.48617744e-02 5.23892283e-01 2.02797234e-01 -1.01336384e+00
-3.52122396e-01 -6.11046672e-01 -3.90928984e-01 -1.47521347e-01
7.57386029e-01 -1.01466107e+00 -6.99417770e-01 5.06108165e-01
-5.54012597e-01 -2.96632759e-02 2.35968783e-01 6.47376359e-01
-2.75179923e-01 2.45907545e-01 -3.71834129e-01 -1.18130982e+00
-6.97805643e-01 -1.09684563e+00 6.70582056e-01 -9.53319594e-02
-7.12624311e-01 -8.91169429e-01 4.18242812e-01 4.22766984e-01
5.62377453e-01 1.21382439e+00 1.46895754e+00 -4.58895475e-01
5.30929789e-02 -4.22755582e-03 2.44921118e-01 3.23531657e-01
5.88826418e-01 4.40072060e-01 -1.18581164e+00 -4.23866659e-01
2.78096050e-01 -6.20503463e-02 6.26869678e-01 4.69016671e-01
6.67297244e-01 -2.78708100e-01 -4.99450117e-01 5.18921554e-01
1.80594957e+00 4.30551082e-01 3.48462760e-01 8.92579794e-01
2.00814635e-01 1.06228316e+00 6.84826374e-01 4.67338741e-01
-7.31687099e-02 3.86153549e-01 6.11059427e-01 1.97693467e-01
-6.76987395e-02 -4.34080996e-02 2.43980691e-01 6.00703597e-01
-4.22724299e-02 2.44361341e-01 -9.18684006e-01 4.57063407e-01
-1.37571597e+00 -9.32703555e-01 -6.30388439e-01 2.34735537e+00
7.58600593e-01 -5.04064083e-01 1.89669639e-01 6.14612103e-01
7.72941291e-01 -5.04589558e-01 -4.58196312e-01 -8.89114678e-01
-1.36437654e-01 1.10348821e-01 1.95246890e-01 -2.67458390e-02
-9.11230445e-01 -8.60098079e-02 6.95620728e+00 -5.23681715e-02
-1.34463477e+00 -1.83324933e-01 4.60147411e-01 -2.78436486e-02
1.09815620e-01 -6.43860921e-02 -5.22702396e-01 5.93339860e-01
7.62787342e-01 2.38527320e-02 4.15200472e-01 1.00153553e+00
1.42850533e-01 -2.72771418e-01 -9.93687034e-01 5.67768037e-01
-2.24523842e-01 -9.68107939e-01 -1.20388947e-01 3.80100965e-01
7.62648225e-01 2.66272783e-01 -1.75917894e-01 1.37509137e-01
8.95761102e-02 -9.70935345e-01 3.15258116e-01 1.87334046e-01
7.27804720e-01 -7.04272747e-01 8.49889159e-01 4.60363179e-01
-7.74337649e-01 -4.96186644e-01 -3.32895607e-01 -5.46459675e-01
-3.71471673e-01 1.06527281e+00 -1.34909141e+00 4.44765210e-01
9.29376364e-01 3.75394821e-01 -4.35088277e-01 1.11446977e+00
3.03276628e-01 3.34813029e-01 -4.02857304e-01 -3.52536380e-01
-4.20199960e-01 -4.31456298e-01 4.88117844e-01 1.20537090e+00
6.68741882e-01 -3.88624251e-01 -1.54326379e-01 3.52404833e-01
7.53859043e-01 7.31636763e-01 -5.42815268e-01 2.58453060e-02
5.51909387e-01 7.95888960e-01 -6.35064960e-01 -5.96915931e-02
-9.78358388e-02 3.22788090e-01 -1.92705970e-02 -9.27425548e-03
-1.03319570e-01 -5.15585532e-03 1.07070100e+00 2.90244251e-01
1.98866963e-01 8.54122732e-03 -5.74406147e-01 -7.65118241e-01
-2.47841671e-01 -1.10994184e+00 6.41498864e-01 -3.30499977e-01
-1.47240651e+00 2.98463553e-01 -2.32930422e-01 -1.24070919e+00
-3.23185772e-02 -8.03452611e-01 -4.63990510e-01 1.11985767e+00
-1.31718600e+00 -6.34696901e-01 -4.98037308e-01 -2.73028892e-02
2.32811645e-01 -1.89879626e-01 1.35138130e+00 1.71213746e-01
-6.85406685e-01 -7.32719302e-02 6.03723168e-01 -6.33394539e-01
6.84803545e-01 -1.09129262e+00 -1.20366625e-01 1.50262311e-01
-7.50486970e-01 1.06212986e+00 9.81492043e-01 -9.32009935e-01
-1.59875286e+00 -9.93812799e-01 6.49676800e-01 -1.03798644e-04
3.53976965e-01 6.53304681e-02 -6.47618771e-01 1.98095545e-01
-2.56722458e-02 -4.89675492e-01 1.74612343e+00 2.88162053e-01
-3.03458750e-01 6.42365264e-03 -1.71060252e+00 3.11234176e-01
3.45118552e-01 -9.50335562e-02 -4.30126756e-01 4.83286500e-01
3.13792437e-01 2.97285970e-02 -1.24781072e+00 6.38025939e-01
1.12620854e+00 -1.17706299e+00 1.12955117e+00 -8.00457776e-01
5.80705345e-01 -5.31828344e-01 -4.52000409e-01 -1.25485575e+00
-2.44587272e-01 5.91887012e-02 -1.01163380e-01 1.01869965e+00
2.85127722e-02 -1.06675184e+00 4.21162993e-01 1.23827472e-01
2.61350334e-01 -9.12561953e-01 -8.64469290e-01 -6.63193643e-01
1.33137340e-02 1.91895097e-01 7.42951632e-01 1.18796408e+00
1.98104918e-01 -5.00631183e-02 -9.42284614e-02 -7.31414333e-02
8.28828931e-01 3.18888247e-01 9.05411601e-01 -1.29582882e+00
-3.71717364e-01 -2.94860184e-01 -5.68215251e-01 2.58942515e-01
-4.13252234e-01 -8.78167510e-01 -3.69435847e-01 -1.52250111e+00
2.04548746e-01 -7.47904897e-01 -2.46558890e-01 5.60565256e-02
-4.18803871e-01 9.63470116e-02 -2.66432256e-01 3.29224467e-01
6.47595823e-01 -9.89055540e-03 9.31247175e-01 1.17624961e-01
-3.58411044e-01 -5.87133802e-02 -8.20559680e-01 3.57533753e-01
1.11344564e+00 -5.87156832e-01 -4.55364510e-02 6.40999715e-05
2.39534140e-01 -2.72280425e-01 2.82443821e-01 -8.40266883e-01
-2.50854254e-01 -3.59561741e-01 8.40688646e-01 -4.42867488e-01
7.26567805e-02 -7.38514245e-01 1.08943546e+00 1.45871007e+00
2.03965962e-01 -1.03042126e-01 2.67273933e-01 5.81188381e-01
-4.48208069e-03 -5.81293628e-02 8.08868289e-01 -3.40053767e-01
-2.53780395e-01 -4.40051317e-01 -4.22264516e-01 -7.69075990e-01
1.26812387e+00 -4.56020325e-01 -4.28963423e-01 5.02530813e-01
-4.24761504e-01 -6.12229370e-02 9.03917611e-01 2.09122360e-01
1.28130987e-01 -1.05666518e+00 -7.65884519e-01 3.60913187e-01
2.56414890e-01 -2.80952156e-01 1.82489291e-01 9.25035894e-01
-1.11933863e+00 3.57749760e-01 -7.01907694e-01 -9.17526305e-01
-1.51735961e+00 6.60192251e-01 2.77878106e-01 4.70410250e-02
-3.51611435e-01 7.00575352e-01 -2.76338905e-01 -6.81523442e-01
-1.89163208e-01 -1.35448743e-02 -4.40216690e-01 2.94780433e-01
4.79811728e-01 9.79642510e-01 2.83000588e-01 -3.94594491e-01
-4.71842229e-01 3.70576411e-01 2.52790779e-01 4.84732866e-01
1.66043639e+00 1.74449682e-01 -5.95239401e-01 5.02777934e-01
1.06740665e+00 -1.14970624e-01 -5.00645757e-01 3.90805662e-01
1.94300953e-02 -6.39820814e-01 -2.70770639e-01 -9.55127597e-01
-3.30371082e-01 6.66219234e-01 8.60054731e-01 3.17494422e-01
1.01887918e+00 -7.09590375e-01 3.35893899e-01 6.23621345e-02
5.79973161e-01 -8.46254349e-01 -4.08167422e-01 -6.24055825e-02
9.61281598e-01 -1.13021874e+00 4.17220324e-01 -6.11626446e-01
-1.03173852e-01 1.06673157e+00 3.36592883e-01 3.39758359e-02
5.61983883e-01 1.68661639e-01 1.69560641e-01 4.07444537e-02
-8.58502328e-01 3.67162615e-01 -2.57308453e-01 1.00474405e+00
9.95015681e-01 4.47473615e-01 -7.78884232e-01 2.57650733e-01
-5.43667614e-01 1.50823504e-01 2.76827216e-01 1.23050475e+00
-6.30289316e-01 -1.15367115e+00 -1.07223964e+00 7.66492307e-01
-6.07477248e-01 7.17481971e-02 -2.96564847e-01 4.74928111e-01
2.54764646e-01 1.14284050e+00 -2.94171453e-01 -3.84942532e-01
1.20741889e-01 3.94579947e-01 4.52769399e-01 -3.25821072e-01
-9.49917376e-01 -6.93682358e-02 3.75785440e-01 -6.94209859e-02
-4.64849561e-01 -9.44697022e-01 -6.98358119e-01 -2.28228614e-01
-6.34019911e-01 2.34253183e-01 1.62936080e+00 4.27485943e-01
3.67063105e-01 5.65510869e-01 7.88116932e-01 -6.97398961e-01
-5.91697931e-01 -1.16575634e+00 -5.20057917e-01 3.06464911e-01
1.84491009e-01 -7.15890765e-01 -6.27599955e-01 1.86877340e-01] | [5.123040199279785, 5.063233375549316] |
40d67bc5-bbdf-4371-9c49-0f9fb0e1a650 | robust-knowledge-adaptation-for-federated | 2301.07320 | null | https://arxiv.org/abs/2301.07320v1 | https://arxiv.org/pdf/2301.07320v1.pdf | Robust Knowledge Adaptation for Federated Unsupervised Person ReID | Person Re-identification (ReID) has been extensively studied in recent years due to the increasing demand in public security. However, collecting and dealing with sensitive personal data raises privacy concerns. Therefore, federated learning has been explored for Person ReID, which aims to share minimal sensitive data between different parties (clients). However, existing federated learning based person ReID methods generally rely on laborious and time-consuming data annotations and it is difficult to guarantee cross-domain consistency. Thus, in this work, a federated unsupervised cluster-contrastive (FedUCC) learning method is proposed for Person ReID. FedUCC introduces a three-stage modelling strategy following a coarse-to-fine manner. In detail, generic knowledge, specialized knowledge and patch knowledge are discovered using a deep neural network. This enables the sharing of mutual knowledge among clients while retaining local domain-specific knowledge based on the kinds of network layers and their parameters. Comprehensive experiments on 8 public benchmark datasets demonstrate the state-of-the-art performance of our proposed method. | ['Zhiyong Wang', 'Jingya Wang', 'Tingting Yao', 'Kun Hu', 'Jianfeng Weng'] | 2023-01-18 | null | null | null | null | ['person-re-identification'] | ['computer-vision'] | [-1.42299727e-01 -2.58048564e-01 -1.27643615e-01 -7.32506454e-01
-7.36686468e-01 -5.17215312e-01 7.38886416e-01 2.08466142e-01
-5.32662511e-01 8.26249778e-01 2.16127083e-01 2.90598065e-01
-3.45779359e-01 -6.40386939e-01 -4.65451568e-01 -8.41557801e-01
5.53569868e-02 4.22492176e-01 -5.53253368e-02 1.42574936e-01
2.49869339e-02 3.69286746e-01 -1.51396573e+00 3.41619432e-01
9.20239925e-01 8.19272041e-01 -4.06337082e-01 1.61431748e-02
-2.81433374e-01 5.28846800e-01 -4.10167158e-01 -1.11153257e+00
2.99506664e-01 -1.88499272e-01 -1.20857871e+00 -4.16716993e-01
3.23189229e-01 -2.42926985e-01 -2.35893086e-01 1.30640578e+00
7.60022581e-01 1.79529637e-01 4.42142695e-01 -1.34437346e+00
-9.91216302e-01 7.83918500e-01 -5.33452392e-01 -4.01985608e-02
4.07231897e-01 -2.15522364e-01 5.82529068e-01 -5.35524309e-01
4.46204126e-01 1.22047758e+00 8.35701227e-01 8.62534106e-01
-1.01270235e+00 -1.00349402e+00 2.95626074e-01 4.74482298e-01
-1.74862409e+00 -3.44021857e-01 6.86806500e-01 -2.41657838e-01
4.56268311e-01 2.02960089e-01 1.54864758e-01 1.06367886e+00
-2.19251618e-01 8.52626085e-01 9.62177098e-01 -2.00668558e-01
2.83200383e-01 5.06705582e-01 2.44682282e-01 1.44325957e-01
5.87887585e-01 4.21695746e-02 -4.56509918e-01 -4.65717286e-01
1.74511686e-01 3.59369695e-01 -2.27489144e-01 -3.56736749e-01
-7.82168508e-01 5.10168433e-01 4.92913097e-01 2.63665766e-01
-1.76551521e-01 -3.25237006e-01 6.84605777e-01 1.90281346e-01
3.63206774e-01 -1.49965897e-01 -3.87374490e-01 1.46311998e-01
-6.27768219e-01 4.92095441e-01 8.40699673e-01 1.18631423e+00
8.78404140e-01 -5.02082109e-01 -3.52067873e-02 9.00311112e-01
1.81691483e-01 1.90947846e-01 5.88653862e-01 -5.79936028e-01
4.37421441e-01 9.33255315e-01 2.27619112e-01 -1.08071041e+00
-2.92391688e-01 -3.90855163e-01 -1.28838408e+00 -1.91482291e-01
2.28294343e-01 -1.31035253e-01 -3.50514233e-01 1.84171593e+00
7.75034726e-01 3.35338980e-01 2.73665577e-01 9.49482083e-01
7.84794986e-01 9.07111913e-02 5.75995743e-01 1.14717044e-01
1.36794722e+00 -5.08075714e-01 -7.17107594e-01 1.99634150e-01
5.26784778e-01 -4.35139149e-01 2.95533508e-01 1.61541030e-02
-7.31438041e-01 -4.55074698e-01 -7.68638670e-01 -1.27557576e-01
-8.83364737e-01 -3.45548928e-01 6.83002174e-01 1.08740985e+00
-7.74338722e-01 5.12644827e-01 -3.90456051e-01 -5.65454960e-01
8.21075857e-01 5.79176962e-01 -7.65884340e-01 -3.21880937e-01
-1.34541392e+00 5.17195880e-01 5.67883611e-01 1.57034740e-01
-4.69328314e-01 -7.48975933e-01 -7.20440149e-01 8.18374157e-02
1.99048325e-01 -7.44099498e-01 1.15376401e+00 -7.25170195e-01
-1.04409540e+00 1.12671137e+00 1.26054035e-02 -3.60751241e-01
8.02386642e-01 -2.15650618e-01 -7.35636950e-01 -2.00249985e-01
2.02112213e-01 3.93768430e-01 5.16048729e-01 -1.33033693e+00
-1.02352667e+00 -8.53010178e-01 -1.52101338e-01 1.71322133e-02
-5.25519848e-01 3.58185321e-01 -5.66858590e-01 -4.01218861e-01
-1.29625767e-01 -6.05143785e-01 1.77195743e-01 -1.95797905e-01
-3.52278709e-01 -6.64186776e-01 8.56833339e-01 -6.95748627e-01
1.01759195e+00 -1.96479094e+00 -3.25678289e-01 2.79602468e-01
1.82690144e-01 4.85488087e-01 6.64963797e-02 2.75992811e-01
-5.75510785e-02 6.53971806e-02 -1.94963813e-01 -8.25372756e-01
2.10027382e-01 3.01201176e-03 1.00567162e-01 6.41175926e-01
-1.91792697e-01 8.32315445e-01 -7.55857170e-01 -7.95872688e-01
-2.85158753e-02 6.09265864e-01 -3.39496672e-01 3.37518543e-01
3.39275986e-01 5.62968612e-01 -6.08220100e-01 9.69820023e-01
1.21556258e+00 -1.04936197e-01 2.63226002e-01 -1.28016755e-01
-1.73139989e-01 -2.65088081e-01 -1.38489556e+00 1.79378164e+00
1.53904026e-02 -1.36484131e-01 2.62739629e-01 -9.14191008e-01
9.53576088e-01 4.25849706e-01 5.09434164e-01 -7.42793500e-01
3.90522271e-01 2.09427476e-01 -6.06867731e-01 -3.79938155e-01
4.82970566e-01 9.11802053e-02 -2.23114461e-01 6.08432829e-01
-1.25984356e-01 1.00286925e+00 -1.90235287e-01 2.45593265e-01
6.42390668e-01 2.28200778e-02 1.18769161e-01 -2.06533238e-01
9.25874054e-01 -1.65764257e-01 8.57122660e-01 7.03806877e-01
-5.61039507e-01 3.94325256e-01 -6.96085766e-03 -6.32599175e-01
-6.57303572e-01 -7.51917362e-01 -1.84008285e-01 9.85425472e-01
3.85914534e-01 -3.94196063e-02 -9.90538657e-01 -7.72430301e-01
3.77717495e-01 3.64599913e-01 -6.36373639e-01 -2.70999253e-01
-4.93458122e-01 -7.01031685e-01 8.33574593e-01 4.89675969e-01
1.00318813e+00 -9.42604661e-01 -3.46463919e-02 1.28881544e-01
-3.62854153e-01 -9.81520593e-01 -5.88628769e-01 -3.33673030e-01
-3.87861878e-01 -1.32300293e+00 -8.78177106e-01 -9.08266187e-01
7.52660573e-01 4.61142838e-01 7.06062376e-01 1.41757205e-01
-3.33884597e-01 4.44691032e-01 -3.42484802e-01 -3.93390983e-01
4.05570157e-02 1.72242790e-01 3.80723149e-01 6.00344121e-01
1.02030039e+00 -6.24369025e-01 -7.69389689e-01 4.08599406e-01
-7.41933465e-01 -3.90309721e-01 4.08607662e-01 6.52278602e-01
3.75023752e-01 1.41648740e-01 8.62544477e-01 -1.05407524e+00
6.24809325e-01 -7.12655067e-01 -5.15635371e-01 5.58144689e-01
-7.01613426e-01 -8.52306485e-02 6.02440596e-01 -2.99575865e-01
-1.57715654e+00 7.63260275e-02 3.33705872e-01 -1.89336464e-01
-5.11597693e-01 8.90707374e-02 -7.71415055e-01 -2.39578217e-01
3.85128736e-01 2.70833611e-01 -1.93592995e-01 -7.67373681e-01
3.07938784e-01 1.09057760e+00 1.09443629e+00 -8.69478762e-01
8.22071493e-01 5.80458283e-01 -4.23692167e-01 -2.36771423e-02
-6.04357064e-01 -7.73937285e-01 -8.32992673e-01 -4.06462178e-02
7.80461490e-01 -1.07264173e+00 -1.34662938e+00 9.44507837e-01
-1.12120581e+00 3.23888510e-01 1.45364031e-01 3.21381450e-01
-9.83757749e-02 5.27944148e-01 -4.39087361e-01 -9.11457181e-01
-8.80441308e-01 -5.72201729e-01 6.77719057e-01 6.42022967e-01
2.31771357e-02 -7.59102166e-01 8.38206634e-02 5.36673903e-01
5.37310958e-01 2.51866817e-01 5.74934840e-01 -1.01407456e+00
-5.93753695e-01 -5.74992239e-01 -7.01418996e-01 4.79290448e-02
3.03185880e-01 -7.35019505e-01 -1.17925978e+00 -4.15026486e-01
-3.56514931e-01 -2.10195124e-01 5.75769722e-01 -1.42562896e-01
1.18780434e+00 -5.15469432e-01 -5.62613726e-01 7.26982892e-01
1.45998299e+00 -6.62041977e-02 4.85135108e-01 5.74205160e-01
7.57091403e-01 9.62081909e-01 2.57211715e-01 6.31785214e-01
9.30654109e-01 4.07878041e-01 1.70498639e-01 1.53345823e-01
3.46952587e-01 -3.05604577e-01 -3.10825169e-01 9.54756811e-02
-2.18322009e-01 -1.00639910e-01 -7.45323479e-01 7.37799644e-01
-2.18883657e+00 -9.72543776e-01 9.03089046e-02 2.29651356e+00
9.23164904e-01 -4.42317307e-01 1.49913892e-01 -4.44707535e-02
1.23193693e+00 -8.89517665e-02 -7.55721450e-01 -1.02607816e-01
-1.65997788e-01 -2.07710877e-01 5.73944092e-01 -1.89221933e-01
-1.22138762e+00 6.70667887e-01 5.03367281e+00 5.50374508e-01
-7.27482378e-01 3.15820277e-01 5.04460454e-01 7.84602016e-02
-2.56644133e-02 -1.78580508e-01 -1.04164839e+00 7.27981985e-01
5.81838906e-01 -4.80478197e-01 3.30028564e-01 9.77816463e-01
-2.12941244e-01 2.39994153e-01 -1.18360949e+00 1.41777372e+00
2.52109929e-03 -1.18866944e+00 -5.21796942e-02 3.74619700e-02
8.69020522e-01 -2.88394421e-01 -1.05184302e-01 3.29243660e-01
4.77031231e-01 -7.62006700e-01 3.92487019e-01 7.51369953e-01
6.93398774e-01 -1.23808336e+00 1.09644699e+00 3.33303392e-01
-1.28922296e+00 -3.22317600e-01 -4.87870634e-01 3.69427115e-01
4.84479554e-02 2.58692205e-01 -2.58759737e-01 9.91246760e-01
1.30790925e+00 4.78380740e-01 -4.19194907e-01 1.11594474e+00
2.30521753e-01 1.19824074e-01 -1.36925027e-01 2.18178228e-01
-2.09253833e-01 1.09742939e-01 -2.83887591e-02 1.32421219e+00
-5.97299524e-02 3.94507498e-01 2.34798908e-01 9.02273417e-01
-4.90376711e-01 1.78474933e-01 -1.70147464e-01 3.85069191e-01
7.54947543e-01 1.24727678e+00 -8.89649242e-02 -1.55171573e-01
-4.61129755e-01 1.16872621e+00 5.03846467e-01 2.34566972e-01
-4.85895425e-01 -3.64401162e-01 8.64487350e-01 -2.09799051e-01
4.65433486e-02 2.64844209e-01 -1.01453796e-01 -1.05953276e+00
1.59450829e-01 -6.50862217e-01 1.04552782e+00 -4.64841872e-02
-1.86271846e+00 3.46933573e-01 -5.22688217e-03 -1.06706071e+00
-4.95197289e-02 -1.11802362e-01 -7.28480339e-01 1.17097199e+00
-1.88440871e+00 -1.57181740e+00 -7.03131258e-01 1.21156740e+00
-2.34422117e-01 -3.26653391e-01 1.00967634e+00 6.42509520e-01
-9.31631327e-01 1.36191547e+00 1.95690989e-01 6.13116264e-01
9.83428419e-01 -9.20485854e-01 4.00532961e-01 7.90294409e-01
-6.07007265e-01 1.01122344e+00 2.45242089e-01 -8.47378850e-01
-1.33302712e+00 -1.38831222e+00 1.19137311e+00 -3.14983845e-01
1.78877547e-01 -2.33073264e-01 -1.21251225e+00 6.37780607e-01
-5.70484176e-02 2.62488306e-01 9.97321963e-01 3.20831925e-01
-9.10763264e-01 -5.16040862e-01 -1.72498965e+00 1.20886996e-01
1.15307724e+00 -6.97270811e-01 -3.53618681e-01 1.14888482e-01
4.73027706e-01 -7.84669816e-02 -8.71310771e-01 8.99117142e-02
7.44524658e-01 -8.79771709e-01 9.33163166e-01 -7.40126193e-01
-3.61565948e-01 -4.05394703e-01 -4.77828719e-02 -7.68564701e-01
-5.03989637e-01 -4.56469148e-01 -4.51796576e-02 1.93880606e+00
-9.28198397e-02 -9.05173600e-01 9.15946305e-01 1.50423145e+00
2.73265690e-01 -1.13507807e-01 -9.23930228e-01 -9.17219043e-01
3.45183462e-02 -9.97995287e-02 1.37577045e+00 1.37432325e+00
5.91935888e-02 -3.19303781e-01 -5.52231491e-01 5.90318799e-01
1.24920166e+00 1.11177741e-02 8.54073107e-01 -1.71415520e+00
1.97923496e-01 -1.85256049e-01 -5.68537533e-01 -4.91233230e-01
3.55949193e-01 -9.57058370e-01 -2.89323241e-01 -1.29509020e+00
6.42521381e-01 -7.60714710e-01 -6.69602096e-01 7.74610758e-01
-2.36418337e-01 -4.05907966e-02 -6.47291467e-02 5.13379812e-01
-9.50938046e-01 4.27337527e-01 5.13930917e-01 -3.11250150e-01
-1.69864550e-01 -5.15565425e-02 -1.06028616e+00 1.91571802e-01
8.27276528e-01 -4.16151673e-01 -8.72017369e-02 -4.91608649e-01
-2.26115569e-01 -4.00490940e-01 6.32412255e-01 -9.62784231e-01
9.22530532e-01 6.67404011e-02 6.20717883e-01 -6.34876072e-01
3.84790711e-02 -9.63669538e-01 3.97992462e-01 1.20695166e-01
-2.67124981e-01 -2.82942235e-01 -2.30089878e-03 7.91637540e-01
-1.47041067e-01 7.25816041e-02 8.03133011e-01 -2.05559865e-01
-8.56014371e-01 6.36502683e-01 3.97927284e-01 -2.95423597e-01
9.07994032e-01 -1.80017784e-01 -4.09263492e-01 -5.23228273e-02
-2.90644944e-01 6.18263483e-01 4.57262844e-01 5.77537596e-01
3.61220092e-01 -1.41380441e+00 -9.11761642e-01 1.78273559e-01
5.11488855e-01 6.14353754e-02 9.19263363e-01 3.87510329e-01
1.33450702e-01 4.90209460e-01 -2.48737991e-01 -2.73233771e-01
-1.36620343e+00 8.26074243e-01 3.72097552e-01 1.67743191e-02
-6.05934858e-01 8.01690876e-01 -5.25036752e-02 -7.67360270e-01
5.63921869e-01 5.39168417e-01 -2.55783170e-01 4.67141122e-02
1.12026608e+00 5.96567512e-01 1.12960160e-01 -8.97055387e-01
-7.19436884e-01 4.20043081e-01 -5.94204962e-01 4.96206224e-01
1.18501306e+00 -3.49184334e-01 -3.12267572e-01 -3.98710310e-01
1.40516710e+00 -4.41249311e-01 -9.56975460e-01 -7.14012384e-01
1.58372730e-01 -5.09401739e-01 -2.69130737e-01 -7.62468159e-01
-1.04076111e+00 2.14032412e-01 8.87084067e-01 -3.15435678e-01
1.05585659e+00 -4.11961563e-02 9.15376365e-01 4.66745794e-01
8.16579163e-01 -1.20148849e+00 -6.99131966e-01 7.90448394e-03
3.88697267e-01 -1.46093941e+00 -6.64709089e-03 -2.17055142e-01
-3.00420135e-01 6.84274673e-01 6.89962208e-01 4.21190739e-01
6.68759584e-01 -1.98542535e-01 1.44185752e-01 8.48621354e-02
-2.73185402e-01 -2.89965570e-02 3.85465845e-02 9.21901166e-01
4.31679562e-03 2.23018497e-01 -2.84694493e-01 1.15914798e+00
-2.51862369e-02 1.28995165e-01 -1.25646397e-01 1.08562791e+00
3.23364437e-02 -1.34074879e+00 -5.45216799e-01 2.99487971e-02
-5.92338860e-01 2.98518121e-01 -4.04855639e-01 4.97606158e-01
4.14577186e-01 1.17301083e+00 -1.29940018e-01 -2.64524162e-01
1.98635519e-01 1.46941498e-01 -1.16438651e-02 -1.33024424e-01
-7.91861236e-01 -4.84174669e-01 -1.96760893e-01 -4.60601330e-01
-6.72662079e-01 -7.02886403e-01 -1.23436189e+00 -7.24768221e-01
-2.67562091e-01 3.28704625e-01 6.31909966e-01 9.55869973e-01
6.53707027e-01 -1.87193111e-01 6.86404943e-01 -4.18053716e-01
-4.24886018e-01 -6.27407968e-01 -6.84706211e-01 7.73409307e-01
1.13274187e-01 -4.86543417e-01 -7.41221383e-02 3.75948697e-02] | [14.754451751708984, 1.0724077224731445] |
513fb0e4-8705-4689-8772-cdb2e7983b72 | distilling-relation-embeddings-from-pre | 2110.15705 | null | https://arxiv.org/abs/2110.15705v1 | https://arxiv.org/pdf/2110.15705v1.pdf | Distilling Relation Embeddings from Pre-trained Language Models | Pre-trained language models have been found to capture a surprisingly rich amount of lexical knowledge, ranging from commonsense properties of everyday concepts to detailed factual knowledge about named entities. Among others, this makes it possible to distill high-quality word vectors from pre-trained language models. However, it is currently unclear to what extent it is possible to distill relation embeddings, i.e. vectors that characterize the relationship between two words. Such relation embeddings are appealing because they can, in principle, encode relational knowledge in a more fine-grained way than is possible with knowledge graphs. To obtain relation embeddings from a pre-trained language model, we encode word pairs using a (manually or automatically generated) prompt, and we fine-tune the language model such that relationally similar word pairs yield similar output vectors. We find that the resulting relation embeddings are highly competitive on analogy (unsupervised) and relation classification (supervised) benchmarks, even without any task-specific fine-tuning. Source code to reproduce our experimental results and the model checkpoints are available in the following repository: https://github.com/asahi417/relbert | ['Steven Schockaert', 'Jose Camacho-Collados', 'Asahi Ushio'] | 2021-09-21 | null | null | null | null | ['relation-classification'] | ['natural-language-processing'] | [-1.70399100e-01 2.60180295e-01 -4.74898070e-01 -3.67900699e-01
-4.25776452e-01 -8.27641249e-01 9.84425366e-01 7.74009526e-01
-4.55674440e-01 6.30485356e-01 5.86607158e-01 -5.50465584e-01
-1.61124021e-01 -1.17331791e+00 -4.43389088e-01 -2.71822691e-01
-2.67686937e-02 5.63321769e-01 1.19037829e-01 -5.07847607e-01
2.36665487e-01 4.89124835e-01 -1.27078307e+00 1.20474584e-01
6.02663159e-01 7.33688951e-01 2.89092958e-02 4.99975353e-01
-3.67469817e-01 8.24743509e-01 -2.67322123e-01 -7.77313709e-01
-7.09999800e-02 -2.99300760e-01 -1.41142964e+00 -4.46151376e-01
4.81592454e-02 1.55506298e-01 -4.18946236e-01 8.47481191e-01
7.89859816e-02 3.98773640e-01 9.25190091e-01 -9.39019263e-01
-1.43014181e+00 8.29662263e-01 -4.61598895e-02 3.88157547e-01
6.40200853e-01 -2.89000779e-01 1.79444873e+00 -9.26852047e-01
7.89015472e-01 1.20816493e+00 5.08903682e-01 2.65553534e-01
-1.35502326e+00 -4.45703357e-01 -5.93331046e-02 4.68526185e-01
-1.67500961e+00 -3.36447299e-01 7.54234910e-01 -4.24131751e-01
1.57538724e+00 5.35629690e-02 5.17464161e-01 1.21694851e+00
1.34283230e-01 3.01575810e-01 8.52128923e-01 -7.47086942e-01
-2.49705114e-03 3.36230308e-01 4.26781744e-01 5.88207483e-01
4.96026874e-01 -3.70975360e-02 -4.73922253e-01 -3.01579535e-01
7.26979077e-01 -2.11488325e-02 -3.81331265e-01 -5.36850631e-01
-1.31711841e+00 1.12101197e+00 7.20843494e-01 9.03670669e-01
-3.51484954e-01 1.89699098e-01 4.53979999e-01 4.69063461e-01
3.27459067e-01 9.51020598e-01 -6.39535546e-01 -7.94344991e-02
-2.34160691e-01 5.22626601e-02 9.47117269e-01 8.86669278e-01
9.99343395e-01 -2.31911913e-01 5.03381751e-02 9.46214378e-01
1.37055457e-01 2.13773757e-01 7.05969751e-01 -6.53770149e-01
2.51701206e-01 5.91499388e-01 -1.90609097e-01 -1.22778761e+00
-2.31669441e-01 -1.11473799e-01 -5.44164360e-01 -3.64974618e-01
2.59948909e-01 1.77423969e-01 -4.29822683e-01 1.79806256e+00
7.12490976e-02 -4.66981158e-02 3.54762286e-01 7.08637834e-01
1.01464510e+00 5.88310361e-01 2.57283866e-01 1.79721322e-02
1.60734868e+00 -5.69463491e-01 -5.78634202e-01 -2.47040927e-01
1.11609304e+00 -6.80373549e-01 1.28902197e+00 -1.14978753e-01
-7.25013435e-01 -4.14067358e-01 -8.87473762e-01 -4.95614946e-01
-9.76383626e-01 -3.80337507e-01 1.14249611e+00 4.20868933e-01
-8.73253703e-01 5.33417940e-01 -6.31349266e-01 -8.04187655e-01
3.53985995e-01 4.82502095e-02 -7.65316308e-01 -5.02453595e-02
-1.66336071e+00 1.29957306e+00 6.03266537e-01 -1.77777573e-01
-3.42813015e-01 -5.74728549e-01 -1.23037410e+00 1.34649957e-02
4.33578104e-01 -6.82501078e-01 9.70072031e-01 -5.55397987e-01
-1.08292627e+00 1.14065003e+00 -2.35413551e-01 -4.08926100e-01
-3.09088796e-01 -5.12143858e-02 -5.08736074e-01 -1.30965307e-01
1.11025207e-01 4.38070953e-01 2.30983943e-01 -1.07508612e+00
-1.55572727e-01 -9.23584178e-02 4.03209835e-01 1.17814019e-01
-6.19056582e-01 4.66864407e-02 -1.24714367e-01 -5.86391628e-01
-1.56955123e-01 -6.89563155e-01 -1.06943818e-02 -7.72718061e-03
-3.92497331e-01 -7.16843605e-01 2.30671555e-01 -3.22324455e-01
1.21325064e+00 -2.01056957e+00 1.60497040e-01 2.10535705e-01
2.84483105e-01 3.18481952e-01 -3.22765291e-01 8.58646452e-01
-3.44092637e-01 3.84345442e-01 -1.84987038e-01 6.40188679e-02
2.10038081e-01 6.89362764e-01 -4.76695359e-01 2.28415728e-01
3.30398828e-01 1.32745993e+00 -1.24196136e+00 -3.32740039e-01
1.68481365e-01 3.81759644e-01 -3.96617055e-01 2.65733838e-01
-2.55646497e-01 -3.31409834e-02 -4.87696171e-01 2.77130038e-01
8.23922008e-02 -4.36577916e-01 4.82833654e-01 -2.91211963e-01
3.84373754e-01 9.32271779e-01 -9.42984045e-01 1.71479201e+00
-9.52875555e-01 7.25339711e-01 -7.15966463e-01 -1.22485149e+00
1.08402658e+00 4.18754578e-01 7.61158690e-02 -5.83173931e-01
2.12860152e-01 1.25606582e-01 2.40581557e-01 -4.92423207e-01
6.56422019e-01 -5.30901074e-01 -2.67208248e-01 6.01989508e-01
3.42940599e-01 -3.29265743e-01 1.40514985e-01 3.38917643e-01
1.18657184e+00 -5.92583567e-02 9.39614236e-01 -2.19735801e-01
4.75266963e-01 -7.75933778e-03 2.10938215e-01 3.59193832e-01
1.12012744e-01 2.90859789e-01 5.16075611e-01 -3.12650174e-01
-9.48260248e-01 -1.30327356e+00 -3.57523859e-01 1.09346259e+00
5.43574896e-03 -7.64024436e-01 -7.40059391e-02 -5.27407587e-01
1.56745210e-01 9.32836652e-01 -7.92268217e-01 -3.24004799e-01
-3.90391171e-01 -2.62524456e-01 5.41155577e-01 8.01607370e-01
-1.88636780e-02 -1.26396120e+00 -9.27288160e-02 1.07788034e-01
1.79826226e-02 -1.27917671e+00 -1.35040581e-01 4.12895620e-01
-6.61341965e-01 -1.11184156e+00 -1.54366121e-01 -8.75391543e-01
4.89883542e-01 8.19249451e-02 1.53206038e+00 2.52807170e-01
-2.11158440e-01 4.07041043e-01 -6.89904451e-01 -1.17699727e-01
-4.29833263e-01 1.02201127e-01 1.66230053e-01 -4.01882291e-01
9.05293047e-01 -8.31450403e-01 -1.64335400e-01 -1.68435931e-01
-8.74200046e-01 -3.40672493e-01 3.42150092e-01 9.01260853e-01
5.36187887e-01 -6.65385574e-02 4.20017272e-01 -1.00200880e+00
1.02622640e+00 -6.30742311e-01 -1.15647644e-01 3.38864684e-01
-5.47635078e-01 4.18936938e-01 7.46042311e-01 -4.46246594e-01
-5.95812082e-01 -4.13614988e-01 -2.10195675e-01 -2.08889589e-01
-3.01748842e-01 6.30612314e-01 -7.39922225e-02 1.30888283e-01
9.32324350e-01 1.41638175e-01 -4.19617444e-01 -3.67911667e-01
1.09099495e+00 5.90987682e-01 3.98353279e-01 -9.80968058e-01
8.43778312e-01 1.78068295e-01 -1.41251758e-01 -8.74948323e-01
-1.09056687e+00 -5.52196860e-01 -7.44172454e-01 5.79769790e-01
9.98160303e-01 -6.45971656e-01 -4.50624496e-01 -4.44331855e-01
-1.44057822e+00 -2.86473691e-01 -5.76999605e-01 5.49845397e-01
-3.69516671e-01 1.96928173e-01 -5.15171170e-01 -3.60650599e-01
-1.55482411e-01 -4.96784061e-01 6.26268625e-01 1.73805133e-02
-9.15813863e-01 -1.50405538e+00 2.64782965e-01 1.23524450e-01
2.35776335e-01 6.65096641e-02 1.41147327e+00 -1.17844760e+00
-2.73213357e-01 -1.41155243e-01 -2.82497793e-01 3.76058340e-01
4.60537642e-01 -1.14218324e-01 -8.21665049e-01 2.10854053e-01
-3.59827340e-01 -5.25048494e-01 7.70640194e-01 -2.22894058e-01
8.06384921e-01 -2.48200685e-01 -3.83587718e-01 3.22422445e-01
1.38067031e+00 -1.31017745e-01 5.39883614e-01 2.50108689e-01
8.96504998e-01 5.14725447e-01 3.16416711e-01 8.50219354e-02
6.60846114e-01 6.84439003e-01 -4.39412445e-02 3.81351084e-01
-1.36816725e-01 -4.97642219e-01 3.14838323e-03 1.07078946e+00
-2.19066203e-01 -1.87879324e-01 -1.16520786e+00 9.18091714e-01
-1.61311448e+00 -1.05401790e+00 -1.24491185e-01 1.99371564e+00
1.29272592e+00 1.96081083e-02 -1.98969364e-01 1.37990788e-01
2.91966230e-01 2.84044802e-01 -9.48919654e-02 -8.36288989e-01
-6.74845651e-02 7.72225678e-01 1.99906081e-01 7.88813293e-01
-8.00362587e-01 1.29881954e+00 5.66545200e+00 6.68945193e-01
-7.20910430e-01 3.51867348e-01 4.10632677e-02 3.08195591e-01
-8.31554294e-01 2.09479868e-01 -4.90085930e-01 2.04899721e-02
1.02541578e+00 -4.78832453e-01 3.36712480e-01 5.99099755e-01
-3.49068373e-01 1.60614341e-01 -1.50059819e+00 9.90038812e-01
-2.34347209e-02 -1.50860202e+00 2.19154939e-01 -1.22727789e-01
4.94573981e-01 -2.79207830e-03 -2.95198619e-01 4.75277513e-01
6.00986898e-01 -1.21908677e+00 2.15160713e-01 3.96002382e-01
6.00670993e-01 -5.26475728e-01 6.54731393e-01 1.23149671e-01
-1.33910358e+00 2.50144929e-01 -5.87220371e-01 -3.15797925e-01
6.59021288e-02 6.01884186e-01 -7.31390297e-01 7.03351617e-01
3.76725495e-01 8.00652921e-01 -6.28306031e-01 4.35912132e-01
-8.35468411e-01 5.34651339e-01 -1.63862869e-01 -3.55778068e-01
5.30077256e-02 1.08314259e-02 1.92456171e-01 1.30134165e+00
4.53695767e-02 2.91869193e-01 -1.05379730e-01 1.02334583e+00
-3.12788755e-01 2.34105423e-01 -9.38205302e-01 -4.79463279e-01
7.55119264e-01 1.26827407e+00 -4.43560660e-01 -3.65968943e-01
-6.87299609e-01 8.69045138e-01 8.79930973e-01 3.28008652e-01
-5.35843790e-01 -5.37670076e-01 9.84475672e-01 -4.16473299e-03
3.01897615e-01 -5.11854768e-01 -1.58930272e-01 -1.32187140e+00
-4.50352393e-02 -5.00876248e-01 3.66013378e-01 -8.27473342e-01
-1.67672336e+00 7.24518239e-01 -8.84732082e-02 -6.74116433e-01
-4.92983222e-01 -8.25150371e-01 -4.85681176e-01 9.73950446e-01
-1.34706569e+00 -9.93564665e-01 1.61565691e-02 4.40898150e-01
1.14477359e-01 7.66716013e-03 1.45234919e+00 7.34034032e-02
-1.56437665e-01 5.63601255e-01 -2.84248203e-01 4.78623688e-01
5.57149529e-01 -1.32718754e+00 4.88002151e-01 4.41586882e-01
9.73021567e-01 1.16337013e+00 6.73860371e-01 -3.96731287e-01
-1.29287398e+00 -8.80004227e-01 1.53186822e+00 -8.82891834e-01
1.30050647e+00 -5.49044907e-01 -1.39526784e+00 8.38404417e-01
3.16097379e-01 4.27177846e-01 1.16019368e+00 7.76611567e-01
-1.09117317e+00 1.22752979e-01 -6.98193908e-01 6.27203763e-01
1.28540123e+00 -1.14094269e+00 -1.29263210e+00 4.08809334e-01
9.55286443e-01 -1.02489581e-02 -1.22710764e+00 4.06550951e-02
3.05166990e-01 -5.45040309e-01 1.12656617e+00 -1.04329908e+00
5.93984187e-01 -1.73296943e-01 -4.28045005e-01 -1.37715089e+00
-4.09487903e-01 -3.01106572e-01 -3.09224129e-01 1.28628159e+00
7.62541473e-01 -9.33943212e-01 3.73776108e-01 4.81661141e-01
2.27893680e-01 -9.94991362e-01 -6.60524666e-01 -1.16783607e+00
5.59718609e-01 -5.83263695e-01 5.66775262e-01 1.44444036e+00
5.30877173e-01 8.96067381e-01 1.93066210e-01 7.81833231e-02
1.98901892e-01 3.98259997e-01 4.70618784e-01 -1.25282061e+00
-4.40933228e-01 -6.28050625e-01 -7.94969618e-01 -9.41492140e-01
6.78178191e-01 -1.52598476e+00 -2.73562431e-01 -1.83043540e+00
1.27537429e-01 -6.08391941e-01 -4.01675999e-01 7.18358457e-01
-2.12974682e-01 1.57643795e-01 -1.13383427e-01 -1.09862639e-02
-3.28041583e-01 5.42548478e-01 1.05180669e+00 9.04821530e-02
5.30069433e-02 -5.01042128e-01 -9.97163832e-01 5.40549219e-01
8.22205365e-01 -5.46989679e-01 -5.84950686e-01 -4.39403921e-01
5.20483255e-01 -2.52260864e-01 5.76148808e-01 -4.18945432e-01
2.13372126e-01 -2.96232760e-01 9.31029916e-02 1.41550288e-01
5.20711482e-01 -6.82136059e-01 -1.28829107e-01 1.06992066e-01
-5.42788744e-01 1.58017799e-02 1.89053893e-01 4.69072133e-01
-3.37721676e-01 -3.59633982e-01 4.54309821e-01 1.04802800e-02
-9.06844616e-01 1.55869573e-01 -1.26543090e-01 4.80307400e-01
6.95471644e-01 2.24910788e-02 -3.45924050e-01 -3.89478832e-01
-4.63094115e-01 -2.02515259e-01 4.44082826e-01 6.50182724e-01
5.10320425e-01 -1.55021298e+00 -5.25894403e-01 -1.30304754e-01
5.68901360e-01 -1.70141235e-01 -2.22418055e-01 4.91302699e-01
-1.76742002e-01 5.35489678e-01 2.44282056e-02 -4.51063178e-02
-8.71399462e-01 8.74551833e-01 1.44079164e-01 -1.60628334e-01
-6.97473943e-01 9.99506533e-01 6.97206184e-02 -6.10365510e-01
-1.92125589e-01 -4.99394059e-01 -1.96843863e-01 1.65010314e-03
3.31588089e-01 -1.41041204e-01 -9.22057591e-03 -7.37522185e-01
-6.85407162e-01 6.34006977e-01 -5.47601320e-02 -2.26977165e-03
1.34403253e+00 1.38091296e-01 -2.54226357e-01 6.82271779e-01
1.53836584e+00 1.38854831e-01 -3.19855928e-01 -6.24069035e-01
3.29961866e-01 -3.86342227e-01 -1.28275633e-01 -3.79889935e-01
-6.56582773e-01 9.53799784e-01 -2.46287391e-01 4.95423555e-01
7.66266286e-01 5.53312600e-01 8.12801957e-01 7.47147322e-01
3.24449062e-01 -5.76790214e-01 5.91607615e-02 6.99332833e-01
8.44401896e-01 -1.09788215e+00 1.51610419e-01 -4.80179340e-01
-6.71756446e-01 1.08605886e+00 2.68387526e-01 -4.08527195e-01
8.45008731e-01 1.55823857e-01 -1.24098465e-01 -3.48765910e-01
-8.95116270e-01 -5.00657558e-01 4.33530599e-01 9.48425591e-01
9.17111397e-01 2.94655114e-01 -3.74962687e-01 7.63899267e-01
-5.56809843e-01 -2.93890089e-01 3.63573968e-01 6.72313035e-01
-2.98723847e-01 -1.40239215e+00 1.42424271e-01 4.50639397e-01
-1.62928417e-01 -5.33442616e-01 -5.13074040e-01 7.80206442e-01
-3.24101113e-02 9.18731928e-01 1.31991833e-01 -3.77654970e-01
3.08498174e-01 1.85178742e-01 6.11139536e-01 -1.09979296e+00
-3.29772383e-01 -6.44601882e-01 4.80997741e-01 -4.65366125e-01
-2.85047889e-01 -3.96420002e-01 -1.44821417e+00 -4.40749139e-01
-2.01676816e-01 2.59782344e-01 2.14738786e-01 1.16820252e+00
1.97788671e-01 3.96601617e-01 1.86354905e-01 -3.65894467e-01
-2.86184132e-01 -9.01535869e-01 -5.16241848e-01 6.08867347e-01
-4.18789685e-02 -7.08386600e-01 -2.61681467e-01 -3.70341912e-02] | [10.07752799987793, 8.73450756072998] |
356d2b6f-da94-4d5a-9a29-58b0a5b6d95e | a-shared-multi-attention-framework-for-multi | null | null | http://openaccess.thecvf.com/content_CVPR_2020/html/Huynh_A_Shared_Multi-Attention_Framework_for_Multi-Label_Zero-Shot_Learning_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Huynh_A_Shared_Multi-Attention_Framework_for_Multi-Label_Zero-Shot_Learning_CVPR_2020_paper.pdf | A Shared Multi-Attention Framework for Multi-Label Zero-Shot Learning | In this work, we develop a shared multi-attention model for multi-label zero-shot learning. We argue that designing attention mechanism for recognizing multiple seen and unseen labels in an image is a non-trivial task as there is no training signal to localize unseen labels and an image only contains a few present labels that need attentions out of thousands of possible labels. Therefore, instead of generating attentions for unseen labels which have unknown behaviors and could focus on irrelevant regions due to the lack of any training sample, we let the unseen labels select among a set of shared attentions which are trained to be label-agnostic and to focus on only relevant/foreground regions through our novel loss. Finally, we learn a compatibility function to distinguish labels based on the selected attention. We further propose a novel loss function that consists of three components guiding the attention to focus on diverse and relevant image regions while utilizing all attention features. By extensive experiments, we show that our method improves the state of the art by 2.9% and 1.4% F1 score on the NUS-WIDE and the large scale Open Images datasets, respectively.
| [' Ehsan Elhamifar', 'Dat Huynh'] | 2020-06-01 | null | null | null | cvpr-2020-6 | ['multi-label-zero-shot-learning'] | ['computer-vision'] | [ 5.22026956e-01 6.94236755e-02 -1.88167363e-01 -6.00564182e-01
-1.20264685e+00 -4.20348972e-01 4.36444968e-01 7.60423094e-02
-4.53374922e-01 6.77989364e-01 -1.80111259e-01 1.43887565e-01
3.95028256e-02 -5.40445507e-01 -8.56064677e-01 -7.99546719e-01
4.40265268e-01 4.75940883e-01 5.56469679e-01 2.56930947e-01
4.68612760e-02 2.05709502e-01 -1.76706755e+00 2.39463240e-01
5.55494845e-01 1.10658026e+00 3.55295479e-01 4.57639813e-01
-3.42941750e-03 9.56275344e-01 -5.40044844e-01 -6.38451651e-02
9.68845859e-02 -5.81166923e-01 -9.52335060e-01 3.08396310e-01
8.60346019e-01 -3.11868489e-01 -9.57430452e-02 1.16414237e+00
4.77199435e-01 3.34989607e-01 8.26890767e-01 -1.22599614e+00
-1.08344877e+00 1.60854131e-01 -7.87079155e-01 3.07599515e-01
-2.60747790e-01 1.28214151e-01 1.35184610e+00 -9.06727433e-01
6.08557284e-01 1.10631359e+00 1.50421247e-01 9.93349791e-01
-1.21291709e+00 -7.63557553e-01 5.00999570e-01 1.77276954e-01
-1.33796179e+00 -4.02090728e-01 5.46719551e-01 -5.73713720e-01
3.78637999e-01 2.90384471e-01 -1.62466019e-01 1.14462245e+00
-1.19768545e-01 7.27375984e-01 1.09050465e+00 -4.27172869e-01
2.79936194e-01 2.33405635e-01 5.28384507e-01 8.55943799e-01
1.06871828e-01 -2.78223515e-01 -3.14871937e-01 -2.17270792e-01
5.04249036e-01 2.89451122e-01 -1.73317656e-01 -3.00946712e-01
-1.24650145e+00 1.00112927e+00 5.71383715e-01 8.15020129e-02
-1.96124539e-01 1.81186005e-01 3.01072478e-01 -1.03293121e-01
6.62659883e-01 3.70651782e-01 -6.08563662e-01 4.73776489e-01
-4.64582264e-01 -3.23881477e-01 2.10354164e-01 8.86384010e-01
1.23063862e+00 -3.79547209e-01 -7.14792788e-01 1.07326400e+00
7.54873753e-02 3.21056247e-01 4.48573351e-01 -8.15514684e-01
1.36042401e-01 7.19867289e-01 1.28733739e-01 -6.75336778e-01
-2.66264021e-01 -4.15796071e-01 -7.02541769e-01 9.86090899e-02
2.67066270e-01 -2.78885454e-01 -1.44158983e+00 1.90161967e+00
4.67963547e-01 5.41076839e-01 -3.36529583e-01 1.00157368e+00
9.46470857e-01 4.90865082e-01 3.16719353e-01 -1.50288939e-02
1.36299646e+00 -1.39969933e+00 -6.15057230e-01 -4.97083932e-01
5.93783259e-01 -4.18765813e-01 1.18073583e+00 -1.08043768e-01
-7.18550146e-01 -6.22795820e-01 -8.45602334e-01 -2.65807152e-01
-3.47795427e-01 1.10861853e-01 3.99888724e-01 1.01856515e-01
-7.77063310e-01 3.84263426e-01 -4.08011168e-01 -4.08131480e-01
6.84427261e-01 2.14976549e-01 -1.36929318e-01 -3.27806592e-01
-1.09936690e+00 6.51811659e-01 1.99101791e-01 -3.24483752e-01
-1.31669593e+00 -5.35347521e-01 -7.36599624e-01 2.67682940e-01
6.55302763e-01 -4.58141506e-01 9.67444181e-01 -1.04096830e+00
-9.64774847e-01 1.12567389e+00 -3.36369336e-01 1.21499812e-02
2.00766340e-01 4.65362221e-02 -3.16923946e-01 2.12247103e-01
6.16446853e-01 9.41594005e-01 6.65152609e-01 -1.40177798e+00
-9.28987622e-01 -1.90973327e-01 1.93943828e-01 -2.86823846e-02
-4.24771219e-01 5.96166216e-02 -3.44024092e-01 -3.42119992e-01
6.40849844e-02 -9.69552875e-01 -1.52336344e-01 1.29325315e-01
-4.40153629e-01 -4.26904082e-01 7.00069726e-01 -2.49574423e-01
9.44810152e-01 -2.14767027e+00 -8.46071020e-02 -8.36307853e-02
3.73108447e-01 1.35879666e-02 -3.82850081e-01 -9.71541479e-02
7.83752650e-02 3.41093063e-01 -8.58333558e-02 -2.12653443e-01
-1.86944887e-01 2.13692486e-01 -2.03725949e-01 5.05426705e-01
5.46650708e-01 8.38495851e-01 -1.14921343e+00 -5.47710419e-01
9.02025998e-02 3.43561143e-01 -2.22988173e-01 2.77292997e-01
-2.29116231e-01 5.02141118e-01 -6.06684923e-01 7.82402217e-01
5.29241025e-01 -9.29709256e-01 -1.18507810e-01 -1.35142682e-02
1.78055868e-01 -1.05366133e-01 -8.92773926e-01 1.46755481e+00
-4.28437233e-01 2.11823627e-01 -1.35112032e-01 -8.93854260e-01
7.47781634e-01 2.49760807e-01 2.97161788e-01 -4.48088676e-01
3.18918377e-01 2.46645138e-02 -2.17403799e-01 -3.53258401e-01
-1.70365702e-02 -2.52102226e-01 -1.09791197e-01 6.50953293e-01
4.77339178e-01 5.56744635e-01 1.45619109e-01 9.08150226e-02
1.26728606e+00 -1.94369897e-01 1.63301587e-01 -2.22639158e-01
4.78629261e-01 -2.83696502e-01 9.48894799e-01 9.92354512e-01
-5.91747642e-01 8.15340757e-01 4.31256473e-01 -3.17890704e-01
-8.48684192e-01 -8.10833454e-01 -1.47849247e-01 1.88598192e+00
3.89433503e-01 2.16432750e-01 -3.88011545e-01 -1.19131899e+00
-2.02052221e-01 5.97680748e-01 -1.02708745e+00 -1.99082702e-01
-1.35979742e-01 -6.50570571e-01 1.02842078e-01 3.33883226e-01
2.15608627e-01 -1.24670196e+00 -3.74217540e-01 -8.61363336e-02
-1.68410927e-01 -9.02899683e-01 -8.35009992e-01 5.54867744e-01
-3.83817255e-01 -1.32971847e+00 -8.25383902e-01 -9.79639292e-01
1.03369296e+00 6.95993125e-01 1.04061544e+00 2.51430273e-01
-2.69122154e-01 2.08881021e-01 -4.71715748e-01 -3.00676733e-01
-7.69032240e-02 5.96682355e-02 -3.14961910e-01 5.73585808e-01
4.56610858e-01 -9.28594023e-02 -6.59829199e-01 4.82235372e-01
-7.37831235e-01 -7.54593909e-02 5.17656505e-01 1.12401688e+00
8.25039446e-01 -2.75927573e-01 9.79332924e-01 -1.25280190e+00
9.49542299e-02 -9.66169775e-01 -3.64381611e-01 4.63556409e-01
-3.53543848e-01 4.59432714e-02 4.59729552e-01 -7.46254981e-01
-9.38489974e-01 2.63318360e-01 2.10796803e-01 -6.96546614e-01
-3.53792846e-01 -1.36588037e-01 -1.39788821e-01 -7.19671100e-02
4.95636702e-01 -3.22356783e-02 -3.62073302e-01 -3.54043007e-01
4.28152055e-01 5.70596874e-01 1.39518112e-01 -5.26464939e-01
3.96811515e-01 6.05159402e-01 -1.32001415e-01 -4.52516884e-01
-1.75846684e+00 -9.18543756e-01 -4.29657757e-01 -2.51774609e-01
1.13292539e+00 -8.93397212e-01 -3.68927002e-01 1.73521504e-01
-1.06739831e+00 -2.10370794e-01 -2.70203143e-01 1.87470168e-01
-3.05162787e-01 1.37405887e-01 -5.92181027e-01 -7.54050612e-01
-1.62959784e-01 -1.31752741e+00 1.35592914e+00 2.87242472e-01
-6.64309459e-03 -7.56994426e-01 -2.14003325e-02 3.53948832e-01
2.02327847e-01 2.99833179e-01 8.75246286e-01 -1.01041842e+00
-7.02682197e-01 -1.64507329e-02 -6.43252194e-01 3.84023458e-01
1.49767235e-01 -3.36533695e-01 -1.30003262e+00 -3.22437555e-01
-8.59416425e-02 -8.90728593e-01 1.16181648e+00 2.21265569e-01
1.34643924e+00 -9.42250118e-02 -5.43265939e-01 3.59129876e-01
1.42259121e+00 1.02013893e-01 1.91649050e-01 1.68263540e-01
8.45551968e-01 6.93666518e-01 5.52761197e-01 1.96691602e-01
1.83267891e-01 4.43147123e-01 6.60230994e-01 -1.92327529e-01
-1.62885800e-01 -8.92916769e-02 -5.94200529e-02 4.73759592e-01
3.68018746e-01 -4.39247906e-01 -6.15978479e-01 8.93600762e-01
-1.90299368e+00 -8.13522398e-01 -1.13600962e-01 2.25761890e+00
6.39261544e-01 2.09233034e-02 -1.56290039e-01 -4.11356598e-01
1.40130448e+00 2.06327975e-01 -9.96737480e-01 -3.08206063e-02
-1.03539124e-01 1.96457371e-01 5.71765482e-01 3.51597667e-01
-1.43949878e+00 8.62187207e-01 5.99455166e+00 8.00033092e-01
-9.43576157e-01 5.11134863e-01 1.12588882e+00 -2.44120285e-01
-2.58919269e-01 1.20124714e-02 -9.73015368e-01 6.21414244e-01
7.55188763e-01 7.36056790e-02 1.04597770e-01 9.02010024e-01
-2.56537735e-01 5.35690524e-02 -9.74967599e-01 6.29015505e-01
2.41114929e-01 -1.01161313e+00 7.28846937e-02 6.19565770e-02
1.10405815e+00 2.75315225e-01 4.04691622e-02 4.47348565e-01
4.22423005e-01 -7.87438571e-01 5.88960528e-01 5.19041598e-01
8.55046391e-01 -5.88400424e-01 6.85566545e-01 4.47671324e-01
-9.17113841e-01 -2.35761434e-01 -5.77314377e-01 4.02792543e-02
6.47960007e-02 5.69625318e-01 -5.05059779e-01 5.00997715e-02
4.55352724e-01 4.59282219e-01 -6.57647908e-01 9.61778045e-01
-3.49779189e-01 5.86640775e-01 -1.67473312e-02 -4.01518047e-02
5.17147601e-01 1.91717833e-01 4.47096638e-02 9.02766407e-01
2.42300659e-01 1.85015962e-01 4.72694337e-01 8.33931148e-01
-4.67512518e-01 1.51227340e-01 -3.91647190e-01 3.05190742e-01
4.27667499e-01 1.51666164e+00 -8.67087960e-01 -4.55793858e-01
-7.69455492e-01 1.13203919e+00 7.66621351e-01 4.97060478e-01
-1.06424010e+00 -2.83766598e-01 4.80674088e-01 -1.45558249e-02
4.64828372e-01 5.15373647e-01 8.35226178e-02 -1.12998807e+00
-3.20596665e-01 -2.47695461e-01 6.13674700e-01 -8.18251669e-01
-1.83427775e+00 5.18118918e-01 -4.12016183e-01 -9.47503030e-01
2.69951940e-01 -3.29722077e-01 -6.34229362e-01 9.18328524e-01
-1.60443330e+00 -1.24038112e+00 -2.78737813e-01 2.23716930e-01
6.83476150e-01 8.36256444e-02 8.01185966e-01 4.07462984e-01
-6.07546151e-01 5.56915820e-01 2.22064137e-01 4.79079001e-02
9.92851436e-01 -1.26164103e+00 7.45956898e-02 7.00800657e-01
4.73989956e-02 2.82704145e-01 2.73458958e-01 -4.66236681e-01
-5.97056031e-01 -1.49621642e+00 9.30090666e-01 -4.53595608e-01
6.26656294e-01 -4.26695883e-01 -1.12119508e+00 8.48073304e-01
3.63260582e-02 7.24568665e-01 7.55541742e-01 1.65760875e-01
-4.21834141e-01 1.55654550e-01 -1.11692202e+00 1.81130007e-01
9.81521785e-01 -3.63676846e-01 -3.99208784e-01 5.58933139e-01
8.90653789e-01 1.81649365e-02 -5.58383167e-01 3.67771536e-01
2.06276268e-01 -7.10917056e-01 8.27426851e-01 -7.01678634e-01
3.68583918e-01 -3.44002396e-01 -1.11432008e-01 -1.02377665e+00
-8.76532674e-01 -9.31323506e-03 4.53607887e-02 1.46116948e+00
4.46694463e-01 -4.71091777e-01 6.80970609e-01 5.89281321e-01
-8.48519057e-02 -7.89037228e-01 -7.99469233e-01 -5.42442799e-01
-5.94112091e-02 1.09825164e-01 3.43571603e-01 1.16753304e+00
-4.34080452e-01 8.89074743e-01 -5.95801234e-01 2.74424583e-01
6.68397188e-01 2.36700296e-01 3.36125970e-01 -1.34491503e+00
-1.42476752e-01 -2.07713261e-01 -1.45355776e-01 -9.58182037e-01
4.56522763e-01 -1.01438046e+00 3.98453742e-01 -1.49636161e+00
9.22273934e-01 -5.32616436e-01 -1.03801262e+00 8.09279561e-01
-6.40935659e-01 4.91225600e-01 6.11277148e-02 2.23863408e-01
-1.27124989e+00 4.65041578e-01 1.37108767e+00 -3.36322993e-01
2.69434541e-01 -2.01476291e-01 -9.11000013e-01 6.37261271e-01
6.25751078e-01 -8.10295522e-01 -3.80014151e-01 -3.54940325e-01
-1.30677968e-01 -2.35554740e-01 3.26778859e-01 -7.29145885e-01
9.70856175e-02 -3.68715614e-01 3.10178161e-01 -4.35419232e-01
2.07994699e-01 -6.77120984e-01 -2.58434176e-01 1.13986738e-01
-6.85712993e-01 -5.43882906e-01 -1.94630370e-01 8.58532310e-01
1.18897364e-01 -4.83383864e-01 1.05992937e+00 -2.34905601e-01
-9.81124699e-01 6.68324947e-01 9.23219789e-03 3.83796930e-01
1.42060149e+00 1.61439121e-01 -6.11975253e-01 -7.31868148e-02
-7.26162851e-01 5.32970667e-01 4.22027260e-01 3.33354086e-01
3.70413780e-01 -1.40259743e+00 -6.37356341e-01 1.58498868e-01
6.58416092e-01 4.29593921e-02 5.75510681e-01 3.11300248e-01
6.82178810e-02 3.29514474e-01 -7.92895332e-02 -5.37669003e-01
-1.13652956e+00 1.06494689e+00 1.97273195e-01 -1.83689684e-01
-3.55730444e-01 1.19034612e+00 9.58185911e-01 -4.25967962e-01
1.15347154e-01 5.89101948e-02 -2.63484269e-01 6.14107810e-02
6.43678069e-01 3.11494827e-01 -2.27714315e-01 -8.57314587e-01
-4.27592844e-01 6.02176845e-01 -2.29299486e-01 4.14541125e-01
1.06828392e+00 -3.33960265e-01 -6.74715918e-03 7.06248403e-01
1.41901553e+00 -3.40360641e-01 -1.52596414e+00 -3.15662295e-01
-1.60161242e-01 -4.40865487e-01 1.36433452e-01 -9.00717139e-01
-1.09980166e+00 8.67608726e-01 6.15705252e-01 1.93162486e-01
9.94878411e-01 5.44328630e-01 7.20273435e-01 5.30041382e-02
1.89705610e-01 -1.02971649e+00 4.12293226e-01 4.70867574e-01
4.37430918e-01 -1.69849920e+00 -3.86905640e-01 -2.98399746e-01
-6.78565800e-01 6.14052176e-01 9.53190506e-01 -1.34684622e-01
4.38756943e-01 -2.25225568e-01 9.01081562e-02 -3.69975924e-01
-8.99270236e-01 -6.31165266e-01 3.67526859e-01 3.13701212e-01
2.69779593e-01 -4.69554104e-02 -2.55038857e-01 4.49272275e-01
7.78294683e-01 -2.48601794e-01 3.94317180e-01 6.99673712e-01
-8.79583895e-01 -7.66216874e-01 -2.21337065e-01 7.91011155e-01
-5.48976481e-01 1.59744769e-02 -1.79559141e-01 4.81222391e-01
4.90006447e-01 8.51741493e-01 3.62735875e-02 -1.36563182e-01
1.09558761e-01 3.01278234e-01 1.69451386e-01 -1.12967598e+00
-2.28916392e-01 6.11222675e-03 -1.90128669e-01 -2.72646964e-01
-3.73244107e-01 -3.05443555e-01 -1.10762918e+00 1.81720197e-01
-7.22001016e-01 -4.88726981e-02 2.45060250e-01 8.76249373e-01
4.19973493e-01 8.15352798e-01 7.19401598e-01 -6.45938456e-01
-4.86934483e-01 -9.33841944e-01 -6.81361258e-01 7.24949896e-01
3.38169277e-01 -1.02482581e+00 -5.98403335e-01 -4.78747860e-02] | [9.799369812011719, 3.7769837379455566] |
cceea0ff-cfa4-48a7-8b31-73f4e9c73847 | lexical-query-modeling-in-session-search | 1608.06656 | null | http://arxiv.org/abs/1608.06656v1 | http://arxiv.org/pdf/1608.06656v1.pdf | Lexical Query Modeling in Session Search | Lexical query modeling has been the leading paradigm for session search. In
this paper, we analyze TREC session query logs and compare the performance of
different lexical matching approaches for session search. Naive methods based
on term frequency weighing perform on par with specialized session models. In
addition, we investigate the viability of lexical query models in the setting
of session search. We give important insights into the potential and
limitations of lexical query modeling for session search and propose future
directions for the field of session search. | ['Maarten de Rijke', 'Evangelos Kanoulas', 'Christophe Van Gysel'] | 2016-08-23 | null | null | null | null | ['session-search'] | ['natural-language-processing'] | [ 8.06931853e-02 -3.39500695e-01 -8.32756519e-01 -4.71168548e-01
-1.28285360e+00 -6.48194671e-01 9.58195210e-01 5.43878675e-01
-9.36055064e-01 3.05782259e-01 4.96077865e-01 -6.40983045e-01
-5.96593738e-01 -4.24932837e-01 -1.75085768e-01 -4.42001075e-02
-1.31894559e-01 5.38469434e-01 8.05948079e-01 -4.74043369e-01
8.19295466e-01 2.24958822e-01 -1.45834816e+00 5.71991026e-01
3.67695451e-01 7.72253871e-01 2.12087017e-02 7.94613421e-01
-4.27508414e-01 4.38513786e-01 -6.51405275e-01 -2.88916171e-01
-8.46414566e-02 -3.17387164e-01 -1.23947465e+00 -7.18714178e-01
4.55275863e-01 -1.58507586e-01 -6.95650578e-01 4.10888195e-01
6.88804626e-01 4.67607409e-01 5.04547179e-01 -1.02041793e+00
2.22179554e-02 5.41991115e-01 -1.78789511e-01 8.58932793e-01
1.08236969e+00 -4.76997256e-01 1.30966139e+00 -7.33744144e-01
5.61861575e-01 1.04305387e+00 1.85794130e-01 3.70331198e-01
-1.26128399e+00 -5.85654736e-01 1.57354623e-01 5.28243661e-01
-1.51480138e+00 -5.79663217e-01 3.08206052e-01 -1.28439829e-01
1.27223504e+00 7.14804173e-01 3.58661264e-01 9.55962896e-01
5.98927140e-02 8.70611846e-01 7.83388555e-01 -7.99125969e-01
1.52343273e-01 2.86007971e-01 6.18563294e-01 1.00831054e-01
-5.22627458e-02 4.33188230e-01 -9.78554547e-01 -1.07923245e+00
1.66169465e-01 -2.70940751e-01 5.96481822e-02 -2.76086599e-01
-5.27932286e-01 1.05528569e+00 1.84623852e-01 3.66280317e-01
-4.86769468e-01 2.41544589e-01 5.28340816e-01 5.39963365e-01
3.56826305e-01 4.25051242e-01 -1.06409445e-01 -4.15762693e-01
-1.12024593e+00 7.33413517e-01 1.15646207e+00 9.77045119e-01
3.97772342e-01 -6.62950635e-01 -5.48998952e-01 7.46517599e-01
4.71471816e-01 3.27231497e-01 2.10576817e-01 -8.48334432e-01
2.25958705e-01 1.65097997e-01 2.19324693e-01 -3.88712436e-01
-1.48961753e-01 -3.18522274e-01 2.86886841e-01 -6.99816465e-01
-1.76590178e-02 4.65131730e-01 -7.16630220e-01 1.23878384e+00
-4.09593014e-03 -2.04355136e-01 -3.59620810e-01 3.01679790e-01
6.20374322e-01 4.03594881e-01 7.03126729e-01 -5.25703490e-01
1.25602674e+00 -4.96030062e-01 -6.04449272e-01 -7.62284994e-02
8.84313703e-01 -1.13472641e+00 8.35418999e-01 -3.35598588e-01
-1.11880636e+00 -3.49187963e-02 -4.86738622e-01 7.97611028e-02
-2.44923025e-01 -7.16340184e-01 7.83743858e-01 9.97853637e-01
-1.17203653e+00 9.68183950e-02 -8.02222490e-01 -9.50198829e-01
-4.22280401e-01 2.68491775e-01 3.74652565e-01 1.15140840e-01
-1.34390700e+00 8.20635974e-01 2.44143233e-01 -5.27671754e-01
-3.04544002e-01 -6.25281394e-01 -4.46425885e-01 2.52153248e-01
2.95815021e-01 -4.71258402e-01 1.84838104e+00 2.54891425e-01
-7.35598564e-01 8.40619624e-01 -8.53424311e-01 -4.93510604e-01
3.71280983e-02 -8.08718354e-02 -5.50187171e-01 -3.46052945e-02
1.21327423e-01 2.02592030e-01 1.02251060e-01 -7.15628386e-01
-8.95400405e-01 -2.57166535e-01 2.01371327e-01 5.18225431e-01
-1.90606058e-01 7.86993206e-01 -8.24035168e-01 -3.08693886e-01
1.04414217e-01 -9.05862749e-01 -1.12234235e-01 -4.78500098e-01
1.99418813e-01 -8.99478376e-01 3.85270476e-01 -2.93443710e-01
2.17692900e+00 -2.20642376e+00 -4.25015479e-01 7.45474219e-01
-1.94528207e-01 -1.16551630e-01 -3.62556018e-02 1.28179622e+00
3.81606400e-01 4.60430384e-01 6.98212743e-01 -1.33822739e-01
3.04611862e-01 -1.41182795e-01 -5.71767390e-01 -1.85103118e-02
-8.68106306e-01 1.07523894e+00 -3.93225700e-01 -8.38676751e-01
-8.59137550e-02 -2.52584349e-02 -3.33470702e-01 -1.01456739e-01
-2.21871480e-01 -3.52732986e-01 -7.49049127e-01 6.17637157e-01
8.35445523e-02 -4.31024164e-01 3.44746113e-01 1.91016495e-01
-1.09595671e-01 1.02458119e+00 -5.52636147e-01 1.55213523e+00
-4.24495161e-01 3.94040823e-01 -4.15007435e-02 -3.59145194e-01
1.81923226e-01 5.27855158e-01 6.46152020e-01 -1.47309375e+00
-1.71279594e-01 4.17073935e-01 -3.37048560e-01 -2.94891745e-01
8.42677057e-01 1.37687698e-01 -4.08767313e-01 9.49988902e-01
-4.24403220e-01 1.43728405e-01 3.70220482e-01 6.57529116e-01
1.04687488e+00 -4.83687252e-01 2.45684266e-01 -2.58532703e-01
3.52152854e-01 2.81674922e-01 -1.89537629e-01 1.40398824e+00
-2.36582965e-01 -1.24250412e-01 -2.00689770e-02 -1.76656082e-01
-6.60287440e-01 -9.82256651e-01 -2.75592774e-01 1.92159593e+00
4.95167941e-01 -7.03321159e-01 -2.60730952e-01 -4.87763315e-01
1.49952292e-01 6.02434099e-01 -4.40179199e-01 -1.82326570e-01
-5.52595854e-01 -4.08372611e-01 7.43022680e-01 5.51415145e-01
-7.58059276e-03 -9.85349000e-01 -2.23305047e-01 2.79779106e-01
-3.98331076e-01 -9.65253890e-01 -8.10463548e-01 2.16870904e-01
-8.33809376e-01 -9.14007306e-01 -5.28786600e-01 -7.39072502e-01
-2.54981399e-01 6.42992854e-01 1.18636620e+00 3.02423179e-01
-3.17938060e-01 1.10957789e+00 -4.10049677e-01 -2.41165906e-01
-9.27569717e-02 6.37889087e-01 -5.45108803e-02 -6.58850670e-01
1.02540851e+00 -1.45025760e-01 -9.59519684e-01 6.55356884e-01
-8.76984596e-01 -8.11454475e-01 2.90855914e-01 4.19811398e-01
1.30656779e-01 -2.68575370e-01 4.04502660e-01 -9.87565041e-01
1.30532312e+00 -5.66142857e-01 -3.71464759e-01 7.90985525e-01
-1.39256501e+00 7.34084025e-02 -6.91664815e-01 -2.74277806e-01
-1.13155591e+00 -4.85801458e-01 -3.57871354e-02 4.25242066e-01
3.28648269e-01 8.24634969e-01 5.79012632e-01 -4.96940076e-01
1.07753742e+00 3.01266849e-01 -2.29872897e-01 -6.56419873e-01
5.60667329e-02 5.39056420e-01 -2.09492296e-01 -5.48139393e-01
5.25399685e-01 1.89813808e-01 -2.93356329e-01 -7.70838678e-01
-4.74248499e-01 -1.57053578e+00 1.65228009e-01 -3.17324437e-02
4.97772962e-01 -7.43684709e-01 -8.33766878e-01 6.59814253e-02
-7.25004673e-01 -2.37938553e-01 -3.05142403e-02 4.67186630e-01
-4.54939157e-01 4.82316613e-01 -8.82386863e-01 -9.23018336e-01
-2.52164245e-01 -5.16699672e-01 1.02060616e+00 2.74852693e-01
-6.03017986e-01 -1.21556127e+00 5.64535201e-01 5.00496864e-01
7.54698932e-01 -7.96198964e-01 1.02794313e+00 -9.71444488e-01
-6.65556014e-01 -7.20308423e-01 -1.88297078e-01 -6.89562261e-01
-1.86080500e-01 -4.81102258e-01 -6.15736961e-01 -3.71204466e-01
-3.24307919e-01 -2.38220766e-01 7.85716712e-01 2.38833323e-01
7.57850468e-01 1.73463866e-01 -7.36983478e-01 2.66140215e-02
1.37922668e+00 4.57470328e-01 4.57745373e-01 6.47140920e-01
-2.71028072e-01 5.79870820e-01 6.90011382e-01 3.42575222e-01
3.33659917e-01 1.16069365e+00 -3.08971882e-01 5.32451630e-01
2.54065216e-01 -4.46858227e-01 -4.75945659e-02 2.52718985e-01
1.45513132e-01 -6.96105182e-01 -8.75341475e-01 3.98117304e-01
-1.79088712e+00 -9.60936487e-01 1.86328739e-01 2.18607473e+00
6.23090088e-01 3.68660897e-01 4.49104160e-01 -1.97975352e-01
3.43327731e-01 2.60474563e-01 -1.25366464e-01 -6.08566761e-01
8.89841467e-02 6.64022923e-01 7.58410871e-01 6.90902710e-01
-7.61891127e-01 7.53334343e-01 8.05380821e+00 9.08649027e-01
-6.43880308e-01 9.28202868e-02 -2.99916625e-01 -2.86337227e-01
-5.43249607e-01 3.43166500e-01 -1.16050398e+00 2.96805829e-01
1.23166549e+00 -9.01846766e-01 2.66447484e-01 6.91336036e-01
2.51652822e-02 -3.18875074e-01 -1.08499324e+00 8.30429256e-01
4.97999489e-02 -1.10245502e+00 -1.01516448e-01 4.58277792e-01
3.00553232e-01 1.08542211e-01 1.18175805e-01 8.66611063e-01
-5.56358024e-02 -5.08153379e-01 9.85425413e-02 2.67582864e-01
4.67220426e-01 -1.70848995e-01 5.45199335e-01 3.58118981e-01
-1.17104983e+00 4.76153521e-03 6.05999678e-02 2.25961030e-01
3.61604422e-01 -2.53518492e-01 -8.41387391e-01 4.87186432e-01
5.69915056e-01 -9.00056437e-02 -7.39974678e-01 1.61120665e+00
5.24278462e-01 6.41403258e-01 -6.97711706e-01 -3.95684004e-01
2.42597312e-01 4.96083796e-01 2.40892410e-01 1.20666802e+00
7.50555396e-02 4.35736440e-02 3.95787209e-01 -1.66291818e-02
2.45281965e-01 4.84283566e-01 -4.51830566e-01 -1.51198328e-01
8.65546107e-01 3.72837365e-01 -5.04378200e-01 -2.75084674e-01
-5.81570268e-01 7.83524513e-01 -1.86392203e-01 6.77889466e-01
-3.35073441e-01 -2.71989554e-01 1.53103977e-01 4.51810420e-01
-1.16410993e-01 -1.13979459e-01 -8.89687166e-02 -8.79848957e-01
2.79673040e-01 -6.75228298e-01 1.04520583e+00 -1.52648598e-01
-9.74040806e-01 3.02142352e-01 6.54208839e-01 -8.18040252e-01
-8.65738928e-01 -7.42206872e-02 -5.90990543e-01 9.43219006e-01
-1.14547563e+00 -5.92756927e-01 -5.29704913e-02 5.67646861e-01
7.46184349e-01 9.24779028e-02 1.00553191e+00 5.34494281e-01
2.79991806e-01 9.03402030e-01 1.23528965e-01 -4.09443915e-01
8.60408843e-01 -8.79750848e-01 5.78248024e-01 2.10708022e-01
3.17934930e-01 1.20430243e+00 6.74533427e-01 -6.26682520e-01
-1.09265304e+00 -1.49886593e-01 1.55712891e+00 -6.20968699e-01
6.64383233e-01 -2.26585150e-01 -8.63928199e-01 5.76192617e-01
1.61994711e-01 -8.48170638e-01 1.04154956e+00 9.34785366e-01
-4.24832642e-01 8.62663314e-02 -6.17371023e-01 4.31366712e-01
9.58930016e-01 -1.43300462e+00 -4.51815397e-01 4.49644238e-01
6.27421916e-01 1.75796617e-02 -5.58436751e-01 5.22487044e-01
1.07913375e+00 -5.90116441e-01 1.15682876e+00 -5.82884014e-01
-6.95818603e-01 3.93481582e-01 -3.00382227e-01 -4.88117844e-01
-2.17292741e-01 -5.62854052e-01 -1.40451059e-01 6.45471394e-01
7.27551997e-01 -3.72135699e-01 1.23952115e+00 1.19038641e+00
4.36047465e-01 -4.50682044e-01 -8.55691314e-01 -9.50417519e-01
5.94005585e-02 -4.41443145e-01 2.21596196e-01 3.35731715e-01
3.84847611e-01 4.42293376e-01 6.42262306e-03 -3.42586130e-01
5.89223504e-01 7.12057874e-02 3.00759196e-01 -1.26166081e+00
-2.32653767e-01 -5.41989565e-01 -1.36722133e-01 -1.41360784e+00
2.40474287e-02 -5.20877898e-01 -1.79647192e-01 -1.48315108e+00
4.01421219e-01 -4.39609498e-01 -5.97798944e-01 -1.87808543e-01
-3.98504138e-02 2.48534586e-02 -1.27069518e-01 5.41250169e-01
-9.28859174e-01 7.45134279e-02 4.32470828e-01 -9.60857794e-02
-3.67238879e-01 6.92367733e-01 -4.88316983e-01 -7.98217207e-02
6.55800045e-01 -3.75491470e-01 -7.64589965e-01 -2.76061863e-01
6.86515987e-01 5.28927445e-01 5.93322739e-02 -4.21085209e-01
1.12621868e+00 -2.28341505e-01 -2.56007671e-01 -8.77791882e-01
3.50904197e-01 -6.64762974e-01 -1.06707320e-01 4.29367363e-01
-1.03769302e+00 2.91703939e-01 4.17062305e-02 8.91865790e-01
-4.89766419e-01 -3.42228234e-01 1.52906463e-01 -8.47527236e-02
-8.33321571e-01 1.55859515e-01 -7.70416975e-01 1.14272892e-01
5.88765025e-01 -2.95304835e-01 -1.20031968e-01 -8.48692119e-01
-9.75534499e-01 5.88084757e-01 2.34755963e-01 6.59257054e-01
3.62360775e-01 -1.08809745e+00 -1.74127445e-01 -4.81633320e-02
7.80308008e-01 -1.02704632e+00 1.10098295e-01 5.72896063e-01
-2.23258078e-01 1.40985775e+00 4.72427607e-01 -2.94306040e-01
-1.58161354e+00 3.96386951e-01 -1.04618676e-01 -3.85372788e-01
1.40336543e-01 7.94245958e-01 7.86452964e-02 -3.40199806e-02
8.02637517e-01 4.92918372e-01 -1.12735406e-01 4.14567478e-02
4.91295636e-01 3.28792453e-01 3.21624994e-01 -4.78118053e-03
-6.68136537e-01 2.74905801e-01 -6.77855194e-01 -8.34136248e-01
4.95055497e-01 -5.63911140e-01 1.46158701e-02 4.25949872e-01
1.28452146e+00 -9.67627317e-02 1.47688985e-01 -9.23191845e-01
9.08422530e-01 -5.03033876e-01 5.15909679e-02 -7.54236698e-01
-3.93104888e-02 2.35244125e-01 7.98110366e-01 5.77176392e-01
9.60133433e-01 3.37449163e-01 7.01367140e-01 7.04697728e-01
7.82768846e-01 -1.17879856e+00 -2.00017482e-01 4.71774131e-01
2.65726894e-01 -1.16942894e+00 9.37487036e-02 -3.28741819e-01
-1.47521347e-01 6.44518316e-01 1.98637113e-01 3.78088593e-01
1.13724542e+00 -3.24581504e-01 1.56109765e-01 -7.35157430e-01
-1.01253045e+00 -3.62746030e-01 5.71369886e-01 -4.57832962e-02
7.32821107e-01 -1.30746335e-01 -1.03736031e+00 -2.03042716e-01
-5.94128482e-02 -1.30885961e-02 -1.80509314e-01 1.53384459e+00
-6.03753209e-01 -1.66758168e+00 1.78328916e-01 4.88088876e-01
-8.26291978e-01 -5.07256210e-01 -5.95272064e-01 6.22044444e-01
-8.64695728e-01 1.22202337e+00 -1.63958576e-02 -3.98712516e-01
4.02683169e-01 5.82447469e-01 4.20015931e-01 -7.77166963e-01
-8.03597987e-01 2.61832893e-01 3.94429296e-01 -7.48477459e-01
-3.96700948e-01 -7.44603395e-01 -5.28859377e-01 -5.03335238e-01
-6.41026676e-01 1.04984725e+00 9.03468251e-01 6.41011894e-01
4.42004502e-01 -2.43496135e-01 6.27109349e-01 2.77828336e-01
-8.76971781e-01 -1.00917888e+00 -4.62282300e-01 2.08572194e-01
1.26719670e-02 -4.96633679e-01 -6.06048584e-01 -2.62795866e-01] | [11.743270874023438, 7.676333904266357] |
e58f8103-be4d-4ed0-b22d-91788a7f0ba0 | face-anti-spoofing-from-the-perspective-of | 2208.13164 | null | https://arxiv.org/abs/2208.13164v1 | https://arxiv.org/pdf/2208.13164v1.pdf | Face Anti-Spoofing from the Perspective of Data Sampling | Without deploying face anti-spoofing countermeasures, face recognition systems can be spoofed by presenting a printed photo, a video, or a silicon mask of a genuine user. Thus, face presentation attack detection (PAD) plays a vital role in providing secure facial access to digital devices. Most existing video-based PAD countermeasures lack the ability to cope with long-range temporal variations in videos. Moreover, the key-frame sampling prior to the feature extraction step has not been widely studied in the face anti-spoofing domain. To mitigate these issues, this paper provides a data sampling approach by proposing a video processing scheme that models the long-range temporal variations based on Gaussian Weighting Function. Specifically, the proposed scheme encodes the consecutive t frames of video sequences into a single RGB image based on a Gaussian-weighted summation of the t frames. Using simply the data sampling scheme alone, we demonstrate that state-of-the-art performance can be achieved without any bells and whistles in both intra-database and inter-database testing scenarios for the three public benchmark datasets; namely, Replay-Attack, MSU-MFSD, and CASIA-FASD. In particular, the proposed scheme provides a much lower error (from 15.2% to 6.7% on CASIA-FASD and 5.9% to 4.9% on Replay-Attack) compared to baselines in cross-database scenarios. | ['Mourad Oussalah', 'Usman Muhammad'] | 2022-08-28 | null | null | null | null | ['face-presentation-attack-detection', 'face-anti-spoofing'] | ['computer-vision', 'computer-vision'] | [ 5.79930067e-01 -4.69942868e-01 -2.70088501e-02 -1.95899591e-01
-5.54646015e-01 -7.17893660e-01 5.27826428e-01 -1.71269938e-01
-2.43747041e-01 3.79772961e-01 -3.87664884e-01 -3.70806038e-01
1.25454599e-02 -6.57675028e-01 -7.30034173e-01 -9.60444868e-01
-1.45295471e-01 -4.15671736e-01 2.47362450e-01 -2.21755821e-02
1.73503488e-01 7.57251978e-01 -1.50341487e+00 2.41662741e-01
4.14222896e-01 1.35995114e+00 -3.68680656e-01 5.49940288e-01
1.83859602e-01 1.22407682e-01 -9.86849189e-01 -7.63212264e-01
2.89514542e-01 -2.82451630e-01 -9.58026424e-02 6.95105344e-02
4.88765210e-01 -5.71138740e-01 -6.68273330e-01 1.22444797e+00
5.99138081e-01 -3.67054552e-01 2.83042967e-01 -1.58108664e+00
-3.35782290e-01 -5.99138290e-02 -1.00065017e+00 8.36153254e-02
5.49431086e-01 2.05700144e-01 -1.04465595e-04 -8.82598698e-01
3.93151045e-01 1.49663925e+00 5.16358852e-01 6.87679589e-01
-9.67279196e-01 -1.23355043e+00 -5.15693009e-01 4.44085002e-01
-1.58911026e+00 -8.37210834e-01 8.91461551e-01 -7.60372505e-02
4.73470777e-01 2.53197789e-01 3.71103078e-01 1.10780597e+00
3.74155045e-01 2.89432585e-01 1.19073403e+00 -3.45114410e-01
-1.91766620e-01 1.45832807e-01 -6.31929860e-02 6.86277151e-01
6.48926139e-01 2.40298763e-01 -7.08198428e-01 -4.82713223e-01
5.35921335e-01 -9.19223726e-02 -4.15509999e-01 1.07232600e-01
-6.24394774e-01 5.71617007e-01 -1.01517260e-01 6.21919148e-03
-1.28893033e-01 4.14094031e-02 3.99586707e-01 2.82265842e-01
1.02792546e-01 -3.24157000e-01 -2.03537829e-02 9.25139189e-02
-1.08449066e+00 1.57961935e-01 5.55859447e-01 6.39417648e-01
4.19333845e-01 1.18262984e-01 4.88893948e-02 4.86344457e-01
5.26915669e-01 9.19687450e-01 2.25019842e-01 -5.27244687e-01
4.62058395e-01 -4.25024144e-02 5.70410192e-02 -1.53245568e+00
2.13597082e-02 1.38621414e-02 -6.57073021e-01 1.05431892e-01
4.96454298e-01 -2.04309180e-01 -7.41423309e-01 1.69506562e+00
5.61533570e-01 7.09466934e-01 1.18686602e-01 4.49417025e-01
5.43432832e-01 5.64383149e-01 -1.99885800e-01 -7.45526612e-01
1.68083072e+00 -4.28235322e-01 -1.09480059e+00 6.82984591e-02
-2.74584908e-03 -1.26367521e+00 5.82973897e-01 6.29490256e-01
-8.74122739e-01 -5.55600345e-01 -1.41995919e+00 3.45123500e-01
-1.32395908e-01 1.30373746e-01 1.68957278e-01 1.62882423e+00
-9.48730052e-01 2.96553791e-01 -6.74691379e-01 -1.09705567e-01
7.42877245e-01 5.04040420e-01 -7.67494082e-01 -3.29789847e-01
-1.08360040e+00 2.83393979e-01 3.19454558e-02 7.41525590e-02
-8.75061452e-01 -5.52624702e-01 -6.53470218e-01 -1.56241596e-01
3.49382311e-01 8.69942922e-03 7.91129112e-01 -5.82407475e-01
-1.54979193e+00 7.35686302e-01 -3.77613723e-01 -4.34730738e-01
4.36255544e-01 -2.24215746e-01 -1.10991681e+00 6.32506788e-01
-3.52749437e-01 2.00122699e-01 1.65588093e+00 -9.79746759e-01
-8.87117237e-02 -7.35596001e-01 -2.76597679e-01 -4.79460955e-01
-5.17727494e-01 4.70511138e-01 -5.04817426e-01 -7.64605939e-01
2.77270330e-04 -1.00622654e+00 5.08622289e-01 2.05108583e-01
-3.14750314e-01 6.63528219e-02 1.72989118e+00 -7.76193976e-01
1.22876489e+00 -2.48106456e+00 -5.37626207e-01 1.75366446e-01
-2.28873044e-01 9.52439904e-01 -7.93439001e-02 3.07794780e-01
-1.35444328e-01 1.51124984e-01 -3.45299803e-02 -2.33198881e-01
-2.99060941e-01 -1.92222998e-01 -3.00919771e-01 1.05300891e+00
2.76108325e-01 2.41238624e-01 -6.75239384e-01 -3.40424359e-01
1.83087230e-01 9.65895355e-01 -4.76262331e-01 5.06270006e-02
3.53614390e-01 2.98407763e-01 -1.18635990e-01 8.19741845e-01
1.62674022e+00 2.70511329e-01 2.06609294e-01 -3.97676021e-01
3.03287029e-01 8.72610584e-02 -1.25754285e+00 1.23429191e+00
-5.25413863e-02 7.27660239e-01 3.76487017e-01 -6.87803030e-01
9.30797040e-01 5.93529463e-01 4.30550784e-01 -4.71880108e-01
3.68894160e-01 2.44893715e-01 -6.43277094e-02 -6.32274091e-01
1.30085409e-01 8.55080113e-02 1.70874104e-01 4.31764215e-01
1.65842623e-01 3.03013623e-01 -1.27459988e-01 -6.89046606e-02
1.07651973e+00 -2.08723500e-01 2.92972848e-02 -1.74349353e-01
8.49613309e-01 -8.07885051e-01 5.48143983e-01 4.02403235e-01
-5.63190639e-01 3.23855013e-01 4.63623762e-01 -3.67359892e-02
-6.17828190e-01 -8.26181173e-01 -3.00035179e-01 4.01625961e-01
2.99950272e-01 -7.34400988e-01 -1.16012132e+00 -8.01544905e-01
1.93197921e-01 2.41570964e-01 -3.30370665e-01 -2.65197664e-01
-5.22910357e-01 -6.13878429e-01 1.21624255e+00 -1.51577428e-01
8.66931617e-01 -5.26555955e-01 -5.96651495e-01 -7.45487586e-02
-8.74766633e-02 -1.56671214e+00 -4.71495330e-01 -6.08547747e-01
-5.12411475e-01 -1.41698933e+00 -4.61383730e-01 -5.57246208e-01
6.06777489e-01 7.29334235e-01 3.16869199e-01 1.80138841e-01
-4.98147547e-01 2.51442760e-01 -1.44302830e-01 -3.06440830e-01
-3.40593874e-01 -7.98945069e-01 3.96294683e-01 7.74139106e-01
5.59484959e-01 -3.19838196e-01 -7.04842269e-01 5.66899896e-01
-1.01888418e+00 -5.00895917e-01 1.36464536e-01 7.33566463e-01
9.84542444e-02 4.64391440e-01 3.07817936e-01 -4.69629586e-01
3.45235050e-01 -3.42980415e-01 -6.12051368e-01 2.67516941e-01
-3.21031272e-01 -2.92002857e-01 2.49231085e-01 -6.04084849e-01
-9.51436818e-01 -1.63339287e-01 -1.50270179e-01 -5.49352050e-01
-2.03637645e-01 -4.13423926e-02 -5.74793935e-01 -6.69664621e-01
3.72604430e-01 2.87027121e-01 4.37049299e-01 -3.57227027e-01
-2.61346418e-02 1.05388761e+00 5.58008790e-01 -3.29595059e-01
1.08343256e+00 7.73704469e-01 2.54082501e-01 -1.32440364e+00
2.07177967e-01 -2.51197517e-01 -2.39737090e-02 -2.71995127e-01
5.59254527e-01 -8.80478323e-01 -1.17533326e+00 1.21113157e+00
-1.25254655e+00 3.76107246e-01 6.47658169e-01 4.56605405e-01
-1.71425775e-01 9.94825721e-01 -6.61629081e-01 -1.07193959e+00
-3.31683010e-01 -1.49751294e+00 1.03869057e+00 3.34505737e-01
2.65634358e-01 -3.58153701e-01 -4.99310166e-01 4.17199433e-01
4.95777428e-01 3.87226105e-01 4.40966427e-01 -3.88370544e-01
-3.12990487e-01 -5.76950252e-01 -2.89215922e-01 5.10136604e-01
4.26631629e-01 2.36320749e-01 -1.34333670e+00 -6.90754890e-01
4.83520240e-01 -1.53464779e-01 5.69928646e-01 1.31196171e-01
1.21987689e+00 -4.69245136e-01 -3.52701277e-01 6.63053691e-01
1.34943521e+00 4.89045233e-01 9.24493611e-01 -1.37345970e-01
3.61953974e-01 7.37265944e-01 5.79079092e-01 6.81233406e-01
-1.92077890e-01 9.11807597e-01 4.78362203e-01 4.78328496e-01
-3.14928964e-02 -1.15311697e-01 7.67827332e-01 9.09873322e-02
2.86263913e-01 -5.18067300e-01 -5.45706451e-01 1.02050014e-01
-1.16095006e+00 -1.00082743e+00 -8.01000968e-02 2.48383570e+00
4.85682607e-01 -3.18878214e-03 2.38371156e-02 6.39583409e-01
9.66041148e-01 3.31188112e-01 -2.53068775e-01 -2.01494202e-01
-1.44368112e-01 3.36883426e-01 6.43923879e-01 3.62430423e-01
-1.29379189e+00 7.11233735e-01 4.97915316e+00 1.09663868e+00
-1.44367993e+00 2.01070830e-01 4.97636139e-01 1.02943610e-02
1.28884733e-01 -2.65302360e-01 -1.00699723e+00 7.85549462e-01
1.12276804e+00 1.44280732e-01 3.55237603e-01 3.56079310e-01
1.87776491e-01 -1.59271464e-01 -6.39226735e-01 1.55085695e+00
4.83644754e-01 -1.09746516e+00 -1.72634691e-01 3.83333951e-01
2.42824033e-01 -6.77943766e-01 4.45259720e-01 -3.83078605e-01
-4.64511961e-01 -9.14634705e-01 5.01835823e-01 -3.41883823e-02
1.07349098e+00 -8.30522299e-01 5.76997101e-01 -2.64596865e-02
-1.19722784e+00 -4.38857451e-02 -3.01122814e-01 4.50544924e-01
2.13416710e-01 5.53475916e-01 -7.46400833e-01 6.36189520e-01
6.30761325e-01 2.51520962e-01 -2.42959827e-01 7.46618748e-01
-1.21785268e-01 7.80336916e-01 -5.59557736e-01 2.48556241e-01
-1.01434372e-01 1.79001257e-01 7.67747283e-01 1.13230181e+00
7.53378451e-01 2.65107695e-02 -3.23379725e-01 2.69231498e-01
-7.90665150e-02 -9.12067816e-02 -6.73923194e-01 -1.60926044e-01
6.44897461e-01 1.02332270e+00 -5.10332942e-01 1.64862406e-02
-4.23488140e-01 1.10912228e+00 -4.83957440e-01 3.24257225e-01
-1.01592267e+00 -5.86738050e-01 9.90245938e-01 2.27309465e-01
4.83707845e-01 -2.64823794e-01 1.88670680e-01 -9.69544888e-01
2.14063212e-01 -1.20941269e+00 2.75417328e-01 -2.66406149e-01
-8.66180539e-01 5.55311382e-01 -1.23274855e-01 -1.11131585e+00
-8.03274438e-02 -6.72042370e-01 -3.87346476e-01 7.67506421e-01
-1.56520915e+00 -9.31020379e-01 -2.26208091e-01 9.75343227e-01
2.09730193e-01 -4.42013353e-01 7.37203956e-01 6.37284100e-01
-6.79390907e-01 1.16388476e+00 -1.59926176e-01 2.88587838e-01
8.75354469e-01 -2.84048736e-01 4.38453764e-01 1.12555587e+00
1.92404203e-02 7.94984639e-01 7.33509898e-01 -6.23045504e-01
-1.86044085e+00 -7.14982033e-01 6.87148213e-01 1.24590307e-01
3.01481068e-01 -4.01335776e-01 -8.38619530e-01 1.52343184e-01
2.77074456e-01 5.54684520e-01 8.51607621e-01 -8.43287826e-01
-7.84096479e-01 -3.91161948e-01 -1.84232700e+00 2.32843116e-01
6.41168356e-01 -8.54873061e-01 2.36923154e-02 7.30706006e-02
3.43652368e-01 -2.59122282e-01 -5.70218265e-01 5.02603054e-01
9.77025211e-01 -1.12088811e+00 1.09098864e+00 -3.03654790e-01
-9.12954286e-02 -4.02561843e-01 -4.61088955e-01 -5.95692992e-01
3.56156409e-01 -1.32188439e+00 -3.18828791e-01 1.56123328e+00
-2.16370761e-01 -6.62894189e-01 8.26371074e-01 2.07078308e-01
4.85609323e-01 -3.59632313e-01 -1.17391455e+00 -8.62326682e-01
-5.53087592e-01 -5.37613869e-01 7.49979675e-01 7.18916416e-01
-1.63127825e-01 -5.34597039e-01 -6.38443708e-01 6.79895341e-01
1.18256342e+00 -5.35319686e-01 9.07848239e-01 -8.53648186e-01
-2.28760183e-01 -5.71843870e-02 -8.09817672e-01 -8.20160329e-01
1.19946599e-01 -3.10952008e-01 -2.63874948e-01 -3.86404306e-01
-1.55271530e-01 -1.41955137e-01 -2.70340234e-01 3.49650979e-02
2.65032426e-03 6.73789382e-01 2.98456460e-01 -6.96515590e-02
1.85252249e-01 2.13055119e-01 8.78438354e-01 -8.57966170e-02
1.35637313e-01 2.49103103e-02 -1.85677871e-01 4.81322676e-01
7.25753248e-01 -5.56628168e-01 -4.56359774e-01 -2.17267528e-01
-4.48535621e-01 4.15043354e-01 4.97104496e-01 -1.29823303e+00
2.20289797e-01 2.66252942e-02 3.14214796e-01 -4.45321381e-01
5.92360139e-01 -9.91016507e-01 2.88199037e-01 7.27122664e-01
1.68292522e-01 9.38446820e-02 5.40995777e-01 7.12286294e-01
-1.67970970e-01 -4.94775511e-02 9.65823591e-01 3.27927262e-01
-2.60499090e-01 3.07456285e-01 -2.58790106e-01 -3.97226334e-01
1.22815907e+00 -4.16933835e-01 -7.03046441e-01 -2.74068534e-01
-2.60539919e-01 -4.50630814e-01 3.80858958e-01 3.68557841e-01
9.23753262e-01 -1.18755150e+00 -5.90028584e-01 1.02901280e+00
-1.13093637e-01 -6.91842854e-01 5.12809575e-01 6.12926781e-01
-6.22175217e-01 3.83513212e-01 -3.61439019e-01 -5.53642452e-01
-1.88476872e+00 6.36402726e-01 1.58881307e-01 1.02112494e-01
-7.47452080e-02 1.00637138e+00 -1.44567445e-01 3.21518630e-01
2.54237652e-01 2.08179310e-01 9.22192782e-02 1.18960328e-01
1.10297740e+00 4.99887198e-01 3.08749944e-01 -1.06661987e+00
-5.75726748e-01 6.31091118e-01 -1.34333326e-02 -1.98326156e-01
8.66960466e-01 -1.20922036e-01 -9.44064781e-02 -3.27603728e-01
1.58486974e+00 3.67408901e-01 -1.20908856e+00 -1.08016580e-01
-1.97243825e-01 -1.06165087e+00 2.99772564e-02 -3.29568535e-01
-1.19115031e+00 9.89414811e-01 1.00876760e+00 1.50642499e-01
1.25697434e+00 -5.06163955e-01 1.01643574e+00 -1.27541170e-01
5.47537088e-01 -5.78723311e-01 1.93789482e-01 -1.22292496e-01
5.59456229e-01 -9.95433807e-01 -5.01748845e-02 -8.02798092e-01
-1.50528550e-01 1.19553769e+00 1.86887532e-01 4.52537611e-02
1.06716800e+00 3.91390622e-01 -6.32226244e-02 7.88845122e-02
-4.04661238e-01 4.52961892e-01 7.83875808e-02 7.56053150e-01
7.01851100e-02 -8.15880969e-02 -2.08080396e-01 2.69411385e-01
8.48110318e-02 1.03432842e-01 3.58891457e-01 1.08163023e+00
-8.60589817e-02 -1.40274763e+00 -7.64199495e-01 -2.20808219e-02
-1.07501614e+00 2.56732017e-01 -4.58775349e-02 4.85802174e-01
1.83166429e-01 1.34125185e+00 -9.41828564e-02 -7.65712321e-01
1.33420257e-02 1.14819422e-01 6.02500916e-01 1.76545791e-02
-2.56657690e-01 3.38986069e-01 -1.91996425e-01 -7.58144200e-01
-5.43626726e-01 -7.04275966e-01 -7.83810496e-01 -8.54314208e-01
-3.62179220e-01 -4.27526459e-02 8.60551119e-01 6.71277225e-01
4.58974361e-01 -3.18869986e-02 1.04394054e+00 -7.20893741e-01
-5.12425900e-01 -6.12137914e-01 -6.74409926e-01 2.90645570e-01
7.70567119e-01 -7.85576522e-01 -4.71060842e-01 -9.20833573e-02] | [13.038737297058105, 1.1489773988723755] |
4fb252c8-2f1d-4449-8a2c-e402195031d2 | k-strip-a-novel-segmentation-algorithm-in-k | 2205.09706 | null | https://arxiv.org/abs/2205.09706v2 | https://arxiv.org/pdf/2205.09706v2.pdf | k-strip: A novel segmentation algorithm in k-space for the application of skull stripping | Objectives: Present a novel deep learning-based skull stripping algorithm for magnetic resonance imaging (MRI) that works directly in the information rich k-space. Materials and Methods: Using two datasets from different institutions with a total of 36,900 MRI slices, we trained a deep learning-based model to work directly with the complex raw k-space data. Skull stripping performed by HD-BET (Brain Extraction Tool) in the image domain were used as the ground truth. Results: Both datasets were very similar to the ground truth (DICE scores of 92\%-98\% and Hausdorff distances of under 5.5 mm). Results on slices above the eye-region reach DICE scores of up to 99\%, while the accuracy drops in regions around the eyes and below, with partially blurred output. The output of k-strip often smoothed edges at the demarcation to the skull. Binary masks are created with an appropriate threshold. Conclusion: With this proof-of-concept study, we were able to show the feasibility of working in the k-space frequency domain, preserving phase information, with consistent results. Future research should be dedicated to discovering additional ways the k-space can be used for innovative image analysis and further workflows. | ['Jens Kleesiek', 'Jan Egger', 'Kevin Kröninger', 'Felix Nensa', 'Johannes Haubold', 'Kelsey L. Pomykala', 'Florian Mentzel', 'Moritz Rempe'] | 2022-05-19 | null | null | null | null | ['skull-stripping'] | ['medical'] | [ 1.91349909e-01 3.08029115e-01 1.65833399e-01 -3.56370270e-01
-6.64102614e-01 -3.14470679e-01 2.43805081e-01 2.64399022e-01
-7.53941000e-01 6.32176995e-01 3.08128327e-01 -8.08984190e-02
-6.67112887e-01 -5.29568434e-01 -2.63945997e-01 -9.56866920e-01
-8.48160565e-01 4.02066737e-01 4.16803330e-01 2.05129310e-01
2.92478770e-01 6.62814677e-01 -9.60416913e-01 4.22691762e-01
5.19073129e-01 8.14575732e-01 1.70462191e-01 5.59512496e-01
3.00661296e-01 5.55297852e-01 -4.72377658e-01 -3.20237875e-01
5.72390914e-01 -1.92253739e-01 -1.03528929e+00 -3.34456861e-01
5.31491280e-01 -5.43748379e-01 -4.18113440e-01 1.03615701e+00
9.14778948e-01 -1.90594882e-01 6.18704259e-01 -6.73032165e-01
-5.90217590e-01 5.68184793e-01 -5.49874008e-01 6.83838844e-01
-5.86422049e-02 3.32512289e-01 9.14208367e-02 -7.15429425e-01
1.02721715e+00 4.09192473e-01 1.08070683e+00 3.19114417e-01
-1.34959328e+00 -8.09253037e-01 -7.37590253e-01 3.96804631e-01
-1.51357436e+00 -3.41943443e-01 6.69394910e-01 -8.16654921e-01
9.38369691e-01 2.22946733e-01 9.57700968e-01 4.56058979e-01
8.86400759e-01 1.91688016e-01 1.71867585e+00 -3.59021455e-01
9.77418721e-02 3.15358602e-02 1.10480323e-01 6.27620101e-01
2.01348886e-01 1.27327368e-01 -4.81111795e-01 -5.93272634e-02
8.85858715e-01 -2.47470230e-01 -5.24563074e-01 -3.07918400e-01
-1.56712365e+00 7.15133727e-01 5.85848331e-01 8.48498702e-01
-5.93956411e-01 -1.48630232e-01 5.23547232e-01 2.38660112e-01
4.92665857e-01 4.97617722e-01 -2.15503380e-01 1.06746227e-01
-1.75784898e+00 1.26150221e-01 2.13243484e-01 4.62155372e-01
1.32398441e-01 -1.15688533e-01 -9.50333104e-02 5.90226948e-01
-3.94317620e-02 1.05318896e-01 1.00565016e+00 -1.15259409e+00
-2.68831477e-02 4.91040349e-02 -4.10992384e-01 -9.71049666e-01
-1.17832053e+00 -6.25798702e-01 -9.88775253e-01 5.75266182e-01
6.08093143e-01 -2.52641886e-01 -1.07123947e+00 1.32516587e+00
1.83183208e-01 4.64453828e-03 -3.25275362e-01 1.09697425e+00
8.40892732e-01 2.29149815e-02 -1.91173255e-01 -4.66696858e-01
1.41715515e+00 -4.69422936e-01 -1.03957975e+00 2.00864092e-01
8.03168595e-01 -5.93744516e-01 9.69339192e-01 5.70964038e-01
-1.43655658e+00 -2.74701893e-01 -1.16030800e+00 5.55026643e-02
-4.03099507e-01 -1.15596049e-01 5.31953216e-01 7.11391449e-01
-1.68575668e+00 9.69588459e-01 -1.01407361e+00 2.66336761e-02
8.88690948e-01 5.45006096e-01 -9.07609880e-01 2.14393232e-02
-1.06446362e+00 1.31934059e+00 3.66289526e-01 9.58564728e-02
-3.48038554e-01 -1.10831225e+00 -4.52726573e-01 -4.30345148e-01
-2.69712150e-01 -4.41144258e-01 6.75655484e-01 -5.61757267e-01
-1.20640552e+00 1.30809546e+00 4.69189584e-01 -7.18049049e-01
6.78829074e-01 1.51283979e-01 -4.58450496e-01 7.03735828e-01
2.21775435e-02 6.25120044e-01 6.85131907e-01 -1.07158041e+00
-7.65296668e-02 -8.74017477e-01 -4.45571542e-01 -1.98809996e-01
1.97308697e-02 2.16878131e-01 -2.58374382e-02 -8.46458435e-01
5.49705446e-01 -8.09491217e-01 -8.36791694e-02 1.23919986e-01
-1.27876759e-01 5.27035713e-01 3.93710971e-01 -1.58777213e+00
1.23323464e+00 -1.95395839e+00 -2.58247405e-01 3.55358541e-01
8.15708816e-01 2.00599626e-01 2.69614011e-01 -6.32035360e-02
-6.74344718e-01 -8.76552910e-02 -7.48718679e-01 6.01599328e-02
-5.61729431e-01 -4.00246412e-01 4.93809253e-01 1.19088924e+00
-3.32354277e-01 8.82234752e-01 -8.28975558e-01 -5.87306023e-01
8.54810029e-02 4.71590370e-01 -3.96998405e-01 -3.15361053e-01
9.30473804e-01 4.57406640e-01 4.06799287e-01 2.97741115e-01
1.02545762e+00 1.69920743e-01 -4.45524491e-02 -3.85802865e-01
-1.46588698e-01 -1.95747405e-01 -1.02499223e+00 1.98255825e+00
-2.21170262e-01 9.58116353e-01 3.38384300e-01 -1.03220117e+00
6.94262445e-01 3.42461020e-01 9.41766381e-01 -9.56573009e-01
-3.37246135e-02 3.89218837e-01 5.96513033e-01 -6.46946251e-01
-3.96867059e-02 -7.17725813e-01 4.80100334e-01 5.53621888e-01
4.06116605e-01 -4.98467714e-01 -2.18851760e-01 7.34152496e-02
1.03980339e+00 -1.91861346e-01 7.91849494e-02 -7.78118551e-01
1.03301302e-01 -6.08622283e-02 1.35591075e-01 5.66134334e-01
-5.83372533e-01 9.86813188e-01 2.69335091e-01 -6.63939476e-01
-1.04124403e+00 -1.21286690e+00 -9.62150276e-01 2.35885128e-01
-3.89028937e-01 -1.28445894e-01 -1.32209694e+00 -4.15071905e-01
-2.84005165e-01 5.06056726e-01 -1.00606430e+00 -1.76896662e-01
-6.09891057e-01 -9.42700684e-01 6.38179719e-01 2.09832594e-01
4.23609912e-01 -9.98873353e-01 -1.06292665e+00 2.37525061e-01
-1.18032813e-01 -9.92851734e-01 -1.53390408e-01 3.15189987e-01
-1.24140465e+00 -9.67290878e-01 -1.29256630e+00 -6.26309931e-01
5.59529126e-01 -2.33158842e-01 8.21020424e-01 -1.80385917e-01
-7.42231309e-01 9.49306190e-02 -4.53901179e-02 -3.66807044e-01
-1.25997066e-01 -3.05957615e-01 1.98142871e-01 -4.63002890e-01
2.04211891e-01 -8.00148010e-01 -8.97567153e-01 -3.99449505e-02
-1.01059198e+00 1.34295449e-01 4.77687061e-01 6.57741666e-01
5.07074952e-01 2.44387910e-01 3.80617023e-01 -4.72684085e-01
6.42483830e-01 -2.50748694e-01 -2.18200654e-01 -9.90488455e-02
-6.35269403e-01 -7.91040286e-02 2.85056144e-01 -3.20375681e-01
-5.69889903e-01 -2.50379205e-01 -1.20576173e-02 -2.57887095e-01
-1.06945023e-01 2.58794397e-01 4.80732411e-01 -3.41836721e-01
8.55158687e-01 8.38567466e-02 4.52177554e-01 -1.92557901e-01
-3.21175419e-02 5.80682993e-01 8.91173661e-01 7.59119391e-02
3.70445162e-01 1.01695323e+00 1.26178011e-01 -8.19795787e-01
-2.66760319e-01 -2.36165613e-01 -1.24931741e+00 -4.40398574e-01
1.19786966e+00 -4.37076151e-01 -9.39391330e-02 5.89130819e-01
-7.03621268e-01 -4.84348059e-01 -4.41542208e-01 1.05107307e+00
-6.56639576e-01 3.69216025e-01 -6.35452390e-01 -1.69637769e-01
-7.88770854e-01 -1.45801497e+00 7.37932026e-01 2.70799063e-02
-2.49150619e-01 -1.03590214e+00 2.74254918e-01 3.76297593e-01
5.10641038e-01 5.78916192e-01 8.49832177e-01 -6.23156250e-01
2.91677173e-02 -1.19729385e-01 -5.34199215e-02 4.39955711e-01
1.65127680e-01 -4.39484417e-01 -9.96744692e-01 -2.26310998e-01
5.25837004e-01 1.94972917e-01 6.70147061e-01 9.62206066e-01
1.32525897e+00 -1.46478474e-01 -4.97950539e-02 9.32323039e-01
1.23518598e+00 2.76324421e-01 1.02610767e+00 6.80217683e-01
2.42304027e-01 7.83390641e-01 5.48028201e-02 1.02555687e-02
1.19181558e-01 7.12071061e-01 1.72805078e-02 -3.22950125e-01
-6.26514018e-01 4.84114766e-01 -1.21515661e-01 9.42883432e-01
-2.03986555e-01 7.03025281e-01 -1.22629118e+00 7.40445256e-01
-1.04287636e+00 -8.99549067e-01 -4.18175191e-01 2.28876090e+00
1.04126477e+00 2.42145613e-01 1.45458683e-01 4.38356876e-01
4.05014932e-01 -2.40937799e-01 -3.69430602e-01 -3.66526037e-01
-1.30967513e-01 6.62548661e-01 8.28595519e-01 5.69392860e-01
-1.00820553e+00 4.00775611e-01 6.88008022e+00 5.92829883e-01
-1.49949074e+00 3.62531245e-01 5.56645632e-01 -4.02473778e-01
5.44812493e-02 -3.61077160e-01 6.71025068e-02 6.65082753e-01
1.14348686e+00 -1.15068741e-01 4.06724155e-01 2.11391211e-01
4.02359664e-01 -4.31946278e-01 -5.37922204e-01 1.27435803e+00
5.12557849e-02 -1.62765074e+00 -4.69996125e-01 3.27957273e-01
4.63669479e-01 2.54607439e-01 1.43783897e-01 -3.45671743e-01
-3.06314588e-01 -1.30331337e+00 6.85599625e-01 9.00289536e-01
1.23931456e+00 -6.97086155e-01 8.89504373e-01 -3.51940729e-02
-5.22479415e-01 1.33428395e-01 -1.14926293e-01 2.04905421e-01
1.20435081e-01 9.71392453e-01 -1.13387764e+00 5.71092188e-01
1.04071784e+00 2.55140901e-01 -5.70611775e-01 1.32042646e+00
2.80855030e-01 3.66247267e-01 -3.56895804e-01 7.99027383e-01
2.76709050e-01 -2.49756306e-01 3.97550941e-01 1.46273518e+00
2.03378946e-01 2.78807163e-01 -5.69322765e-01 5.93934178e-01
2.85750598e-01 1.06355444e-01 -5.29626369e-01 4.65143234e-01
3.21702696e-02 1.27976000e+00 -1.25206566e+00 -3.19668829e-01
-3.40368718e-01 8.67753446e-01 -1.45562127e-01 1.55528545e-01
-5.16776621e-01 -5.41921914e-01 3.30457985e-01 7.60785639e-01
-1.28578350e-01 -2.29806378e-01 -7.93773293e-01 -7.24587381e-01
1.16454296e-01 -6.19971931e-01 2.64245957e-01 -9.44378257e-01
-9.30819631e-01 5.32510221e-01 3.67096335e-01 -7.99587548e-01
3.60775888e-02 -5.56562483e-01 -3.42793852e-01 1.03051758e+00
-1.17968214e+00 -6.18233204e-01 -7.52520934e-02 6.74097538e-01
-6.85949549e-02 1.09103359e-02 7.86610961e-01 3.65636259e-01
1.20022945e-01 4.49082583e-01 3.67412478e-01 3.71175587e-01
7.30211616e-01 -1.26079869e+00 5.92561960e-02 6.53223217e-01
-9.70448777e-02 8.36858928e-01 5.45610845e-01 -6.23782218e-01
-9.17376399e-01 -6.39984667e-01 6.68920398e-01 -2.36606702e-01
5.50570309e-01 -7.61811212e-02 -1.04678619e+00 4.30090934e-01
2.74288833e-01 2.48404548e-01 1.04481971e+00 -5.81097424e-01
2.08296284e-01 9.22800526e-02 -1.78825271e+00 2.17565075e-01
7.55075514e-01 -5.29167473e-01 -8.45577002e-01 4.00411636e-01
2.92036057e-01 -5.50445557e-01 -1.44087219e+00 3.88552696e-01
7.77889967e-01 -1.25042033e+00 1.15554881e+00 -3.46386462e-01
3.30263138e-01 -9.16099101e-02 2.20334649e-01 -1.51217186e+00
-4.24324065e-01 -3.22947890e-01 1.85242236e-01 4.52440679e-01
6.89641535e-02 -5.53045690e-01 5.23839474e-01 3.75471830e-01
-3.31457645e-01 -8.79506052e-01 -1.46461248e+00 -5.50165236e-01
5.28550565e-01 -5.05045831e-01 4.10163671e-01 1.22247481e+00
1.60145685e-01 -3.73223037e-01 2.69838095e-01 -3.13058831e-02
8.96961570e-01 -4.21709985e-01 -5.81558645e-02 -1.25747943e+00
2.26344958e-01 -6.77006900e-01 -8.27138841e-01 2.96401121e-02
-1.01870097e-01 -1.13644898e+00 -4.90484565e-01 -1.72433150e+00
1.89481601e-01 -2.94926554e-01 -2.20295474e-01 3.40091258e-01
2.64462888e-01 6.46849453e-01 8.95177573e-02 2.20629573e-01
3.41529846e-01 -6.03193454e-02 1.65521884e+00 -2.55947039e-02
-1.07080482e-01 -5.24842978e-01 -4.74509865e-01 9.25461769e-01
8.11576545e-01 -4.59743649e-01 -2.11350456e-01 -3.58093619e-01
-1.73214808e-01 -5.96083626e-02 3.61535251e-01 -1.34902203e+00
2.00815007e-01 3.46814692e-01 7.29030371e-01 -4.40180212e-01
7.96561763e-02 -8.73684347e-01 3.82447302e-01 5.93504190e-01
-1.90000162e-01 1.72100723e-01 4.17394966e-01 -2.87902385e-01
3.10626309e-02 -2.18426049e-01 1.28429461e+00 -4.16139990e-01
-2.93153048e-01 1.83919832e-01 -3.86401147e-01 -4.73185107e-02
1.03604019e+00 -6.14723802e-01 6.68518096e-02 -1.49691865e-01
-1.23809528e+00 -3.62338305e-01 5.21823049e-01 -1.00601614e-01
6.71651602e-01 -1.20979273e+00 -7.75712430e-01 2.95883030e-01
-2.18361378e-01 -2.05021024e-01 5.98070562e-01 1.54210258e+00
-9.73052800e-01 5.10884106e-01 -7.06838667e-01 -6.54934108e-01
-1.10481977e+00 4.45397824e-01 9.12009716e-01 -2.24839091e-01
-1.33596659e+00 7.67437339e-01 -1.67365253e-01 -3.96185368e-01
-2.97121219e-02 -6.47270977e-01 -1.38416275e-01 2.99899459e-01
9.03492868e-01 3.30688953e-01 8.39033842e-01 -7.45168746e-01
-6.18188262e-01 4.98968571e-01 7.03482181e-02 -4.02364224e-01
1.75049651e+00 -4.07575332e-02 -2.22463891e-01 1.42396182e-01
1.45423412e+00 -5.57447709e-02 -1.04918623e+00 7.63784200e-02
-8.93108249e-02 -4.05948639e-01 6.85578823e-01 -1.07353151e+00
-1.36546886e+00 9.52243805e-01 1.43589091e+00 1.19293973e-01
1.10809064e+00 -5.05107194e-02 5.93004405e-01 -3.39734793e-01
3.95777255e-01 -1.27111077e+00 -3.77150953e-01 1.36370761e-02
9.14045334e-01 -6.85052276e-01 1.82689533e-01 1.99283123e-01
-4.37127709e-01 1.25998557e+00 -1.12409033e-01 -3.75370570e-02
9.23840225e-01 6.37858927e-01 1.72164589e-01 -6.81594193e-01
1.45395666e-01 2.14432642e-01 2.15123877e-01 1.04679894e+00
4.93007869e-01 7.69662037e-02 -5.60725033e-01 5.36266446e-01
-7.13435292e-01 3.36481959e-01 6.66019142e-01 8.19765449e-01
-2.48591289e-01 -5.21258354e-01 -7.35326946e-01 7.93973863e-01
-7.78254211e-01 -1.93392858e-01 -2.30091698e-02 8.53202403e-01
3.46297741e-01 5.70965648e-01 1.26006916e-01 -1.51759759e-01
2.26894990e-01 1.50321156e-01 8.86727154e-01 -1.87967405e-01
-6.40590191e-01 -9.03879032e-02 -2.06302777e-01 -6.87225461e-01
-4.35229897e-01 -8.22785318e-01 -1.38218033e+00 -2.51966059e-01
-1.09008439e-01 -1.01892300e-01 8.46386790e-01 8.83563757e-01
1.64738968e-01 6.71850979e-01 1.25807554e-01 -8.41254711e-01
-1.10677052e-02 -9.60112512e-01 -1.05651796e+00 2.91579604e-01
4.15599376e-01 -7.39354670e-01 -2.47773618e-01 2.23406538e-01] | [14.038758277893066, -2.3446056842803955] |
f829f854-377e-4172-a60e-c9134401d85a | searching-for-legal-clauses-by-analogy-few | 1911.03911 | null | https://arxiv.org/abs/1911.03911v2 | https://arxiv.org/pdf/1911.03911v2.pdf | Contract Discovery: Dataset and a Few-Shot Semantic Retrieval Challenge with Competitive Baselines | We propose a new shared task of semantic retrieval from legal texts, in which a so-called contract discovery is to be performed, where legal clauses are extracted from documents, given a few examples of similar clauses from other legal acts. The task differs substantially from conventional NLI and shared tasks on legal information extraction (e.g., one has to identify text span instead of a single document, page, or paragraph). The specification of the proposed task is followed by an evaluation of multiple solutions within the unified framework proposed for this branch of methods. It is shown that state-of-the-art pretrained encoders fail to provide satisfactory results on the task proposed. In contrast, Language Model-based solutions perform better, especially when unsupervised fine-tuning is applied. Besides the ablation studies, we addressed questions regarding detection accuracy for relevant text fragments depending on the number of examples available. In addition to the dataset and reference results, LMs specialized in the legal domain were made publicly available. | ['Filip Graliński', 'Gabriela Pałka', 'Łukasz Szałkiewicz', 'Dawid Wiśniewski', 'Łukasz Borchmann', 'Andrzej Gretkowski', 'Agnieszka Kaliska', 'Izabela Kosmala', 'Dawid Jurkiewicz', 'Karol Kaczmarek'] | 2019-11-10 | null | https://aclanthology.org/2020.findings-emnlp.380 | https://aclanthology.org/2020.findings-emnlp.380.pdf | findings-of-the-association-for-computational | ['semantic-retrieval'] | ['natural-language-processing'] | [ 6.59057617e-01 5.90459466e-01 -6.36667728e-01 -5.21690369e-01
-1.81218719e+00 -7.50433683e-01 9.10732508e-01 7.41358027e-02
-4.23127443e-01 9.66755986e-01 6.08868420e-01 -3.69976193e-01
-7.75884867e-01 -4.30107892e-01 -6.57885432e-01 -3.13718110e-01
2.17956692e-01 8.40336919e-01 1.52819067e-01 -2.36566082e-01
6.16436124e-01 1.46612734e-01 -1.34002948e+00 9.39410865e-01
1.07812071e+00 6.98313773e-01 8.74444768e-02 2.52808183e-01
-5.10737479e-01 1.10781634e+00 -9.29656804e-01 -9.69140828e-01
1.28539518e-01 -2.76479959e-01 -1.29047620e+00 -2.45843887e-01
9.36724842e-01 -6.16721749e-01 -3.09795111e-01 1.18470120e+00
3.27679127e-01 -2.52987772e-01 8.36263299e-01 -8.55615854e-01
-8.50507379e-01 1.01740873e+00 -2.31800362e-01 2.11926609e-01
5.88669538e-01 -4.23876315e-01 1.47969580e+00 -6.05350554e-01
1.55303371e+00 1.24369621e+00 3.81098717e-01 7.86628127e-01
-9.81480420e-01 -4.39884633e-01 -1.66648012e-02 4.01604891e-01
-1.05270791e+00 -5.94765484e-01 6.53982103e-01 -4.84900892e-01
1.46473849e+00 6.70587942e-02 -2.55466849e-01 1.35819769e+00
1.11932963e-01 1.14029479e+00 6.63714349e-01 -6.60985589e-01
2.62993146e-02 4.02144462e-01 3.89054537e-01 3.20850462e-01
8.84639442e-01 -7.49514028e-02 -4.04310912e-01 -3.90124977e-01
6.60930425e-02 -4.50696528e-01 -5.16842753e-02 -3.21657479e-01
-5.05152106e-01 9.95459795e-01 -3.70768577e-01 8.84794712e-01
-4.83937152e-02 4.49011847e-02 8.34536672e-01 3.34875017e-01
5.62446415e-01 4.80337948e-01 -6.98848128e-01 -2.07647130e-01
-1.24883866e+00 9.43407416e-01 1.00140464e+00 1.30197728e+00
6.29753649e-01 -6.44404471e-01 -4.29814219e-01 7.56049097e-01
1.19616844e-01 3.23585957e-01 2.94972181e-01 -1.10368824e+00
1.36667359e+00 6.96858644e-01 1.81176305e-01 -4.23839152e-01
-1.22261278e-01 -2.62751162e-01 -3.51400338e-02 1.27798781e-01
5.35273194e-01 -2.07654193e-01 -8.58994007e-01 1.37716031e+00
-9.69891325e-02 -4.24540341e-01 4.33941483e-01 6.22221231e-01
7.14961469e-01 2.50800371e-01 7.27290288e-02 -1.62617937e-01
1.41164291e+00 -7.39471614e-01 -1.07911062e+00 -3.51541817e-01
9.51153815e-01 -8.32152247e-01 6.95771039e-01 3.19288313e-01
-1.09835398e+00 -1.74045265e-01 -8.61379385e-01 -5.71240187e-01
-6.82571590e-01 3.00038606e-01 6.84658051e-01 5.84181786e-01
-5.24491251e-01 6.37387991e-01 -1.54818788e-01 -4.02182758e-01
6.25379026e-01 -3.06745142e-01 -2.75928169e-01 -1.53520018e-01
-1.45892239e+00 1.05188572e+00 4.73679364e-01 -2.88041025e-01
-4.97783840e-01 -5.47466576e-01 -1.04521084e+00 3.40323150e-01
1.01764905e+00 -5.30250609e-01 1.48392177e+00 -7.37716675e-01
-9.76655424e-01 1.33573210e+00 -2.46617705e-01 -8.46329510e-01
5.77561855e-01 -6.68336689e-01 -5.16757607e-01 3.80484819e-01
9.59067941e-01 1.81893229e-01 7.35930264e-01 -1.02623510e+00
-7.26487100e-01 -3.05813879e-01 4.42610174e-01 -1.88248992e-01
-5.89937009e-02 7.16216743e-01 -4.14485842e-01 -7.72963941e-01
-3.87303442e-01 -5.29529214e-01 -9.73396294e-04 -5.09408712e-01
-4.54559505e-01 -5.86187959e-01 6.32346690e-01 -8.60194087e-01
1.45123291e+00 -1.90597498e+00 4.47248034e-02 1.68276668e-01
-2.45075494e-01 2.27439582e-01 1.42439606e-03 8.17521214e-01
-1.05661869e-01 5.56009233e-01 -4.06721711e-01 -1.73592836e-01
5.23818374e-01 1.90055236e-01 -8.01548779e-01 1.35801241e-01
1.45289958e-01 6.71412528e-01 -6.69516206e-01 -8.25929701e-01
-2.50910014e-01 -4.89273928e-02 -5.54919720e-01 -2.90927082e-01
-4.69891995e-01 -2.92779878e-02 -7.77744889e-01 6.11384451e-01
3.89044642e-01 -2.32284278e-01 3.45844299e-01 7.65864775e-02
2.09406037e-02 1.07505059e+00 -9.80028331e-01 2.08585954e+00
-2.91239083e-01 4.83359158e-01 1.28912777e-01 -1.07577348e+00
5.05568266e-01 7.54789352e-01 4.16539967e-01 -9.63266671e-01
1.67518072e-02 5.71785212e-01 -1.37565941e-01 -9.05790865e-01
4.54422355e-01 3.52640674e-02 -3.39317769e-01 7.10071504e-01
1.85382694e-01 2.08094254e-01 1.12484217e+00 5.93013585e-01
1.30387437e+00 3.98448497e-01 4.91185546e-01 -1.59266695e-01
6.96372211e-01 4.48973477e-01 5.26254654e-01 9.31934953e-01
1.25676498e-01 2.90294766e-01 8.12496483e-01 -2.22216532e-01
-9.23884869e-01 -6.51960015e-01 -6.37707591e-01 8.16344500e-01
-2.86362953e-02 -8.91264796e-01 -8.96025479e-01 -1.16693103e+00
2.22896099e-01 8.57112885e-01 -5.37718415e-01 2.67490447e-01
-8.55442822e-01 -3.45689744e-01 6.79768682e-01 4.95690733e-01
3.26000184e-01 -1.28286290e+00 -7.05200016e-01 3.42722505e-01
-2.77252913e-01 -1.41334975e+00 -3.27687383e-01 -9.77980718e-02
-6.15172505e-01 -1.64533663e+00 -4.76047665e-01 -5.67378283e-01
3.61982226e-01 -3.22150737e-01 9.59231794e-01 6.42561167e-02
-4.02539372e-01 2.72951514e-01 -4.48923558e-01 -3.96885723e-01
-5.49288094e-01 1.94570214e-01 -6.65866375e-01 -2.98491597e-01
7.05217957e-01 -7.34798610e-02 -3.90525162e-03 -4.30424005e-01
-1.05834472e+00 -3.95862877e-01 7.24213719e-01 7.99725175e-01
1.42881274e-01 -4.55998518e-02 6.97113514e-01 -1.61614561e+00
1.09761822e+00 -2.60751456e-01 -5.67265749e-01 6.43792629e-01
-9.69042361e-01 4.68962163e-01 3.86110872e-01 2.01739982e-01
-1.51895976e+00 -2.75569379e-01 -1.08631112e-01 1.06592558e-01
-4.19322520e-01 7.20703661e-01 -5.12279831e-02 4.73165214e-01
6.84861362e-01 -6.62051663e-02 -2.92723984e-01 -7.55923450e-01
4.96485174e-01 8.84166420e-01 4.07226712e-01 -7.06745148e-01
7.05351055e-01 4.80460286e-01 -4.73681331e-01 -2.32981414e-01
-1.06510139e+00 -6.72670424e-01 -7.41929531e-01 7.39959255e-02
9.14501250e-01 -7.25452125e-01 -6.99426234e-02 -1.67560101e-01
-1.50525939e+00 2.64142081e-02 -3.37284088e-01 2.23188922e-01
-7.93680012e-01 5.80989063e-01 -8.48730445e-01 -8.41842175e-01
-4.27586466e-01 -9.63286757e-01 1.43437886e+00 -3.59610140e-01
-5.44376373e-01 -8.49670649e-01 1.42322138e-01 1.07656920e+00
1.30182385e-01 3.46397869e-02 1.27813673e+00 -1.12782025e+00
-5.72211981e-01 -3.66611391e-01 -2.29416996e-01 2.98600495e-01
7.86341876e-02 -2.17376322e-01 -9.31905746e-01 -1.29757553e-01
-5.73405027e-02 -3.76382530e-01 1.13348699e+00 2.41444334e-01
6.99352026e-01 -6.31814718e-01 -6.64943397e-01 4.40919846e-02
1.40483713e+00 5.37583768e-01 6.82941198e-01 7.30606079e-01
-2.56368401e-03 8.87698412e-01 7.99348474e-01 1.52741402e-01
9.48203504e-02 8.21668327e-01 7.59052336e-02 2.11370856e-01
-1.78665161e-01 -9.39679369e-02 2.89778829e-01 -6.97700307e-02
7.43119651e-03 -1.28536433e-01 -9.05131102e-01 7.43173540e-01
-1.96390414e+00 -1.42579865e+00 9.17370319e-02 1.85665190e+00
7.14062512e-01 3.99791956e-01 -2.14304268e-01 6.43447191e-02
5.32206178e-01 1.93568304e-01 -3.07403445e-01 -3.89032453e-01
-1.94705978e-01 4.09914404e-01 6.12571537e-01 6.64550781e-01
-1.26781917e+00 1.12529898e+00 6.44603539e+00 1.03337741e+00
-4.28567439e-01 4.12154347e-01 1.88350752e-01 -1.68400615e-01
-5.82285523e-01 4.28580463e-01 -1.12466598e+00 4.07170564e-01
8.68062079e-01 -3.65168840e-01 2.23334447e-01 9.42296922e-01
-1.24617424e-02 -1.09283946e-01 -1.15176511e+00 5.05610466e-01
3.36339414e-01 -1.67250645e+00 2.05847725e-01 2.27390096e-01
6.39565885e-01 -1.03457034e-01 -3.15504938e-01 4.71056968e-01
3.20217520e-01 -7.72045374e-01 8.17625880e-01 2.51613319e-01
6.70512438e-01 -3.44736099e-01 9.64866459e-01 2.68836856e-01
-7.30273962e-01 -4.09396738e-01 -1.93745255e-01 1.97118163e-01
3.89167517e-01 4.53065425e-01 -5.62118351e-01 1.03993464e+00
6.15378022e-01 8.36266637e-01 -4.57236320e-01 6.74782395e-01
-6.09340549e-01 3.09542626e-01 6.52270392e-02 6.73888996e-02
3.36601257e-01 -1.61384251e-02 6.22617960e-01 1.39133275e+00
1.39979407e-01 -1.90909579e-01 -1.37519971e-01 1.17083323e+00
-1.73547626e-01 2.95178533e-01 -8.81647170e-01 7.97573775e-02
3.28340292e-01 8.61730397e-01 -2.90661365e-01 -8.03583980e-01
-6.43493652e-01 7.92654753e-01 4.18455780e-01 5.95602632e-01
-6.31124020e-01 -6.57301128e-01 4.55340743e-01 1.95498690e-01
4.92390066e-01 2.58438349e-01 -1.21544287e-01 -1.25742388e+00
7.46702731e-01 -8.34818482e-01 8.79442692e-01 -6.61481082e-01
-1.10278118e+00 4.27186519e-01 4.71477598e-01 -1.07120454e+00
-9.26764250e-01 -5.48661053e-01 -3.21567386e-01 6.28807127e-01
-1.73968470e+00 -7.96389937e-01 5.45735717e-01 3.69457692e-01
6.87228262e-01 -3.60667884e-01 7.51014411e-01 7.08403945e-01
-2.81374186e-01 2.95364618e-01 1.83730006e-01 4.29604262e-01
8.73861372e-01 -1.25612867e+00 1.67655900e-01 1.04603922e+00
3.65925372e-01 1.06023502e+00 5.45112610e-01 -9.54365373e-01
-8.05561662e-01 -8.32186103e-01 1.73426986e+00 -4.56273228e-01
7.49531388e-01 -3.80874038e-01 -7.33928502e-01 9.51443195e-01
5.27502954e-01 -6.86147690e-01 5.35760999e-01 2.30237976e-01
-5.30806661e-01 -2.42088605e-02 -1.09322941e+00 3.46275032e-01
1.34089756e+00 -8.50517631e-01 -1.24997485e+00 8.55406523e-01
4.24714893e-01 -3.68028253e-01 -6.53548717e-01 1.54252231e-01
3.26266021e-01 -6.85775936e-01 7.77195752e-01 -9.43461537e-01
5.99669278e-01 -2.86638122e-02 1.37658278e-02 -5.31983435e-01
3.68712693e-02 -4.91527200e-01 3.43497209e-02 1.29725528e+00
9.29523945e-01 -6.23523355e-01 6.48025334e-01 8.01925957e-01
-3.60352278e-01 -4.05097663e-01 -1.33298671e+00 -1.01958156e+00
1.06883198e-01 -4.86812443e-01 4.07549441e-01 7.99851596e-01
4.89942908e-01 5.02920508e-01 -6.85815588e-02 -5.28027862e-02
5.60110807e-01 5.96757770e-01 6.18129849e-01 -1.26076615e+00
-1.68411300e-01 -4.37937707e-01 -5.23812287e-02 -9.20480967e-01
5.68397760e-01 -1.19104207e+00 -2.45697692e-01 -1.90711594e+00
3.71404260e-01 -1.44057691e-01 3.27209360e-03 6.37012064e-01
1.62792906e-01 -4.09286797e-01 1.66320279e-01 3.74175936e-01
-7.88064420e-01 7.26531669e-02 7.73349285e-01 -6.07394934e-01
1.15860559e-01 -1.64956495e-01 -9.84867930e-01 5.73281884e-01
4.71357703e-01 -7.78685153e-01 -4.80318755e-01 -4.77710873e-01
4.18744594e-01 2.33426929e-01 2.02792823e-01 -7.40718663e-01
4.16144967e-01 6.96291775e-02 -2.93605447e-01 -5.99115312e-01
4.91197780e-02 -7.37115145e-01 -2.78994799e-01 2.94022441e-01
-6.75943434e-01 -3.55771333e-01 -2.39359420e-02 4.55230117e-01
-6.00169778e-01 -1.06515348e+00 2.44356487e-02 -3.49598885e-01
-5.82015514e-01 -1.99494705e-01 -5.45394182e-01 2.50103831e-01
8.95299196e-01 -1.93588123e-01 -7.05750048e-01 -2.91040808e-01
-7.22878575e-01 1.29860654e-01 -1.36864521e-02 5.14130175e-01
4.05793369e-01 -8.44372571e-01 -9.18918133e-01 -1.75008997e-01
3.77693325e-01 -1.58617899e-01 2.22841680e-01 5.23141563e-01
-2.37106174e-01 1.33368742e+00 2.70083427e-01 -9.05001443e-03
-1.36716568e+00 5.29972672e-01 9.96139497e-02 -9.31859732e-01
-6.87669694e-01 1.43778846e-01 -1.09904058e-01 -2.20120624e-01
2.00705782e-01 -4.82007444e-01 -4.32508975e-01 7.26612210e-02
5.42638719e-01 4.53072339e-02 4.43784177e-01 -4.51945543e-01
-3.15223277e-01 5.10740876e-01 -5.25313079e-01 -2.26601899e-01
1.43768740e+00 1.63836673e-01 -1.93382338e-01 -2.74771620e-02
1.10181212e+00 2.33087212e-01 -5.78495383e-01 -2.96199650e-01
7.92003989e-01 -3.06483924e-01 -3.62036049e-01 -1.09552455e+00
-7.74787009e-01 4.28598523e-01 1.29620701e-01 1.52449548e-01
6.83541238e-01 3.97691309e-01 6.98916376e-01 5.92924654e-01
5.32595396e-01 -1.62722468e+00 -4.31193382e-01 5.83867967e-01
1.08885252e+00 -1.14789593e+00 2.15547070e-01 -5.63462138e-01
-3.52880865e-01 1.13026464e+00 2.55054057e-01 1.32684991e-01
3.24708492e-01 2.69927591e-01 7.10795820e-02 -8.51238310e-01
-8.37636232e-01 -3.03995103e-01 2.55640805e-01 3.34729910e-01
4.88721162e-01 -4.81078863e-01 -1.13334167e+00 6.01387501e-01
-1.07057288e-01 7.43168741e-02 5.15738964e-01 1.14473867e+00
-2.60164440e-01 -1.45694888e+00 -3.88174802e-02 5.61868370e-01
-1.07292807e+00 -2.94321954e-01 -6.74534142e-01 1.20545876e+00
7.39619136e-02 1.04336023e+00 -1.58773825e-01 1.59888655e-01
3.19533080e-01 4.83342886e-01 2.59129077e-01 -1.04588103e+00
-6.54634178e-01 -1.08562596e-01 8.31395924e-01 -6.97448134e-01
-7.86893010e-01 -9.74054337e-01 -1.23354435e+00 4.22093362e-01
-2.80340850e-01 4.14485335e-01 6.71360731e-01 1.37201238e+00
3.63489419e-01 2.29655460e-01 -1.89821169e-01 8.71971920e-02
-7.34421372e-01 -1.09498942e+00 -5.37701547e-01 8.91938090e-01
8.74610543e-02 -5.47226489e-01 -4.01513606e-01 1.63613394e-01] | [9.954813957214355, 9.232892036437988] |
27419e7a-3333-457d-86ad-3e05e06b63c2 | unsupervised-translation-of-german-lower | 2109.12012 | null | https://arxiv.org/abs/2109.12012v1 | https://arxiv.org/pdf/2109.12012v1.pdf | Unsupervised Translation of German--Lower Sorbian: Exploring Training and Novel Transfer Methods on a Low-Resource Language | This paper describes the methods behind the systems submitted by the University of Groningen for the WMT 2021 Unsupervised Machine Translation task for German--Lower Sorbian (DE--DSB): a high-resource language to a low-resource one. Our system uses a transformer encoder-decoder architecture in which we make three changes to the standard training procedure. First, our training focuses on two languages at a time, contrasting with a wealth of research on multilingual systems. Second, we introduce a novel method for initializing the vocabulary of an unseen language, achieving improvements of 3.2 BLEU for DE$\rightarrow$DSB and 4.0 BLEU for DSB$\rightarrow$DE. Lastly, we experiment with the order in which offline and online back-translation are used to train an unsupervised system, finding that using online back-translation first works better for DE$\rightarrow$DSB by 2.76 BLEU. Our submissions ranked first (tied with another team) for DSB$\rightarrow$DE and third for DE$\rightarrow$DSB. | ['Gertjan van Noord', 'Antonio Toral', 'Ahmet Üstün', 'Lukas Edman'] | 2021-09-24 | null | null | null | null | ['unsupervised-machine-translation'] | ['natural-language-processing'] | [-1.18580759e-01 2.23232910e-01 -3.32255483e-01 -4.21231091e-01
-1.36945152e+00 -8.17711234e-01 8.55257034e-01 -1.15856886e-01
-7.92178333e-01 1.31875002e+00 2.74467349e-01 -8.65064979e-01
2.42914334e-01 -5.86064696e-01 -7.52487183e-01 -2.12917462e-01
2.79914320e-01 1.15950584e+00 -1.47651091e-01 -8.97284150e-01
-9.92852077e-02 1.63112402e-01 -8.01815450e-01 3.41009319e-01
6.28712535e-01 1.87420756e-01 2.93694884e-01 5.65372109e-01
-8.23548436e-03 2.69932270e-01 -5.39554954e-01 -1.03440893e+00
4.55431610e-01 -6.28625095e-01 -1.22440505e+00 -5.60640275e-01
5.42443275e-01 -2.35735461e-01 -2.20644638e-01 8.26109886e-01
6.01799011e-01 6.83352202e-02 5.43526888e-01 -7.28650689e-01
-5.82739413e-01 1.15119541e+00 -3.14327538e-01 2.82849073e-01
5.11402667e-01 -8.91496316e-02 1.21198952e+00 -1.08193457e+00
1.10901499e+00 1.17903960e+00 6.64618254e-01 7.25311279e-01
-1.19483602e+00 -8.88893306e-01 -1.29266843e-01 -1.52973905e-01
-1.44242597e+00 -7.70956278e-01 8.59606341e-02 -5.69860280e-01
1.65112793e+00 2.71375149e-01 3.87873173e-01 1.33380330e+00
1.62283584e-01 4.35287207e-01 1.28397381e+00 -7.93278158e-01
-4.59181219e-01 5.07768393e-01 -1.57064289e-01 5.34244776e-01
2.21527055e-01 1.53910398e-01 -6.28228486e-01 2.27850944e-01
5.51381588e-01 -6.29668176e-01 6.30374625e-02 1.89391360e-01
-1.59886503e+00 8.31916988e-01 -2.70214658e-02 5.07420719e-01
7.05455467e-02 -1.30954251e-01 4.65009719e-01 8.13259602e-01
5.12599945e-01 7.06106961e-01 -8.15432489e-01 -4.39159632e-01
-1.12688947e+00 1.02405034e-01 7.98371494e-01 1.29524732e+00
7.07745612e-01 -1.12179175e-01 2.81599555e-02 1.09400606e+00
1.42232329e-01 6.58930242e-01 4.72971112e-01 -6.04301870e-01
1.00542200e+00 7.37165473e-03 2.59531051e-01 -4.96225178e-01
-1.61712483e-01 -3.22111249e-01 -5.58926642e-01 -3.43203366e-01
5.91061592e-01 -2.56967425e-01 -8.40789378e-01 1.78086543e+00
-1.55594543e-01 -8.76103818e-01 2.59079665e-01 7.89989412e-01
7.54232228e-01 8.61257017e-01 -1.24520533e-01 -3.04039538e-01
1.37404943e+00 -1.32977974e+00 -6.95913613e-01 -4.45347935e-01
9.20954108e-01 -1.35908866e+00 1.12183702e+00 4.34871405e-01
-1.39040196e+00 -5.46930850e-01 -9.30763364e-01 -3.82825404e-01
-7.00911462e-01 3.75409454e-01 3.39567244e-01 7.06081808e-01
-1.44334662e+00 4.95049179e-01 -5.81953228e-01 -8.88224840e-01
-3.58899832e-01 3.94976467e-01 -4.97643799e-01 -9.40713212e-02
-1.64020050e+00 1.40248919e+00 4.32469994e-01 -1.51895672e-01
-8.53232980e-01 -3.07107776e-01 -8.36300433e-01 -4.67878133e-01
-4.84881848e-02 -4.86811191e-01 1.47391224e+00 -7.99286425e-01
-1.71174848e+00 1.44012344e+00 -3.14287663e-01 -4.26133186e-01
7.79426754e-01 -2.39700139e-01 -7.66781867e-01 -2.75828183e-01
5.31114340e-01 7.37821102e-01 3.38515580e-01 -7.11100042e-01
-7.16305077e-01 -1.49382174e-01 -3.00245360e-02 4.40661967e-01
-1.62682801e-01 5.88197529e-01 -4.72472280e-01 -8.03639770e-01
9.85812098e-02 -9.60130930e-01 2.59275548e-02 -8.75857472e-01
-3.17550242e-01 -1.92022711e-01 1.40588596e-01 -1.18850720e+00
1.27055192e+00 -1.80942965e+00 3.19261760e-01 2.66458634e-02
2.25394033e-02 1.26625523e-01 -1.31520838e-01 9.45149183e-01
-2.66301364e-01 2.58094370e-01 4.53854427e-02 -3.99755448e-01
1.77300908e-02 2.07232088e-01 -3.49254578e-01 1.53648213e-01
6.23883754e-02 7.37410545e-01 -1.08143747e+00 -2.87452787e-01
-2.34630331e-01 1.73991188e-01 -2.66926587e-01 -8.29375291e-04
8.76549482e-02 4.52668279e-01 2.00223159e-02 5.66888034e-01
5.49246728e-01 1.20241515e-01 4.59239274e-01 1.35844603e-01
-4.54828054e-01 9.45425510e-01 -6.95985973e-01 1.84778321e+00
-7.11172342e-01 8.64187896e-01 9.85515118e-02 -7.89933920e-01
1.04186225e+00 5.00839412e-01 2.37987250e-01 -8.32340002e-01
-8.74195341e-03 1.03361464e+00 4.04765923e-03 -6.38958365e-02
8.42702568e-01 -5.88681281e-01 -3.66351902e-01 5.39152622e-01
2.76647687e-01 -4.11833525e-01 5.65821290e-01 2.40637839e-01
7.75135517e-01 3.86419624e-01 2.45794207e-01 -6.66035235e-01
4.17766154e-01 1.75827608e-01 5.22645235e-01 5.86290240e-01
-1.20343819e-01 5.30629218e-01 1.81468830e-01 -3.76824826e-01
-1.36622024e+00 -9.63729441e-01 -1.23878449e-01 1.27866328e+00
-1.19873106e-01 -6.26223624e-01 -7.49975920e-01 -6.21678770e-01
-2.87823945e-01 8.31191063e-01 -1.86961606e-01 1.07495829e-01
-9.02241051e-01 -6.33043528e-01 8.78750741e-01 2.71438539e-01
4.87765372e-01 -9.00447428e-01 2.43428480e-02 2.79247165e-01
-7.21387684e-01 -1.12997329e+00 -7.31200635e-01 3.58629376e-01
-7.28821456e-01 -4.26092029e-01 -9.83635068e-01 -1.13343191e+00
4.52118188e-01 -1.49178907e-01 1.57641792e+00 -2.01011360e-01
1.10704869e-01 -1.33708585e-02 -3.20528567e-01 -2.87808031e-01
-6.97825909e-01 5.87273359e-01 3.28752071e-01 -6.52767420e-01
7.21787930e-01 -9.06884074e-02 -2.19850287e-01 5.12480438e-01
-4.42586422e-01 1.73285127e-01 5.76801062e-01 9.99538660e-01
3.86250228e-01 -4.83422250e-01 2.38035917e-01 -9.30006623e-01
6.65610433e-01 -2.14663550e-01 -5.88752866e-01 2.19216913e-01
-8.10268939e-01 -4.27016476e-03 6.16392970e-01 -1.33400053e-01
-7.26436257e-01 -2.42150784e-01 -3.81946653e-01 1.42325565e-01
1.63035050e-01 4.34914172e-01 -1.33251235e-01 2.41702318e-01
6.72151804e-01 1.04812272e-01 -3.01012248e-01 -7.18835890e-01
3.75527412e-01 1.04858613e+00 1.87949926e-01 -7.20487595e-01
7.02775478e-01 -2.67878287e-02 -7.70882607e-01 -5.86712897e-01
-3.82841676e-01 -4.75756466e-01 -6.90653801e-01 1.07286766e-01
7.67925322e-01 -1.18340826e+00 -1.63102046e-01 3.98012519e-01
-1.18680322e+00 -5.57191610e-01 -2.47292757e-01 7.87040591e-01
-6.11054063e-01 4.21375260e-02 -8.14949155e-01 -3.31086189e-01
-5.43661952e-01 -1.40273690e+00 1.17992973e+00 -2.41849601e-01
-6.29984200e-01 -9.00291264e-01 3.29114556e-01 6.29838765e-01
2.70691842e-01 -4.27710950e-01 8.02392423e-01 -7.08405852e-01
-8.89113024e-02 4.78409305e-02 -2.65041053e-01 4.40958112e-01
6.40364662e-02 -2.79723316e-01 -4.90682364e-01 -7.00031519e-01
-3.02094907e-01 -3.47806305e-01 4.42003638e-01 -6.94978684e-02
8.19943398e-02 -3.14175695e-01 -3.38902473e-01 4.54169512e-01
1.29542422e+00 1.69869110e-01 6.59666359e-01 7.35943675e-01
3.92664939e-01 4.43168610e-01 5.70639133e-01 -1.76956877e-02
5.21329641e-01 9.17419851e-01 -4.19865549e-01 -2.07651094e-01
-3.12368959e-01 -3.78423065e-01 8.88215065e-01 1.50057077e+00
-1.43076032e-01 -1.65464684e-01 -1.15152717e+00 9.51766312e-01
-1.21851599e+00 -5.45126855e-01 3.98108130e-03 2.34551907e+00
1.42821181e+00 2.50548661e-01 2.22181603e-01 -3.76241833e-01
7.18448222e-01 -1.50299491e-03 2.44371980e-01 -1.10206211e+00
-2.11097136e-01 6.43162251e-01 7.40496278e-01 8.67739201e-01
-7.51083314e-01 1.60410392e+00 6.61777449e+00 7.76851118e-01
-1.06987190e+00 2.52906412e-01 3.98611993e-01 -1.29579857e-01
-3.42713803e-01 3.02257031e-01 -1.11995471e+00 4.59138393e-01
1.31408954e+00 -1.84837103e-01 7.37620473e-01 4.36258733e-01
1.39955729e-01 8.62202421e-02 -1.26685655e+00 9.01660264e-01
6.50321022e-02 -9.79882836e-01 8.41531605e-02 8.31700712e-02
8.82285357e-01 5.51251411e-01 -7.50860646e-02 7.56694376e-01
7.21627116e-01 -9.86692250e-01 8.20977926e-01 1.29885867e-01
1.27028012e+00 -5.01980841e-01 7.29579210e-01 3.29891145e-01
-9.98124599e-01 5.03976047e-01 -3.47936153e-01 -6.68668300e-02
1.24407597e-01 4.19626802e-01 -1.10820746e+00 7.86125302e-01
8.17020237e-01 5.05207181e-01 -3.29028487e-01 2.93128759e-01
-3.69775534e-01 5.82384288e-01 -3.65091473e-01 -5.47984429e-02
4.06848550e-01 -4.25224096e-01 6.46294951e-01 1.67831314e+00
5.01235902e-01 -1.97963864e-01 1.01361074e-01 3.37707728e-01
-4.40886378e-01 4.16627079e-01 -7.80447125e-01 -2.05178604e-01
3.55189800e-01 8.46797824e-01 -4.12258446e-01 -5.58867753e-01
-1.75405145e-01 1.22662354e+00 3.44417453e-01 3.02958786e-01
-7.69831002e-01 -6.84476018e-01 5.52897096e-01 1.88257501e-01
2.06153125e-01 -4.37639058e-01 -7.50212818e-02 -1.43553817e+00
3.90694737e-02 -1.50286770e+00 3.59213710e-01 -7.39055812e-01
-1.02497995e+00 9.53249335e-01 -3.55622508e-02 -1.09819448e+00
-4.69275713e-01 -6.41216636e-01 1.31820768e-01 1.41227138e+00
-1.32751095e+00 -9.91618156e-01 3.41617972e-01 3.30767304e-01
8.08987975e-01 -4.73541737e-01 1.25485420e+00 6.32876217e-01
-4.35395539e-01 9.18938160e-01 5.17309487e-01 3.41749758e-01
1.37092817e+00 -1.21163809e+00 7.58875251e-01 8.56922209e-01
4.21352446e-01 1.00684953e+00 6.80532455e-01 -7.31567800e-01
-1.05673647e+00 -6.60749376e-01 1.98749268e+00 -7.37764180e-01
8.05840790e-01 -6.61842942e-01 -2.71320134e-01 9.99060392e-01
7.44847178e-01 -5.96755803e-01 6.48291707e-01 3.76597404e-01
-4.19353426e-01 -1.17109664e-01 -9.88957822e-01 7.76063085e-01
1.04105461e+00 -8.11146796e-01 -8.08359325e-01 5.85323036e-01
6.33075356e-01 -6.62978530e-01 -1.02974391e+00 1.41646326e-01
6.09870970e-01 -4.51394111e-01 7.11008251e-01 -7.82625794e-01
5.58471441e-01 -1.86600760e-02 -2.76159197e-01 -1.47793531e+00
-2.29181841e-01 -1.08323371e+00 5.25124669e-01 1.00510788e+00
1.00972748e+00 -8.06440592e-01 3.91007453e-01 5.70128188e-02
-9.78670418e-02 -3.58317524e-01 -1.04531753e+00 -8.95204365e-01
6.14590824e-01 -2.05009133e-01 4.87028807e-01 1.05197358e+00
2.06923395e-01 6.16635144e-01 -3.39870155e-01 -3.48914236e-01
2.11375356e-01 -3.62874828e-02 7.41993427e-01 -7.12272584e-01
-3.64924401e-01 -5.31657457e-01 -5.14370464e-02 -1.29176617e+00
-1.62586480e-01 -1.23255730e+00 1.10277936e-01 -1.47245741e+00
1.06291428e-01 -3.73954743e-01 -1.16526678e-01 5.09118259e-01
6.31328449e-02 5.61520755e-01 9.17012915e-02 3.75945240e-01
-2.12830886e-01 4.06432934e-02 1.04088664e+00 -2.49188304e-01
-7.40975589e-02 -1.86435297e-01 -7.25253701e-01 4.13181216e-01
6.35345995e-01 -5.46871424e-01 -1.51152596e-01 -1.12168717e+00
4.72556084e-01 -1.02224171e-01 -2.64393657e-01 -5.39276540e-01
-2.56246645e-02 1.11954346e-01 2.31066160e-02 -5.15772402e-01
1.72487050e-01 -4.59478319e-01 6.07723370e-02 2.60816872e-01
-2.74143994e-01 6.19031906e-01 4.24237192e-01 -9.20112431e-02
-2.84825712e-01 -2.63231039e-01 6.76400363e-01 -2.85549074e-01
-4.41259563e-01 -1.56789288e-01 -5.59441566e-01 3.15227598e-01
5.90393305e-01 -1.49469987e-01 -1.06809914e-01 -3.86472017e-01
-9.05976892e-01 7.66778663e-02 3.78336638e-01 6.16221011e-01
1.87317461e-01 -1.22918916e+00 -1.13473558e+00 3.47922504e-01
1.28216773e-01 -4.29815471e-01 -4.07737732e-01 9.13241982e-01
-9.25962925e-01 7.05038607e-01 -1.47331297e-01 -2.82334447e-01
-9.45587575e-01 7.99090117e-02 3.57310593e-01 -7.63691962e-01
-3.27174842e-01 1.24089968e+00 -3.04849893e-01 -9.93884861e-01
-7.38523304e-02 -6.50594151e-03 1.94349691e-01 1.94736212e-01
7.21131563e-02 2.27115929e-01 5.99809468e-01 -8.19211483e-01
-5.16101420e-01 4.45493221e-01 -5.22832274e-01 -8.09833407e-01
1.05208158e+00 -2.45813161e-01 -3.95226002e-01 4.60051864e-01
1.29844904e+00 3.97410363e-01 -3.39110851e-01 -2.51082540e-01
1.56982318e-01 -1.75020263e-01 -1.66575164e-01 -1.13420999e+00
-5.62278271e-01 7.42305338e-01 2.84864396e-01 -1.84058845e-01
8.43533278e-01 -2.03502372e-01 1.06206310e+00 5.06628454e-01
7.28341937e-01 -1.52670085e+00 -4.76542950e-01 9.93692815e-01
7.46231019e-01 -1.16662705e+00 3.00969407e-02 -1.20138273e-01
-5.63974023e-01 1.07332242e+00 2.54204661e-01 2.18067124e-01
2.73043990e-01 1.40602738e-01 3.81485522e-01 1.10512063e-01
-4.49716747e-01 -2.31875442e-02 1.54009208e-01 2.73722649e-01
9.34142053e-01 2.66898125e-01 -8.99254799e-01 4.76893276e-01
-8.41238618e-01 -1.28718019e-01 2.96198368e-01 7.68422008e-01
1.08739128e-02 -1.77161491e+00 -2.42419809e-01 3.13320458e-02
-5.73077261e-01 -6.85270488e-01 -6.61430180e-01 1.10772502e+00
1.01938836e-01 8.86776268e-01 -1.33957013e-01 -3.92574549e-01
3.14617127e-01 3.55354697e-01 5.76436937e-01 -8.80476654e-01
-9.06847417e-01 3.46987009e-01 6.21656716e-01 -2.53661901e-01
-1.82492316e-01 -7.53768682e-01 -8.99602890e-01 -4.81995285e-01
-2.44204804e-01 6.97030902e-01 7.99182475e-01 8.43659163e-01
6.82102069e-02 1.14320770e-01 5.83329558e-01 -2.82382667e-01
-6.10163569e-01 -1.19360268e+00 -3.28877270e-01 1.39024690e-01
2.68350858e-02 -2.52107441e-01 -2.37367064e-01 2.28729680e-01] | [11.504048347473145, 10.360682487487793] |
141905c5-01a6-4128-8685-bfbedaa40b70 | truveta-mapper-a-zero-shot-ontology-alignment | 2301.09767 | null | https://arxiv.org/abs/2301.09767v2 | https://arxiv.org/pdf/2301.09767v2.pdf | Truveta Mapper: A Zero-shot Ontology Alignment Framework | In this paper, a new perspective is suggested for unsupervised Ontology Matching (OM) or Ontology Alignment (OA) by treating it as a translation task. Ontologies are represented as graphs, and the translation is performed from a node in the source ontology graph to a path in the target ontology graph. The proposed framework, Truveta Mapper (TM), leverages a multi-task sequence-to-sequence transformer model to perform alignment across multiple ontologies in a zero-shot, unified and end-to-end manner. Multi-tasking enables the model to implicitly learn the relationship between different ontologies via transfer-learning without requiring any explicit cross-ontology manually labeled data. This also enables the formulated framework to outperform existing solutions for both runtime latency and alignment quality. The model is pre-trained and fine-tuned only on publicly available text corpus and inner-ontologies data. The proposed solution outperforms state-of-the-art approaches, Edit-Similarity, LogMap, AML, BERTMap, and the recently presented new OM frameworks in Ontology Alignment Evaluation Initiative (OAEI22), offers log-linear complexity in contrast to quadratic in the existing end-to-end methods, and overall makes the OM task efficient and more straightforward without much post-processing involving mapping extension or mapping repair. | ['Saman Zarandioon', 'Sadra Naddaf-sh', 'Alireza Bahramali', 'Sina Ehsani', 'Mahsa Eslamialishah', 'Murchana Baruah', 'Mariyam Amir'] | 2023-01-24 | null | null | null | null | ['ontology-matching'] | ['knowledge-base'] | [ 6.62286162e-01 3.65956217e-01 -1.67299509e-02 -4.32186395e-01
-7.98993945e-01 -5.15110672e-01 5.83891809e-01 5.14466107e-01
-4.48241532e-01 4.71311629e-01 -3.36087513e-04 -3.04941356e-01
-6.12703741e-01 -6.98147535e-01 -7.72520185e-01 -5.33027165e-02
1.79624371e-02 1.28211260e+00 1.91947386e-01 -3.60369802e-01
1.80351794e-01 5.88912554e-02 -1.58567739e+00 3.09639931e-01
1.43303573e+00 7.22560823e-01 3.18045825e-01 4.00734633e-01
-6.37674928e-01 6.35691285e-01 -9.86289978e-03 -1.11996031e+00
3.19983810e-01 -2.33999819e-01 -1.17526329e+00 -2.94893205e-01
8.16063523e-01 3.46138775e-01 -2.97057163e-02 1.30584610e+00
6.49025679e-01 -1.42414019e-01 3.73836190e-01 -1.46549785e+00
-7.50024199e-01 4.67974097e-01 -2.06857309e-01 -4.16128002e-02
5.43705761e-01 -3.19493085e-01 1.29597926e+00 -8.37115288e-01
8.02664459e-01 1.22585416e+00 7.08351851e-01 1.51099280e-01
-1.01410031e+00 -5.28058469e-01 -2.71459967e-01 6.49116933e-01
-1.41716480e+00 -6.19342446e-01 2.28721261e-01 -5.73852241e-01
1.48598361e+00 2.21030638e-01 2.76880890e-01 6.47803366e-01
4.40634996e-01 1.76752523e-01 7.52022564e-01 -8.18879247e-01
-6.86641186e-02 -1.33278146e-01 -1.51399627e-01 9.39913571e-01
1.08674347e-01 -2.45974168e-01 -6.54646814e-01 -3.77626330e-01
2.60046590e-02 -1.67114079e-01 3.07515059e-02 -7.12753117e-01
-1.18401778e+00 2.74653107e-01 2.21422866e-01 1.79705784e-01
-3.67522478e-01 -1.12389065e-01 6.50781870e-01 8.89183521e-01
4.76266891e-01 3.19780111e-01 -5.23930013e-01 -1.58717006e-01
-8.20372701e-01 2.47673184e-01 8.24788153e-01 1.51861608e+00
1.11319959e+00 -5.01607239e-01 1.18385300e-01 8.33089173e-01
3.49174738e-01 1.32985398e-01 5.87152064e-01 -4.31116730e-01
9.37973022e-01 9.82527792e-01 1.62862763e-01 -8.63788128e-01
-3.48393589e-01 -3.65648180e-01 -3.70664150e-01 -2.17181414e-01
7.65915960e-02 3.50280166e-01 -6.57270968e-01 1.70176935e+00
5.00782192e-01 3.93682092e-01 3.08529377e-01 6.10257447e-01
5.99563599e-01 1.79319128e-01 1.79082870e-01 1.99765526e-02
1.79529977e+00 -1.51237619e+00 -8.26072693e-01 -5.27696967e-01
1.13235366e+00 -1.12776041e+00 1.18607485e+00 5.58391586e-03
-1.03129780e+00 -5.36264122e-01 -1.07998490e+00 -3.61675262e-01
-6.89147592e-01 -8.12520608e-02 3.81046116e-01 4.01002198e-01
-1.06332386e+00 6.26422644e-01 -6.42270267e-01 -1.04900372e+00
2.18359366e-01 5.14070451e-01 -8.37520421e-01 -1.90711170e-01
-1.33905470e+00 1.35706866e+00 8.83262217e-01 -1.10679768e-01
-2.89145678e-01 -7.27000892e-01 -8.84431481e-01 1.72426090e-01
5.25740802e-01 -1.21073472e+00 1.09771240e+00 -1.03498244e+00
-1.66638839e+00 1.18580222e+00 -3.62583160e-01 -5.41508377e-01
4.81920958e-01 -6.18619740e-01 -7.85816431e-01 -7.28584379e-02
2.26294190e-01 2.44512320e-01 4.59735870e-01 -5.80375910e-01
-7.02689767e-01 -4.48569983e-01 1.44234649e-03 2.81781703e-01
-4.86430258e-01 4.90562469e-01 -5.90834975e-01 -4.69555408e-01
2.29510665e-03 -1.04391372e+00 1.93665754e-02 -8.05421993e-02
-1.17929712e-01 -2.01266378e-01 5.02033651e-01 -9.97156024e-01
1.41304708e+00 -1.67719257e+00 5.87549746e-01 2.46561512e-01
-6.09228052e-02 2.70977020e-01 -4.52002108e-01 1.04980171e+00
-3.00756812e-01 -2.70715833e-01 -5.54190695e-01 -3.63147378e-01
3.98296356e-01 2.69778550e-01 3.45705673e-02 3.66409570e-01
1.65583327e-01 1.02383113e+00 -1.08791041e+00 -7.48719335e-01
6.41795248e-02 6.64450154e-02 -6.26950443e-01 4.34831738e-01
-1.46050170e-01 3.70694607e-01 -6.39388710e-02 5.88758588e-01
4.98320252e-01 -2.70979553e-01 5.74542880e-01 -2.34724641e-01
2.72221230e-02 4.79898691e-01 -1.04339087e+00 2.45508528e+00
-8.67519557e-01 1.73793420e-01 -3.54781479e-01 -1.27525902e+00
1.15688789e+00 6.37417734e-01 4.62864518e-01 -8.49077821e-01
-2.77540535e-01 7.53896296e-01 2.47619133e-02 -7.51518190e-01
3.96931916e-01 2.08938360e-01 -4.17670235e-02 2.00311691e-01
4.76692885e-01 4.15207326e-01 4.44999486e-01 -9.50228050e-02
1.12188232e+00 7.49007642e-01 8.03476512e-01 -4.26201552e-01
9.88256812e-01 1.07285112e-01 5.03828287e-01 4.03506696e-01
3.19295824e-01 -7.64989033e-02 4.44577672e-02 -4.91700977e-01
-1.23868370e+00 -8.63910258e-01 1.04184877e-02 1.41802454e+00
1.99421704e-01 -7.00660467e-01 -7.63597906e-01 -5.60226262e-01
6.38572946e-02 7.30497539e-01 -2.72123992e-01 -2.83082247e-01
-6.92928910e-01 -2.87595034e-01 7.36757755e-01 4.25296612e-02
2.83258617e-01 -8.02639425e-01 -2.98143744e-01 6.59837604e-01
-5.32013416e-01 -1.69418192e+00 -4.65043098e-01 -4.01707590e-02
-8.81581247e-01 -9.56769347e-01 -2.79051095e-01 -8.18094671e-01
5.82653642e-01 -2.13309273e-01 1.30928624e+00 -8.84290561e-02
-2.76962578e-01 4.74277250e-02 -3.67385566e-01 -2.72243857e-01
-5.69866955e-01 5.66889048e-01 3.94493565e-02 1.52313471e-01
8.08932066e-01 -1.04493475e+00 -1.43109962e-01 4.78010923e-01
-7.86824167e-01 2.23597229e-01 6.35103464e-01 8.09223354e-01
5.75830519e-01 -5.12538493e-01 6.88879073e-01 -9.83109057e-01
3.80993962e-01 -5.69895387e-01 -7.88772285e-01 8.53532970e-01
-9.11269724e-01 3.04591715e-01 6.31266773e-01 -5.13406433e-02
-9.29437280e-01 2.25435998e-02 3.17545161e-02 -3.74295801e-01
9.52679738e-02 9.02153850e-01 -2.78906196e-01 -2.39840224e-01
5.99146724e-01 1.88334703e-01 -4.82710861e-02 -7.59452939e-01
5.22231519e-01 9.41203475e-01 8.50248516e-01 -7.14989126e-01
9.81165409e-01 1.72054470e-01 1.33090280e-02 -2.24233747e-01
-8.08075964e-01 -7.35131145e-01 -1.16479146e+00 1.05064670e-02
8.07013750e-01 -1.09548938e+00 -5.08325696e-01 1.14997737e-02
-1.27987099e+00 1.34878337e-01 1.05141990e-01 3.92987251e-01
-8.00625861e-01 6.06003463e-01 -9.15639326e-02 -4.63675618e-01
-8.41520905e-01 -1.03416169e+00 1.20056009e+00 -1.12590045e-01
-2.75271386e-01 -1.10816586e+00 1.95537895e-01 6.47122324e-01
5.74599087e-01 8.81598443e-02 1.27508962e+00 -1.07372224e+00
-4.99888092e-01 -3.20738435e-01 -2.49978840e-01 4.28137369e-02
2.40731597e-01 -3.65013123e-01 -6.43594682e-01 -4.75740761e-01
-5.60476780e-01 -2.30431736e-01 1.86648056e-01 -4.51308280e-01
6.67514384e-01 -3.12201887e-01 -4.01241094e-01 7.13998497e-01
1.58194196e+00 -1.19912572e-01 6.56173050e-01 8.05982709e-01
7.92280912e-01 7.07911134e-01 8.84649634e-01 1.79811954e-01
7.87711799e-01 1.41538393e+00 4.62660670e-01 -4.78644222e-02
2.41467617e-02 -2.61511207e-01 4.03247654e-01 1.43746710e+00
-2.03489318e-01 -2.88314611e-01 -1.27989626e+00 5.71084678e-01
-2.33909917e+00 -8.53965223e-01 -4.09125909e-02 2.38833022e+00
6.60783410e-01 -2.13482946e-01 -1.06260039e-01 -2.88326621e-01
7.75497377e-01 -3.22899312e-01 -2.50202149e-01 -4.64277565e-01
3.85411642e-02 5.71349084e-01 6.46034002e-01 6.92377925e-01
-8.92528415e-01 1.36730540e+00 5.17174578e+00 6.18051112e-01
-7.20943332e-01 5.29417872e-01 -5.02290249e-01 2.04305500e-01
-1.19504645e-01 5.75101495e-01 -5.48013151e-01 3.33163917e-01
1.13720214e+00 -6.01280332e-01 6.03892803e-01 7.49212980e-01
-1.77311018e-01 4.71832871e-01 -1.42934954e+00 8.89858067e-01
9.93160252e-03 -1.18711257e+00 2.42338836e-01 8.22579954e-03
4.21180457e-01 3.30898583e-01 -6.37590587e-01 2.88743585e-01
4.04516637e-01 -7.96944618e-01 6.20608032e-01 6.65723324e-01
7.59495020e-01 -6.48340702e-01 1.02932680e+00 7.93436468e-02
-1.61978173e+00 -1.15167528e-01 -6.15859807e-01 9.67570245e-02
3.72058243e-01 3.41765583e-01 -8.10271502e-01 1.77917516e+00
5.34252524e-01 6.73221290e-01 -5.50119340e-01 1.00581026e+00
-3.75481993e-02 -7.67056569e-02 -8.32417309e-02 4.00384605e-01
2.55895078e-01 -5.91399550e-01 7.09967017e-01 1.30182958e+00
7.53883362e-01 -5.22077560e-01 4.83278394e-01 3.99266273e-01
-1.44788280e-01 7.29976714e-01 -4.10134614e-01 -1.26158819e-01
7.20801234e-01 1.15618992e+00 -8.86363313e-02 -5.26540518e-01
-5.97246647e-01 1.41183364e+00 7.28254437e-01 1.08963862e-01
-8.98797750e-01 -6.11964166e-01 7.71681249e-01 -1.50263846e-01
1.68968692e-01 9.40620005e-02 1.35744914e-01 -1.09621763e+00
3.76583725e-01 -1.16681755e+00 7.44339228e-01 -5.96505225e-01
-1.17094159e+00 6.70920551e-01 -4.77208011e-02 -1.43644178e+00
-3.22663993e-01 -3.53720367e-01 -2.65981078e-01 1.03347480e+00
-1.84900856e+00 -1.57596135e+00 -4.67054218e-01 3.92643780e-01
3.35794687e-01 -5.18171668e-01 1.26554465e+00 1.06367958e+00
-4.06512082e-01 8.94641995e-01 1.15710855e-01 -2.41373554e-01
1.02799380e+00 -1.05990565e+00 8.89571309e-01 7.34425724e-01
1.81118265e-01 5.85557163e-01 6.80576921e-01 -5.28304756e-01
-1.43719757e+00 -1.32209551e+00 1.58546817e+00 -3.26525003e-01
9.05507863e-01 -3.58415902e-01 -1.01540363e+00 8.64579618e-01
4.36781496e-01 -6.73524141e-02 8.85402203e-01 1.30358875e-01
-4.99391526e-01 -7.16635510e-02 -8.68295133e-01 5.69852293e-01
1.70336080e+00 -8.31739962e-01 -8.33275795e-01 6.75638318e-01
7.06435263e-01 -4.61256146e-01 -1.40532529e+00 4.70636129e-01
6.36410892e-01 -6.20485604e-01 9.26573217e-01 -1.08872354e+00
2.06019938e-01 -5.95930874e-01 -2.06709698e-01 -1.15714681e+00
-2.87543446e-01 -1.02473819e+00 -2.38688570e-02 1.27539051e+00
4.38772380e-01 -8.31556380e-01 2.72962600e-01 -6.22474262e-03
-5.77306390e-01 -4.32349354e-01 -1.20497811e+00 -1.13504481e+00
-3.73225749e-01 -8.57370049e-02 1.05952215e+00 1.45216405e+00
2.73268551e-01 4.20838207e-01 -4.91212457e-01 6.25779212e-01
6.27146602e-01 1.36173934e-01 1.18236983e+00 -1.57656610e+00
-2.82091528e-01 -2.16009885e-01 -8.37076008e-01 -7.37828910e-01
3.80372018e-01 -1.64802861e+00 -3.15225184e-01 -1.72907948e+00
-7.82671198e-02 -3.81998748e-01 -3.84772032e-01 7.59238183e-01
-3.24044764e-01 -1.00600876e-01 -7.80124515e-02 4.97465283e-01
-4.59679425e-01 6.05763555e-01 6.90515161e-01 -1.21828970e-02
1.59117848e-01 -5.34855843e-01 -2.41785407e-01 3.95650476e-01
5.17756224e-01 -9.28559482e-01 -4.78330880e-01 -7.67685235e-01
4.54688013e-01 -6.79412857e-02 -2.11437000e-03 -1.00550151e+00
5.57268977e-01 1.36210322e-01 -5.19983590e-01 -2.03902781e-01
1.95273478e-02 -1.02723098e+00 7.43544400e-01 3.73181373e-01
-1.73786432e-01 7.20830798e-01 8.75641406e-02 4.22848046e-01
-4.00054932e-01 -3.86974424e-01 5.57804942e-01 8.23454261e-02
-7.57399261e-01 2.53541291e-01 2.54460990e-01 4.12748978e-02
9.36599314e-01 -3.16281766e-01 -2.20309362e-01 1.45836264e-01
-6.87917888e-01 2.86429912e-01 5.47695220e-01 7.63392448e-01
-7.83044938e-03 -1.51651061e+00 -8.57244611e-01 -1.93743572e-01
8.33182037e-01 -1.78129062e-01 9.28300321e-02 1.08257985e+00
-5.81341267e-01 3.06006014e-01 -5.55158079e-01 -6.39431775e-01
-1.36075640e+00 4.86641169e-01 3.47632408e-01 -6.91853523e-01
-6.97280228e-01 5.30500174e-01 -1.80710211e-01 -1.23569536e+00
4.69200909e-02 1.70827582e-01 1.53049110e-02 1.20409317e-02
2.32165083e-01 2.87523746e-01 6.02682054e-01 -6.87930942e-01
-5.31898499e-01 7.82699764e-01 -2.35691413e-01 1.24618791e-01
1.22341883e+00 -1.78145498e-01 -6.50578499e-01 2.55608857e-01
1.00791097e+00 -1.96638018e-01 -2.56288707e-01 -8.12123716e-01
9.25239027e-01 -4.07428056e-01 -3.20541024e-01 -6.76536918e-01
-6.28113627e-01 6.39429152e-01 6.54465854e-01 -3.75075728e-01
8.74784648e-01 -3.25566351e-01 8.08200896e-01 8.13643873e-01
4.86945480e-01 -1.05653214e+00 -1.47279754e-01 4.64443415e-01
6.63781166e-01 -9.54951525e-01 -7.05451816e-02 -7.10121751e-01
-2.86031961e-01 1.25563228e+00 5.52741170e-01 2.93808639e-01
1.35149419e-01 3.11334468e-02 9.60283503e-02 -3.39300364e-01
-7.93231428e-01 -3.59737247e-01 4.78584498e-01 4.94987637e-01
5.54066062e-01 -4.03568298e-01 -3.91914159e-01 2.11419433e-01
-3.60080212e-01 2.70020813e-01 -6.84191436e-02 9.29670870e-01
-2.35754013e-01 -1.59609997e+00 1.46452799e-01 1.12035200e-01
-2.89968908e-01 -4.69146162e-01 -1.61381841e-01 6.47364020e-01
8.56743231e-02 5.23809671e-01 2.11074039e-01 -2.36019507e-01
7.29289174e-01 6.66405082e-01 5.68432391e-01 -5.85170090e-01
-7.28089035e-01 -3.24718326e-01 5.69005609e-01 -6.79461479e-01
-4.89651561e-01 -3.97629887e-01 -1.04957151e+00 -7.68473223e-02
-6.32535398e-01 2.15226650e-01 5.60152590e-01 1.24018383e+00
9.42398489e-01 6.13164783e-01 2.37344712e-01 -4.32252556e-01
-6.05779767e-01 -1.08206844e+00 2.32032076e-01 5.76361656e-01
-3.17866296e-01 -7.54965603e-01 2.17229828e-01 2.87563920e-01] | [9.195852279663086, 8.129453659057617] |
7e1b95cd-6a60-4160-ada2-2866a0ebefe0 | an-analysis-on-ensemble-learning-optimized | 2201.11440 | null | https://arxiv.org/abs/2201.11440v2 | https://arxiv.org/pdf/2201.11440v2.pdf | An Analysis on Ensemble Learning optimized Medical Image Classification with Deep Convolutional Neural Networks | Novel and high-performance medical image classification pipelines are heavily utilizing ensemble learning strategies. The idea of ensemble learning is to assemble diverse models or multiple predictions and, thus, boost prediction performance. However, it is still an open question to what extent as well as which ensemble learning strategies are beneficial in deep learning based medical image classification pipelines. In this work, we proposed a reproducible medical image classification pipeline for analyzing the performance impact of the following ensemble learning techniques: Augmenting, Stacking, and Bagging. The pipeline consists of state-of-the-art preprocessing and image augmentation methods as well as 9 deep convolution neural network architectures. It was applied on four popular medical imaging datasets with varying complexity. Furthermore, 12 pooling functions for combining multiple predictions were analyzed, ranging from simple statistical functions like unweighted averaging up to more complex learning-based functions like support vector machines. Our results revealed that Stacking achieved the largest performance gain of up to 13% F1-score increase. Augmenting showed consistent improvement capabilities by up to 4% and is also applicable to single model based pipelines. Cross-validation based Bagging demonstrated significant performance gain close to Stacking, which resulted in an F1-score increase up to +11%. Furthermore, we demonstrated that simple statistical pooling functions are equal or often even better than more complex pooling functions. We concluded that the integration of ensemble learning techniques is a powerful method for any medical image classification pipeline to improve robustness and boost performance. | ['Frank Kramer', 'Iñaki Soto-Rey', 'Dominik Müller'] | 2022-01-27 | null | null | null | null | ['image-augmentation'] | ['computer-vision'] | [ 4.18765187e-01 5.88827245e-02 2.73358732e-01 -5.05342960e-01
-9.71060455e-01 -2.64310002e-01 3.40314388e-01 6.08460486e-01
-6.02746785e-01 5.70762753e-01 -1.06459312e-01 -4.11857843e-01
-2.60749549e-01 -5.46576798e-01 -4.97298479e-01 -9.91941094e-01
-3.47715229e-01 3.84416223e-01 1.37526676e-01 -2.32227862e-01
1.43873453e-01 6.71665490e-01 -1.59246039e+00 9.00170505e-01
9.16832805e-01 1.12994421e+00 -1.11939378e-01 9.64781404e-01
-9.48026851e-02 8.06846738e-01 -7.00444698e-01 -3.31357986e-01
-3.60161229e-03 4.33034971e-02 -5.86134434e-01 -2.60532051e-01
5.09964705e-01 -2.50649661e-01 1.57368273e-01 4.23654526e-01
8.75909150e-01 -1.51447624e-01 4.01542813e-01 -1.02054930e+00
-2.70481884e-01 4.82985407e-01 -2.83055425e-01 3.07002783e-01
4.10606014e-03 4.81160253e-01 6.14081323e-01 -7.17135370e-01
2.50845939e-01 7.09394038e-01 1.06523371e+00 4.80387390e-01
-1.30516028e+00 -7.67076731e-01 -2.65595198e-01 6.65864116e-03
-8.71365309e-01 -3.56294155e-01 2.48743594e-01 -7.09221065e-01
1.35109794e+00 3.69947731e-01 2.84225613e-01 7.45614827e-01
5.99328876e-01 7.12141395e-01 1.50144315e+00 -4.32339132e-01
5.73208258e-02 3.66744161e-01 5.00473619e-01 8.11121881e-01
2.94927180e-01 -1.42013445e-01 -3.84356171e-01 -3.01843107e-01
3.43545198e-01 -4.18262668e-02 -5.72220199e-02 3.44778001e-01
-1.31581271e+00 6.57256901e-01 5.01507223e-01 4.07431543e-01
-6.11929595e-01 1.49318175e-02 5.52026808e-01 -4.80309464e-02
6.24484301e-01 1.02296340e+00 -9.25628066e-01 1.59328923e-01
-1.03504407e+00 1.95669338e-01 4.84093428e-01 3.02454293e-01
4.65179443e-01 -1.60555035e-01 -4.25206512e-01 7.49079227e-01
-6.14430122e-02 1.58246920e-01 5.18148720e-01 -7.31606185e-01
1.62182420e-01 7.90928662e-01 -1.47516295e-01 -7.59873390e-01
-9.71174896e-01 -8.83802891e-01 -9.88677442e-01 6.13621771e-01
4.31455016e-01 -2.37701595e-01 -1.09071136e+00 1.38109803e+00
3.35668996e-02 1.37752861e-01 -7.58275241e-02 2.98430234e-01
1.12137771e+00 8.61828253e-02 4.65362906e-01 -5.93036003e-02
1.64187765e+00 -7.42975891e-01 -5.61506867e-01 9.78783295e-02
9.02933598e-01 -7.37395227e-01 7.19186962e-01 4.80848730e-01
-1.03808093e+00 -6.70477986e-01 -1.13444746e+00 2.20992893e-01
-4.83135432e-01 2.05815956e-01 6.98349178e-01 9.73573327e-01
-1.11536860e+00 9.68964219e-01 -1.13563848e+00 -1.15950674e-01
8.76773238e-01 7.80733287e-01 -7.33876169e-01 -1.04795527e-02
-7.35981405e-01 1.11555731e+00 4.87015039e-01 -1.50994450e-01
-1.95197731e-01 -1.38107443e+00 -7.45768666e-01 -4.88564765e-05
-2.33164832e-01 -1.05934143e+00 9.54912007e-01 -8.76461327e-01
-1.20988750e+00 8.67445230e-01 -2.04196826e-01 -6.55577898e-01
4.51648921e-01 -2.41623104e-01 -3.58472496e-01 -6.50842935e-02
-1.74922660e-01 7.06858277e-01 3.73394728e-01 -1.05946076e+00
-4.75246578e-01 -4.99690175e-01 -3.45029771e-01 -1.21690065e-01
-3.99112135e-01 2.37926587e-01 1.53106481e-01 -5.25985718e-01
8.02488625e-03 -8.49534452e-01 -4.80934083e-01 -2.42790267e-01
-1.11162134e-01 1.37945125e-02 5.25670052e-01 -1.01355612e+00
1.12486815e+00 -1.95733440e+00 -3.05277437e-01 1.23689987e-01
4.23892945e-01 4.20667380e-01 -6.90384284e-02 1.32731885e-01
-3.94818664e-01 3.80678385e-01 -3.66614908e-01 -3.60272706e-01
-4.44015235e-01 8.16674978e-02 1.27882764e-01 1.57476291e-01
6.91289723e-01 9.72215176e-01 -7.15521038e-01 -4.56351191e-01
3.93435180e-01 7.20581114e-01 -4.53030109e-01 1.38374999e-01
1.90054744e-01 7.86429048e-01 1.13808259e-01 7.75996149e-01
6.93040371e-01 -4.92400199e-01 6.40334655e-03 -4.99593318e-01
4.56576981e-02 1.68809239e-02 -7.46802688e-01 1.50394523e+00
-4.05952871e-01 4.67331022e-01 -3.68895888e-01 -7.96523631e-01
9.49640930e-01 2.79103518e-01 7.32051313e-01 -2.48759031e-01
1.97192386e-01 3.35389495e-01 4.09103543e-01 -5.63877106e-01
2.98644245e-01 -2.12913454e-01 2.84315020e-01 2.53580660e-02
4.79858309e-01 1.63292170e-01 6.90305978e-02 -2.10685581e-01
1.36944473e+00 7.70361945e-02 5.05825341e-01 -2.64218032e-01
4.97004539e-01 2.27854308e-02 3.86871815e-01 8.59251678e-01
-2.63108373e-01 7.84521997e-01 3.16464752e-01 -5.24010658e-01
-8.85867774e-01 -8.58513176e-01 -4.11840796e-01 1.23587549e+00
-5.11619866e-01 -2.99200982e-01 -6.87206805e-01 -6.59191430e-01
-1.36355221e-01 5.85610747e-01 -8.17798078e-01 5.61383218e-02
-5.58790028e-01 -1.50618720e+00 7.83829868e-01 7.63523281e-01
5.66545725e-01 -9.81551707e-01 -6.62838399e-01 3.43309700e-01
-7.82792177e-03 -1.10744226e+00 1.61180899e-01 4.46236938e-01
-1.24846900e+00 -9.91019368e-01 -7.19570696e-01 -2.85058379e-01
5.72374105e-01 -2.86798626e-01 1.26204491e+00 2.77213186e-01
-5.25274277e-01 6.05091974e-02 -2.89799571e-01 -9.77584839e-01
-5.74637532e-01 1.36953801e-01 -1.57697424e-01 -1.20680116e-01
2.28954166e-01 -3.80547613e-01 -7.96189606e-01 5.19707091e-02
-6.87576473e-01 1.77352339e-01 9.48442280e-01 1.08686829e+00
5.60417652e-01 -2.13424474e-01 5.41108787e-01 -9.85071778e-01
3.97133797e-01 -4.07818407e-01 -2.50573214e-02 3.28856856e-01
-6.23979092e-01 7.91617632e-02 2.97606617e-01 -2.08362371e-01
-9.18966949e-01 2.43397146e-01 -5.19580543e-01 -1.13079615e-01
-4.58568156e-01 4.66641188e-01 2.68156141e-01 -1.98484734e-01
9.04890120e-01 2.37053744e-02 5.15691161e-01 -2.89674908e-01
-2.20154282e-02 5.82852364e-01 3.47710490e-01 -2.07197711e-01
9.17112604e-02 4.46451455e-01 1.51582122e-01 -7.79407680e-01
-5.77734292e-01 -5.26181459e-01 -8.01404059e-01 -1.57446861e-01
1.12758982e+00 -7.79435754e-01 -6.06907606e-01 7.20410407e-01
-9.75187063e-01 -3.00747305e-01 8.67712721e-02 5.00447392e-01
-2.10462272e-01 1.31761745e-01 -6.51583672e-01 -7.87532628e-01
-9.58445489e-01 -1.42297399e+00 9.88928199e-01 2.21033722e-01
-4.37771827e-01 -9.25241292e-01 -8.95913616e-02 5.28797805e-01
7.90165901e-01 7.29337096e-01 8.72151673e-01 -1.30603731e+00
-1.55754671e-01 -4.08141524e-01 -1.13963850e-01 4.77967024e-01
6.88676983e-02 1.90356374e-01 -1.41413271e+00 -2.60483623e-01
-2.80391604e-01 9.54008847e-02 1.03810823e+00 4.92370754e-01
1.42664814e+00 2.16071669e-04 -3.68023574e-01 4.94673401e-01
1.29964888e+00 3.13350677e-01 9.21042860e-01 4.14190650e-01
3.62959772e-01 5.74172020e-01 1.46791980e-01 7.91271701e-02
6.57129586e-02 6.18330836e-01 3.09343308e-01 -2.78234124e-01
-1.17549293e-01 4.52753693e-01 7.26536140e-02 5.07556617e-01
-4.59713072e-01 2.49159425e-01 -1.19120085e+00 1.02755830e-01
-1.55304289e+00 -8.89532685e-01 -4.41789478e-01 2.15871263e+00
5.56247532e-01 1.53903008e-01 4.41424847e-02 2.93632209e-01
3.85209084e-01 -3.90545905e-01 -3.71294171e-01 -4.15420234e-01
-1.91115528e-01 7.03441203e-01 5.15893340e-01 -2.65117045e-02
-1.45431113e+00 2.81487435e-01 6.76705027e+00 6.26967728e-01
-1.35442948e+00 2.82062113e-01 1.03975356e+00 -6.35154396e-02
2.62710452e-01 -4.47374612e-01 -7.00164378e-01 5.01255214e-01
1.12431574e+00 3.68607968e-01 -2.34754249e-01 7.60418952e-01
-5.93722053e-02 -1.30551383e-01 -7.10368812e-01 8.01382244e-01
1.28709167e-01 -1.66929805e+00 -2.65048474e-01 5.31757772e-02
6.77081048e-01 4.49134201e-01 -3.67805585e-02 2.53393471e-01
1.50955588e-01 -1.23006701e+00 1.61430407e-02 5.65489054e-01
6.37906849e-01 -4.59461063e-01 1.39529264e+00 2.21749693e-01
-6.50068104e-01 -1.47759750e-01 1.73714459e-01 -7.28008943e-03
-7.60254040e-02 7.68237114e-01 -1.11316562e+00 7.94144809e-01
9.67593968e-01 4.59560126e-01 -9.04053569e-01 1.25513089e+00
2.19604641e-01 6.48226380e-01 -2.39924788e-01 1.50439978e-01
-2.73679872e-03 4.44614619e-01 1.73315525e-01 1.62627208e+00
3.34440529e-01 1.26021057e-01 -1.27567202e-01 2.55264342e-01
2.33969972e-01 2.97819912e-01 -2.55920976e-01 2.17370942e-01
1.59801655e-02 1.70807898e+00 -9.43747759e-01 -3.72490168e-01
-2.13370800e-01 6.68199241e-01 3.02202124e-02 -9.89061669e-02
-7.60191739e-01 -1.23450741e-01 6.29172146e-01 1.40212357e-01
1.14341013e-01 9.90468338e-02 -8.80974233e-01 -8.21836948e-01
-3.10918629e-01 -8.45120251e-01 5.18630803e-01 -6.68413937e-01
-1.37406349e+00 8.97554755e-01 -1.86997965e-01 -1.08494365e+00
5.50098531e-02 -1.06905377e+00 -7.15245962e-01 1.07856905e+00
-1.21795368e+00 -1.41420460e+00 -6.82213783e-01 2.25672692e-01
2.49865562e-01 -4.00158018e-01 1.29817200e+00 2.03658432e-01
-5.60534179e-01 6.75588369e-01 7.21320733e-02 1.26032263e-01
7.73580134e-01 -1.34598804e+00 -8.82830769e-02 5.36983609e-01
-1.24507464e-01 6.05237067e-01 6.08062923e-01 -3.65387261e-01
-8.04589033e-01 -1.06708407e+00 6.93854153e-01 -7.36744046e-01
2.99512237e-01 1.33367464e-01 -1.27305305e+00 3.96150112e-01
3.45669448e-01 1.86281830e-01 1.41880667e+00 2.28133440e-01
-3.13279063e-01 -1.59832641e-01 -1.44213104e+00 8.54284167e-02
3.47103268e-01 7.92025402e-03 -3.47066641e-01 2.07186416e-01
3.85251731e-01 -5.74385345e-01 -1.30714834e+00 9.83415842e-01
8.68833721e-01 -1.12359917e+00 1.02099550e+00 -8.20942581e-01
8.34003448e-01 -2.33070865e-01 -1.03881165e-01 -1.28867733e+00
-4.43766475e-01 2.63496451e-02 6.49405420e-02 1.00834680e+00
6.92618370e-01 -8.07030082e-01 6.65669620e-01 7.51277864e-01
-3.62622291e-01 -1.31992912e+00 -7.64973581e-01 -3.18165690e-01
2.43715540e-01 -5.90201139e-01 4.27337795e-01 8.73744249e-01
-5.55267222e-02 3.32398452e-02 3.29204425e-02 1.87411770e-01
4.82858509e-01 -4.11353201e-01 7.03604221e-01 -1.16014779e+00
-4.73864168e-01 -7.15258062e-01 -7.59111941e-01 1.46498203e-01
-1.48196459e-01 -1.11536252e+00 -2.60279655e-01 -1.51842439e+00
1.95835635e-01 -6.08151078e-01 -9.66132164e-01 7.46408284e-01
-5.43976963e-01 5.31227350e-01 2.27719948e-01 1.31897509e-01
-1.95735022e-01 -2.83121973e-01 9.05333042e-01 -4.08058688e-02
-2.47895986e-01 1.18984878e-01 -8.36867750e-01 7.04376042e-01
1.13017464e+00 -2.21142352e-01 4.49950919e-02 -3.09630424e-01
-1.21134654e-01 -2.37288684e-01 4.61033642e-01 -1.25451446e+00
2.10114673e-01 3.63719314e-01 8.17539752e-01 -5.89203060e-01
1.44896120e-01 -5.65957367e-01 3.38512510e-01 8.02359402e-01
-2.54839510e-01 1.14802076e-02 7.06802189e-01 9.08848494e-02
-3.36993262e-02 -1.01523519e-01 7.96810031e-01 -2.07296997e-01
-6.82499290e-01 7.80227259e-02 -2.07876846e-01 -6.31904840e-01
1.12927890e+00 -3.02228391e-01 -3.14156473e-01 1.63767353e-01
-1.06838357e+00 -4.65106145e-02 -1.56137332e-01 2.26414397e-01
4.14306730e-01 -8.80948246e-01 -1.06771612e+00 1.39437914e-01
2.42781386e-01 -2.20297918e-01 6.86927617e-01 1.35774684e+00
-6.39954805e-01 2.07683519e-01 -4.81484532e-01 -1.01763022e+00
-1.84028482e+00 2.23563790e-01 5.72710514e-01 -7.33762980e-01
-3.10669094e-01 9.37594414e-01 -9.02819037e-02 -5.29905379e-01
-1.58251375e-01 -2.58581102e-01 -4.02914941e-01 2.14546844e-01
8.79932106e-01 5.41140676e-01 7.67368019e-01 -3.19113433e-01
-5.62701404e-01 3.21423709e-01 -1.83463261e-01 3.17189872e-01
1.52549577e+00 4.81801331e-01 -2.81996340e-01 3.24642658e-01
1.10531187e+00 -4.52527642e-01 -6.66525483e-01 1.00646898e-01
1.09636020e-02 -7.70908073e-02 9.67418626e-02 -1.38800108e+00
-9.59411204e-01 9.34802413e-01 9.91557956e-01 -1.90244932e-02
1.41545701e+00 -2.00694025e-01 2.33704135e-01 2.57658303e-01
2.53122151e-01 -6.10521257e-01 -1.60173088e-01 2.80307353e-01
7.94562817e-01 -1.65030646e+00 1.89183444e-01 -3.26348960e-01
-6.01262271e-01 1.36493111e+00 5.71823120e-01 -6.24046214e-02
8.29102933e-01 6.73312187e-01 5.43718673e-02 -4.83566225e-02
-7.10945308e-01 -6.21912479e-02 5.37617683e-01 6.14771903e-01
1.06680715e+00 1.00486763e-01 -2.05631837e-01 8.13589573e-01
-1.63350254e-01 3.86174738e-01 1.24209225e-01 7.58827090e-01
-1.59727633e-01 -9.95781243e-01 -4.80161935e-01 1.05315936e+00
-9.46193576e-01 -2.16785192e-01 1.86361834e-01 7.43192315e-01
5.41527510e-01 8.04219127e-01 2.19759271e-01 -6.64549291e-01
2.97097534e-01 3.39897424e-01 4.30891573e-01 -4.57424879e-01
-1.30404747e+00 -1.85077608e-01 2.29725428e-02 -4.34665769e-01
-5.59003890e-01 -6.19728863e-01 -1.01236773e+00 -1.15857497e-01
-4.13338363e-01 -3.54574770e-01 9.38434064e-01 1.05916059e+00
5.06699443e-01 1.03226697e+00 1.90086931e-01 -9.57633853e-01
-2.87742049e-01 -1.28092229e+00 -8.14083219e-02 2.05108881e-01
1.98392510e-01 -4.72261786e-01 -1.64584592e-01 1.79108605e-01] | [15.17789077758789, -2.630194902420044] |
7f936abd-d952-48d9-b8e2-f920b2bb05e2 | towards-better-document-level-relation | 2211.14470 | null | https://arxiv.org/abs/2211.14470v1 | https://arxiv.org/pdf/2211.14470v1.pdf | Towards Better Document-level Relation Extraction via Iterative Inference | Document-level relation extraction (RE) aims to extract the relations between entities from the input document that usually containing many difficultly-predicted entity pairs whose relations can only be predicted through relational inference. Existing methods usually directly predict the relations of all entity pairs of input document in a one-pass manner, ignoring the fact that predictions of some entity pairs heavily depend on the predicted results of other pairs. To deal with this issue, in this paper, we propose a novel document-level RE model with iterative inference. Our model is mainly composed of two modules: 1) a base module expected to provide preliminary relation predictions on entity pairs; 2) an inference module introduced to refine these preliminary predictions by iteratively dealing with difficultly-predicted entity pairs depending on other pairs in an easy-to-hard manner. Unlike previous methods which only consider feature information of entity pairs, our inference module is equipped with two Extended Cross Attention units, allowing it to exploit both feature information and previous predictions of entity pairs during relational inference. Furthermore, we adopt a two-stage strategy to train our model. At the first stage, we only train our base module. During the second stage, we train the whole model, where contrastive learning is introduced to enhance the training of inference module. Experimental results on three commonly-used datasets show that our model consistently outperforms other competitive baselines. | ['Xiaodong Shi', 'Qingguo Hu', 'Zijun Min', 'Zhongjian Miao', 'Yidong Chen', 'Jinsong Su', 'Liang Zhang'] | 2022-11-26 | null | null | null | null | ['document-level-relation-extraction'] | ['natural-language-processing'] | [ 1.10114276e-01 7.28145361e-01 -3.43877107e-01 -3.69827658e-01
-7.09327757e-01 -3.43474805e-01 7.54022419e-01 3.61654580e-01
-4.15715903e-01 8.17446172e-01 2.90381424e-02 -3.74328226e-01
-1.67644303e-02 -1.24939275e+00 -9.57307816e-01 -2.75538206e-01
-4.83119889e-04 9.22309399e-01 4.54996079e-01 -3.34618628e-01
-1.38921723e-01 3.13207656e-01 -1.19766295e+00 4.31009918e-01
9.52172220e-01 1.02998042e+00 1.90048739e-01 3.25497121e-01
-2.71312475e-01 1.08611000e+00 -3.55333865e-01 -9.37047720e-01
2.26149466e-02 -2.04064175e-01 -1.16857481e+00 -1.99253887e-01
-1.39525250e-01 -2.58968711e-01 -3.48062932e-01 9.40498650e-01
8.14981908e-02 8.94213468e-02 5.97375572e-01 -9.22761381e-01
-5.32356441e-01 1.12453544e+00 -4.39867944e-01 1.08673602e-01
4.49967831e-01 -3.87182832e-01 1.49990785e+00 -1.28830707e+00
6.63186789e-01 1.07126474e+00 5.28841257e-01 1.55922279e-01
-1.02039647e+00 -6.75959408e-01 5.48329413e-01 4.82077122e-01
-1.60790539e+00 -4.06130105e-01 7.76551068e-01 -1.15839504e-01
1.28166234e+00 1.16367273e-01 3.12775195e-01 6.20020032e-01
-1.60300836e-01 8.44723403e-01 4.23038810e-01 -2.86027461e-01
-2.56543368e-01 2.73462474e-01 3.54333133e-01 4.64980513e-01
4.87359203e-02 -3.34312856e-01 -1.09262809e-01 4.45063859e-02
4.72808510e-01 -8.22633058e-02 -3.78724277e-01 -1.57670621e-02
-1.18194509e+00 4.69614327e-01 7.97615588e-01 4.53058004e-01
-5.05782604e-01 -3.29369724e-01 1.18447080e-01 1.80462196e-01
5.63511372e-01 4.36494887e-01 -8.15129638e-01 3.09913158e-01
-5.73066771e-01 2.65954196e-01 1.01177669e+00 1.37563789e+00
9.73786652e-01 -8.67830873e-01 -3.21849436e-01 8.55332792e-01
3.31773192e-01 -5.97268008e-02 3.91661435e-01 -1.44222170e-01
1.22805083e+00 1.03657985e+00 1.87577307e-01 -1.19543147e+00
-3.40379208e-01 -4.12550539e-01 -1.01019502e+00 -7.05912173e-01
-2.85862554e-02 -2.47794151e-01 -5.96554935e-01 1.42800152e+00
5.63246608e-01 3.24305385e-01 2.90662825e-01 6.17695689e-01
1.37582123e+00 7.74105787e-01 7.90579990e-02 -3.86083156e-01
1.26378679e+00 -1.39772117e+00 -8.51410985e-01 -2.86638469e-01
8.95666420e-01 -6.62638903e-01 4.45534676e-01 -3.49076241e-02
-1.06296504e+00 -7.69031942e-01 -9.86894190e-01 -4.15160298e-01
-6.25158310e-01 4.39402878e-01 5.20166278e-01 -1.91474948e-02
-6.29923284e-01 6.89541101e-01 -6.45056665e-01 2.93053258e-02
3.13312292e-01 5.29966354e-01 -5.77211022e-01 1.09283030e-01
-1.66245914e+00 1.17392302e+00 9.23851013e-01 6.48529351e-01
-1.01189055e-01 -5.68290293e-01 -1.18512166e+00 4.08143163e-01
8.32111180e-01 -7.02702880e-01 1.05250883e+00 -4.19088244e-01
-1.25408089e+00 6.50489569e-01 -5.08279383e-01 -3.09073865e-01
4.55191314e-01 -5.15312195e-01 -4.72869813e-01 -2.06484973e-01
6.59691095e-02 3.75803024e-01 1.91930354e-01 -1.25145078e+00
-8.11339200e-01 -1.16193421e-01 3.23091269e-01 3.76431972e-01
-3.38491835e-02 9.11525451e-03 -9.95081842e-01 -3.90515149e-01
2.34689012e-01 -6.69544697e-01 -3.14807564e-01 -5.97108364e-01
-8.98082674e-01 -7.95320511e-01 4.65389013e-01 -6.10192358e-01
1.39305639e+00 -1.76683509e+00 2.43362173e-01 1.92085296e-01
4.08271343e-01 5.49979806e-01 -5.14588989e-02 4.93193120e-01
-2.75248379e-01 1.61601618e-01 -1.82621345e-01 -5.89108884e-01
-9.63950381e-02 1.32864848e-01 -5.06807446e-01 -1.73728466e-01
9.59519804e-01 1.11639202e+00 -9.93901730e-01 -7.37854779e-01
-4.57625203e-02 2.83481389e-01 -3.33061635e-01 6.03685975e-01
-3.13219398e-01 3.38400453e-01 -5.93956232e-01 3.87217283e-01
9.00905252e-01 -4.07721728e-01 5.71849585e-01 -5.12792647e-01
1.02829270e-01 8.51515949e-01 -1.28712392e+00 1.16850865e+00
-5.44816554e-01 3.27167735e-02 -3.78259838e-01 -1.17862594e+00
9.94466364e-01 3.42041343e-01 1.96423009e-01 -3.05894256e-01
4.88897935e-02 1.08328551e-01 1.11941606e-01 -5.70130944e-01
5.56943536e-01 2.14697868e-02 -3.37462164e-02 1.32398605e-01
3.05091709e-01 9.54845846e-02 4.21695024e-01 3.03049505e-01
9.50767219e-01 3.03915620e-01 4.93001372e-01 3.37478697e-01
1.02332449e+00 -2.27994919e-01 8.40694010e-01 6.18289709e-01
3.46810579e-01 3.92730802e-01 8.07601988e-01 -3.71146739e-01
-6.09330714e-01 -7.73123026e-01 -7.45496750e-02 7.92248309e-01
4.97408301e-01 -8.34294081e-01 -2.79835582e-01 -1.31008053e+00
-1.99914634e-01 5.98210812e-01 -6.61151946e-01 -1.65069699e-02
-7.51272142e-01 -8.03652883e-01 1.02577142e-01 7.44879365e-01
6.37872756e-01 -1.10304344e+00 2.16223270e-01 3.83081257e-01
-2.98124760e-01 -1.58740473e+00 -2.31106177e-01 3.70998174e-01
-6.02033854e-01 -1.11809182e+00 -1.72884285e-01 -9.23536181e-01
7.58677125e-01 -8.16233456e-03 1.30563188e+00 4.27823871e-01
4.51176941e-01 -4.50077772e-01 -5.49999833e-01 -3.11876863e-01
-1.66371241e-01 4.95298952e-01 -2.64068156e-01 7.03715980e-02
6.75303996e-01 -5.16620219e-01 -1.89257920e-01 2.85218179e-01
-5.38326621e-01 4.00429189e-01 9.81718063e-01 9.45755959e-01
7.00729012e-01 4.43129092e-01 5.07443190e-01 -1.41394973e+00
4.30198848e-01 -8.12016964e-01 -3.33478302e-01 6.77083552e-01
-4.69244421e-01 1.43487886e-01 9.20517623e-01 -3.77128124e-01
-1.26695478e+00 1.03507794e-01 -3.13003570e-01 -8.37521106e-02
-9.24435630e-02 9.57236469e-01 -7.01527715e-01 5.47356009e-01
2.25902218e-02 6.65077046e-02 -5.50850213e-01 -4.94876355e-01
3.36692005e-01 7.12307215e-01 6.73501253e-01 -5.95956981e-01
1.12759542e+00 -1.03371181e-01 -1.61201835e-01 -1.45709231e-01
-1.29423332e+00 -3.40943187e-01 -1.05000675e+00 2.68369228e-01
5.82438946e-01 -1.01239765e+00 -6.54694498e-01 2.40201727e-01
-1.42980313e+00 -1.27017811e-01 -3.04111857e-02 4.51500565e-01
3.27415951e-03 1.64663091e-01 -7.46005058e-01 -6.08642578e-01
-4.28679734e-01 -9.64388669e-01 1.07998979e+00 3.54802936e-01
-1.05965137e-01 -8.88473928e-01 -5.07198572e-02 3.30862075e-01
-1.49816722e-01 -1.20957904e-01 8.93388152e-01 -1.20754015e+00
-6.83427870e-01 -2.78173327e-01 -5.81982493e-01 1.04145966e-01
2.55233586e-01 -1.93421915e-03 -8.01038206e-01 1.57182932e-01
-2.85351425e-01 -2.21766934e-01 9.02611792e-01 -3.23784471e-01
1.11428666e+00 -3.85823309e-01 -8.92965853e-01 4.46076721e-01
1.09436154e+00 8.39215070e-02 6.70891881e-01 1.20655015e-01
1.02173901e+00 7.11542010e-01 9.87346053e-01 2.27366928e-02
1.00358689e+00 6.56322002e-01 1.41748682e-01 -1.24196149e-01
2.38850966e-01 -5.41962147e-01 -5.83875328e-02 9.90907013e-01
-2.17773050e-01 -2.93389440e-01 -7.51947522e-01 4.81579393e-01
-2.11021233e+00 -7.66553044e-01 -1.42785862e-01 1.90752530e+00
1.25151598e+00 5.59741318e-01 -1.30786553e-01 1.46393880e-01
7.53935039e-01 -8.12635422e-02 -2.79670060e-01 -1.30024374e-01
1.71203360e-01 2.72712141e-01 -3.04456651e-02 5.66828132e-01
-1.39706993e+00 1.31011045e+00 4.74111128e+00 7.34204292e-01
-6.49477541e-01 -2.44566143e-01 5.98664105e-01 3.26454848e-01
-3.99858236e-01 2.04480529e-01 -1.27397156e+00 4.25580025e-01
5.38236141e-01 -4.33512330e-02 -4.97246049e-02 6.88538253e-01
-3.38412732e-01 1.11249633e-01 -1.51952267e+00 5.34899950e-01
-1.74697712e-01 -1.19905257e+00 1.43566597e-02 -2.72198588e-01
4.18594420e-01 -2.91064322e-01 -4.82094347e-01 8.54928076e-01
3.54811311e-01 -9.44324970e-01 3.24969739e-01 6.79338157e-01
6.04832470e-01 -9.14644659e-01 1.27891934e+00 6.06240392e-01
-1.55777001e+00 8.43148679e-02 -3.58002335e-01 -1.87780827e-01
1.78654626e-01 7.93669105e-01 -9.24631894e-01 1.16511369e+00
4.80738252e-01 1.03799343e+00 -5.12250841e-01 6.96689606e-01
-8.08301866e-01 2.57955313e-01 -2.70794272e-01 -2.72856932e-02
3.42119709e-02 -2.12690577e-01 2.23585844e-01 1.26279128e+00
9.28234905e-02 4.41131413e-01 1.14145167e-01 9.01510358e-01
-2.90704399e-01 2.46916264e-01 -4.22299773e-01 2.16254607e-01
7.32866585e-01 1.56484413e+00 -4.85913455e-01 -5.77065051e-01
-7.50394881e-01 8.76104712e-01 1.12533462e+00 1.54082909e-01
-7.81101525e-01 -7.78643012e-01 1.17847264e-01 -1.01426981e-01
7.11503863e-01 7.68425912e-02 -1.75836310e-01 -1.45695698e+00
4.07415986e-01 -5.21267951e-01 4.57405657e-01 -6.23699725e-01
-1.29757166e+00 9.08595622e-01 -3.81113291e-02 -1.05287433e+00
-3.01973909e-01 -3.66474181e-01 -7.22737908e-01 1.13417876e+00
-1.64464641e+00 -1.44243717e+00 -1.53622910e-01 2.21607506e-01
2.88295597e-01 1.63257986e-01 8.01589549e-01 2.87838101e-01
-1.03312683e+00 7.91696608e-01 -4.39602882e-01 6.31042004e-01
6.42648339e-01 -1.41441000e+00 6.14646375e-01 9.01999295e-01
2.56387830e-01 9.64897633e-01 3.05091172e-01 -7.70497918e-01
-8.68088186e-01 -1.17169619e+00 1.74219394e+00 -3.61909837e-01
7.19647765e-01 -3.97039294e-01 -1.18448830e+00 1.19381499e+00
-3.86486994e-04 1.88040361e-01 6.30447447e-01 6.15769863e-01
-3.35111290e-01 -2.10142419e-01 -7.07948565e-01 6.10871851e-01
1.05053723e+00 -3.94247353e-01 -9.65227962e-01 2.48236910e-01
8.07448864e-01 -7.52488732e-01 -1.22266448e+00 9.27806556e-01
2.41615266e-01 -5.07588923e-01 9.88564312e-01 -6.07780457e-01
7.75170863e-01 -3.68801624e-01 2.72569001e-01 -1.20965815e+00
-3.40651780e-01 -4.15754169e-01 -6.80987239e-01 1.77532327e+00
8.73368561e-01 -5.75766027e-01 5.08470237e-01 8.29883516e-01
-3.61392051e-02 -1.25273788e+00 -4.51712519e-01 -3.80961001e-01
-5.35558425e-02 -2.17024520e-01 8.66001666e-01 9.84735727e-01
2.91101873e-01 1.04520965e+00 -1.42284125e-01 4.82320607e-01
5.91596887e-02 6.53590322e-01 8.87609363e-01 -1.17304122e+00
-5.37601352e-01 -2.28832543e-01 -1.08802266e-01 -1.55942941e+00
4.95949209e-01 -9.24381495e-01 1.81723446e-01 -1.52134693e+00
4.17413384e-01 -7.31662095e-01 -3.09181750e-01 6.59605443e-01
-9.90371644e-01 -1.81335792e-01 3.85190621e-02 1.57400981e-01
-7.05190599e-01 6.10183299e-01 1.22510958e+00 -1.24841265e-01
-2.51904547e-01 3.00938040e-01 -9.03619468e-01 6.89166129e-01
4.27896589e-01 -4.81724054e-01 -4.17572200e-01 -2.83301771e-01
4.35560703e-01 9.84986871e-02 1.55451313e-01 -6.25403821e-01
4.79820788e-01 -3.28538567e-02 4.65106189e-01 -9.26276028e-01
1.39479920e-01 -7.99863040e-01 -1.28702268e-01 2.28955988e-02
-3.75527412e-01 -2.28852496e-01 -7.50542358e-02 4.39826638e-01
-5.26034832e-01 -3.60999465e-01 2.51956582e-01 8.91662091e-02
-4.28547382e-01 4.76094335e-01 1.56067565e-01 -1.15717225e-01
8.73590410e-01 3.21854144e-01 -3.28348219e-01 -1.10371754e-01
-8.01039100e-01 6.72601879e-01 -6.35364801e-02 4.79425788e-01
3.69709998e-01 -1.31695652e+00 -7.59567320e-01 6.05717748e-02
1.28442407e-01 8.94104779e-01 -8.92661326e-03 7.64299095e-01
-7.06251245e-03 3.48235160e-01 4.38958257e-01 -1.57627672e-01
-1.13403726e+00 9.42609489e-01 2.11608097e-01 -1.09298694e+00
-4.51004714e-01 1.13560688e+00 1.64558917e-01 -6.00433350e-01
1.23385107e-02 -4.83081311e-01 -7.29159713e-01 1.56574905e-01
5.40301919e-01 -2.25826487e-01 1.60911337e-01 -7.18764007e-01
-4.58834708e-01 3.83615613e-01 -7.27240622e-01 2.57505685e-01
1.40690076e+00 -9.55588371e-02 -4.45183575e-01 3.67321938e-01
1.04385805e+00 7.05534965e-02 -9.52956319e-01 -6.96312189e-01
2.31751502e-01 -1.87113687e-01 -2.45462194e-01 -6.27843499e-01
-1.06052828e+00 6.83495104e-01 -4.14559305e-01 3.22398692e-01
1.11250544e+00 2.37098381e-01 9.42983747e-01 4.70869571e-01
1.70712039e-01 -7.95855165e-01 -3.82298738e-01 7.29257941e-01
8.37772489e-01 -1.39319789e+00 1.55211076e-01 -1.06429589e+00
-5.18996656e-01 1.01217890e+00 1.04258537e+00 1.05561040e-01
7.88337111e-01 3.47500533e-01 -3.24167103e-01 -4.87333424e-02
-1.01141226e+00 -3.81738007e-01 6.27649546e-01 3.90188485e-01
6.36145175e-01 -1.54109657e-01 -4.68748689e-01 1.30095470e+00
-2.29651749e-01 -1.46178499e-01 1.72209721e-02 6.64144516e-01
-6.38621673e-02 -1.44882727e+00 1.57307729e-01 4.29910779e-01
-4.34553951e-01 -4.57106382e-01 -5.45054376e-01 6.96685255e-01
3.55576992e-01 8.56460631e-01 1.29660800e-01 -7.00169146e-01
4.49349761e-01 -4.12912667e-02 2.08695769e-01 -8.13821793e-01
-6.71476901e-01 -1.75901636e-01 5.06567299e-01 -3.45942140e-01
-4.13116217e-01 -4.71958250e-01 -1.29484117e+00 -2.24111811e-03
-7.79128850e-01 3.40054423e-01 9.34643596e-02 1.44207311e+00
4.00329769e-01 6.69302583e-01 6.86185181e-01 -5.49136102e-01
-1.41930386e-01 -1.22822714e+00 -1.97090715e-01 4.07924294e-01
1.39516369e-01 -6.72662497e-01 1.24141343e-01 3.15677524e-02] | [9.235584259033203, 8.600212097167969] |
936a8817-de02-4157-b35e-1c4b3022c52c | assessing-the-effectiveness-of-syntactic | 2106.06110 | null | https://arxiv.org/abs/2106.06110v1 | https://arxiv.org/pdf/2106.06110v1.pdf | Assessing the Effectiveness of Syntactic Structure to Learn Code Edit Representations | In recent times, it has been shown that one can use code as data to aid various applications such as automatic commit message generation, automatic generation of pull request descriptions and automatic program repair. Take for instance the problem of commit message generation. Treating source code as a sequence of tokens, state of the art techniques generate commit messages using neural machine translation models. However, they tend to ignore the syntactic structure of programming languages. Previous work, i.e., code2seq has used structural information from Abstract Syntax Tree (AST) to represent source code and they use it to automatically generate method names. In this paper, we elaborate upon this state of the art approach and modify it to represent source code edits. We determine the effect of using such syntactic structure for the problem of classifying code edits. Inspired by the code2seq approach, we evaluate how using structural information from AST, i.e., paths between AST leaf nodes can help with the task of code edit classification on two datasets of fine-grained syntactic edits. Our experiments shows that attempts of adding syntactic structure does not result in any improvements over less sophisticated methods. The results suggest that techniques such as code2seq, while promising, have a long way to go before they can be generically applied to learning code edit representations. We hope that these results will benefit other researchers and inspire them to work further on this problem. | ['Rahul Kumar', 'Ranjita Bhagwan', 'Sonu Mehta', 'Syed Arbaaz Qureshi'] | 2021-06-11 | null | null | null | null | ['program-repair', 'program-repair'] | ['computer-code', 'reasoning'] | [ 4.22486901e-01 4.41400498e-01 -1.84784800e-01 -4.62521493e-01
-6.56658471e-01 -4.74343568e-01 5.77541053e-01 4.83000368e-01
-1.02604575e-01 4.22668785e-01 3.59288245e-01 -7.66487658e-01
3.90795529e-01 -8.88949096e-01 -8.84239435e-01 -9.60014984e-02
-1.13916457e-01 1.33719757e-01 4.91563290e-01 -3.93304944e-01
7.23911405e-01 -2.12466381e-02 -1.69655252e+00 7.22098172e-01
9.63570118e-01 1.38353974e-01 2.54533619e-01 7.09840477e-01
-7.95826852e-01 1.08581471e+00 -7.09573507e-01 -5.19522130e-01
-5.94920292e-02 -6.29783988e-01 -1.28342223e+00 -4.89823848e-01
2.51628697e-01 -6.57825246e-02 1.61077544e-01 1.05801988e+00
1.63006738e-01 -1.95001364e-01 4.41065013e-01 -1.07573295e+00
-7.00837731e-01 1.27544093e+00 -3.62346411e-01 1.19377472e-01
5.40262580e-01 6.63593933e-02 1.06979334e+00 -6.82387471e-01
9.14572477e-01 1.13486433e+00 9.55450833e-01 7.24408388e-01
-1.48505020e+00 -2.31281161e-01 -3.15521181e-01 1.15449965e-01
-9.88395989e-01 -3.28479499e-01 6.63367808e-01 -8.58909786e-01
1.47216260e+00 3.51000458e-01 1.53679594e-01 9.92358983e-01
4.35668439e-01 8.10889065e-01 8.53051007e-01 -7.37767279e-01
7.69343451e-02 2.45919272e-01 3.56863350e-01 9.89823520e-01
1.18299358e-01 4.97753546e-02 -1.64337665e-01 -3.94785255e-01
7.43041858e-02 -3.33641559e-01 -2.09979698e-01 -1.64546415e-01
-1.12278306e+00 1.21215451e+00 2.85858840e-01 5.89137197e-01
1.09321274e-01 6.50595725e-01 9.32709932e-01 6.93697512e-01
3.06106150e-01 8.45825493e-01 -4.76788193e-01 -6.28081143e-01
-1.11353588e+00 3.87801915e-01 1.13302732e+00 1.03763151e+00
9.56412077e-01 -2.67455745e-02 -1.66407511e-01 8.59237731e-01
1.31675050e-01 -1.23455495e-01 7.44416595e-01 -9.59772885e-01
5.65081716e-01 8.07341754e-01 -2.03647256e-01 -9.73442078e-01
-1.92246914e-01 1.53547466e-01 -1.71876401e-01 3.66540313e-01
3.53124410e-01 -9.39061195e-02 -4.86448795e-01 1.49331307e+00
-5.20333461e-02 -3.89014304e-01 -1.77249417e-01 3.90316010e-01
5.88722229e-01 5.41843832e-01 -3.06767017e-01 2.17728823e-01
1.04007971e+00 -9.96442139e-01 -2.20064700e-01 -5.55681027e-02
1.35869694e+00 -9.49881315e-01 1.13476562e+00 2.89170921e-01
-1.07464731e+00 -5.11537790e-01 -9.61512387e-01 -3.24099176e-02
-4.01077181e-01 1.21546529e-01 7.82582521e-01 6.64924204e-01
-1.23465002e+00 8.66951108e-01 -8.99730086e-01 -6.19304895e-01
-1.09252138e-02 9.82229784e-02 -1.45300940e-01 1.21130265e-01
-9.07929003e-01 8.99718523e-01 5.25684297e-01 -3.57411891e-01
-7.68391728e-01 -5.56763470e-01 -9.97324824e-01 5.77630103e-02
3.33345115e-01 -4.81747627e-01 1.62799251e+00 -1.44954813e+00
-1.23962522e+00 8.22598755e-01 -2.08950818e-01 -5.34363210e-01
1.60924926e-01 -2.01851670e-02 -6.12616204e-02 -3.90040964e-01
1.98373422e-01 5.16051948e-01 4.21439409e-01 -9.93015885e-01
-4.98145103e-01 9.83551797e-03 4.45392609e-01 -5.36299527e-01
-3.06055725e-01 6.68474376e-01 3.05610672e-02 -8.83562446e-01
-4.41847950e-01 -1.11509204e+00 -1.87875345e-01 -1.49358556e-01
-5.21190763e-01 -4.49727565e-01 4.38960969e-01 -7.66893864e-01
1.36053562e+00 -2.06508946e+00 4.40898448e-01 1.10554591e-01
6.65007681e-02 1.56382367e-01 -2.80716211e-01 9.60587680e-01
-1.59692273e-01 6.49814010e-01 -5.93175173e-01 -9.62586924e-02
2.65578121e-01 2.38884211e-01 -4.37061578e-01 1.48971481e-02
3.62952888e-01 8.14863145e-01 -9.32893813e-01 -3.22055787e-01
-4.94488925e-01 2.20580548e-01 -1.05582428e+00 2.26785883e-01
-6.46794081e-01 2.60291295e-03 -4.76211071e-01 1.98477894e-01
2.13834554e-01 -3.16214226e-02 2.80026682e-02 3.76600772e-01
-4.10769612e-01 7.49745011e-01 -8.85918617e-01 1.81564820e+00
-6.97533369e-01 9.95277882e-01 -2.51511693e-01 -9.96601880e-01
1.00317621e+00 1.90457478e-01 9.71726328e-02 -5.29814839e-01
-1.75760850e-01 4.19659853e-01 2.34679416e-01 -8.60410571e-01
7.35665679e-01 1.50526926e-01 -3.47924471e-01 8.43670726e-01
-1.76236078e-01 -1.38160273e-01 4.47878957e-01 3.55087966e-01
1.54011905e+00 5.61650872e-01 2.62895137e-01 -1.76835805e-01
5.88484526e-01 2.94493228e-01 1.75530344e-01 8.08734179e-01
3.64381939e-01 5.01314223e-01 8.46103609e-01 -4.19214100e-01
-1.21065462e+00 -6.31474614e-01 4.64407690e-02 1.35060990e+00
-4.87419754e-01 -1.02536213e+00 -1.12760937e+00 -9.13564742e-01
-1.03348359e-01 8.78212631e-01 -6.14338994e-01 -2.45304599e-01
-1.01969421e+00 -5.20078659e-01 1.09342003e+00 6.20287240e-01
-1.24995010e-02 -1.26053619e+00 -8.27364624e-01 3.81854713e-01
-1.57319948e-01 -5.44290841e-01 -5.44574320e-01 2.06175685e-01
-8.69134784e-01 -9.82009292e-01 -3.94770861e-01 -7.91980803e-01
6.13380373e-01 -2.40413919e-01 1.35321224e+00 7.49179363e-01
-3.80494088e-01 2.89857894e-01 -7.47897506e-01 -2.35923737e-01
-1.37624955e+00 4.20743048e-01 -4.87812638e-01 -4.61625069e-01
3.26669335e-01 -6.22076571e-01 2.96296980e-02 -2.82684900e-03
-1.14825892e+00 1.94481581e-01 5.56829035e-01 6.98566318e-01
-1.19864583e-01 -2.78817892e-01 4.15679753e-01 -1.51597810e+00
8.78127635e-01 -6.23219967e-01 -4.61937249e-01 1.26410827e-01
-6.65845037e-01 7.12627709e-01 8.08315098e-01 -5.45777753e-02
-1.04437816e+00 2.04112865e-02 -4.15691793e-01 1.68686181e-01
-1.47803262e-01 1.03436887e+00 3.46465319e-01 -9.40316841e-02
1.04204834e+00 2.83616573e-01 -4.16195802e-02 -6.18507206e-01
2.84061700e-01 8.15035641e-01 1.93680346e-01 -9.40363884e-01
7.04876244e-01 -8.82928893e-02 -2.28751600e-01 -5.12239814e-01
-1.96140885e-01 -1.56295240e-01 -5.38562655e-01 2.11315945e-01
7.55788147e-01 -3.56721401e-01 -1.34895861e-01 2.38548473e-01
-1.62416136e+00 -5.69544792e-01 -1.62568033e-01 -1.43910229e-01
-8.09535384e-01 6.69548750e-01 -7.67947435e-01 -3.64291579e-01
-4.10050116e-02 -1.39820457e+00 1.04035246e+00 -1.52566001e-01
-6.86601460e-01 -9.97835517e-01 3.86864334e-01 1.22929953e-01
8.75495076e-01 3.04019034e-01 1.50816846e+00 -6.96590126e-01
-8.06033432e-01 8.26913491e-02 4.00749855e-02 2.78030992e-01
5.90493381e-02 3.87644261e-01 -7.18107462e-01 -1.02965072e-01
-2.06875786e-01 -2.73835778e-01 8.37618887e-01 -3.11750472e-01
1.19789147e+00 -6.63963616e-01 -3.43968034e-01 4.22440052e-01
1.46112502e+00 5.51145896e-02 6.97596490e-01 4.02050346e-01
6.44532382e-01 7.89210320e-01 2.60057211e-01 3.51423174e-01
5.11012197e-01 9.18230772e-01 4.32038277e-01 3.90953690e-01
-2.88578004e-01 -2.03621462e-01 8.55644166e-01 9.99436319e-01
1.23230383e-01 3.14731412e-02 -1.30104053e+00 6.62764788e-01
-1.66575074e+00 -9.99047458e-01 -6.31631076e-01 2.04838252e+00
1.32620728e+00 7.66784325e-02 5.84637150e-02 3.75642329e-02
4.73273516e-01 -1.00601852e-01 5.05733676e-02 -1.17676568e+00
4.90744114e-01 3.20576280e-01 3.23290080e-01 5.09828568e-01
-6.76237643e-01 6.78637326e-01 5.62343597e+00 5.60519636e-01
-1.16544282e+00 2.30533779e-01 1.10605985e-01 3.94969255e-01
-6.58407509e-01 4.87254977e-01 -7.25453734e-01 6.98895454e-01
1.29234874e+00 -3.13874662e-01 6.89958036e-01 9.00196612e-01
1.35928812e-02 -9.25478563e-02 -1.54713082e+00 1.81669787e-01
2.07234904e-01 -1.41815019e+00 -1.43088520e-01 -1.52468085e-01
7.45822012e-01 3.01699996e-01 -4.79916245e-01 6.07723713e-01
6.55359626e-01 -8.71100783e-01 7.20649302e-01 5.87235451e-01
3.61792684e-01 -4.06227022e-01 7.45629311e-01 4.57485944e-01
-1.00027859e+00 -6.57653809e-02 -3.20284128e-01 -7.41558149e-02
-7.76142180e-02 4.42632467e-01 -1.06248057e+00 5.46747327e-01
4.29616809e-01 7.06691027e-01 -1.17760074e+00 1.21773779e+00
-1.94503024e-01 6.58403158e-01 2.22999752e-02 -3.02345842e-01
2.34798208e-01 1.93520665e-01 5.12065351e-01 1.74260032e+00
4.88236725e-01 -5.53104520e-01 9.83473733e-02 1.34549463e+00
-2.11112387e-02 1.89090505e-01 -9.45235729e-01 -2.86298215e-01
6.55554682e-02 9.00000155e-01 -6.92581534e-01 -3.84161741e-01
-6.33756578e-01 8.99453044e-01 3.87670785e-01 6.54479414e-02
-8.63601446e-01 -9.44948852e-01 2.92903066e-01 1.64072931e-01
3.97063673e-01 -1.48181260e-01 -2.55788803e-01 -1.09757376e+00
2.47322053e-01 -1.15050423e+00 8.18942934e-02 -8.70569587e-01
-9.28202927e-01 6.47220552e-01 3.27869244e-02 -8.76675069e-01
-6.78921759e-01 -4.68192905e-01 -1.01929939e+00 6.79317474e-01
-1.32876420e+00 -8.19803715e-01 -7.93101862e-02 -5.20813279e-02
8.47126126e-01 -7.99013451e-02 8.79176080e-01 3.30929041e-01
-2.13651225e-01 7.18349695e-01 6.76483214e-02 4.45674837e-01
6.29968345e-01 -1.53687882e+00 8.67047310e-01 1.00265574e+00
2.79569864e-01 9.52037930e-01 8.68484199e-01 -6.72184587e-01
-1.50536299e+00 -1.23425376e+00 1.23353863e+00 -8.65518868e-01
7.51030207e-01 -3.33859384e-01 -1.09878600e+00 7.95684099e-01
5.01596808e-01 -3.79005194e-01 3.54834527e-01 1.13860629e-02
-5.49461365e-01 3.27666342e-01 -6.74691379e-01 4.28039610e-01
8.54565740e-01 -6.32393420e-01 -7.79415309e-01 2.83325553e-01
8.41782868e-01 -4.02800649e-01 -8.32978189e-01 6.92401826e-02
2.77772635e-01 -9.94299591e-01 4.58889157e-01 -6.83571160e-01
1.08977687e+00 -1.92490265e-01 1.22926747e-02 -1.45555305e+00
7.91438892e-02 -5.62400877e-01 8.07336196e-02 1.63602602e+00
6.10417783e-01 -4.00142640e-01 5.03621995e-01 3.88965309e-01
-5.91514587e-01 -5.88366210e-01 -5.93513489e-01 -7.96701074e-01
3.73410553e-01 -4.22196090e-01 4.72554117e-01 9.41657960e-01
4.18713450e-01 2.54171520e-01 -9.31268185e-02 -3.94822150e-01
5.29902168e-02 2.33085737e-01 9.66133237e-01 -1.29518116e+00
-7.01262891e-01 -7.23992348e-01 -3.07714999e-01 -7.14041889e-01
5.50280809e-01 -1.57630217e+00 3.40452909e-01 -1.42752302e+00
2.13839144e-01 -6.42073989e-01 3.45886290e-01 7.97654629e-01
-4.96737510e-02 -4.83146422e-02 2.95136034e-01 1.10940047e-01
-2.56392568e-01 7.29388818e-02 5.78921258e-01 -4.15843517e-01
1.13333523e-01 4.56579961e-02 -6.31804824e-01 6.07731700e-01
8.29627454e-01 -1.10870624e+00 -2.33555317e-01 -6.35056138e-01
7.96016514e-01 1.50077626e-01 2.47097865e-01 -8.59689355e-01
1.26623362e-01 -1.00656204e-01 -4.36568290e-01 1.26999572e-01
-2.37922534e-01 -5.98403335e-01 2.17890605e-01 7.61209965e-01
-7.20212579e-01 4.48181808e-01 1.69373900e-01 3.25379401e-01
-2.66332835e-01 -1.18676460e+00 4.46537703e-01 -5.22507429e-01
-6.58081830e-01 -3.07712078e-01 -8.42728555e-01 3.05345535e-01
8.25603664e-01 -1.50373369e-01 -3.46601725e-01 -6.66692182e-02
-3.77688825e-01 -1.59679532e-01 9.41312551e-01 8.09199452e-01
2.63177246e-01 -1.05130446e+00 -7.22663224e-01 2.53106683e-01
3.18232089e-01 -4.91386056e-01 -3.77462476e-01 9.62406456e-01
-7.12864697e-01 4.19234693e-01 -1.86740339e-01 -5.11854172e-01
-1.32075655e+00 5.73501050e-01 2.56032586e-01 -1.81166604e-01
-5.15800655e-01 6.59639120e-01 -2.21910432e-01 -7.35952914e-01
2.88112890e-02 -6.77367032e-01 -4.48006429e-02 -1.07613087e-01
4.06085670e-01 1.81470543e-01 3.46641839e-01 -2.55500078e-01
-1.73931524e-01 4.29820806e-01 -1.26539439e-01 2.04926938e-01
1.39016473e+00 3.60519320e-01 -6.77166164e-01 5.92791796e-01
1.32740808e+00 4.28738803e-01 -6.22589648e-01 1.30776912e-01
6.78980410e-01 -5.12771249e-01 -5.04173338e-01 -7.33518004e-01
-8.83593798e-01 1.16126490e+00 2.39931732e-01 5.95841408e-01
6.49368227e-01 1.25688285e-01 6.91962898e-01 6.60302699e-01
6.42786503e-01 -7.79177904e-01 6.77792206e-02 8.06780398e-01
7.18117774e-01 -1.09706080e+00 -4.86016363e-01 -9.31589007e-02
-3.68845105e-01 1.53192329e+00 6.58051729e-01 -5.34859449e-02
1.03115983e-01 5.69444716e-01 -3.37187707e-01 -1.72426090e-01
-9.28773284e-01 6.65980354e-02 -3.13944258e-02 5.13998270e-01
1.13323438e+00 -1.67100206e-01 -5.31369746e-01 2.53098607e-01
-1.89235881e-01 2.05522060e-01 1.35939789e+00 1.13630319e+00
-5.26311159e-01 -1.81347144e+00 -2.21010804e-01 7.49647260e-01
-6.34133816e-01 -4.49499130e-01 -6.33437574e-01 6.53067112e-01
7.84389004e-02 5.92192173e-01 -1.62231177e-01 -3.45363706e-01
1.89182922e-01 3.55104834e-01 4.06998992e-01 -1.34672868e+00
-1.13439965e+00 -7.83335865e-01 3.47868204e-01 -5.38619816e-01
-2.34938934e-01 -7.47478426e-01 -1.39074302e+00 -3.97969007e-01
-2.60903507e-01 3.69635582e-01 6.50425792e-01 7.17064977e-01
5.67800462e-01 5.15982389e-01 3.15141618e-01 -7.15480924e-01
-7.33484030e-01 -7.49081790e-01 1.57281637e-01 3.88186842e-01
3.24805260e-01 -1.99734464e-01 -2.17309251e-01 5.00832260e-01] | [7.697798728942871, 7.834532737731934] |
62d8b63a-d95f-4015-9c52-dc4168267657 | invariant-priors-for-bayesian-quadrature | 2112.01578 | null | https://arxiv.org/abs/2112.01578v1 | https://arxiv.org/pdf/2112.01578v1.pdf | Invariant Priors for Bayesian Quadrature | Bayesian quadrature (BQ) is a model-based numerical integration method that is able to increase sample efficiency by encoding and leveraging known structure of the integration task at hand. In this paper, we explore priors that encode invariance of the integrand under a set of bijective transformations in the input domain, in particular some unitary transformations, such as rotations, axis-flips, or point symmetries. We show initial results on superior performance in comparison to standard Bayesian quadrature on several synthetic and one real world application. | ['Maren Mahsereci', 'Javier Gonzalez', 'Masha Naslidnyk'] | 2021-12-02 | null | null | null | null | ['numerical-integration'] | ['miscellaneous'] | [ 1.68016069e-02 -8.76059383e-02 -7.32718334e-02 -4.00607109e-01
-8.53113353e-01 -4.66687173e-01 1.00518978e+00 -2.49742597e-01
-3.65882128e-01 1.19726598e+00 7.16027468e-02 -2.17876464e-01
-3.65067542e-01 -8.70223284e-01 -5.29943466e-01 -6.02344871e-01
-1.48539960e-01 7.19862938e-01 2.67862529e-01 -1.52242184e-01
5.44982374e-01 7.99620330e-01 -1.00741208e+00 -2.45050445e-01
7.34271944e-01 1.01426852e+00 -5.39722383e-01 6.95979416e-01
3.29762787e-01 6.25967205e-01 -3.41487586e-01 -2.88650870e-01
3.16391140e-01 -3.10122550e-01 -7.16840565e-01 -3.75455469e-01
2.97923356e-01 -1.54203326e-01 -3.53841990e-01 1.32733512e+00
4.59587842e-01 3.84531587e-01 1.28541028e+00 -9.02598679e-01
-2.27251604e-01 8.95092040e-02 -4.88142818e-01 4.17812824e-01
5.72175086e-01 1.07138738e-01 5.44602394e-01 -9.75841641e-01
8.51234853e-01 1.33152366e+00 1.00219095e+00 -8.44567455e-03
-1.70536816e+00 -4.56172287e-01 -7.68158913e-01 -1.38898134e-01
-1.53072643e+00 -6.49115860e-01 6.52518988e-01 -6.27781391e-01
1.00599945e+00 2.77994096e-01 5.14983416e-01 6.39527023e-01
9.77893174e-01 1.54813007e-01 1.32414472e+00 -5.01567960e-01
7.10201681e-01 -1.82193235e-01 8.03424120e-02 4.64408755e-01
2.83544004e-01 4.15542722e-01 -3.29280108e-01 -6.24071777e-01
9.66831565e-01 -6.37131929e-01 -2.42916003e-01 -8.16904545e-01
-1.10967827e+00 1.18890941e+00 2.99609601e-01 -6.13308884e-02
-4.43359166e-01 5.25582075e-01 2.50157356e-01 2.35578604e-02
5.91256022e-01 4.61683661e-01 -2.91894376e-01 -2.37524584e-01
-1.10739601e+00 6.76897168e-01 9.12403286e-01 8.42809081e-01
7.17704535e-01 1.40703276e-01 -3.06229472e-01 4.64899331e-01
6.57000422e-01 8.34541082e-01 3.05494875e-01 -1.54051995e+00
2.32605368e-01 -1.35256514e-01 1.81010127e-01 -8.69103491e-01
-3.70710343e-01 -2.50044495e-01 -6.65847361e-01 6.41386032e-01
4.56230909e-01 -2.67080843e-01 -8.32604647e-01 1.54912472e+00
4.65872318e-01 -1.04646228e-01 -4.06193584e-02 5.56838870e-01
2.77892858e-01 5.76279044e-01 -1.83586672e-01 -3.49684000e-01
1.20688546e+00 -3.41398180e-01 -1.01658559e+00 1.74545139e-01
2.43150800e-01 -1.12043047e+00 2.24903077e-01 6.07094765e-01
-1.19820154e+00 -4.48163450e-01 -1.27091074e+00 -4.58366051e-02
-4.37361628e-01 -2.46419191e-01 8.04140389e-01 9.68876541e-01
-8.92603397e-01 7.46053874e-01 -1.00543261e+00 2.88624227e-01
3.43473643e-01 3.93512875e-01 -3.32962215e-01 -1.67896360e-01
-1.18584085e+00 1.15964150e+00 1.13232426e-01 3.22420821e-02
-5.44306278e-01 -8.40719521e-01 -1.14529383e+00 -1.27092347e-01
-2.06756908e-02 -7.78784752e-01 1.38355529e+00 -2.34506220e-01
-1.77240741e+00 4.01748359e-01 -7.14304075e-02 -6.00487709e-01
7.71923184e-01 6.67304844e-02 -1.35855392e-01 1.38551936e-01
6.02110736e-02 8.19217861e-01 1.09788668e+00 -6.32275760e-01
7.08365217e-02 -2.97424376e-01 -1.27577156e-01 -1.38242960e-01
5.00542223e-01 7.96994716e-02 1.61296219e-01 -6.69667959e-01
3.13529760e-01 -8.01231682e-01 -3.04265171e-01 8.39398056e-02
-1.86615855e-01 -2.78242044e-02 6.48400247e-01 -7.17225730e-01
8.29496503e-01 -1.96798444e+00 1.37939125e-01 6.01505101e-01
-1.80487603e-01 1.59135684e-02 4.12515253e-01 5.94470263e-01
-1.40527427e-01 -6.48797676e-02 -6.17229581e-01 -5.38159870e-02
2.45538071e-01 2.47301087e-01 -4.78112698e-01 9.38264668e-01
3.12295169e-01 7.60948420e-01 -8.40698600e-01 -6.78945661e-01
3.18493754e-01 6.76361144e-01 -7.61134326e-01 -1.55748278e-01
-1.56508729e-01 4.34168100e-01 -2.55725354e-01 2.76155263e-01
9.61625636e-01 1.99567646e-01 2.86734551e-02 -2.69229174e-01
-1.46456137e-01 4.62392122e-01 -1.68376112e+00 1.55089915e+00
-3.55432153e-01 6.72037959e-01 7.94428587e-02 -8.69738102e-01
7.29158878e-01 5.16973794e-01 3.75232220e-01 -5.83824635e-01
1.18121035e-01 2.94365525e-01 1.30220875e-01 -1.33109763e-01
4.03241932e-01 -5.90886116e-01 4.85177292e-03 3.03155333e-01
4.59539622e-01 -1.10025311e+00 3.98745477e-01 6.47443905e-02
8.19412112e-01 3.39608312e-01 8.85034382e-01 -8.42846453e-01
4.04476911e-01 -2.02605844e-01 4.43919867e-01 7.59517789e-01
-1.37329131e-01 7.71180034e-01 5.31217933e-01 -2.45371848e-01
-9.23316717e-01 -1.15070331e+00 -8.58508050e-01 2.24919453e-01
-2.45872498e-01 -1.47846147e-01 -5.41044831e-01 -1.76291212e-01
1.56464517e-01 7.40222812e-01 -5.72425604e-01 -1.55939192e-01
-5.53504705e-01 -9.59456086e-01 5.40451050e-01 3.72409850e-01
6.38965666e-01 -8.90900373e-01 -7.78379738e-01 4.86989886e-01
1.68167517e-01 -8.73155415e-01 -2.71642804e-01 3.08540523e-01
-1.14184558e+00 -1.01229131e+00 -7.19995975e-01 -9.74726006e-02
3.13962668e-01 -4.08022612e-01 9.82556701e-01 -7.88536906e-01
-3.84780586e-01 3.51323545e-01 2.89935708e-01 -2.95689970e-01
-1.94670618e-01 -3.61019015e-01 -2.05265526e-02 -4.21417952e-01
-3.10357690e-01 -5.08612990e-01 -2.73446470e-01 -8.78037768e-04
-7.85734653e-01 -4.27413583e-01 6.57072514e-02 8.04364026e-01
3.63574624e-01 9.10874531e-02 2.23425448e-01 -5.68984807e-01
8.55826080e-01 -2.75550097e-01 -9.22208488e-01 -6.66806102e-02
-2.67147660e-01 4.17698979e-01 6.44900948e-02 -1.82766438e-01
-1.25600684e+00 -4.12243791e-02 -2.16654420e-01 4.88493294e-02
2.17791364e-01 6.53300762e-01 8.52556303e-02 -6.15989506e-01
8.41773152e-01 -2.52604067e-01 -1.31136654e-02 -3.17029417e-01
3.87278497e-01 3.46186727e-01 3.91417354e-01 -1.07277524e+00
3.79926741e-01 6.06204867e-01 6.43624485e-01 -8.11496496e-01
-3.87270212e-01 -2.10596144e-01 -6.05474830e-01 -1.08604822e-02
7.07260251e-01 -5.48300862e-01 -4.45565552e-01 4.50083256e-01
-1.23379517e+00 -2.51295716e-01 -6.60618365e-01 8.18382204e-01
-9.26944792e-01 3.85092437e-01 -4.89529073e-01 -9.39602137e-01
-3.49711999e-02 -1.26241839e+00 1.08738637e+00 2.27061763e-01
-4.19221520e-01 -1.18328071e+00 6.04313612e-01 -2.46976495e-01
6.97296798e-01 2.53708452e-01 7.42640018e-01 -3.54105979e-01
-4.97277975e-01 -5.15800044e-02 -1.72037408e-01 4.43478197e-01
-2.37606585e-01 3.78651470e-01 -8.64576638e-01 -1.96825862e-01
4.53589052e-01 -1.73105881e-01 6.26680851e-01 5.62871456e-01
5.71552098e-01 -1.09323241e-01 -3.24407071e-01 6.63422346e-01
1.25158334e+00 2.73925453e-01 6.98384404e-01 -5.94547801e-02
-8.63626599e-02 2.05120936e-01 4.69736665e-01 3.01271468e-01
-2.27458507e-01 8.51581991e-01 8.33248049e-02 6.86537921e-02
-1.25144377e-01 9.62709188e-02 -3.74473445e-02 5.54890811e-01
-1.39992982e-01 3.22793841e-01 -9.27288115e-01 3.18760037e-01
-1.52738202e+00 -7.20956206e-01 -1.74304359e-02 2.12053347e+00
9.83575940e-01 2.73165703e-01 -2.73846924e-01 2.62730271e-01
3.52155000e-01 7.85792619e-02 -1.41367868e-01 -5.64006269e-01
2.09176615e-01 1.02090645e+00 5.58894992e-01 9.62868392e-01
-1.14768684e+00 3.19892287e-01 8.20623589e+00 9.66798306e-01
-8.75157714e-01 1.08148560e-01 2.36186162e-01 4.68679070e-01
-1.79317936e-01 1.62328809e-01 -6.33515596e-01 3.87344867e-01
7.27584243e-01 -1.00540988e-01 3.72064024e-01 5.35338402e-01
-1.58605859e-01 -4.10948932e-01 -9.75135207e-01 5.69765270e-01
-5.69313280e-02 -1.26850808e+00 -2.59826094e-01 2.30647072e-01
9.97501254e-01 -4.60011549e-02 -3.26318108e-02 2.71329701e-01
4.51243728e-01 -9.48396087e-01 6.03856921e-01 1.05730915e+00
5.96709073e-01 -6.66350722e-01 8.90511096e-01 2.75548548e-01
-7.51864374e-01 4.29926485e-01 -8.88049677e-02 -2.35205203e-01
3.48647267e-01 7.54647195e-01 -5.79046726e-01 6.21271789e-01
3.18367481e-01 2.74710506e-01 -1.86110333e-01 1.37901878e+00
-3.71120572e-01 5.11534810e-01 -7.45336831e-01 2.67960876e-01
1.54859170e-01 -5.99584639e-01 7.38556504e-01 1.28403080e+00
5.19658744e-01 2.16500804e-01 -2.91629463e-01 8.88245463e-01
3.88970613e-01 -2.66712517e-01 -8.59528124e-01 3.46892208e-01
2.72990316e-01 9.04199719e-01 -6.96510315e-01 -4.03512448e-01
4.86476086e-02 2.23602220e-01 8.23833048e-03 7.40317941e-01
-7.57589340e-01 -7.19434738e-01 5.38692534e-01 -3.14311273e-02
5.54285049e-01 -6.20278418e-01 -3.69083673e-01 -1.05971420e+00
-1.61940098e-01 -7.20604181e-01 2.42855251e-01 -1.03453612e+00
-1.00033498e+00 1.01355882e-02 7.11092889e-01 -8.20952952e-01
-7.20991254e-01 -7.83023715e-01 -5.05952775e-01 1.05682933e+00
-1.19861627e+00 -7.23405421e-01 2.38858014e-01 2.67677426e-01
-1.21809892e-01 1.04026861e-01 8.24586511e-01 8.56725276e-02
1.95047587e-01 7.18227103e-02 2.61876404e-01 -7.07090423e-02
4.41879541e-01 -1.18372679e+00 1.85597017e-01 7.81730354e-01
2.08693683e-01 1.02024055e+00 1.00140667e+00 -5.71744263e-01
-1.11740482e+00 -3.88626695e-01 8.72489691e-01 -1.55069426e-01
7.37846673e-01 -2.44017959e-01 -7.41113186e-01 8.27802360e-01
2.41460085e-01 3.88805509e-01 3.06651622e-01 -1.65151536e-01
-3.11007738e-01 7.93208461e-03 -1.60968971e+00 2.51754135e-01
5.27112722e-01 -6.47538781e-01 -9.29605305e-01 2.41593316e-01
1.85355827e-01 -7.10925281e-01 -1.08706641e+00 9.08140242e-01
6.81165218e-01 -9.17012870e-01 1.46658349e+00 -5.65510631e-01
7.91015700e-02 -3.25799882e-01 -3.48593563e-01 -1.25623858e+00
-9.97471586e-02 -8.48248959e-01 -1.85260698e-01 9.36874926e-01
-7.29671344e-02 -1.15824652e+00 2.16512814e-01 4.37998474e-01
6.18474819e-02 -5.30560851e-01 -1.60219741e+00 -7.91670442e-01
4.48680609e-01 -4.81942296e-01 5.01346588e-01 7.85323739e-01
2.12334871e-01 1.77516758e-01 2.40979604e-02 -1.27770364e-01
8.52734327e-01 -5.64845987e-02 5.51821589e-01 -1.07822943e+00
-2.37076506e-01 -4.90725994e-01 -6.30843997e-01 -9.53416705e-01
-8.00553933e-02 -7.00537205e-01 3.23609412e-02 -1.20204377e+00
-1.51288345e-01 -4.01293695e-01 7.88640976e-02 1.87579170e-03
1.78284287e-01 3.55676085e-01 -8.55606943e-02 -5.86565733e-02
-2.65621394e-01 7.55421162e-01 1.33607078e+00 -7.05871508e-02
2.18981847e-01 8.34025890e-02 -3.79080623e-02 9.14083838e-01
5.95300078e-01 -7.42775202e-01 -1.84296638e-01 5.37103266e-02
5.32853067e-01 4.39233571e-01 4.31634396e-01 -1.25716662e+00
1.95755869e-01 -8.99965614e-02 3.59459728e-01 -5.81257701e-01
6.67163670e-01 -5.60873687e-01 3.78413141e-01 5.78119934e-01
-6.19365536e-02 -1.18539736e-01 3.89767617e-01 3.87870669e-01
-3.77432913e-01 -8.18348289e-01 9.38492239e-01 1.00514308e-01
-1.50840476e-01 -1.68316141e-01 -5.20935118e-01 3.01294029e-01
7.09209025e-01 5.87608926e-02 -1.14134073e-01 -4.14144278e-01
-7.92625070e-01 -2.62369514e-01 3.26311767e-01 -4.33845371e-01
2.49958068e-01 -1.57507324e+00 -4.78846729e-01 5.10141671e-01
-3.53710175e-01 -1.89110875e-01 -2.58860350e-01 9.83896494e-01
-8.65898669e-01 7.41809309e-01 -3.88570249e-01 -5.89828730e-01
-8.58823836e-01 2.12021351e-01 5.91531873e-01 -4.68423426e-01
-2.76062787e-01 6.09284699e-01 -2.24828079e-01 -7.07676291e-01
-1.16343692e-01 -7.59917796e-01 1.18802674e-01 -4.79241610e-02
1.80265903e-01 6.15045071e-01 1.91220835e-01 -5.20476043e-01
-3.15190375e-01 7.11944580e-01 5.16798437e-01 -8.21519852e-01
8.43452215e-01 3.37114424e-01 -1.61396399e-01 4.46844757e-01
1.18398237e+00 2.11960059e-02 -1.27776444e+00 -1.82389673e-02
-1.50180578e-01 -5.02120793e-01 7.33311996e-02 -5.32326639e-01
-6.08445108e-01 1.12511945e+00 3.67796689e-01 2.37290695e-01
5.94019473e-01 -4.45807546e-01 2.06221640e-01 4.33583528e-01
6.48575425e-01 -9.47817445e-01 -2.64529228e-01 6.67029202e-01
1.02503729e+00 -9.21747327e-01 5.87669253e-01 -4.64765072e-01
7.55650084e-03 1.19222450e+00 -1.25197992e-01 -5.53311169e-01
1.22930181e+00 4.50204343e-01 -2.97703266e-01 -3.55090827e-01
-4.18871820e-01 1.12918615e-01 5.15139997e-01 6.29219353e-01
3.55920732e-01 3.77393961e-02 -5.51536918e-01 7.45768845e-02
-1.77957997e-01 2.35964432e-01 2.75587052e-01 1.03557861e+00
-3.33140135e-01 -9.49457586e-01 -7.29286551e-01 2.28047237e-01
-3.11226696e-01 9.30424258e-02 1.30862653e-01 1.14321065e+00
1.27605097e-02 5.02489209e-01 1.19090624e-01 5.75895190e-01
2.71414399e-01 2.57557899e-01 1.08075142e+00 -3.54684860e-01
-2.84741491e-01 1.74587622e-01 1.55473024e-01 -5.28673172e-01
-5.00252783e-01 -1.10420597e+00 -1.12992501e+00 2.27329247e-02
-3.39042336e-01 2.49870941e-01 8.08846235e-01 1.15304863e+00
1.82461664e-02 5.63524008e-01 2.09983513e-02 -1.09241092e+00
-1.11888075e+00 -1.31388128e+00 -6.30201459e-01 2.36350615e-02
3.54838878e-01 -1.06946003e+00 -5.19175708e-01 -3.81226162e-03] | [6.812225341796875, 3.7793571949005127] |
21464efc-5eab-459a-b6a9-56c8f11631cd | matching-disparate-image-pairs-using-shape | 1811.09889 | null | http://arxiv.org/abs/1811.09889v1 | http://arxiv.org/pdf/1811.09889v1.pdf | Matching Disparate Image Pairs Using Shape-Aware ConvNets | An end-to-end trainable ConvNet architecture, that learns to harness the
power of shape representation for matching disparate image pairs, is proposed.
Disparate image pairs are deemed those that exhibit strong affine variations in
scale, viewpoint and projection parameters accompanied by the presence of
partial or complete occlusion of objects and extreme variations in ambient
illumination. Under these challenging conditions, neither local nor global
feature-based image matching methods, when used in isolation, have been
observed to be effective. The proposed correspondence determination scheme for
matching disparate images exploits high-level shape cues that are derived from
low-level local feature descriptors, thus combining the best of both worlds. A
graph-based representation for the disparate image pair is generated by
constructing an affinity matrix that embeds the distances between feature
points in two images, thus modeling the correspondence determination problem as
one of graph matching. The eigenspectrum of the affinity matrix, i.e., the
learned global shape representation, is then used to further regress the
transformation or homography that defines the correspondence between the source
image and target image. The proposed scheme is shown to yield state-of-the-art
results for both, coarse-level shape matching as well as fine point-wise
correspondence determination. | ['Suchendra M. Bhandarkar', 'Deepak Sharma', 'Abhimanyu Chopra', 'Arun CS Kumar', 'Shefali Srivastava'] | 2018-11-24 | null | null | null | null | ['matching-disparate-images'] | ['computer-vision'] | [ 2.00931832e-01 -2.61220366e-01 1.91686377e-01 -4.32501823e-01
-8.55525017e-01 -6.96740985e-01 6.32995725e-01 9.69148055e-02
-1.52872220e-01 1.74650148e-01 5.24244085e-03 2.27938712e-01
-3.62299711e-01 -7.42666483e-01 -8.28256130e-01 -8.24674726e-01
2.89127588e-01 3.62971544e-01 -9.79172587e-02 -2.78310269e-01
2.25729138e-01 9.67067599e-01 -1.50636184e+00 -8.36193040e-02
4.03842032e-01 1.16596901e+00 -4.04598974e-02 4.48527843e-01
1.94115072e-01 1.20834276e-01 -2.03475833e-01 -5.31798720e-01
8.41315389e-01 -3.23315322e-01 -2.44872317e-01 3.50188404e-01
1.39699447e+00 -1.31395325e-01 -6.75811768e-01 1.10309613e+00
5.20665228e-01 2.22106025e-01 4.92192090e-01 -1.12218070e+00
-9.45299268e-01 -2.28060186e-01 -6.19253099e-01 3.52442823e-02
5.02743125e-01 2.99275458e-01 1.14167237e+00 -1.22594059e+00
8.56421351e-01 1.10381985e+00 6.39098108e-01 8.88519287e-02
-1.46944070e+00 -4.10809159e-01 -1.99831590e-01 -7.34674186e-02
-1.38161087e+00 -5.50258696e-01 1.15927601e+00 -5.12305021e-01
8.64342093e-01 5.30847423e-02 7.31315017e-01 6.84117377e-01
4.69785124e-01 2.10496873e-01 8.54980886e-01 -3.72035265e-01
8.25107284e-03 -1.82817042e-01 -3.07801038e-01 1.01287580e+00
3.26772481e-01 4.25244033e-01 -5.78105390e-01 -3.14431608e-01
9.96251643e-01 1.21405005e-01 -3.49458903e-01 -1.21089280e+00
-1.23056018e+00 7.08010674e-01 8.17179203e-01 3.18019539e-01
-5.11947870e-01 1.30374581e-01 -9.57243331e-03 4.59067941e-01
1.13126613e-01 4.22927797e-01 -4.44790460e-02 3.20726246e-01
-5.90145230e-01 2.66624600e-01 5.58168232e-01 1.02751446e+00
9.58462358e-01 1.91977546e-01 -1.43892199e-01 6.21555507e-01
2.39584357e-01 6.79819167e-01 4.80915047e-02 -7.02532828e-01
5.74726164e-01 8.76637101e-01 1.52385691e-02 -1.75861537e+00
-3.45350176e-01 -5.57964265e-01 -7.30431855e-01 2.28514686e-01
3.93042654e-01 3.85836780e-01 -7.83396184e-01 2.00595164e+00
3.45652372e-01 2.74777293e-01 7.01959431e-03 1.05963123e+00
8.17478955e-01 1.04646578e-01 -2.85530031e-01 1.47193536e-01
1.09783769e+00 -4.88511175e-01 -2.35154837e-01 -2.21336052e-01
1.72702238e-01 -8.71503711e-01 7.70779669e-01 -2.79194772e-01
-1.28038883e+00 -5.95041931e-01 -1.38490844e+00 -1.37110606e-01
-2.78030217e-01 -4.76051271e-02 1.24191597e-01 2.87972957e-01
-1.16647613e+00 4.80756670e-01 -5.32403648e-01 -3.07567120e-01
1.69772238e-01 4.23868507e-01 -6.87457442e-01 -2.32522950e-01
-8.53548884e-01 7.96965182e-01 -1.74314290e-01 2.01534316e-01
-6.37817919e-01 -8.03429008e-01 -1.04734468e+00 2.04994977e-01
1.43136978e-01 -1.10002601e+00 4.51284856e-01 -8.03082347e-01
-9.48884308e-01 1.31316900e+00 5.45313698e-04 -7.15262741e-02
5.73140502e-01 1.74592584e-01 -6.36820570e-02 1.86867371e-01
2.31574699e-01 4.58742589e-01 1.13756514e+00 -1.46786475e+00
-2.16431215e-01 -8.70876372e-01 -7.18011782e-02 4.51457649e-01
-2.48989258e-02 -1.80676579e-01 -4.59384829e-01 -5.37597775e-01
4.09472644e-01 -9.50918376e-01 -2.63483357e-02 3.59933197e-01
-1.19348682e-01 1.54552042e-01 8.71401548e-01 -4.55073535e-01
4.57757086e-01 -2.17038631e+00 3.38351279e-01 5.74660897e-01
4.35481131e-01 1.25552146e-02 -5.96029520e-01 5.09732246e-01
-9.72533524e-02 -2.97677070e-01 -1.87300161e-01 -1.73110723e-01
4.54444624e-02 -1.25883073e-01 -1.77962914e-01 9.21683788e-01
1.86659440e-01 1.21396720e+00 -8.22242558e-01 -7.56865218e-02
3.84342432e-01 7.13604569e-01 -2.22336754e-01 4.30212140e-01
1.80765778e-01 4.97927338e-01 -3.94207180e-01 5.40724099e-01
6.43498659e-01 -1.80462345e-01 -2.82255281e-02 -5.51924706e-01
2.68270791e-01 -2.46286273e-01 -1.11084914e+00 1.92676365e+00
-4.88312602e-01 6.06141508e-01 1.62532870e-02 -9.50910985e-01
1.28153932e+00 1.90435857e-01 7.35160530e-01 -8.50930810e-01
3.29263151e-01 2.97407448e-01 -1.25028910e-02 -1.63357273e-01
4.92770895e-02 4.02048565e-02 8.42059497e-03 2.08078176e-01
9.55444574e-02 -2.96075076e-01 -1.03158250e-01 1.32405730e-02
1.05205154e+00 1.39734671e-01 4.11402196e-01 -1.97395384e-01
7.45453238e-01 -3.82156849e-01 2.72512972e-01 5.78526318e-01
-9.06407088e-02 7.54401267e-01 3.54036987e-02 -6.81896091e-01
-1.24438787e+00 -1.25414348e+00 -3.60362269e-02 6.67755723e-01
5.72943091e-01 1.96019150e-02 -5.70400536e-01 -2.79436052e-01
3.00702959e-01 -7.57566690e-02 -5.00178814e-01 -4.95876282e-01
-7.81729102e-01 -1.10167213e-01 2.30229214e-01 4.59496588e-01
5.88778496e-01 -8.25836062e-01 -6.01500034e-01 2.25487828e-01
8.41278024e-03 -1.38435256e+00 -8.05670679e-01 -1.37785748e-01
-7.39354312e-01 -1.19655955e+00 -5.40878475e-01 -9.31743741e-01
8.61578584e-01 5.96744955e-01 1.08209658e+00 1.68735445e-01
-6.92489803e-01 8.18940580e-01 2.04035401e-01 2.08503976e-01
-1.71396494e-01 -4.63881403e-01 1.11166425e-02 6.12715721e-01
9.67416689e-02 -8.53365660e-01 -7.79681206e-01 4.45769340e-01
-7.13051260e-01 -2.04111144e-01 5.62488139e-01 1.11328220e+00
7.33258724e-01 -5.67939341e-01 2.90683270e-01 -3.72184306e-01
5.51200509e-01 -1.92275736e-02 -7.36114979e-01 4.06062007e-01
-3.67945850e-01 2.01802850e-02 6.44641280e-01 -4.12567645e-01
-5.46621144e-01 2.12917909e-01 2.40148470e-01 -9.32014763e-01
-2.38062456e-01 3.22417647e-01 -2.31134087e-01 -7.29384959e-01
6.57356560e-01 3.39234293e-01 -4.63083237e-02 -1.78661972e-01
4.27238286e-01 1.11456536e-01 9.18633580e-01 -6.09618306e-01
1.21958244e+00 6.15792274e-01 3.57937127e-01 -7.40269959e-01
-5.64557791e-01 -6.94112003e-01 -9.96018469e-01 -1.01245649e-01
8.72750580e-01 -8.16053450e-01 -5.96885383e-01 4.59197372e-01
-1.06923270e+00 1.67568445e-01 -1.68278366e-01 2.35239327e-01
-8.64668310e-01 3.73467118e-01 -1.40444383e-01 -3.32182974e-01
-3.48895431e-01 -1.11181617e+00 1.37451410e+00 1.49026528e-01
3.72216366e-02 -1.12024820e+00 7.82897323e-02 3.83583724e-01
2.94572532e-01 7.70791471e-01 1.14890456e+00 -5.38415492e-01
-1.02372742e+00 -6.71844721e-01 -3.01044285e-01 -5.15417606e-02
2.15252474e-01 -2.42293090e-01 -8.84202123e-01 -8.05626392e-01
9.26775709e-02 -1.73129171e-01 3.92088741e-01 2.43843764e-01
6.37308419e-01 -9.75794718e-02 -1.04190454e-01 1.05763030e+00
1.77568018e+00 9.73879471e-02 3.64333570e-01 2.38554165e-01
1.11365020e+00 4.93394345e-01 1.97908089e-01 2.37547129e-01
1.63959801e-01 1.11648822e+00 5.43401062e-01 -3.96620214e-01
-3.32803369e-01 -3.06799322e-01 -3.87589149e-02 6.46340132e-01
-6.10720851e-02 5.10039292e-02 -9.23055887e-01 4.51914221e-01
-1.84606564e+00 -8.41403246e-01 1.31750852e-01 2.48949647e+00
2.18992338e-01 -1.47259980e-01 -2.02777311e-01 -2.51965046e-01
7.42098153e-01 2.36519888e-01 -7.92041183e-01 6.41493686e-03
-2.36644655e-01 1.11636572e-01 3.62075359e-01 3.81597638e-01
-9.33096409e-01 6.93412662e-01 5.28156042e+00 3.47763002e-01
-1.27868354e+00 -2.35422149e-01 2.51922190e-01 3.10595810e-01
-2.97739297e-01 1.04405664e-01 -2.78840333e-01 5.00026084e-02
2.87297666e-01 -1.62160188e-01 5.00445843e-01 5.73268175e-01
-3.14224772e-02 4.80610967e-01 -1.28018177e+00 1.29373670e+00
4.16129172e-01 -1.33605540e+00 2.27715120e-01 2.11873934e-01
8.97261083e-01 -6.06591441e-02 4.17233855e-01 -2.84282506e-01
1.61248483e-02 -8.82664859e-01 5.47993362e-01 5.40277362e-01
7.59178817e-01 -5.92050016e-01 6.36848748e-01 8.59723911e-02
-1.44636881e+00 4.69972054e-03 -4.70822275e-01 4.03221339e-01
-1.76320225e-01 9.68475342e-02 -5.21746635e-01 6.55488372e-01
5.00843465e-01 7.26365507e-01 -6.64165020e-01 9.63236094e-01
1.62979752e-01 -2.28351682e-01 -2.42935091e-01 4.59498882e-01
3.38213116e-01 -6.03881717e-01 7.84923136e-01 7.57385433e-01
2.22273737e-01 1.71440979e-03 3.10784966e-01 1.13188839e+00
-2.09328219e-01 1.69568226e-01 -1.06895530e+00 2.54795909e-01
3.44313771e-01 1.38764369e+00 -5.13767779e-01 1.15673980e-02
-5.28185844e-01 8.20426106e-01 5.22214830e-01 5.97055078e-01
-4.52561855e-01 -5.80996163e-02 8.50370944e-01 1.38833672e-01
3.54763001e-01 -2.82121390e-01 -2.58653164e-01 -1.18001676e+00
3.07278395e-01 -8.92525792e-01 2.07477957e-01 -7.71393955e-01
-1.45233560e+00 6.41230881e-01 -2.19814420e-01 -1.28234124e+00
-3.20388496e-01 -4.97041345e-01 -9.14915264e-01 9.88452673e-01
-1.29981148e+00 -1.55507851e+00 -7.32500792e-01 9.21189070e-01
1.99564695e-01 -4.48892295e-01 7.84500778e-01 1.33218348e-01
-6.05045296e-02 8.32577825e-01 1.52224407e-01 3.81380528e-01
5.26001632e-01 -9.73690212e-01 4.43234801e-01 8.67771327e-01
4.59286183e-01 7.21166074e-01 4.87101436e-01 -4.05389547e-01
-1.92048764e+00 -8.80863130e-01 6.32505119e-01 -4.58853215e-01
5.23068428e-01 -5.19020915e-01 -9.78391469e-01 3.57669353e-01
-1.20357573e-01 6.05623424e-01 3.05225343e-01 -2.98107147e-01
-9.08891022e-01 -2.83142149e-01 -1.11605978e+00 5.34914613e-01
1.17949545e+00 -9.67987478e-01 -3.71697575e-01 1.52731776e-01
4.64575529e-01 -5.64339876e-01 -1.04682612e+00 3.13486367e-01
6.89988434e-01 -1.00414896e+00 1.29419184e+00 -7.34076858e-01
2.26882681e-01 -1.47146493e-01 -5.07807314e-01 -1.14510262e+00
-5.20677626e-01 -5.15513718e-01 2.80616999e-01 9.11415398e-01
1.12978816e-01 -7.06927896e-01 7.44083703e-01 5.69934726e-01
-2.04529800e-02 -8.48510861e-01 -9.50517833e-01 -8.16864252e-01
1.17964651e-02 1.37935862e-01 6.42342567e-01 9.42663491e-01
-4.72691804e-01 3.41863811e-01 -5.72597757e-02 3.15242618e-01
9.58420932e-01 6.65047050e-01 9.85014141e-01 -1.19481206e+00
-1.87177807e-01 -6.02018774e-01 -1.10776329e+00 -7.06188083e-01
3.34774792e-01 -1.09861493e+00 1.05931414e-02 -1.17903674e+00
1.34951800e-01 -3.98789018e-01 -2.64708132e-01 1.82286710e-01
-1.56779047e-02 4.50985909e-01 5.35944760e-01 4.35622752e-01
-8.49243775e-02 5.72865844e-01 1.34871113e+00 -2.86015093e-01
-6.46917447e-02 -1.78599387e-01 -4.55539793e-01 5.67099333e-01
4.32987422e-01 -3.04097861e-01 -3.82965595e-01 -4.96119022e-01
9.73242447e-02 3.47280473e-01 6.69823647e-01 -9.23293054e-01
4.16639119e-01 -1.53313577e-01 5.46987891e-01 -2.48819813e-01
5.00756264e-01 -1.07866049e+00 2.99439222e-01 1.73991695e-01
-5.35628259e-01 3.26436907e-01 6.14905311e-03 8.11671734e-01
-3.30406725e-01 1.93583399e-01 9.96093810e-01 -5.22587784e-02
-4.16483372e-01 9.08914447e-01 4.98853207e-01 1.85834378e-01
8.93448055e-01 -6.11502528e-01 -1.02792986e-01 -5.42851985e-01
-5.46903372e-01 -1.78898782e-01 7.73477316e-01 6.30046546e-01
9.43424165e-01 -1.72956240e+00 -8.96309078e-01 5.92255771e-01
4.77022201e-01 -1.93408370e-01 7.35322759e-02 7.46255994e-01
-2.71987289e-01 2.14607760e-01 -5.51165938e-01 -8.68361712e-01
-1.28161645e+00 4.86985177e-01 6.55409515e-01 -1.05202839e-01
-8.00191998e-01 5.90145290e-01 5.45856893e-01 -5.05269766e-01
9.88261998e-02 -1.52486572e-02 1.12551555e-01 -2.47193098e-01
5.22930734e-02 -1.41447792e-02 3.14876765e-01 -1.12029040e+00
-3.36541563e-01 1.38558388e+00 1.98751569e-01 8.33399370e-02
1.22265756e+00 -1.85149178e-01 -7.75329396e-02 5.95692284e-02
1.87649798e+00 -1.42748296e-01 -1.24593735e+00 -6.23598993e-01
-2.20449492e-01 -9.82540667e-01 -7.79479295e-02 -4.73217517e-01
-1.26883590e+00 8.19963634e-01 7.78719187e-01 -5.28858379e-02
9.81637836e-01 -3.95453535e-02 7.60675550e-01 4.36743855e-01
4.04167026e-01 -5.33982694e-01 2.65988231e-01 2.90968269e-01
1.21593130e+00 -1.35903132e+00 -3.79742011e-02 -2.61180580e-01
-1.62622660e-01 1.15532649e+00 5.58023870e-01 -5.53296566e-01
5.34531593e-01 -3.85117978e-02 2.30688155e-01 -6.31259084e-01
-4.61155027e-01 -1.81564018e-01 8.76834750e-01 7.26515234e-01
4.24083881e-02 -5.64571507e-02 2.08882108e-01 -2.15154439e-01
-9.27933231e-02 -5.31237781e-01 7.86036178e-02 5.89111984e-01
-1.50093362e-01 -6.90790296e-01 -4.52139735e-01 2.60083675e-01
-2.43241224e-03 1.65598229e-01 -6.66965067e-01 7.51290441e-01
-8.85466114e-02 5.84785402e-01 1.92979112e-01 -3.64562631e-01
5.64414561e-01 -2.65980035e-01 7.81018615e-01 -2.60592192e-01
-5.35724044e-01 -2.21516509e-02 -3.36765140e-01 -8.38275373e-01
-3.67406815e-01 -3.74041170e-01 -9.77115691e-01 -2.20242456e-01
-1.67948559e-01 -4.52999383e-01 6.68750644e-01 8.98776412e-01
3.82124603e-01 -1.14602502e-02 7.17678785e-01 -1.08780301e+00
-7.72085786e-01 -2.74641991e-01 -4.11096394e-01 1.09326315e+00
5.12389600e-01 -8.19396853e-01 -3.91287565e-01 -3.42146531e-02] | [8.367376327514648, -2.244828224182129] |
6c76c129-89a0-4ef6-8720-c4c0ef73aa11 | exploration-and-exploitation-two-ways-to | 2105.14813 | null | https://arxiv.org/abs/2105.14813v2 | https://arxiv.org/pdf/2105.14813v2.pdf | Exploration and Exploitation: Two Ways to Improve Chinese Spelling Correction Models | A sequence-to-sequence learning with neural networks has empirically proven to be an effective framework for Chinese Spelling Correction (CSC), which takes a sentence with some spelling errors as input and outputs the corrected one. However, CSC models may fail to correct spelling errors covered by the confusion sets, and also will encounter unseen ones. We propose a method, which continually identifies the weak spots of a model to generate more valuable training instances, and apply a task-specific pre-training strategy to enhance the model. The generated adversarial examples are gradually added to the training set. Experimental results show that such an adversarial training method combined with the pretraining strategy can improve both the generalization and robustness of multiple CSC models across three different datasets, achieving stateof-the-art performance for CSC task. | ['Xuanjing Huang', 'Xiaoqing Zheng', 'Cenyuan Zhang', 'Chong Li'] | 2021-05-31 | null | https://aclanthology.org/2021.acl-short.56 | https://aclanthology.org/2021.acl-short.56.pdf | acl-2021-5 | ['spelling-correction'] | ['natural-language-processing'] | [ 7.46829867e-01 -3.58457476e-01 1.07373118e-01 -1.91842288e-01
-6.85827374e-01 -7.01641560e-01 5.02126515e-01 -7.90412910e-03
-5.65164804e-01 8.65123451e-01 9.01770666e-02 -4.96124625e-01
3.42582017e-01 -5.83462715e-01 -7.38052070e-01 -7.14814126e-01
3.15295339e-01 3.04550171e-01 3.55761647e-01 -3.90577227e-01
4.98967379e-01 3.49739105e-01 -1.17954075e+00 7.12952316e-01
1.57331562e+00 5.65745771e-01 4.01470989e-01 9.90249932e-01
-2.96367496e-01 9.68313634e-01 -1.19157434e+00 -6.68067694e-01
3.36310267e-02 -5.61975718e-01 -7.22770214e-01 -2.55665064e-01
1.76474541e-01 3.21992598e-02 -2.80458510e-01 1.36749744e+00
4.75998878e-01 1.21879324e-01 6.19263530e-01 -9.06217635e-01
-1.03890896e+00 5.29901445e-01 -8.68827701e-02 2.67393678e-01
2.54070431e-01 3.60309333e-02 4.02022958e-01 -1.06833375e+00
3.93941373e-01 7.41777241e-01 6.89673781e-01 1.24909174e+00
-5.18151522e-01 -8.23574841e-01 -7.02487491e-03 2.70998657e-01
-1.18645275e+00 -1.72173277e-01 6.82767689e-01 3.74371260e-02
7.33845413e-01 4.34946448e-01 2.77541578e-01 1.21717918e+00
2.02624753e-01 1.11013889e+00 1.02073240e+00 -7.43840635e-01
1.57980338e-01 2.23968122e-02 1.49350956e-01 2.36732230e-01
1.25198588e-01 2.02109516e-01 -2.13913053e-01 1.59325346e-01
2.97661602e-01 1.27353847e-01 -4.73971933e-01 2.86368787e-01
-7.39735603e-01 5.10824084e-01 4.54628855e-01 6.21990621e-01
-9.21056196e-02 -2.09940180e-01 4.22731191e-01 3.20367813e-01
4.25777763e-01 9.65396106e-01 -6.40245736e-01 -2.56923705e-01
-8.67885113e-01 2.39420861e-01 5.40762365e-01 1.04782867e+00
4.43230778e-01 2.06497073e-01 -4.74987000e-01 9.49315786e-01
-3.27047050e-01 6.05247617e-01 8.26282799e-01 -3.19220275e-01
8.72261703e-01 7.39618659e-01 2.19644323e-01 -9.17865515e-01
-1.37470841e-01 -7.29199886e-01 -9.13523972e-01 -4.07010764e-02
2.53522038e-01 -3.01821858e-01 -1.08445847e+00 1.40259886e+00
-2.78200150e-01 4.19443637e-01 3.93822402e-01 8.02909613e-01
6.63080931e-01 7.05875814e-01 1.10149190e-01 -1.78223521e-01
5.88583469e-01 -1.26993918e+00 -8.07229578e-01 -5.58136106e-01
7.76248395e-01 -7.92253852e-01 1.21015072e+00 3.43615234e-01
-9.43342209e-01 -6.24818563e-01 -1.21122062e+00 3.38141799e-01
-4.22540724e-01 2.99487412e-01 2.08576545e-01 5.91640294e-01
-5.86796403e-01 9.29325521e-01 -3.20796847e-01 1.26971290e-01
4.61682767e-01 3.14241499e-02 -1.25653461e-01 -2.96214432e-01
-1.46409702e+00 1.02897394e+00 7.05842316e-01 8.68621394e-02
-8.76536012e-01 -6.99465156e-01 -6.57749176e-01 1.86496213e-01
2.91922987e-01 4.23435960e-03 1.38725781e+00 -1.45689142e+00
-1.28350103e+00 4.30432796e-01 1.31157385e-02 -4.53099132e-01
6.56636596e-01 -4.66933668e-01 -1.02490342e+00 -1.13866031e-01
-1.56301856e-01 3.57612073e-01 9.03478682e-01 -1.26821351e+00
-7.24814355e-01 -4.23094146e-02 -3.43326658e-01 2.35495865e-01
-6.72842264e-01 3.87076795e-01 -4.96302158e-01 -1.02684247e+00
-1.97045520e-01 -7.79127300e-01 -1.94610164e-01 -5.42402685e-01
-3.92281950e-01 -3.77001464e-02 8.97256911e-01 -8.70171547e-01
1.71038687e+00 -2.06558299e+00 -1.52558915e-03 1.09552659e-01
-2.53724843e-01 1.47120452e+00 -5.80650270e-01 3.70841533e-01
-1.00273147e-01 3.11480910e-01 -5.65913379e-01 -3.46413076e-01
-5.64408839e-01 1.09968610e-01 -3.58736724e-01 -3.59746404e-02
4.84560013e-01 9.04954195e-01 -1.26557803e+00 -8.31138119e-02
-1.30063087e-01 1.21248528e-01 -2.99829632e-01 5.53572536e-01
-2.44602412e-01 1.77049488e-01 -1.58200219e-01 3.25577199e-01
9.55511689e-01 -4.85834070e-02 -5.18151000e-02 5.56705534e-01
2.47688800e-01 1.90914914e-01 -1.07507038e+00 1.07289469e+00
-3.30727547e-01 5.69609404e-01 -5.11792660e-01 -6.58262014e-01
1.16834211e+00 9.54557434e-02 -3.14066231e-01 -6.77500308e-01
4.97233868e-02 4.28201973e-01 5.36200292e-02 -6.76854134e-01
6.02102339e-01 -4.40655127e-02 -5.28433509e-02 4.18473035e-01
-1.73176914e-01 -9.11743790e-02 1.16960511e-01 1.62858501e-01
1.02072120e+00 1.21445343e-01 1.36475086e-01 1.88596338e-01
8.67244780e-01 -1.44052077e-02 8.40464175e-01 1.00109017e+00
-3.29442114e-01 8.79657567e-01 1.44581556e-01 -4.82920945e-01
-1.16586375e+00 -7.78998256e-01 2.05500141e-01 8.61246705e-01
3.67640555e-02 -2.96192050e-01 -1.11097574e+00 -1.27533889e+00
-2.60968506e-01 1.02209771e+00 -5.62922299e-01 -7.15535402e-01
-6.87923253e-01 -5.69582641e-01 1.01767671e+00 8.32879364e-01
6.87574267e-01 -1.52630413e+00 -1.79292798e-01 2.22809345e-01
-1.40981182e-01 -8.79177332e-01 -7.43044078e-01 1.85904220e-01
-5.71751118e-01 -1.16817701e+00 -8.51490140e-01 -1.10282922e+00
9.85226095e-01 2.74997801e-01 7.35813320e-01 8.18450809e-01
1.64567843e-01 -4.91066903e-01 -7.38500059e-01 -6.08290732e-01
-1.14096212e+00 1.26803992e-03 1.42227098e-01 -5.25489859e-02
5.92377484e-01 1.06591962e-01 -1.50499970e-01 1.68579981e-01
-1.05433166e+00 2.57900748e-02 4.80380803e-01 1.09333277e+00
1.93368196e-01 1.25500068e-01 9.06378746e-01 -1.05223072e+00
1.05884469e+00 -1.99687093e-01 -4.01574016e-01 5.16650021e-01
-6.30712509e-01 -4.76964638e-02 1.32416761e+00 -6.31089926e-01
-1.18579364e+00 -2.91499272e-02 -2.26549804e-01 -4.46938872e-01
-2.50717878e-01 5.83570540e-01 -1.81600258e-01 -1.24375992e-01
9.19668615e-01 6.63519204e-01 -1.89976498e-01 -5.27512610e-01
-3.70949097e-02 1.02031350e+00 8.88341188e-01 -1.63476110e-01
9.39584851e-01 -2.62178659e-01 -4.93366361e-01 -2.41857022e-01
-9.94324088e-01 -3.68113041e-01 -8.34596574e-01 -2.58423597e-01
5.18122613e-01 -5.54772854e-01 -1.17009051e-01 1.00308967e+00
-1.30552733e+00 -2.91818947e-01 2.01870576e-01 1.78927451e-01
-5.76259196e-02 6.14417791e-01 -6.19608283e-01 -8.33593845e-01
-5.79189122e-01 -8.42980385e-01 5.34080029e-01 6.14461184e-01
-3.70082306e-03 -7.77966738e-01 -1.86013989e-02 1.21213637e-01
3.93428564e-01 -5.89652620e-02 8.12849581e-01 -1.06585050e+00
-1.77072242e-01 -5.63269913e-01 5.56137674e-02 8.91592324e-01
3.65320668e-02 7.21118078e-02 -9.06071723e-01 -2.43259862e-01
-9.76371765e-02 -1.71337709e-01 7.46306181e-01 -2.10313186e-01
1.53460872e+00 -5.34077287e-01 -2.74087340e-01 6.70706451e-01
1.27141190e+00 5.68310738e-01 1.03780901e+00 4.34887260e-01
7.37913966e-01 2.42293492e-01 6.64608419e-01 4.65038121e-02
-2.18466550e-01 4.11762953e-01 2.54020095e-01 1.04030371e-02
6.04407713e-02 -5.04453540e-01 3.26211035e-01 9.29181874e-01
9.81319025e-02 -4.61394042e-01 -1.00490069e+00 4.76263523e-01
-1.63152015e+00 -1.05594134e+00 -2.68451363e-01 2.22994876e+00
1.11748540e+00 3.38348895e-01 -2.99495816e-01 4.95340914e-01
1.01582205e+00 -3.60036157e-02 -3.47726703e-01 -6.05974793e-01
-4.20309007e-01 2.69037813e-01 3.60314310e-01 3.57693881e-01
-1.24472833e+00 1.38610780e+00 6.27879095e+00 9.71912801e-01
-1.28624249e+00 8.24532192e-03 5.77478468e-01 3.04565996e-01
-4.15766597e-01 -3.23365211e-01 -6.47893786e-01 8.48833084e-01
8.37773919e-01 -5.66154979e-02 5.82149446e-01 8.35983396e-01
9.55951288e-02 2.47261524e-01 -7.42375433e-01 5.86288393e-01
4.73759085e-01 -1.44636929e+00 2.85535723e-01 -5.69311857e-01
1.18907416e+00 -1.64459676e-01 -6.84180036e-02 6.09165668e-01
2.68137813e-01 -1.11334717e+00 5.79644144e-01 4.27778721e-01
7.93499529e-01 -9.24178600e-01 1.04542565e+00 8.22117984e-01
-5.72323024e-01 -3.23760241e-01 -5.04982352e-01 -1.86609790e-01
-1.04548037e-01 1.58901170e-01 -9.54153121e-01 4.32187349e-01
5.33665001e-01 5.38115025e-01 -1.01749456e+00 1.39825034e+00
-7.82069027e-01 8.05659652e-01 3.65509301e-01 -7.40223169e-01
2.98818767e-01 6.77342638e-02 4.73955393e-01 1.44025588e+00
1.83012962e-01 7.27188662e-02 -2.63100147e-01 6.66387320e-01
-1.88433677e-01 1.34134457e-01 -4.97411668e-01 -1.29009515e-01
7.75467575e-01 8.35211813e-01 -2.89628804e-01 -4.58176404e-01
-1.15273036e-01 1.37996852e+00 6.04033530e-01 4.23872441e-01
-7.56021857e-01 -8.08925986e-01 3.26301634e-01 -3.60912532e-01
3.52643520e-01 1.67392895e-01 -5.82583189e-01 -1.19482720e+00
9.10766572e-02 -1.23725009e+00 2.83355832e-01 -9.64742422e-01
-1.12524009e+00 6.96259797e-01 -5.60652673e-01 -1.60490048e+00
-1.59860000e-01 -6.03571832e-01 -1.12367606e+00 1.04178178e+00
-1.48387098e+00 -7.72962987e-01 -3.04306269e-01 3.68308455e-01
8.46815646e-01 -5.73665321e-01 8.36844146e-01 2.24120781e-01
-6.41898036e-01 1.09605050e+00 3.86204243e-01 4.87513840e-01
6.95900381e-01 -1.38110352e+00 6.16618693e-01 1.58199978e+00
4.87258248e-02 6.91693127e-01 6.26849353e-01 -9.79995251e-01
-8.71823251e-01 -1.51016748e+00 1.45734990e+00 -4.15028632e-01
2.09271476e-01 -3.46092194e-01 -1.28286743e+00 3.87090802e-01
-8.13040789e-03 -1.79532766e-01 4.21228737e-01 -2.81701267e-01
-1.32373482e-01 5.96191399e-02 -1.09048402e+00 8.16770971e-01
8.63315225e-01 -4.88593817e-01 -9.25116003e-01 2.38360465e-01
8.29694927e-01 -7.40719795e-01 -1.82169035e-01 3.16888720e-01
1.04895018e-01 -5.86376607e-01 5.46042323e-01 -1.20112455e+00
8.87067676e-01 -4.54663932e-01 2.04401985e-01 -1.88208902e+00
-5.11304200e-01 -6.90841496e-01 6.11003377e-02 1.30862916e+00
6.50578916e-01 -1.87263668e-01 6.19297385e-01 3.85591805e-01
-5.60995877e-01 -6.89652503e-01 -5.62319994e-01 -9.29512143e-01
4.11126912e-01 -3.67340267e-01 1.01339853e+00 1.03657508e+00
-1.13760009e-02 5.06932735e-02 -5.43862462e-01 2.11947322e-01
1.13788970e-01 -3.45544010e-01 5.65280914e-01 -8.58737409e-01
4.80667725e-02 -3.16324562e-01 -6.73508644e-03 -7.84465373e-01
2.56690353e-01 -8.75259042e-01 3.82360905e-01 -1.16976321e+00
6.15483429e-03 -4.60060388e-01 -4.97677922e-01 3.54179859e-01
-1.08861065e+00 1.16472088e-01 4.83181298e-01 4.76906560e-02
-5.92648804e-01 5.03449559e-01 1.36391449e+00 -1.62383556e-01
-1.89768314e-01 4.02719557e-01 -7.82690287e-01 6.53232515e-01
8.49191546e-01 -6.49342418e-01 -7.53945261e-02 -5.22113144e-01
8.29080045e-02 -1.22979790e-01 9.29983705e-03 -1.17216480e+00
2.41365850e-01 -3.48170042e-01 7.57629514e-01 -4.83993262e-01
-2.24721208e-01 -6.42480731e-01 -2.46060953e-01 8.20021570e-01
-7.76138008e-01 2.01622576e-01 2.07749754e-01 4.95232821e-01
-9.13617462e-02 -8.45954776e-01 8.69025707e-01 -6.81421682e-02
-8.95547926e-01 1.68689668e-01 -3.11460495e-01 1.60467118e-01
9.50458288e-01 -2.46473074e-01 -5.27542591e-01 -2.65588045e-01
-4.33984578e-01 2.23312333e-01 3.97963971e-01 7.90844321e-01
7.26415396e-01 -1.44125628e+00 -8.16718936e-01 4.73648608e-01
2.86621779e-01 -3.43382865e-01 1.60101593e-01 8.70295465e-02
-6.22159362e-01 2.92709768e-01 -2.01726586e-01 -1.92458585e-01
-1.33335543e+00 9.33354378e-01 5.45738816e-01 -2.68234938e-01
-2.85159051e-01 9.58700418e-01 -4.41660821e-01 -6.50433004e-01
3.01700145e-01 -1.10844687e-01 -5.92268825e-01 -4.73764122e-01
9.75476086e-01 2.18358576e-01 3.29806805e-01 -4.42805320e-01
-4.01217073e-01 2.08541527e-01 -3.29865813e-01 2.24883169e-01
1.07452476e+00 1.97189838e-01 2.66100317e-01 7.30026141e-02
1.13795900e+00 1.27852544e-01 -1.23949409e+00 -3.52871656e-01
9.01220739e-02 -5.87264836e-01 -3.91522378e-01 -1.36928713e+00
-8.63601565e-01 8.53496075e-01 4.82064068e-01 -5.92700690e-02
1.14096463e+00 -5.63672662e-01 9.21225727e-01 5.13416111e-01
1.53378114e-01 -1.43574631e+00 2.70048827e-01 1.03435814e+00
9.25510228e-01 -1.06771088e+00 -5.36358237e-01 -3.31536382e-01
-9.48535562e-01 1.27404332e+00 1.05339241e+00 -1.75574154e-01
1.06841952e-01 2.19113976e-01 2.79185444e-01 4.92480069e-01
-5.64348340e-01 1.79200113e-01 2.66278446e-01 9.26827610e-01
1.96596161e-01 1.73476323e-01 -4.87671256e-01 9.24124777e-01
-1.65718555e-01 1.67602018e-01 7.43227303e-01 8.37786436e-01
-5.27358949e-01 -1.20053136e+00 -3.89571130e-01 4.54446018e-01
-5.25414109e-01 -4.67623591e-01 -6.47057950e-01 3.96287918e-01
1.51145503e-01 1.03094411e+00 1.89540107e-02 -7.60986745e-01
3.25786829e-01 1.31152287e-01 8.07469562e-02 -7.06607044e-01
-1.11351168e+00 -3.94583702e-01 2.67106593e-02 -1.24602355e-01
1.37818143e-01 -5.84286451e-01 -1.27365983e+00 -1.20506972e-01
-4.64025527e-01 1.98258385e-01 2.46732607e-01 1.23258424e+00
3.56187522e-01 5.34363151e-01 9.06915009e-01 -2.54600734e-01
-8.99050415e-01 -1.23279381e+00 -5.84220327e-02 8.14591646e-01
3.92590046e-01 -2.13764742e-01 -3.27339679e-01 7.15871006e-02] | [10.949370384216309, 10.824515342712402] |
2e086f96-2ab4-4841-a4c5-2bc5d32d69c0 | features-compression-based-on-counterfactual | 2211.09894 | null | https://arxiv.org/abs/2211.09894v3 | https://arxiv.org/pdf/2211.09894v3.pdf | Supervised Feature Compression based on Counterfactual Analysis | Counterfactual Explanations are becoming a de-facto standard in post-hoc interpretable machine learning. For a given classifier and an instance classified in an undesired class, its counterfactual explanation corresponds to small perturbations of that instance that allows changing the classification outcome. This work aims to leverage Counterfactual Explanations to detect the important decision boundaries of a pre-trained black-box model. This information is used to build a supervised discretization of the features in the dataset with a tunable granularity. Using the discretized dataset, an optimal Decision Tree can be trained that resembles the black-box model, but that is interpretable and compact. Numerical results on real-world datasets show the effectiveness of the approach in terms of accuracy and sparsity. | ['Cecilia Salvatore', 'Dolores Romero Morales', 'Veronica Piccialli'] | 2022-11-17 | null | null | null | null | ['feature-compression', 'interpretable-machine-learning', 'counterfactual-explanation'] | ['computer-vision', 'methodology', 'miscellaneous'] | [ 5.41503012e-01 1.18915236e+00 -8.36324692e-01 -5.42106390e-01
-3.43881994e-01 -3.92550409e-01 5.43211401e-01 3.26735348e-01
8.55199397e-02 1.33340633e+00 2.36649990e-01 -5.55371881e-01
-4.16999847e-01 -9.10627961e-01 -8.96566391e-01 -6.06425583e-01
-4.32309866e-01 4.72763419e-01 -4.56638783e-01 3.03738028e-01
3.82540137e-01 3.18676233e-01 -1.66850984e+00 8.78159404e-01
9.31552172e-01 1.02134979e+00 -5.47556758e-01 2.76363611e-01
1.81420669e-01 6.75175488e-01 -6.49465561e-01 -4.91062343e-01
4.65303451e-01 -4.92491335e-01 -9.76659298e-01 3.49950552e-01
3.33236396e-01 1.84791088e-02 1.62183523e-01 1.09450924e+00
-1.46433324e-01 1.66103229e-01 8.29929829e-01 -1.50635898e+00
-4.30026919e-01 1.09715796e+00 -2.57857710e-01 -5.51469587e-02
1.68421268e-01 1.85412720e-01 1.12732303e+00 -4.82970893e-01
4.93051767e-01 1.44173503e+00 4.92301375e-01 6.10319436e-01
-1.71270907e+00 -6.68063641e-01 4.11268353e-01 2.30778530e-01
-9.35864151e-01 -1.80841580e-01 8.96936178e-01 -3.02306324e-01
6.48712397e-01 8.04553270e-01 8.13299298e-01 9.24405098e-01
6.41994298e-01 4.78269219e-01 1.41482377e+00 -5.46231389e-01
1.01658237e+00 4.75750238e-01 1.70080185e-01 5.17775059e-01
9.27877128e-01 6.01996481e-01 -4.06630576e-01 -6.67388737e-01
1.47551179e-01 1.64428309e-01 -2.97624379e-01 -7.61759520e-01
-8.29960942e-01 1.22534442e+00 4.84278470e-01 8.74974802e-02
-4.32892263e-01 1.34983569e-01 4.58988309e-01 5.66055417e-01
7.99512923e-01 1.12152398e+00 -9.47117865e-01 4.01136935e-01
-7.99431086e-01 5.16413748e-01 7.68305004e-01 3.82051975e-01
6.82969272e-01 -7.73334876e-02 -1.68461666e-01 2.91952491e-01
8.98881480e-02 1.44586831e-01 6.48155034e-01 -9.34751272e-01
3.94449890e-01 9.26436484e-01 3.21823061e-01 -7.45333374e-01
-3.86630565e-01 -6.06443763e-01 -8.14172089e-01 7.08891392e-01
4.96952742e-01 -3.37644726e-01 -7.42588103e-01 1.48914206e+00
5.85074842e-01 1.26926959e-01 9.19929221e-02 8.18868935e-01
-2.75351945e-02 4.45124775e-01 -5.98497577e-02 -6.83782339e-01
9.86664653e-01 -4.72143650e-01 -5.99712729e-01 -2.62478054e-01
8.94257009e-01 -3.44847292e-02 9.82018590e-01 5.50480664e-01
-5.69819570e-01 -3.37850064e-01 -1.30222702e+00 5.70545316e-01
-2.89907396e-01 -2.28972673e-01 9.86530602e-01 7.89951324e-01
-3.01162601e-01 9.73201990e-01 -3.66352767e-01 7.96865821e-02
7.78792024e-01 4.96482790e-01 -1.86685681e-01 1.74966693e-01
-1.21392977e+00 8.20805609e-01 8.93240392e-01 -2.79865205e-01
-8.16991270e-01 -1.12569952e+00 -7.81698585e-01 3.47295463e-01
6.24998093e-01 -7.91771889e-01 1.10890472e+00 -1.61507237e+00
-1.23814750e+00 7.39333212e-01 8.65931064e-02 -1.26650405e+00
8.52635443e-01 -3.05983014e-02 -4.36750263e-01 -1.61136538e-01
1.02419220e-01 4.32539672e-01 1.19315898e+00 -1.30982208e+00
-7.24556446e-01 -5.51708996e-01 2.12371573e-01 -5.21049835e-02
1.78901136e-01 -6.10886157e-01 9.74797189e-01 -7.57540882e-01
2.43721902e-02 -9.03883994e-01 -5.89712441e-01 -8.73553455e-02
-9.27842319e-01 1.87652167e-02 9.89542842e-01 -5.17509915e-02
1.22398674e+00 -1.60862243e+00 -4.73893702e-01 4.04468000e-01
2.03988716e-01 -1.02640949e-01 4.12908942e-01 2.29722373e-02
-6.95311904e-01 5.92589438e-01 -4.93940771e-01 2.73808569e-01
-1.18949160e-01 1.98921114e-01 -9.77324009e-01 7.36848831e-01
1.08030386e-01 5.39079785e-01 -7.70075023e-01 -2.46460453e-01
4.76704895e-01 -3.91231984e-01 -8.43155444e-01 1.57942586e-02
-6.55246615e-01 2.95409948e-01 -5.97942412e-01 8.45137760e-02
7.28775203e-01 3.77699062e-02 5.15601218e-01 3.29042882e-01
3.56640434e-03 4.49220031e-01 -1.37007236e+00 8.62355709e-01
-4.54045236e-01 5.65802872e-01 -7.03483164e-01 -1.45807874e+00
8.61144423e-01 2.54946262e-01 1.11438364e-01 -1.66146740e-01
1.77017599e-01 1.90896824e-01 1.52021766e-01 -3.74473870e-01
2.74281367e-03 -6.88399613e-01 -3.23234051e-01 5.51091135e-01
-2.14370385e-01 -2.28202999e-01 -2.39789158e-01 -2.57432044e-01
7.68181503e-01 -3.18497688e-01 1.37715709e+00 -7.35658169e-01
3.74515086e-01 1.27638459e-01 6.63522065e-01 9.42529261e-01
8.15352648e-02 4.77837265e-01 7.57750213e-01 -1.13163137e+00
-7.98267484e-01 -7.65854001e-01 -3.62124771e-01 4.31376934e-01
-2.32219800e-01 2.06426680e-02 -7.57546723e-01 -1.34601951e+00
3.77741128e-01 1.46977115e+00 -1.27344429e+00 -5.04476070e-01
-2.44737938e-01 -7.55518675e-01 3.11310925e-02 6.04211837e-02
4.68103290e-01 -8.35236430e-01 -1.17540216e+00 -8.98555946e-03
8.10436681e-02 -4.01157498e-01 5.01562878e-02 3.73590708e-01
-1.30454636e+00 -1.44435143e+00 5.89935519e-02 5.08248340e-03
9.31396365e-01 -1.54451266e-01 1.03190744e+00 -3.85793187e-02
-3.34568731e-02 -3.70138347e-01 -3.01208552e-02 -7.78729737e-01
-6.40772820e-01 -2.74809450e-01 2.71534711e-01 4.10302371e-01
2.97362357e-01 -5.54787219e-01 -4.19768423e-01 -3.29306573e-02
-6.99633300e-01 1.74846798e-01 4.66111451e-01 1.14644623e+00
4.71716255e-01 4.04017240e-01 7.69633174e-01 -1.53120971e+00
4.97668833e-01 -6.61522090e-01 -6.99431241e-01 3.01070035e-01
-1.05667520e+00 6.23899043e-01 1.22549248e+00 -5.66896379e-01
-1.31168318e+00 2.17379570e-01 7.22779572e-01 -1.94391772e-01
-4.35544580e-01 4.75324899e-01 -5.33344269e-01 3.94562900e-01
1.02841568e+00 -2.34333783e-01 -4.65015680e-01 -4.11951900e-01
6.28859639e-01 7.27131963e-01 3.44218761e-01 -4.67268169e-01
6.66436195e-01 7.35220969e-01 2.78840184e-01 -4.40174222e-01
-1.29045868e+00 1.82817757e-01 -7.06256509e-01 8.94544572e-02
3.57403606e-01 -2.43319064e-01 -6.12024009e-01 -4.21627492e-01
-9.91510808e-01 -1.12971976e-01 -1.09221292e+00 4.58206326e-01
-8.29891622e-01 -2.53907412e-01 4.82253611e-01 -1.12824833e+00
3.63168158e-02 -6.40408039e-01 6.38467371e-01 4.61692177e-02
-8.88274312e-01 -1.04475439e+00 2.13783211e-03 2.95490146e-01
-3.43266398e-01 7.56967843e-01 1.30093205e+00 -1.07575631e+00
-5.16008317e-01 -4.67707962e-01 1.60129830e-01 1.84036210e-01
1.73556745e-01 -1.89838763e-02 -1.25390172e+00 -1.56878129e-01
3.32327336e-01 -1.07622541e-01 8.88655484e-01 6.73866630e-01
1.55727041e+00 -1.38928401e+00 -6.37925208e-01 3.45860988e-01
1.25159919e+00 2.46730119e-01 2.98931211e-01 2.69799471e-01
-5.47884479e-02 7.03169525e-01 1.01295543e+00 6.77493751e-01
-2.06728399e-01 4.65351582e-01 6.69955611e-01 9.24031138e-02
5.35792530e-01 -5.59751809e-01 3.30729596e-02 -4.96962070e-01
1.64241210e-01 5.74846156e-02 -6.50169551e-01 5.81065297e-01
-1.90525866e+00 -1.32480693e+00 -4.43043327e-03 2.44042110e+00
6.73702300e-01 4.14547265e-01 -1.41228080e-01 6.26358628e-01
6.79957271e-01 6.78330362e-02 -8.37781429e-01 -1.03853357e+00
7.37073794e-02 1.17898904e-01 4.75519419e-01 5.76543808e-01
-1.23200154e+00 3.51525515e-01 6.24402332e+00 5.77502847e-01
-1.10464406e+00 -1.46097809e-01 1.16684759e+00 -3.03412706e-01
-5.38410068e-01 3.84306431e-01 -3.85668904e-01 5.14575899e-01
1.04383445e+00 -8.74285221e-01 2.09926903e-01 1.18720973e+00
7.00002670e-01 -3.01225096e-01 -1.69470584e+00 3.67973447e-01
-4.04249460e-01 -1.89391553e+00 4.82312620e-01 2.22517252e-01
1.05596697e+00 -6.09500110e-01 1.68338388e-01 1.47798494e-01
4.33892787e-01 -1.23626089e+00 9.45251048e-01 3.01076531e-01
7.19929218e-01 -9.93397117e-01 7.51724362e-01 7.35149384e-01
-5.14494121e-01 -7.35902131e-01 -3.71926546e-01 -8.13172400e-01
-5.31589389e-01 8.53557229e-01 -1.23967206e+00 4.04014021e-01
3.80461216e-01 5.09869277e-01 -1.71315670e-01 6.77900791e-01
-5.06072879e-01 8.40482473e-01 -1.30348638e-01 -2.26596743e-02
7.58700669e-02 -7.57301971e-02 6.34625196e-01 9.04412568e-01
2.03471500e-02 3.03550094e-01 2.55153682e-02 1.10491025e+00
-1.84116572e-01 -1.58206627e-01 -1.11776888e+00 4.36848998e-01
4.88069564e-01 5.46879828e-01 -4.81944233e-01 -6.50511146e-01
3.64002436e-02 6.94346309e-01 2.22013354e-01 3.04011315e-01
-7.49457240e-01 -8.33746418e-03 9.42318082e-01 1.95174903e-01
4.91921902e-02 7.66328573e-01 -1.08998299e+00 -1.20118904e+00
6.30448684e-02 -9.08670545e-01 9.99478340e-01 -3.72868448e-01
-1.01453090e+00 6.11148663e-02 2.56243825e-01 -1.41355109e+00
-5.28332710e-01 -4.21141028e-01 -8.38343859e-01 5.37743509e-01
-9.20491457e-01 -7.00539529e-01 2.81275541e-01 8.76057521e-02
5.34595490e-01 -3.86567251e-03 1.04161751e+00 -7.22539902e-01
-3.95157278e-01 4.78074342e-01 1.38597712e-01 -2.53693193e-01
-1.05801979e-02 -1.35717297e+00 9.08596665e-02 7.56741226e-01
3.70446652e-01 6.83438003e-01 1.22783685e+00 -6.08332396e-01
-7.64208674e-01 -1.49705982e+00 8.70461583e-01 -3.28682393e-01
4.12786752e-01 -1.59108356e-01 -7.11404383e-01 9.83501434e-01
2.48933327e-03 1.78223446e-01 7.12160110e-01 2.24997610e-01
-2.57183045e-01 -2.29212314e-01 -1.76397741e+00 7.64320433e-01
7.69431889e-01 -1.89260513e-01 -1.28362203e+00 3.42721552e-01
5.44301510e-01 -2.01465771e-01 -4.34114963e-01 2.52973676e-01
5.20875156e-01 -9.99323308e-01 7.41908491e-01 -1.44334519e+00
5.44288337e-01 -6.42890930e-02 1.15641303e-01 -1.82344496e+00
-3.08700912e-02 -8.79685581e-01 -2.25963444e-01 7.94938743e-01
6.30127788e-01 -9.29754496e-01 8.94399762e-01 9.24636424e-01
3.42920303e-01 -1.00734138e+00 -1.33603382e+00 -7.85835087e-01
5.85061014e-02 -4.32329476e-01 9.71186519e-01 1.21443939e+00
6.47658885e-01 2.19798297e-01 -2.22915083e-01 3.50095510e-01
9.33714688e-01 8.39711547e-01 7.11009681e-01 -1.32029176e+00
-2.41686389e-01 -1.47974432e-01 -3.72849166e-01 -4.32275504e-01
4.36771870e-01 -9.78634238e-01 -4.25495446e-01 -9.12256539e-01
2.45191410e-01 -3.59674752e-01 -1.64794773e-01 4.85986739e-01
-1.74955770e-01 -3.48556012e-01 9.97251347e-02 -1.77936152e-01
-9.17540491e-02 5.67308903e-01 8.18308711e-01 -2.72037059e-01
-3.22768182e-01 4.84381557e-01 -1.03223562e+00 1.26143897e+00
1.01086628e+00 -9.13485229e-01 -4.79365706e-01 3.21397454e-01
-1.42616197e-01 1.33017525e-01 5.39295614e-01 -5.39691865e-01
-4.02987868e-01 -6.80722654e-01 5.26510715e-01 -2.22126752e-01
4.77954894e-02 -1.06420410e+00 3.26622069e-01 9.81026351e-01
-1.10016847e+00 -5.65645695e-01 8.98366943e-02 7.51126051e-01
-1.25749141e-01 -2.54663616e-01 1.03491890e+00 -6.10905848e-02
-2.33324453e-01 -4.96218093e-02 -2.78477520e-01 1.62767917e-01
1.40327919e+00 -3.25911671e-01 -2.17308193e-01 -3.07295412e-01
-6.95859432e-01 8.51219427e-03 2.86743581e-01 1.48703754e-01
5.26813269e-01 -1.14306557e+00 -6.41480923e-01 1.33294776e-01
2.17878968e-01 -1.64399117e-01 2.33977214e-02 1.61107227e-01
-3.28888297e-02 7.59445906e-01 -1.87176079e-01 -1.25913396e-01
-1.12056434e+00 8.55687678e-01 6.58116460e-01 -5.68280399e-01
-5.08486271e-01 3.43002796e-01 4.57586706e-01 -3.46443117e-01
-2.22035244e-01 -8.44803274e-01 2.71599414e-03 -2.48052049e-02
6.24772310e-01 5.61451972e-01 -1.48827136e-02 1.76528692e-02
-2.22503409e-01 -2.00335622e-01 1.89469874e-01 6.59897551e-02
1.39808893e+00 1.43499777e-01 5.45881316e-02 6.34488285e-01
1.10545695e+00 -1.29832000e-01 -1.26318097e+00 8.72898102e-02
4.63560432e-01 -8.74830723e-01 -2.68746987e-02 -1.02179718e+00
-7.51472831e-01 6.97335482e-01 3.42072040e-01 5.48749149e-01
9.56649780e-01 -1.20209277e-01 -1.16070867e-01 5.42024434e-01
4.20497566e-01 -1.01638544e+00 -5.80475688e-01 -2.79580146e-01
1.24368668e+00 -1.09983814e+00 3.04361314e-01 -5.17148376e-01
-6.40922308e-01 1.14274287e+00 3.55924964e-01 -3.04643303e-01
5.72186768e-01 5.55530377e-02 -2.55967528e-01 -6.59636557e-02
-1.10651863e+00 3.54440600e-01 2.81784922e-01 5.24030566e-01
6.03505000e-02 6.68312728e-01 -4.29614007e-01 8.78179252e-01
-5.64943135e-01 2.14950055e-01 7.70431161e-01 3.39667559e-01
-3.16956222e-01 -4.69123274e-01 -6.84839249e-01 9.58188176e-01
-4.62907314e-01 7.42884949e-02 -5.82138717e-01 9.55175698e-01
3.12936425e-01 9.79921043e-01 1.31796196e-03 -1.70769885e-01
3.26163888e-01 2.38028526e-01 1.67171344e-01 -8.58827531e-01
-4.62608427e-01 -6.92363501e-01 3.58395070e-01 -6.84839666e-01
-1.05313025e-01 -8.85287702e-01 -1.29405320e+00 -1.95335180e-01
-3.48928720e-01 4.88807410e-01 1.98270410e-01 1.22725952e+00
2.67244905e-01 2.92957276e-01 1.00727379e+00 -3.87273997e-01
-1.05837274e+00 -8.42334092e-01 -6.02912664e-01 4.66265678e-01
7.69522130e-01 -7.14018226e-01 -7.08090067e-01 8.30689743e-02] | [8.668357849121094, 5.6085333824157715] |
c4f4a8d0-af9b-4a22-a7c7-0d4e70308c60 | med-unic-unifying-cross-lingual-medical | 2305.19894 | null | https://arxiv.org/abs/2305.19894v1 | https://arxiv.org/pdf/2305.19894v1.pdf | Med-UniC: Unifying Cross-Lingual Medical Vision-Language Pre-Training by Diminishing Bias | The scarcity of data presents a critical obstacle to the efficacy of medical visionlanguage pre-training (VLP). A potential solution lies in the combination of datasets from various language communities. Nevertheless, the main challenge stems from the complexity of integrating diverse syntax and semantics, language-specific medical terminology, and culture-specific implicit knowledge. Therefore, one crucial aspect to consider is the presence of community bias caused by different languages. This paper presents a novel framework named Unifying Cross-Lingual Medical Vision-Language Pre-Training (Med-UniC), designed to integrate multimodal medical data from the two most prevalent languages, English and Spanish. Specifically, we propose Cross-lingual Text Alignment Regularization (CTR) to explicitly unify cross-lingual semantic representations of medical reports originating from diverse language communities. CTR is optimized through latent language disentanglement, rendering our optimization objective to not depend on negative samples, thereby significantly mitigating the bias from determining positive-negative sample pairs within analogous medical reports. Furthermore, it ensures that the cross-lingual representation is not biased toward any specific language community. Med-UniC reaches superior performance across 5 medical image tasks and 10 datasets encompassing over 30 diseases, offering a versatile framework for unifying multi-modal medical data within diverse linguistic communities. The experimental outcomes highlight the presence of community bias in cross-lingual VLP. Reducing this bias enhances the performance not only in vision-language tasks but also in uni-modal visual tasks. | ['Rossella Arcucci', 'César Quilodrán-Casas', 'Lei Ma', 'Sibo Cheng', 'Benyou Wang', 'Jie Fu', 'Mi Zhang', 'Che Liu', 'Zhongwei Wan'] | 2023-05-31 | null | null | null | null | ['disentanglement'] | ['methodology'] | [ 6.38124123e-02 -2.11447731e-01 -4.43856239e-01 -1.72746420e-01
-1.09294569e+00 -4.28958803e-01 5.42149663e-01 3.85255218e-01
-7.07148910e-01 4.85649645e-01 4.52353746e-01 -2.43286639e-01
4.57671620e-02 -3.89526337e-01 -3.01102757e-01 -7.71175206e-01
4.11705047e-01 3.54170412e-01 -3.19576353e-01 -4.39278670e-02
-1.62237883e-01 3.05685252e-01 -1.17063916e+00 6.92045689e-01
1.19410133e+00 4.99106675e-01 5.02870977e-01 2.86485344e-01
-4.23210770e-01 5.30798316e-01 -3.29758674e-01 -7.10509777e-01
-6.52118251e-02 -3.11300129e-01 -4.33562011e-01 2.65430003e-01
7.56120563e-01 2.70795226e-01 1.59958795e-01 1.28210187e+00
6.88462555e-01 -5.04203796e-01 7.53566742e-01 -9.73412335e-01
-1.04734898e+00 3.47196251e-01 -8.04108560e-01 1.31596312e-01
1.50816083e-01 2.08497882e-01 1.19252753e+00 -9.71363902e-01
9.23232555e-01 1.30214953e+00 7.22195804e-01 5.42557418e-01
-1.39276445e+00 -4.79907632e-01 2.17267498e-01 1.02305263e-01
-1.57634306e+00 -2.70050794e-01 6.95586979e-01 -8.93767416e-01
7.09944844e-01 2.81934500e-01 4.03810173e-01 1.39991045e+00
1.00532174e-01 7.46887863e-01 1.39841735e+00 -4.02762890e-01
-2.06148595e-01 4.89801824e-01 7.09974095e-02 8.98149490e-01
3.88576061e-01 -2.61077374e-01 -4.97103631e-01 -3.71603698e-01
3.31687808e-01 -1.12106591e-01 -5.65561533e-01 -5.45341253e-01
-1.48259926e+00 9.61118639e-01 3.78220052e-01 6.04588389e-01
-3.41026127e-01 -5.13250351e-01 5.98126173e-01 4.51791547e-02
4.19076592e-01 2.47948438e-01 -2.97414124e-01 6.26758873e-01
-8.21256220e-01 -3.46023768e-01 4.17886704e-01 5.43953776e-01
6.12586319e-01 -1.47344068e-01 -2.85473019e-01 1.12963402e+00
4.59062040e-01 9.10815418e-01 6.20094001e-01 -5.56326151e-01
7.22842991e-01 7.14271963e-01 -4.70559835e-01 -1.35034955e+00
-3.21633369e-01 -7.07160890e-01 -1.20203245e+00 -2.02054605e-01
2.11036786e-01 3.24387252e-02 -7.35682905e-01 1.93689132e+00
2.03458369e-01 -1.86494276e-01 3.90572160e-01 8.34426165e-01
1.11884487e+00 1.03859261e-01 3.23006094e-01 -3.03530127e-01
1.78238845e+00 -7.41422236e-01 -6.73781455e-01 -4.03498411e-01
6.77641511e-01 -1.01210797e+00 1.10369098e+00 5.45194373e-02
-5.21701872e-01 -3.36598635e-01 -6.09095573e-01 -2.73850888e-01
-3.37482899e-01 5.78252256e-01 4.43925947e-01 4.20326650e-01
-1.08129489e+00 -3.90080780e-01 -5.64695776e-01 -5.52653849e-01
3.45425606e-01 1.56722099e-01 -6.54742360e-01 -4.48532671e-01
-9.64051545e-01 8.67410004e-01 3.35023403e-01 1.26776144e-01
-3.13342631e-01 -7.05254793e-01 -9.90476251e-01 -3.64282668e-01
3.37594330e-01 -1.07060564e+00 3.53898793e-01 -1.18319452e+00
-7.08561182e-01 1.60798633e+00 -3.99466246e-01 -2.47560591e-01
7.09786415e-01 8.09075087e-02 -5.50742745e-01 3.27330202e-01
6.52301610e-01 6.60706103e-01 7.62299001e-01 -1.43123364e+00
-3.40387821e-01 -4.33265299e-01 -3.62134874e-01 2.82129586e-01
-3.96540880e-01 4.10601646e-02 -8.55285227e-01 -1.00550675e+00
-5.08813299e-02 -1.00751817e+00 -2.12765649e-01 9.21126306e-02
-4.42207456e-01 -1.36656418e-01 2.27280021e-01 -7.94575691e-01
9.02016521e-01 -2.27706861e+00 2.76420057e-01 2.04448298e-01
3.94321620e-01 1.69898584e-01 -4.95080441e-01 1.08289689e-01
-3.77661288e-02 7.52225742e-02 -2.30889916e-01 -5.99571824e-01
-3.38108420e-01 3.46476495e-01 -2.63542309e-03 4.68407869e-01
3.50783616e-01 9.62442577e-01 -8.62069428e-01 -9.46578503e-01
6.73320219e-02 6.55982256e-01 -6.48905993e-01 -4.06686552e-02
-4.62144315e-02 7.51117706e-01 -4.28425044e-01 8.00193131e-01
5.76589406e-01 -6.43135786e-01 4.70081329e-01 -5.91763020e-01
-4.17037075e-03 -2.91685909e-02 -6.84410930e-01 1.78304517e+00
-4.34492677e-01 4.32551771e-01 3.32162976e-01 -9.78001297e-01
7.80414224e-01 3.24723631e-01 6.85631692e-01 -7.78693080e-01
3.68500128e-02 4.06713873e-01 1.59796269e-03 -9.14256513e-01
1.35964021e-01 -2.26741984e-01 -4.29034866e-02 4.92135696e-02
-1.60073657e-02 2.22854227e-01 2.09406674e-01 2.13390484e-01
4.73820001e-01 -3.78930330e-01 2.87903756e-01 -2.43339285e-01
9.51458693e-01 1.84730813e-01 6.88427210e-01 5.98137081e-01
-4.67562377e-01 5.28125525e-01 3.09471101e-01 -4.95485170e-03
-6.02301478e-01 -1.20233548e+00 -3.65025997e-01 9.16925848e-01
-4.33801189e-02 -4.12637860e-01 -2.78467923e-01 -7.41524041e-01
8.36756919e-03 3.37757707e-01 -6.36964500e-01 6.54675672e-03
-3.87961656e-01 -1.13054049e+00 6.16380394e-01 7.97879770e-02
2.78062612e-01 -6.31014228e-01 -1.15854472e-01 -2.06576705e-01
-8.37878406e-01 -1.67860317e+00 -7.29267716e-01 -1.96003556e-01
-6.33403838e-01 -1.21944892e+00 -1.05798197e+00 -1.01266158e+00
8.71404409e-01 2.60784537e-01 1.29985046e+00 -1.33999124e-01
-6.02043748e-01 8.15638423e-01 -5.19111939e-02 -1.88857973e-01
-6.37883842e-01 -5.18056145e-03 -1.84446983e-02 4.35640782e-01
6.68937325e-01 -8.99420213e-03 -4.54092056e-01 8.93949240e-04
-5.30687213e-01 2.55352318e-01 8.49125624e-01 9.73093092e-01
8.46986175e-01 -4.58940893e-01 4.34934288e-01 -7.24149287e-01
5.16205966e-01 -4.37225640e-01 -1.37506753e-01 7.93252468e-01
-3.48542720e-01 -2.70810463e-02 3.44529510e-01 -5.77399611e-01
-8.38844419e-01 -1.63202345e-01 1.15255222e-01 -3.68544191e-01
-1.92949459e-01 8.11650217e-01 -2.95384806e-02 1.57309204e-01
5.37633121e-01 3.30799937e-01 1.38533324e-01 -3.21676880e-01
3.55078131e-01 6.42610788e-01 5.25078535e-01 -5.48314810e-01
4.98709112e-01 7.02091098e-01 -2.49252424e-01 -1.16315520e+00
-8.51278186e-01 -7.27062941e-01 -7.24871993e-01 -2.10450292e-02
1.28112197e+00 -1.37821674e+00 -4.01962370e-01 2.34888077e-01
-1.14070404e+00 1.96413159e-01 1.55188188e-01 7.71316707e-01
5.28976806e-02 6.56715333e-01 -5.37723482e-01 -5.42878866e-01
-3.98490518e-01 -1.41741574e+00 1.02703559e+00 -2.32053757e-01
-3.12712818e-01 -1.27127123e+00 1.34883136e-01 1.02351999e+00
-2.41690874e-02 1.97106957e-01 1.13594711e+00 -5.33172607e-01
-4.21362311e-01 9.86653343e-02 -5.70634544e-01 5.22961736e-01
5.43422937e-01 -8.69955122e-02 -9.29792225e-01 -4.17432994e-01
-1.36229336e-01 -3.94379884e-01 9.90530968e-01 3.49372953e-01
6.12333119e-01 -8.00967738e-02 -2.68579036e-01 5.86002529e-01
1.46121883e+00 -3.70355815e-01 9.40731391e-02 1.84332207e-01
1.11567128e+00 9.53136444e-01 3.56878877e-01 -3.14603746e-02
8.18233907e-01 5.35009801e-01 8.31441730e-02 -5.81650317e-01
-2.90695459e-01 4.07059416e-02 4.02966082e-01 1.38706291e+00
-2.32751295e-02 1.87983558e-01 -1.15674508e+00 8.10125053e-01
-1.63380361e+00 -6.55273795e-01 -3.06696147e-01 2.17846227e+00
1.15068221e+00 -4.32519823e-01 -4.33284603e-02 -5.41402698e-01
7.25706756e-01 5.03611602e-02 -4.01395500e-01 4.11932766e-02
-7.18246579e-01 -1.47481576e-01 3.21401507e-01 5.58864653e-01
-1.09495783e+00 8.12891483e-01 5.47022772e+00 7.94727266e-01
-1.59515727e+00 3.42739850e-01 4.12971109e-01 -3.27894427e-02
-5.04893661e-01 -4.86865968e-01 -7.20090091e-01 3.80938172e-01
4.44271594e-01 6.60903677e-02 -1.15001537e-01 4.73556787e-01
2.85452276e-01 1.55935641e-02 -7.83563375e-01 1.26199186e+00
5.85535347e-01 -1.24998856e+00 4.70673829e-01 1.41517967e-01
6.69660568e-01 5.29101610e-01 3.12662333e-01 -1.51917078e-02
3.08282720e-03 -8.81984770e-01 4.89757627e-01 2.94749886e-01
9.58715737e-01 -3.54260713e-01 7.12224662e-01 1.90475896e-01
-1.04570067e+00 2.30028853e-01 -1.76256642e-01 7.11382389e-01
6.92648888e-02 6.16194010e-01 -9.83384967e-01 8.92137825e-01
7.87558556e-01 8.92445028e-01 -7.90166438e-01 6.41075075e-01
5.97115718e-02 2.36746639e-01 -6.19512051e-02 4.08100665e-01
2.28632599e-01 -3.83885413e-01 4.87373322e-01 1.62867033e+00
6.41514957e-02 -3.86214584e-01 5.19323468e-01 8.36332381e-01
1.59277078e-02 7.20666170e-01 -7.28208482e-01 -1.66575432e-01
1.45372227e-01 9.93475795e-01 -4.94267076e-01 -3.03377956e-01
-8.28760266e-01 9.08158600e-01 4.24350947e-01 5.08143783e-01
-5.87696850e-01 3.69715303e-01 6.27945423e-01 -2.22833440e-01
7.21220970e-02 -1.52328312e-01 -4.55013812e-01 -1.54237163e+00
8.85104313e-02 -1.12609994e+00 6.80557013e-01 -4.75775868e-01
-1.81876028e+00 6.94832921e-01 -3.06929201e-01 -1.19479358e+00
1.50144264e-01 -7.20769823e-01 1.85715824e-01 1.09631968e+00
-1.98360729e+00 -1.82574642e+00 -1.05783693e-01 1.11250472e+00
3.12121272e-01 -4.41548318e-01 8.92391562e-01 7.70390272e-01
-6.23296201e-01 7.91512549e-01 -7.42668584e-02 2.95208454e-01
1.19150662e+00 -9.45750594e-01 -4.68046367e-01 7.36150146e-01
2.45051399e-01 9.88148570e-01 3.72725636e-01 -7.76295185e-01
-1.12552977e+00 -1.07504082e+00 1.27253532e+00 -5.14967918e-01
8.01280916e-01 -2.44030267e-01 -1.09195590e+00 4.05148476e-01
1.45140633e-01 -1.19475350e-01 1.18891418e+00 2.11958647e-01
-8.19643676e-01 4.62860614e-02 -8.67694795e-01 9.05174017e-01
6.79000974e-01 -1.15210450e+00 -5.61365008e-01 4.65636283e-01
4.96748358e-01 -2.97949966e-02 -8.56055915e-01 5.45357823e-01
3.55928510e-01 -7.19249129e-01 1.26086414e+00 -4.95324731e-01
3.41828883e-01 -3.24200749e-01 -3.54371399e-01 -8.71381342e-01
-1.37299344e-01 -1.55161738e-01 4.81142998e-01 1.20317996e+00
4.41036880e-01 -8.94086361e-01 3.47539842e-01 2.38453388e-01
1.59258455e-01 -6.34930253e-01 -8.94073188e-01 -4.51538175e-01
2.54807889e-01 -4.77683991e-01 -1.62514642e-01 1.59332669e+00
-3.80362004e-01 4.91406858e-01 -3.39243114e-01 5.15484452e-01
7.72751749e-01 2.11704120e-01 6.32776439e-01 -1.12352717e+00
-2.32983232e-01 -5.78645349e-01 -1.27455607e-01 -5.86264670e-01
2.79281110e-01 -1.47641182e+00 -2.72153556e-01 -1.53044629e+00
5.71989357e-01 -5.18792927e-01 -3.35742742e-01 5.46903193e-01
-3.77329141e-01 4.36827987e-01 1.44800857e-01 5.06598830e-01
-2.96249956e-01 3.98609996e-01 1.41510296e+00 -4.54067469e-01
-3.16659436e-02 -4.39991057e-01 -7.71498144e-01 8.53329360e-01
5.50217032e-01 -4.55220312e-01 -3.32626075e-01 -5.94089270e-01
2.02054709e-01 -2.75506377e-01 5.81358790e-01 -4.44014430e-01
1.38606265e-01 -3.01340193e-01 1.87415406e-02 -5.84026992e-01
2.44241714e-01 -7.63829052e-01 -7.30471835e-02 5.82121015e-01
-3.04199636e-01 8.91853347e-02 1.82378367e-01 6.79831326e-01
-3.63688201e-01 1.01596020e-01 7.42652833e-01 -2.38551602e-01
-6.74165428e-01 5.60569242e-02 -2.00614214e-01 4.68116671e-01
7.25129962e-01 -3.14522758e-02 -4.19507861e-01 1.76443771e-01
-8.91159594e-01 3.14420730e-01 4.20002162e-01 6.27365351e-01
3.85120928e-01 -1.11032581e+00 -1.05237162e+00 2.84347624e-01
6.94815159e-01 -3.73311520e-01 5.54792047e-01 1.40120506e+00
-3.99726242e-01 4.54963803e-01 5.96563593e-02 -1.14728808e+00
-1.63165462e+00 5.14912367e-01 4.39101189e-01 -4.71907407e-01
-5.45784950e-01 7.41345584e-01 6.76198840e-01 -7.48256087e-01
1.06730305e-01 -2.08000302e-01 -3.29625249e-01 3.77720892e-01
3.01881492e-01 -1.88469678e-01 -3.26060839e-02 -1.15684915e+00
-6.86941087e-01 1.01297534e+00 -1.09108351e-01 3.57550420e-02
9.98514771e-01 -3.48108590e-01 -4.61989552e-01 6.44873142e-01
1.23863935e+00 4.78656054e-01 -5.34829855e-01 -5.95708132e-01
7.67429471e-02 -1.36614949e-01 -1.63241565e-01 -8.92430246e-01
-1.11354065e+00 9.24055338e-01 7.26688981e-01 -4.22254086e-01
8.89511585e-01 1.67138070e-01 4.85138357e-01 8.15987736e-02
2.79334486e-01 -1.07430899e+00 1.38214856e-01 3.15766364e-01
9.19390142e-01 -1.72235858e+00 -4.87043783e-02 -5.84392548e-01
-1.01986194e+00 6.35244727e-01 3.41559738e-01 2.86630929e-01
4.98594970e-01 -6.97085932e-02 7.10854232e-01 -3.68816406e-01
-3.76820952e-01 -4.82414365e-01 9.58616912e-01 6.62245214e-01
6.97898507e-01 2.23101810e-01 -3.18795651e-01 4.26302761e-01
-1.04440590e-02 -1.61473051e-01 2.34549213e-02 4.84947890e-01
2.28044078e-01 -1.12463069e+00 -5.38750648e-01 5.96719310e-02
-6.97525084e-01 -3.67917866e-01 -4.58329231e-01 6.57132447e-01
5.44035971e-01 8.78472865e-01 -5.80716804e-02 4.56857122e-02
-2.45481767e-02 4.40554544e-02 3.28186393e-01 -6.70498669e-01
-5.87268233e-01 3.66451800e-01 3.66390087e-02 -2.03421116e-01
-8.17321599e-01 -5.97373307e-01 -9.34988141e-01 1.59470722e-01
-5.00004999e-02 -5.76680489e-02 5.00975668e-01 9.38023388e-01
5.70746779e-01 4.76404309e-01 2.07757398e-01 -2.19239607e-01
-1.38989404e-01 -5.45564711e-01 -1.69008002e-01 6.45106316e-01
5.93499184e-01 -5.05913556e-01 -2.73987114e-01 1.57634109e-01] | [10.908919334411621, 1.505889892578125] |
81492974-28d0-49a1-bcf6-62f467e8b0bf | instance-segmentation-by-deep-coloring | 1807.10007 | null | http://arxiv.org/abs/1807.10007v1 | http://arxiv.org/pdf/1807.10007v1.pdf | Instance Segmentation by Deep Coloring | We propose a new and, arguably, a very simple reduction of instance
segmentation to semantic segmentation. This reduction allows to train
feed-forward non-recurrent deep instance segmentation systems in an end-to-end
fashion using architectures that have been proposed for semantic segmentation.
Our approach proceeds by introducing a fixed number of labels (colors) and then
dynamically assigning object instances to those labels during training
(coloring). A standard semantic segmentation objective is then used to train a
network that can color previously unseen images. At test time, individual
object instances can be recovered from the output of the trained convolutional
network using simple connected component analysis. In the experimental
validation, the coloring approach is shown to be capable of solving diverse
instance segmentation tasks arising in autonomous driving (the Cityscapes
benchmark), plant phenotyping (the CVPPP leaf segmentation challenge), and
high-throughput microscopy image analysis.
The source code is publicly available:
https://github.com/kulikovv/DeepColoring. | ['Victor Yurchenko', 'Victor Kulikov', 'Victor Lempitsky'] | 2018-07-26 | null | null | null | null | ['plant-phenotyping'] | ['computer-vision'] | [ 7.27597415e-01 4.44927096e-01 5.66153489e-02 -5.38360238e-01
-7.47623265e-01 -1.12432635e+00 4.14591104e-01 1.43870130e-01
-1.41273513e-01 4.88411963e-01 -7.68476188e-01 -3.99724662e-01
-5.40782558e-03 -8.29887211e-01 -9.32237804e-01 -7.71111012e-01
6.30064905e-02 8.85978758e-01 2.69700706e-01 5.20732328e-02
2.56336898e-01 5.38636744e-01 -1.65423644e+00 3.51352096e-01
7.78992891e-01 1.10228896e+00 5.25786340e-01 1.10028660e+00
-2.98964620e-01 4.58105415e-01 -5.69754362e-01 3.99514437e-02
2.77566642e-01 -5.44771791e-01 -1.40378511e+00 4.81387883e-01
3.95322591e-01 -5.16402163e-02 3.68560940e-01 9.28653181e-01
1.45665243e-01 -5.25198393e-02 3.24301690e-01 -1.30456960e+00
-6.74589813e-01 5.80106437e-01 -3.81786257e-01 -7.62461498e-02
-2.63890624e-01 2.48063818e-01 9.30147350e-01 -6.01983726e-01
7.69901693e-01 9.27802682e-01 4.38862503e-01 7.13975132e-01
-1.60500097e+00 -1.66566670e-01 3.14066023e-01 -5.43202087e-02
-1.03913450e+00 -1.12339512e-01 6.56411171e-01 -3.27198654e-01
4.91809785e-01 3.96037221e-01 8.57833922e-01 7.06363380e-01
-5.11874735e-01 9.22603488e-01 1.09002268e+00 -2.43550003e-01
5.58100641e-01 -6.02419861e-02 3.27457160e-01 9.10142124e-01
-9.73585472e-02 -1.54421806e-01 -5.65312207e-02 1.06331445e-01
5.23169816e-01 -7.06327111e-02 -2.06352577e-01 -4.69263732e-01
-9.89862740e-01 6.33428872e-01 8.24299395e-01 1.92700848e-01
-2.11488858e-01 2.81421095e-01 2.99014539e-01 1.06446547e-02
5.00702202e-01 5.39066136e-01 -9.74498630e-01 1.80532396e-01
-9.75286722e-01 1.16721019e-01 9.26748931e-01 7.61738956e-01
9.55750108e-01 -1.11543767e-01 -2.22525403e-01 7.23540485e-01
2.94154156e-02 3.39978665e-01 2.36559696e-02 -1.39849401e+00
-2.45582163e-01 6.93540514e-01 6.54659197e-02 -3.61579716e-01
-5.02455473e-01 -7.28385448e-02 -5.41363835e-01 5.26955307e-01
7.18446076e-01 2.35503688e-02 -1.44839549e+00 1.75399148e+00
4.85921890e-01 3.87889951e-01 1.33687839e-01 9.88215029e-01
8.20787430e-01 6.90312445e-01 3.69883150e-01 2.66898990e-01
1.25704980e+00 -9.52096999e-01 -1.22052029e-01 -2.53708929e-01
7.59550035e-01 -4.24187273e-01 1.14609838e+00 4.57330793e-01
-9.85363424e-01 -5.02987802e-01 -9.03731167e-01 -1.77052602e-01
-8.67817104e-01 5.03114462e-01 7.71311164e-01 4.92219180e-01
-1.28247762e+00 7.63138771e-01 -9.56314147e-01 -4.15004939e-01
9.79857624e-01 4.59260911e-01 -4.79753055e-02 -1.62760586e-01
-5.93815923e-01 3.71489584e-01 6.47042572e-01 4.21404302e-01
-1.33074141e+00 -6.58368051e-01 -4.50822204e-01 8.08995664e-02
4.49594885e-01 -4.32122976e-01 1.19648778e+00 -1.51902628e+00
-1.68024099e+00 1.39282048e+00 -9.98219103e-02 -5.04553497e-01
2.50368059e-01 -2.07205787e-02 2.66418934e-01 2.43361831e-01
2.97681633e-02 1.33144832e+00 7.03224838e-01 -1.47149324e+00
-6.58026040e-01 -5.35660148e-01 2.90321112e-01 -5.26842587e-02
2.21285060e-01 -3.84914756e-01 -4.69514132e-01 -8.19817558e-02
2.62843907e-01 -1.35913658e+00 -3.38662058e-01 2.87438303e-01
-8.01269531e-01 -1.64162740e-01 1.15531933e+00 -6.19872034e-01
2.73987353e-01 -1.95523167e+00 5.87596357e-01 1.34229183e-01
1.90159440e-01 4.95140851e-01 -5.81501365e-01 1.85449682e-02
-2.29875386e-01 3.06351125e-01 -1.01490545e+00 -2.36350432e-01
-2.15069398e-01 2.89012939e-01 -6.64533302e-02 2.30048865e-01
7.30859518e-01 1.20805681e+00 -8.79806519e-01 -3.20126489e-02
3.45270753e-01 4.50763434e-01 -2.90243298e-01 2.44961560e-01
-9.97436225e-01 7.45563745e-01 -4.32745278e-01 8.18976462e-01
7.40433514e-01 -5.28237641e-01 2.52059042e-01 2.22415701e-01
-8.17332044e-02 -1.97200533e-02 -6.95274770e-01 1.98114550e+00
-2.42748201e-01 6.60765171e-01 1.11979365e-01 -1.40170586e+00
8.40609074e-01 -5.99427335e-02 5.17917812e-01 -5.30741513e-01
-5.12154028e-02 1.28985181e-01 -1.47027478e-01 -7.13801831e-02
1.65334180e-01 5.95522702e-01 9.87777337e-02 1.52538940e-01
1.43888533e-01 -4.18505579e-01 5.02535641e-01 3.11797857e-01
9.97696579e-01 7.26599395e-01 -3.32016468e-01 -3.73399377e-01
4.82723981e-01 6.14534557e-01 4.24220145e-01 5.26901066e-01
-1.95849732e-01 7.29736507e-01 7.95505643e-01 -4.69763845e-01
-1.08401728e+00 -9.90353763e-01 -9.07939300e-02 1.19442070e+00
7.91571215e-02 -1.20348945e-01 -1.13392103e+00 -7.55848706e-01
-1.19321525e-01 5.80736458e-01 -7.16934502e-01 -1.05933458e-01
-2.78351188e-01 -7.91291535e-01 4.36612964e-01 3.47778171e-01
4.45531458e-01 -1.50901127e+00 -8.51887643e-01 7.83153772e-02
6.57204017e-02 -1.07575440e+00 3.27701747e-01 8.69217873e-01
-8.77569616e-01 -1.26388550e+00 -6.84450805e-01 -8.84664953e-01
7.71092772e-01 1.77480832e-01 1.23413050e+00 3.82456541e-01
-8.09827745e-01 3.26091439e-01 -3.30004275e-01 -2.49289095e-01
-2.94907063e-01 5.31579852e-01 -8.56842518e-01 -1.65747285e-01
2.56003559e-01 -3.01574886e-01 -6.92582607e-01 -4.13442999e-02
-1.06932545e+00 2.76810408e-01 3.09873521e-01 8.41827095e-01
1.11095798e+00 -3.26693386e-01 3.93907636e-01 -1.50383818e+00
-8.93176475e-04 -3.78035933e-01 -1.24473763e+00 3.50170910e-01
-3.94087166e-01 -1.24817379e-02 7.75317132e-01 -7.81328529e-02
-6.92751110e-01 7.74366856e-01 -1.24401994e-01 -2.01093808e-01
-7.99529374e-01 3.19211841e-01 -3.16432826e-02 -1.08093888e-01
4.46779341e-01 1.68802008e-01 -1.01650786e-02 -2.61759758e-01
8.20688605e-01 2.23134190e-01 6.50207043e-01 -4.51660752e-01
7.00955451e-01 5.39282262e-01 1.55003697e-01 -6.49603248e-01
-1.04734671e+00 -3.11631858e-01 -9.97529209e-01 -2.54198343e-01
1.18427336e+00 -6.09885693e-01 -7.98735976e-01 5.38199306e-01
-1.00319612e+00 -1.16631269e+00 -4.97249067e-01 -4.27881747e-01
-6.71179056e-01 -6.45555109e-02 -4.47424144e-01 -5.73284626e-01
-1.12319767e-01 -1.16638982e+00 1.41361701e+00 5.98365963e-01
1.88976809e-01 -8.49530756e-01 -1.40445307e-01 3.12166423e-01
2.24679098e-01 5.37613750e-01 1.01610589e+00 -4.47188765e-01
-8.22707593e-01 -1.90940931e-01 -5.17219961e-01 3.08864653e-01
-1.13892108e-01 5.55048704e-01 -1.30908740e+00 -8.20949897e-02
-6.78067148e-01 -7.38446295e-01 1.09338200e+00 5.81776381e-01
1.77428401e+00 8.13870728e-02 -3.65617871e-01 7.98673987e-01
1.62012649e+00 3.44617032e-02 7.05015182e-01 2.59073853e-01
8.13570678e-01 8.04348946e-01 4.21105206e-01 6.57887310e-02
2.27770120e-01 3.78983438e-01 9.06345189e-01 -6.31324232e-01
-1.15794674e-01 2.01969355e-01 -1.32788911e-01 -5.30074351e-02
1.96208254e-01 -2.77258426e-01 -9.81843710e-01 5.84972382e-01
-1.87833774e+00 -7.34409451e-01 -4.20474410e-01 1.94811225e+00
6.31854296e-01 -2.71832198e-02 1.74072713e-01 1.44827545e-01
5.86986005e-01 -2.78309938e-02 -9.71419513e-01 -4.29192841e-01
-1.66271165e-01 4.33464766e-01 6.66860819e-01 4.62744534e-01
-1.32052028e+00 1.53449714e+00 5.53029156e+00 3.34208816e-01
-1.26623893e+00 -8.13856795e-02 1.35575902e+00 1.18696861e-01
-3.57411578e-02 3.18064958e-01 -4.73597109e-01 3.66212219e-01
1.05237198e+00 4.65049773e-01 7.55056620e-01 8.18674922e-01
5.83991408e-03 -3.43199551e-01 -8.78745675e-01 3.55410129e-01
-3.19037229e-01 -1.43934870e+00 -1.26454175e-01 -2.26723358e-01
6.60502017e-01 2.56090254e-01 1.62843674e-01 6.94018081e-02
7.16037929e-01 -9.86048460e-01 5.40013194e-01 2.63961643e-01
7.23981142e-01 -5.26898384e-01 1.96148798e-01 1.47820041e-01
-1.00899792e+00 -8.43923390e-02 -5.77208936e-01 4.37656194e-01
-2.40090653e-01 5.96206725e-01 -7.94204056e-01 3.94176304e-01
7.13851988e-01 7.22500861e-01 -9.32972252e-01 8.66945028e-01
-3.01914304e-01 9.22155917e-01 -4.95015562e-01 2.68408954e-01
4.95682299e-01 -4.21854317e-01 1.78471461e-01 1.00717819e+00
-7.73981512e-02 -2.01392341e-02 4.34552491e-01 1.33059883e+00
-8.54128599e-02 -3.48740339e-01 -4.39832717e-01 -3.13817680e-01
2.24555224e-01 1.80376041e+00 -1.48648000e+00 -3.29518735e-01
-1.49886291e-02 1.47731984e+00 5.24810016e-01 4.96607512e-01
-5.79773426e-01 -2.50461310e-01 3.10651928e-01 -2.95583814e-01
5.08629084e-01 -2.61228997e-02 -3.41136813e-01 -6.43204570e-01
-2.63031334e-01 -4.07159150e-01 2.56942421e-01 -9.49139833e-01
-7.74591923e-01 4.12843287e-01 -3.81136775e-01 -6.10523343e-01
-4.28624749e-02 -8.68969560e-01 -6.04129255e-01 7.62051761e-01
-1.66305196e+00 -1.28238702e+00 -6.85553968e-01 2.11949244e-01
6.07880890e-01 1.90158561e-01 1.08306491e+00 -1.41338691e-01
-8.07661474e-01 -7.91404024e-02 1.51998982e-01 1.22525590e-02
8.90931413e-02 -1.64787471e+00 5.62187016e-01 7.50773370e-01
2.16126844e-01 -1.18375070e-01 3.64129543e-01 -3.01569581e-01
-1.22799444e+00 -1.45400524e+00 3.79979312e-01 -3.93521756e-01
3.58944893e-01 -4.55879986e-01 -1.01149452e+00 6.20609999e-01
1.50637850e-01 3.57494473e-01 3.64668816e-01 -9.37137306e-02
-2.87960261e-01 6.23838976e-02 -1.13268137e+00 3.70198071e-01
9.15763378e-01 -3.28982234e-01 3.56790513e-01 8.20340633e-01
6.35916710e-01 -2.40669414e-01 -6.12767100e-01 1.37374684e-01
2.69814342e-01 -6.71402216e-01 8.32271755e-01 -7.61289835e-01
5.96830130e-01 -5.15681744e-01 9.77834454e-04 -1.17580926e+00
-3.14546555e-01 -5.32415211e-01 2.66217977e-01 1.10745621e+00
6.70551777e-01 -3.33591580e-01 9.95729923e-01 4.56914544e-01
-1.92347959e-01 -7.00195909e-01 -5.55195570e-01 -4.29086268e-01
1.27137050e-01 3.97041589e-02 3.76539409e-01 7.59184837e-01
-7.33246028e-01 1.89565495e-01 2.45604113e-01 3.63287449e-01
4.85081166e-01 4.02133673e-01 5.36132872e-01 -1.56668019e+00
-2.36372381e-01 -5.52165031e-01 -2.67808408e-01 -7.13299990e-01
3.64078403e-01 -1.21021497e+00 3.28113735e-01 -1.67495716e+00
1.00515440e-01 -7.22957492e-01 -3.94936025e-01 8.17823410e-01
-2.15907827e-01 4.87831950e-01 1.95104584e-01 1.19135231e-01
-7.58032143e-01 5.66643775e-02 1.04223907e+00 -3.83396178e-01
-1.89715564e-01 -2.73738876e-02 -4.94474053e-01 4.17001784e-01
1.21018302e+00 -5.83635747e-01 -3.26947004e-01 -5.59407949e-01
-2.01954097e-02 -1.28606737e-01 8.60107422e-01 -9.47303593e-01
-2.66048193e-01 4.51235945e-04 5.38186669e-01 -4.66152847e-01
3.13758016e-01 -7.03870773e-01 -2.41623316e-02 5.43358862e-01
-6.76701486e-01 -1.90414146e-01 3.95738304e-01 4.31227893e-01
1.97191134e-01 -3.98415327e-01 1.05874252e+00 -4.04723853e-01
-9.28340793e-01 3.55030835e-01 -3.73728395e-01 1.42624453e-02
1.17948806e+00 -1.15358040e-01 -4.33286697e-01 1.15098678e-01
-1.01265061e+00 3.23613346e-01 5.35769284e-01 2.62306780e-01
3.08772147e-01 -7.39482999e-01 -6.17541313e-01 8.91781300e-02
1.81541249e-01 3.26025367e-01 8.82840827e-02 6.01502955e-01
-8.52808177e-01 1.81687742e-01 -3.30235362e-01 -1.03078771e+00
-9.90951896e-01 3.71058822e-01 5.10638297e-01 -8.80545750e-02
-5.99357188e-01 1.10559845e+00 1.41697362e-01 -7.89408088e-01
1.97263241e-01 -2.93416679e-01 -1.18654661e-01 -2.44287029e-01
9.17935520e-02 -9.44308937e-02 5.88722527e-02 -3.92227769e-01
-2.00556219e-01 4.52631831e-01 4.43680631e-03 7.35807866e-02
1.46946931e+00 9.69944671e-02 -1.56699091e-01 5.40595889e-01
1.06562054e+00 -8.34331930e-01 -1.53282607e+00 1.21944383e-01
3.75933647e-01 3.04822270e-02 1.23085529e-01 -1.19770741e+00
-1.45939112e+00 9.22137022e-01 9.95689154e-01 5.96079051e-01
1.25286591e+00 -5.32468827e-03 5.47396004e-01 5.90346277e-01
-4.31747660e-02 -9.88024950e-01 -1.75385341e-01 3.40229154e-01
4.61583167e-01 -1.37240374e+00 -4.51642543e-01 -5.26056588e-01
-4.37914193e-01 1.11669672e+00 5.74894667e-01 -3.39229703e-01
5.20274818e-01 3.29780966e-01 1.81147948e-01 -3.09443742e-01
-7.42067397e-01 -6.88048124e-01 -8.64102393e-02 8.18064749e-01
4.17774439e-01 2.31161058e-01 9.30019096e-02 4.14184965e-02
2.42726594e-01 3.28929983e-02 4.30950552e-01 9.03265417e-01
-5.10413885e-01 -1.10875356e+00 6.97772577e-02 4.42236096e-01
-2.84435362e-01 -8.10981318e-02 -1.00404227e+00 5.58001637e-01
1.06367230e-01 6.70560837e-01 1.37098610e-01 2.92771328e-02
5.90140522e-02 3.58110785e-01 3.40728074e-01 -7.27965236e-01
-6.34482443e-01 -1.66589752e-01 -1.71385482e-01 -8.06344330e-01
-4.77956474e-01 -5.54962695e-01 -1.56516600e+00 8.91831145e-02
-1.19837590e-01 -9.76331383e-02 9.68850791e-01 7.55044103e-01
5.00776350e-01 8.62793744e-01 4.99740362e-01 -1.03406262e+00
8.84508789e-02 -6.32903874e-01 -3.66869241e-01 4.97098207e-01
2.34608576e-01 -2.48417705e-01 1.07820481e-02 4.92514819e-01] | [9.526178359985352, 0.14816851913928986] |
1090f896-00da-4379-9679-d3599adb561a | an-exploration-into-the-performance-of | 2305.05443 | null | https://arxiv.org/abs/2305.05443v1 | https://arxiv.org/pdf/2305.05443v1.pdf | An Exploration into the Performance of Unsupervised Cross-Task Speech Representations for "In the Wild'' Edge Applications | Unsupervised speech models are becoming ubiquitous in the speech and machine learning communities. Upstream models are responsible for learning meaningful representations from raw audio. Later, these representations serve as input to downstream models to solve a number of tasks, such as keyword spotting or emotion recognition. As edge speech applications start to emerge, it is important to gauge how robust these cross-task representations are on edge devices with limited resources and different noise levels. To this end, in this study we evaluate the robustness of four different versions of HuBERT, namely: base, large, and extra-large versions, as well as a recent version termed Robust-HuBERT. Tests are conducted under different additive and convolutive noise conditions for three downstream tasks: keyword spotting, intent classification, and emotion recognition. Our results show that while larger models can provide some important robustness to environmental factors, they may not be applicable to edge applications. Smaller models, on the other hand, showed substantial accuracy drops in noisy conditions, especially in the presence of room reverberation. These findings suggest that cross-task speech representations are not yet ready for edge applications and innovations are still needed. | ['Tiago H. Falk', 'Mehdi Rezagholizadeh', 'Anderson Avila', 'Arthur Pimentel', 'Heitor Guimarães'] | 2023-05-09 | null | null | null | null | ['intent-classification', 'keyword-spotting'] | ['natural-language-processing', 'speech'] | [ 1.78315759e-01 -2.18527645e-01 1.45040661e-01 -2.63092309e-01
-1.06048155e+00 -5.62000811e-01 3.17989975e-01 2.71986932e-01
-1.53376281e-01 3.85816693e-01 6.32575929e-01 -3.51185828e-01
-2.02082932e-01 -2.26657122e-01 -3.46755058e-01 -5.18587112e-01
1.91955920e-02 -3.16538036e-01 -5.79701960e-02 -2.50750065e-01
-6.02341704e-02 2.62789786e-01 -1.82247174e+00 4.79095399e-01
5.16588926e-01 1.04987943e+00 2.69701272e-01 9.35138762e-01
-3.81267108e-02 7.96822906e-01 -1.01677036e+00 1.03337802e-01
-1.12443700e-01 -1.51900873e-01 -4.08367038e-01 -1.49084896e-01
-7.33894780e-02 -1.89160675e-01 -3.44521999e-01 6.31299555e-01
1.00461936e+00 4.33278143e-01 4.22681481e-01 -1.09635830e+00
-6.65452003e-01 5.77055693e-01 -4.32651974e-02 3.99963111e-01
5.24594486e-01 1.19827554e-01 9.59692955e-01 -8.61595213e-01
7.55587891e-02 1.07692325e+00 8.95664573e-01 4.75253314e-01
-1.17226017e+00 -6.51377201e-01 2.42036849e-01 2.02212185e-01
-1.32879746e+00 -1.26622319e+00 6.30945444e-01 -3.38100106e-01
1.32735729e+00 5.96736312e-01 1.24089338e-01 1.48363960e+00
-6.98730946e-02 6.10965431e-01 1.06042349e+00 -3.58133644e-01
4.15645480e-01 2.72216231e-01 1.25468671e-01 2.76599657e-02
-9.96905416e-02 9.35762078e-02 -1.01847482e+00 -3.02556396e-01
1.65667504e-01 -2.74194092e-01 -4.78319585e-01 5.07148862e-01
-8.94771278e-01 5.00099301e-01 1.24496959e-01 5.61182082e-01
-5.12678325e-01 3.13967913e-02 5.18324256e-01 4.67932075e-01
8.94803464e-01 4.60628986e-01 -5.46150565e-01 -7.11671114e-01
-8.18502128e-01 -3.12959105e-01 7.93048263e-01 8.03637505e-01
2.51117498e-01 4.44610327e-01 -2.02569276e-01 1.18553579e+00
2.43727654e-01 3.12594473e-01 4.99202132e-01 -5.12099266e-01
4.55157340e-01 -3.44052203e-02 -7.97200128e-02 -9.13407624e-01
-3.65639329e-01 -6.22345686e-01 -5.11262178e-01 -5.74258603e-02
1.71142351e-02 -3.64881188e-01 -8.14520776e-01 1.67555666e+00
-1.16727486e-01 2.77379692e-01 2.90490121e-01 7.70211816e-01
9.98152614e-01 8.55014443e-01 1.23321615e-01 -2.53278673e-01
1.07999349e+00 -7.95728743e-01 -9.34556425e-01 -6.40465379e-01
5.32035649e-01 -9.94836807e-01 1.13753331e+00 5.65964103e-01
-8.73759508e-01 -5.79341829e-01 -9.47140217e-01 2.07365841e-01
-3.95563185e-01 1.07448801e-01 3.78324449e-01 1.12163031e+00
-1.18798625e+00 3.42295825e-01 -5.20659626e-01 -4.05267507e-01
-3.53079587e-02 1.50862888e-01 -3.27996373e-01 1.09963179e-01
-1.16912580e+00 6.56349957e-01 -1.75360128e-01 2.63487667e-01
-7.82289147e-01 -3.68009001e-01 -7.90398061e-01 2.76035070e-01
8.62040222e-02 -2.48295352e-01 1.39433503e+00 -9.67444301e-01
-1.55288661e+00 3.45503628e-01 -4.02951628e-01 -3.08709860e-01
3.77198979e-02 -3.98280084e-01 -9.66681182e-01 -8.32607970e-02
-4.67932075e-02 1.46299526e-01 9.53364611e-01 -1.16016984e+00
-3.56146991e-01 -2.43298367e-01 -2.02928469e-01 2.03781009e-01
-6.41956687e-01 3.34698737e-01 -2.71173090e-01 -8.03095281e-01
6.52752444e-02 -8.49158883e-01 1.29617050e-01 -2.46828392e-01
-2.44825259e-01 -1.61345247e-02 8.32642913e-01 -9.49465573e-01
1.41639876e+00 -2.56152844e+00 -3.73676479e-01 6.83294982e-02
-7.00178221e-02 2.30506569e-01 -1.57037467e-01 5.76524556e-01
-2.50906348e-01 4.58609432e-01 1.98244765e-01 -5.91426373e-01
-6.20520599e-02 -7.53260553e-02 -2.14571029e-01 1.87264189e-01
1.87181637e-01 4.72348064e-01 -6.91320419e-01 -6.93172440e-02
8.77742171e-02 5.85513294e-01 -4.30451006e-01 2.88229138e-02
1.67240232e-01 2.71672904e-01 -2.12384835e-01 5.99678695e-01
1.85130388e-01 1.63647115e-01 -2.27642715e-01 -2.45367028e-02
-1.33397385e-01 6.58908427e-01 -1.07126641e+00 1.41187191e+00
-8.81960154e-01 1.13235176e+00 4.96119142e-01 -8.58040571e-01
9.94968474e-01 7.17010498e-01 1.92634255e-01 -6.92991138e-01
7.68751726e-02 1.39737472e-01 1.25053793e-01 -6.12143815e-01
6.83051407e-01 -2.10217535e-01 -3.15645151e-02 3.56083304e-01
-1.62251949e-01 -1.12770766e-01 -2.93824643e-01 6.06976040e-02
1.17173278e+00 -5.03418028e-01 -1.59680862e-02 5.85553907e-02
8.61036777e-02 -4.76837784e-01 3.94206017e-01 7.12627888e-01
-3.15982401e-01 8.21568131e-01 2.23433077e-01 2.14697123e-01
-4.88361657e-01 -9.70161557e-01 1.00424923e-01 1.38035667e+00
-1.93988115e-01 -7.93269396e-01 -5.46909809e-01 -2.20265031e-01
-2.77250469e-01 1.05389214e+00 -2.74848014e-01 -5.92260897e-01
9.66781378e-02 -4.50954348e-01 8.36411655e-01 6.68848455e-01
1.30538091e-01 -8.80288899e-01 -3.19421679e-01 1.95418447e-01
-3.34340602e-01 -1.19989097e+00 -5.14177322e-01 5.13036013e-01
-4.70742941e-01 -4.30840105e-01 -6.07201874e-01 -5.28186738e-01
1.98373914e-01 5.43176770e-01 7.87758350e-01 -2.85185188e-01
-1.60383228e-02 8.43467474e-01 -6.74353600e-01 -6.84543669e-01
-5.11869431e-01 -7.25802109e-02 2.96232045e-01 1.70557648e-01
1.58074647e-01 -5.88286936e-01 -4.16641623e-01 4.39533770e-01
-6.97190583e-01 -2.67799199e-01 5.52019060e-01 6.82291090e-01
1.07849188e-01 1.86869740e-01 9.35007811e-01 -2.92745441e-01
1.19419134e+00 -6.04822695e-01 1.55996069e-01 1.33387685e-01
-4.10555720e-01 -1.86045215e-01 4.74184275e-01 -5.38454592e-01
-1.14150918e+00 -2.73741066e-01 -4.01990175e-01 -4.16568458e-01
-2.75856376e-01 6.70231640e-01 -2.56620497e-01 3.42193842e-01
9.43819165e-01 2.95308568e-02 -1.68317646e-01 -5.12356460e-01
1.40479058e-01 1.45767164e+00 2.49373168e-01 -4.71792310e-01
4.45363671e-01 1.41627431e-01 -6.48148179e-01 -1.56872535e+00
-3.62012744e-01 -5.88763237e-01 1.37030751e-01 -3.16474825e-01
7.43520021e-01 -9.93964434e-01 -3.17674428e-01 4.29027915e-01
-1.03560483e+00 -4.63215411e-01 -2.84908116e-01 5.97949386e-01
-2.82636613e-01 2.33082727e-01 -5.82814038e-01 -1.27602398e+00
-2.49004319e-01 -1.20132434e+00 1.18539703e+00 1.53138936e-01
-6.17525458e-01 -7.12918758e-01 -3.82128447e-01 5.79833567e-01
7.27043748e-01 -3.69437933e-01 7.46835351e-01 -8.74336481e-01
1.11234404e-01 -3.23155642e-01 1.39538318e-01 6.10314131e-01
4.83293116e-01 -2.31879042e-03 -1.66750050e+00 -1.84286803e-01
2.18081534e-01 -2.16160476e-01 6.28156722e-01 3.42894971e-01
1.10978436e+00 -1.69332564e-01 -1.01893112e-01 3.08818281e-01
5.98442674e-01 5.58709383e-01 5.89642227e-01 -1.37636038e-02
2.05227941e-01 6.68636918e-01 4.58164275e-01 3.90010327e-01
-1.12965517e-02 5.66171050e-01 8.29740316e-02 -9.09513459e-02
-2.85429507e-01 -2.86631018e-01 7.04383016e-01 1.17863071e+00
3.85783762e-01 -5.52426040e-01 -9.86712277e-01 5.34420848e-01
-1.42262840e+00 -8.20869684e-01 2.56404001e-02 2.25222611e+00
3.89294833e-01 3.21793199e-01 -6.84129819e-03 5.16982675e-01
7.31158137e-01 1.91292346e-01 -4.88462031e-01 -5.59707046e-01
-1.54475465e-01 1.21835470e-01 9.56733674e-02 2.93626010e-01
-8.55670631e-01 6.55154526e-01 6.78016567e+00 6.39990807e-01
-1.34087765e+00 2.72334725e-01 9.04482067e-01 -1.59835771e-01
-3.23883057e-01 -1.33203313e-01 -2.70777881e-01 3.74112546e-01
1.34346318e+00 -6.20324872e-02 3.96165192e-01 7.94071436e-01
7.28782773e-01 -1.77336633e-01 -9.72181678e-01 1.31074727e+00
9.72403809e-02 -8.88940513e-01 -4.99377400e-01 -1.47055686e-01
3.42907101e-01 1.05814822e-01 1.37423515e-01 5.08298278e-01
-1.26382500e-01 -1.05981493e+00 7.77678907e-01 2.90126979e-01
8.10185850e-01 -5.53967118e-01 4.90698457e-01 3.07844430e-01
-1.17884231e+00 -2.92660564e-01 -1.25312470e-02 -2.71487087e-01
1.21344887e-01 5.64027727e-01 -9.76410151e-01 2.64266640e-01
7.76443481e-01 2.82558531e-01 -3.91415179e-01 8.04333806e-01
-6.67189807e-02 1.01245654e+00 -4.00583267e-01 -1.05720177e-01
-8.61516893e-02 3.83644342e-01 5.55368125e-01 1.24382424e+00
6.15247428e-01 -6.29796609e-02 -1.09603085e-01 6.32401228e-01
6.31441101e-02 5.36205526e-03 -8.33564878e-01 -4.71142203e-01
8.01614344e-01 1.03647482e+00 -6.51948035e-01 3.92764062e-02
-4.26544279e-01 1.08955073e+00 -1.53608218e-01 7.59780705e-01
-7.10658669e-01 -3.48057806e-01 9.67504859e-01 -2.45299488e-02
-5.68140931e-02 -4.49124873e-01 -2.78143257e-01 -8.07690203e-01
1.95945054e-02 -1.02965510e+00 4.83740941e-02 -1.17253864e+00
-1.06592560e+00 5.30242741e-01 -1.77510828e-01 -1.02508664e+00
-2.40190178e-01 -4.58006173e-01 -7.62425959e-01 8.06721091e-01
-9.83873129e-01 -7.02185512e-01 -2.07456067e-01 3.97513449e-01
1.05221438e+00 -1.45080268e-01 1.00002146e+00 4.60756451e-01
-6.84939027e-01 6.11548007e-01 2.45339736e-01 -3.01450454e-02
8.39704096e-01 -7.63398826e-01 4.76310402e-01 8.67410183e-01
3.81486714e-01 7.03768253e-01 8.33905816e-01 -4.74742383e-01
-1.43430197e+00 -8.56831372e-01 7.17022777e-01 -1.79930478e-01
5.38067997e-01 -6.44670665e-01 -7.41182745e-01 4.60380197e-01
6.25329465e-02 -1.09860882e-01 9.03781056e-01 4.59465384e-01
-2.21137404e-01 -2.18476325e-01 -8.79706025e-01 7.24314928e-01
9.67795730e-01 -1.10476613e+00 -2.99588025e-01 1.86218962e-01
9.15274858e-01 -8.10581222e-02 -6.70307755e-01 1.33652121e-01
4.04167861e-01 -9.92419839e-01 7.91878104e-01 -2.94421464e-01
-8.54349509e-03 -2.41756998e-02 -4.86054838e-01 -1.61158156e+00
-6.84410706e-02 -9.93729413e-01 1.28135741e-01 1.74348605e+00
8.74199092e-01 -8.69822264e-01 4.28844482e-01 7.22438514e-01
-5.35137177e-01 -4.78907228e-01 -1.02413058e+00 -9.43994105e-01
-3.59586358e-01 -1.28164268e+00 2.63008475e-01 7.11326718e-01
2.55211771e-01 5.82191169e-01 -3.03987861e-01 2.99969375e-01
2.17939280e-02 -4.29813683e-01 6.33819997e-01 -1.01474226e+00
-3.48373711e-01 -3.99957091e-01 -3.82264197e-01 -9.46367443e-01
1.18089452e-01 -6.77970529e-01 3.56532812e-01 -1.56084168e+00
-4.39841062e-01 -4.77487355e-01 -3.14116210e-01 4.22253311e-01
-1.53701589e-01 -1.64360210e-01 1.82881802e-01 -3.57263237e-02
-1.70041740e-01 7.32661664e-01 3.42680991e-01 -2.85918802e-01
-4.10902113e-01 3.30674380e-01 -1.04275835e+00 6.00431144e-01
1.07673466e+00 -3.51954818e-01 -7.55979776e-01 -5.16980946e-01
1.12322845e-01 -6.35761917e-02 7.24997744e-02 -1.27018487e+00
3.28902870e-01 1.61792874e-01 9.52293351e-02 -1.22482954e-02
8.61993253e-01 -7.27777898e-01 1.05992012e-01 -2.95598563e-02
-3.80933344e-01 6.59848601e-02 6.39221728e-01 6.08592749e-01
-2.90303737e-01 -2.02549949e-01 3.04569483e-01 2.39236221e-01
-6.91402614e-01 -2.41163030e-01 -7.54765511e-01 -6.56073838e-02
7.79087663e-01 -3.72833908e-01 -3.06977928e-01 -1.07896912e+00
-8.58891070e-01 -1.38145074e-01 -4.58433032e-02 8.87285709e-01
8.48099411e-01 -9.86792803e-01 -5.19436777e-01 3.19452614e-01
1.70470402e-01 -4.18354720e-01 3.53177339e-01 9.40894246e-01
6.07000366e-02 3.11031103e-01 4.72419828e-01 -5.68585515e-01
-1.41522026e+00 2.98843175e-01 2.37491831e-01 3.68966043e-01
-3.29872131e-01 9.30937827e-01 2.37896755e-01 -1.67174324e-01
5.22296906e-01 -5.31630337e-01 3.04077417e-02 2.16125906e-01
4.58703458e-01 4.40255165e-01 5.44449866e-01 -6.55499518e-01
-4.12893891e-01 1.79515228e-01 2.54815996e-01 -4.08033937e-01
1.10822988e+00 -3.65232050e-01 4.06340450e-01 8.93289387e-01
1.00471258e+00 2.60415584e-01 -8.57076049e-01 1.33608151e-02
1.02415428e-01 -1.58860847e-01 3.86365056e-01 -6.58182085e-01
-7.90752470e-01 9.21944559e-01 8.73493910e-01 4.64162737e-01
1.27925158e+00 4.21059206e-02 6.02308631e-01 3.49074066e-01
9.91187394e-02 -1.22134411e+00 1.16942294e-01 4.86816794e-01
9.21395779e-01 -8.91612113e-01 -4.63340759e-01 -2.29246765e-01
-7.00258136e-01 7.32103348e-01 3.03606093e-01 6.49792373e-01
6.70936942e-01 5.90959549e-01 2.45849311e-01 3.52800358e-03
-9.72277820e-01 -4.28901851e-01 1.14838690e-01 8.09083939e-01
7.38642931e-01 -7.22945407e-02 1.82453737e-01 8.29234958e-01
-2.23751098e-01 -3.36588472e-01 2.61004388e-01 9.14740920e-01
-3.63279372e-01 -7.65493870e-01 -6.65556788e-01 5.61258972e-01
-4.06956583e-01 -2.97876030e-01 -7.44234025e-01 3.75549674e-01
-1.50766224e-01 1.82565808e+00 -1.29006401e-01 -8.83161366e-01
5.55467606e-01 4.75761443e-01 -1.47447243e-01 -6.87255621e-01
-7.12920725e-01 1.34719223e-01 6.34371281e-01 -2.34465450e-01
-6.37572706e-02 -5.71056902e-01 -9.37077522e-01 3.89428437e-02
-6.85145438e-01 2.21638799e-01 9.99526262e-01 8.64028871e-01
6.50360227e-01 8.93306732e-01 5.76799273e-01 -9.36530888e-01
-4.68942046e-01 -1.20823264e+00 -4.13961500e-01 9.05238166e-02
4.22372490e-01 -3.29305530e-01 -5.26771784e-01 -6.98507503e-02] | [14.597381591796875, 6.185009956359863] |
b383fcf3-fd46-45c9-9d62-452003401d2b | unsupervised-intra-domain-adaptation-for | 2004.07703 | null | https://arxiv.org/abs/2004.07703v4 | https://arxiv.org/pdf/2004.07703v4.pdf | Unsupervised Intra-domain Adaptation for Semantic Segmentation through Self-Supervision | Convolutional neural network-based approaches have achieved remarkable progress in semantic segmentation. However, these approaches heavily rely on annotated data which are labor intensive. To cope with this limitation, automatically annotated data generated from graphic engines are used to train segmentation models. However, the models trained from synthetic data are difficult to transfer to real images. To tackle this issue, previous works have considered directly adapting models from the source data to the unlabeled target data (to reduce the inter-domain gap). Nonetheless, these techniques do not consider the large distribution gap among the target data itself (intra-domain gap). In this work, we propose a two-step self-supervised domain adaptation approach to minimize the inter-domain and intra-domain gap together. First, we conduct the inter-domain adaptation of the model; from this adaptation, we separate the target domain into an easy and hard split using an entropy-based ranking function. Finally, to decrease the intra-domain gap, we propose to employ a self-supervised adaptation technique from the easy to the hard split. Experimental results on numerous benchmark datasets highlight the effectiveness of our method against existing state-of-the-art approaches. The source code is available at https://github.com/feipan664/IntraDA.git. | ['Inkyu Shin', 'Fei Pan', 'Seokju Lee', 'In So Kweon', 'Francois Rameau'] | 2020-04-16 | unsupervised-intra-domain-adaptation-for-1 | http://openaccess.thecvf.com/content_CVPR_2020/html/Pan_Unsupervised_Intra-Domain_Adaptation_for_Semantic_Segmentation_Through_Self-Supervision_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Pan_Unsupervised_Intra-Domain_Adaptation_for_Semantic_Segmentation_Through_Self-Supervision_CVPR_2020_paper.pdf | cvpr-2020-6 | ['synthetic-to-real-translation'] | ['computer-vision'] | [ 3.64100665e-01 1.15288273e-01 -2.44477019e-01 -5.00979364e-01
-6.75581157e-01 -4.80584413e-01 3.10373425e-01 2.18138788e-02
-5.53416610e-01 7.03049123e-01 -2.38536000e-01 -1.22115031e-01
6.01243600e-02 -8.96937847e-01 -6.78768635e-01 -5.35204232e-01
5.80977798e-01 5.58274806e-01 6.48042977e-01 4.12374176e-03
2.52790809e-01 -1.91559717e-02 -1.35036552e+00 1.48416638e-01
1.50605559e+00 1.05836236e+00 3.94287020e-01 9.93049964e-02
-5.14790297e-01 3.91125202e-01 -4.95069206e-01 -1.32303059e-01
1.85470447e-01 -7.38858402e-01 -9.46825624e-01 2.00256452e-01
-7.89930150e-02 -2.41020113e-01 -1.49896340e-02 1.27604115e+00
3.60756785e-01 1.80081859e-01 5.91259122e-01 -1.21224332e+00
-5.19411623e-01 3.13291818e-01 -7.42375433e-01 -3.09544485e-02
3.92947905e-02 -2.80967038e-02 6.67179823e-01 -7.87788093e-01
6.81939900e-01 9.06309783e-01 4.73714650e-01 6.12922430e-01
-1.05593550e+00 -7.37328708e-01 9.37996954e-02 2.50759006e-01
-1.28962684e+00 -1.39681816e-01 1.09408927e+00 -4.56892967e-01
3.32999110e-01 -1.58239946e-01 5.47739506e-01 9.87365842e-01
-2.76171595e-01 8.19319963e-01 1.26127112e+00 -5.21791816e-01
3.88437241e-01 2.54792154e-01 1.02429204e-01 4.63904858e-01
6.92229271e-02 -1.22738518e-01 -9.15942788e-02 1.52894437e-01
6.92580760e-01 -1.81683183e-01 -2.13081598e-01 -4.78226423e-01
-7.54884779e-01 7.95811713e-01 3.74655902e-01 4.84712094e-01
-1.77716777e-01 -4.43381160e-01 5.44344246e-01 1.27454316e-02
7.33576298e-01 1.71485871e-01 -6.07732117e-01 -5.88914193e-02
-1.08623302e+00 8.88944641e-02 6.69928670e-01 9.48879480e-01
9.23320174e-01 -2.63024002e-01 7.78114423e-02 1.22847402e+00
3.32448661e-01 1.33896425e-01 7.58055925e-01 -7.23093092e-01
6.72613859e-01 8.65186155e-01 1.03683686e-02 -9.13824499e-01
-2.85966337e-01 -2.96007067e-01 -7.79678464e-01 2.61266511e-02
7.48601854e-01 -2.18092307e-01 -1.22721541e+00 1.62559164e+00
6.37722790e-01 2.24549696e-02 6.72988221e-02 9.63263273e-01
8.50593090e-01 7.33916759e-01 2.15965480e-01 -1.12197734e-02
1.02694476e+00 -1.30130196e+00 -5.50592721e-01 -3.78737122e-01
6.34248018e-01 -8.18163097e-01 1.24253094e+00 2.25124151e-01
-9.49095547e-01 -7.14952528e-01 -1.00909305e+00 -1.46787837e-02
-5.17328680e-01 4.14272130e-01 1.57568380e-01 4.51318800e-01
-6.64765894e-01 5.77385485e-01 -7.91820765e-01 -3.40878308e-01
6.67723477e-01 1.95515171e-01 -2.45334476e-01 -1.38401657e-01
-1.39303529e+00 6.42616272e-01 1.04907942e+00 7.02284202e-02
-5.00964224e-01 -5.07806838e-01 -8.63236725e-01 -1.51906163e-01
6.76659763e-01 -3.48992705e-01 1.22946393e+00 -1.46557260e+00
-1.66661274e+00 8.58669817e-01 7.14262575e-02 -2.10860997e-01
6.58699691e-01 -8.71894881e-02 -2.31253549e-01 1.44265264e-01
1.04206569e-01 7.30138242e-01 6.24237061e-01 -1.52540934e+00
-6.48602903e-01 -2.11235672e-01 -8.26642197e-03 3.75777453e-01
-5.50340652e-01 -2.72502691e-01 -7.75309026e-01 -5.93324780e-01
2.32382238e-01 -9.79853332e-01 -2.32132167e-01 -5.56774959e-02
-4.95122850e-01 -2.32917726e-01 9.68748331e-01 -7.58818269e-01
1.18348300e+00 -2.16970515e+00 -8.15591365e-02 1.43769294e-01
3.63683291e-02 7.84279823e-01 -1.09531008e-01 1.26044333e-01
1.66495591e-02 1.62987873e-01 -6.87555373e-01 -2.86772579e-01
-1.93267792e-01 1.69858873e-01 7.14926422e-02 9.90476608e-02
2.38323510e-01 5.58878124e-01 -8.45714569e-01 -9.07449901e-01
2.58529216e-01 3.51426721e-01 -4.15314525e-01 3.08605671e-01
-3.85535210e-01 6.89549327e-01 -6.98713720e-01 4.64074463e-01
9.61450517e-01 -2.25275248e-01 1.41987890e-01 -1.89186752e-01
3.29922475e-02 2.29754850e-01 -9.36943173e-01 1.85858941e+00
-3.81483942e-01 2.87461191e-01 -9.86779481e-02 -1.36776793e+00
1.02101409e+00 1.02490380e-01 5.29917955e-01 -7.62437642e-01
3.36320549e-01 4.46090072e-01 3.37421894e-03 -5.20909071e-01
3.43705893e-01 -1.71772614e-01 -3.56299728e-02 2.72193074e-01
4.67223413e-02 -1.71764135e-01 3.62783790e-01 -5.62022962e-02
7.00802565e-01 5.75454950e-01 2.20028952e-01 -7.77155012e-02
7.02186704e-01 2.75504321e-01 9.06779528e-01 2.08786950e-01
-4.94361401e-01 9.68673825e-01 4.72503543e-01 -3.36801469e-01
-1.04444420e+00 -7.73461998e-01 5.80979027e-02 7.77520180e-01
5.51214099e-01 -1.58170462e-01 -1.35310972e+00 -1.06700730e+00
-3.06083143e-01 5.99072754e-01 -5.18254161e-01 -2.33026519e-01
-5.66686869e-01 -7.57149279e-01 4.27054703e-01 6.18167758e-01
1.01796460e+00 -1.22060263e+00 -4.75954562e-01 2.72567570e-01
-4.37859386e-01 -1.18833041e+00 -4.94544566e-01 6.43565282e-02
-9.28406775e-01 -8.53109360e-01 -9.43988085e-01 -9.72014010e-01
8.73300374e-01 2.15643458e-02 9.33130324e-01 1.13948323e-01
-7.15936273e-02 -1.32906258e-01 -4.70149726e-01 -2.50966668e-01
-3.81778598e-01 3.70741606e-01 -4.23681110e-01 -8.51158947e-02
5.58921993e-01 -4.92314726e-01 -7.21237481e-01 6.19855940e-01
-1.08612335e+00 3.22206676e-01 5.94257414e-01 8.24503899e-01
9.59844172e-01 1.58649787e-01 6.52528226e-01 -9.19243574e-01
5.47897637e-01 -5.60059726e-01 -5.72191060e-01 2.48959020e-01
-6.49770200e-01 -6.11600727e-02 9.14620161e-01 -5.96880794e-01
-1.30149758e+00 2.86886781e-01 -2.07551673e-01 -3.65602553e-01
-4.79370534e-01 5.69150865e-01 -4.73800957e-01 1.46027595e-01
3.82740527e-01 2.90125251e-01 -2.60054469e-02 -4.50019538e-01
1.30335286e-01 8.51073325e-01 2.85318941e-01 -6.27213240e-01
7.16565609e-01 2.00597733e-01 -3.88755292e-01 -4.89443809e-01
-8.83550823e-01 -3.31486970e-01 -7.68384516e-01 -1.55874431e-01
1.04054165e+00 -6.49775505e-01 6.37342557e-02 6.67840242e-01
-9.86298561e-01 -5.85335255e-01 -2.86283404e-01 3.69830847e-01
-4.93971497e-01 4.86830115e-01 -4.23479557e-01 -5.00164270e-01
-3.60993534e-01 -1.17669702e+00 7.87340462e-01 7.23371565e-01
-3.69723663e-02 -1.03052282e+00 6.97692633e-02 6.38884902e-01
3.52815062e-01 3.58666629e-01 7.04372764e-01 -9.67648149e-01
-2.60730058e-01 -1.11778215e-01 -4.87016857e-01 5.96790493e-01
2.80265808e-01 -7.42582977e-02 -5.94735146e-01 -5.69897629e-02
-4.14378121e-02 -6.07084930e-01 7.40541399e-01 2.44836748e-01
1.42181122e+00 -7.94631708e-03 -3.22357863e-01 5.27322948e-01
1.36207366e+00 4.36432868e-01 6.40950382e-01 5.59644639e-01
7.22786129e-01 6.62610531e-01 9.99388039e-01 1.94694474e-01
5.99022448e-01 5.84005594e-01 1.42839536e-01 -2.27378473e-01
-1.67905375e-01 -4.02870536e-01 5.49908280e-02 8.61587286e-01
2.58229505e-02 -4.26176697e-01 -1.07591188e+00 5.90138376e-01
-1.82868016e+00 -4.28693444e-01 5.28341010e-02 2.13633442e+00
1.10429549e+00 4.50920612e-01 1.15215734e-01 7.00545013e-02
8.96104753e-01 -7.36208493e-03 -7.09755242e-01 -2.99334317e-01
1.58712626e-01 1.08580263e-02 3.49584699e-01 2.03404978e-01
-1.25971234e+00 1.13935208e+00 4.73884344e+00 1.23720598e+00
-1.20365608e+00 1.47882521e-01 9.08443987e-01 2.10896611e-01
-1.80439085e-01 3.61422971e-02 -5.72586894e-01 8.41393292e-01
7.64806449e-01 -8.07627738e-02 1.86826319e-01 9.58983541e-01
1.15184680e-01 -2.64068782e-01 -7.84554541e-01 6.89578176e-01
-4.84953299e-02 -7.22593606e-01 -1.48568600e-01 -3.36317085e-02
8.25269163e-01 -1.36316538e-01 -1.56600699e-01 3.18495154e-01
6.78186566e-02 -7.07836509e-01 5.37724495e-01 1.88396841e-01
8.52938771e-01 -7.24326849e-01 8.19650888e-01 5.01926720e-01
-1.27131033e+00 3.22283477e-01 -3.19354892e-01 2.67423958e-01
1.32105112e-01 7.77746379e-01 -6.38373375e-01 7.36680031e-01
6.41570449e-01 5.04174531e-01 -4.45786953e-01 1.02607358e+00
-2.61019289e-01 5.98816872e-01 -3.19077641e-01 1.92459196e-01
2.02275500e-01 -5.27729273e-01 3.53369236e-01 1.04403555e+00
2.30139673e-01 4.64248434e-02 3.37834418e-01 1.03505337e+00
-1.80085525e-01 3.30119669e-01 -4.25726354e-01 -1.35426119e-01
4.05559331e-01 1.15353739e+00 -1.17003047e+00 -3.65370780e-01
-3.79918635e-01 1.13842428e+00 3.11137170e-01 2.76987076e-01
-1.00066817e+00 -6.66766107e-01 1.30224511e-01 3.90233621e-02
2.48248711e-01 -5.32970615e-02 -5.07860065e-01 -1.07549703e+00
2.31407911e-01 -8.76110196e-01 3.24395627e-01 -4.76553321e-01
-1.19603312e+00 7.32926905e-01 1.30267084e-01 -1.53003061e+00
-9.63103250e-02 -3.14580679e-01 -4.79127318e-01 7.76718080e-01
-1.57901013e+00 -1.09224582e+00 -3.64843130e-01 3.21093768e-01
7.97446907e-01 4.62291017e-02 5.01523912e-01 4.52394366e-01
-6.84745729e-01 6.79604232e-01 1.63034752e-01 2.68784046e-01
6.99761391e-01 -1.12724602e+00 3.12181801e-01 6.98857605e-01
-2.58709580e-01 3.71009320e-01 3.58147651e-01 -7.25318074e-01
-5.56353509e-01 -1.17347276e+00 6.35207176e-01 5.54856248e-02
3.20833683e-01 -1.98818326e-01 -1.28033006e+00 4.12943661e-01
7.70486221e-02 1.52324423e-01 6.15610778e-01 -2.63540983e-01
-1.52822688e-01 -5.23786470e-02 -1.29363048e+00 4.24915552e-01
8.48904908e-01 -1.47214308e-01 -5.18426180e-01 1.27845407e-01
5.54761350e-01 -5.71719646e-01 -8.37933600e-01 5.91787398e-01
4.40281034e-01 -9.39071894e-01 6.29753649e-01 -1.46998927e-01
8.91497254e-01 -3.57431650e-01 2.73472995e-01 -1.40629673e+00
9.58354846e-02 -1.31996736e-01 2.54332393e-01 1.43914938e+00
4.61071193e-01 -6.89313889e-01 9.59824800e-01 5.98467886e-01
-1.12912886e-01 -9.91945922e-01 -7.83245087e-01 -8.07748556e-01
3.36070329e-01 2.94117350e-02 4.84131485e-01 1.09300196e+00
-1.20502621e-01 1.25212207e-01 -1.98975086e-01 -1.46563292e-01
4.94715095e-01 1.65206283e-01 5.90809405e-01 -1.19066811e+00
-1.34781063e-01 -4.01078254e-01 -6.77152872e-02 -1.03408122e+00
1.04006074e-01 -6.67059600e-01 4.32150245e-01 -1.70114112e+00
2.78212756e-01 -6.74596131e-01 -3.26367646e-01 5.20860851e-01
-3.77252877e-01 2.11040616e-01 1.62580401e-01 3.18280607e-01
-6.47208214e-01 7.10620522e-01 1.46250319e+00 1.34410970e-02
-4.03956443e-01 -1.24169379e-01 -5.45779586e-01 9.43595529e-01
1.21147549e+00 -6.15322351e-01 -6.63546264e-01 -4.95872259e-01
-3.09373736e-01 -1.20114446e-01 7.97668565e-03 -1.05004382e+00
1.86196081e-02 -2.74537534e-01 2.58078098e-01 -5.15706360e-01
7.64333606e-02 -8.68757904e-01 -1.11413509e-01 2.96934277e-01
-2.79827595e-01 -3.72378439e-01 3.08135778e-01 4.07522261e-01
-4.69183713e-01 -4.69875932e-01 1.01631093e+00 -1.47072211e-01
-6.78146720e-01 2.04986811e-01 -3.57318930e-02 3.71313840e-01
1.22047198e+00 -2.38559365e-01 -2.22419187e-01 -1.65926039e-01
-5.50922811e-01 3.87764335e-01 6.45201027e-01 3.24248344e-01
4.55316037e-01 -1.17807984e+00 -4.11316782e-01 6.06391355e-02
7.89813846e-02 5.50029397e-01 3.20953429e-01 7.14399517e-01
-5.72585106e-01 1.19363837e-01 -3.57933253e-01 -4.88105208e-01
-1.06770337e+00 3.80020440e-01 3.54126751e-01 -5.75009763e-01
-3.54275107e-01 7.60971606e-01 1.85756683e-01 -7.18719840e-01
1.06284648e-01 -1.48467034e-01 -2.95710951e-01 8.01810920e-02
1.65698662e-01 1.89183325e-01 -1.98478382e-02 -6.37310386e-01
-2.79403299e-01 7.32516229e-01 -2.47958153e-01 3.38069140e-03
1.15881908e+00 -2.47225970e-01 6.60325363e-02 2.78422207e-01
1.24727333e+00 -4.32420731e-01 -1.45750725e+00 -2.72094548e-01
6.02639169e-02 -3.67018133e-01 -9.34188440e-02 -8.25480700e-01
-1.22698998e+00 9.89347279e-01 6.65826321e-01 7.79925287e-02
1.40089023e+00 -1.58835843e-01 1.18801093e+00 2.20600646e-02
8.21598917e-02 -1.48991275e+00 1.32311970e-01 3.89921993e-01
5.74941397e-01 -1.46484184e+00 -1.81762189e-01 -6.89324915e-01
-7.87678719e-01 9.69555199e-01 9.89897847e-01 -2.71537974e-02
5.81864893e-01 4.04866226e-02 1.67496473e-01 1.38413504e-01
-2.25122437e-01 -1.67132005e-01 3.03390056e-01 5.33818245e-01
3.62993807e-01 -1.28617808e-01 -7.23291516e-01 7.43913829e-01
7.38643631e-02 4.48796391e-01 3.09027344e-01 1.08266950e+00
-3.25246304e-01 -1.45467675e+00 -2.22388878e-01 2.74996966e-01
-3.89549851e-01 8.38256478e-02 -3.61553878e-01 7.89259911e-01
4.00164187e-01 8.96195233e-01 -1.39642626e-01 -4.40937459e-01
3.20719063e-01 1.25494912e-01 1.30037591e-01 -5.49568117e-01
-2.66856462e-01 2.36561075e-01 -1.01953343e-01 -2.73769379e-01
-4.97573674e-01 -4.78216678e-01 -1.43944097e+00 1.63471222e-01
-3.45493317e-01 1.29784375e-01 6.33952737e-01 1.00346720e+00
2.96698600e-01 5.56869626e-01 5.42161524e-01 -8.14877808e-01
-4.69335973e-01 -1.01796031e+00 -2.39267141e-01 5.93122363e-01
-3.71374153e-02 -6.91313744e-01 -3.18546444e-01 1.92664400e-01] | [9.658743858337402, 1.3513282537460327] |
51cb12dd-6a1d-45fc-86e3-c4e2be78df41 | simmatch-semi-supervised-learning-with | 2203.06915 | null | https://arxiv.org/abs/2203.06915v2 | https://arxiv.org/pdf/2203.06915v2.pdf | SimMatch: Semi-supervised Learning with Similarity Matching | Learning with few labeled data has been a longstanding problem in the computer vision and machine learning research community. In this paper, we introduced a new semi-supervised learning framework, SimMatch, which simultaneously considers semantic similarity and instance similarity. In SimMatch, the consistency regularization will be applied on both semantic-level and instance-level. The different augmented views of the same instance are encouraged to have the same class prediction and similar similarity relationship respected to other instances. Next, we instantiated a labeled memory buffer to fully leverage the ground truth labels on instance-level and bridge the gaps between the semantic and instance similarities. Finally, we proposed the \textit{unfolding} and \textit{aggregation} operation which allows these two similarities be isomorphically transformed with each other. In this way, the semantic and instance pseudo-labels can be mutually propagated to generate more high-quality and reliable matching targets. Extensive experimental results demonstrate that SimMatch improves the performance of semi-supervised learning tasks across different benchmark datasets and different settings. Notably, with 400 epochs of training, SimMatch achieves 67.2\%, and 74.4\% Top-1 Accuracy with 1\% and 10\% labeled examples on ImageNet, which significantly outperforms the baseline methods and is better than previous semi-supervised learning frameworks. Code and pre-trained models are available at https://github.com/KyleZheng1997/simmatch. | ['Chang Xu', 'Chen Qian', 'Fei Wang', 'Lang Huang', 'Shan You', 'Mingkai Zheng'] | 2022-03-14 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Zheng_SimMatch_Semi-Supervised_Learning_With_Similarity_Matching_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Zheng_SimMatch_Semi-Supervised_Learning_With_Similarity_Matching_CVPR_2022_paper.pdf | cvpr-2022-1 | ['semi-supervised-image-classification'] | ['computer-vision'] | [ 2.62195855e-01 1.88485205e-01 -5.66371858e-01 -8.84777486e-01
-7.61824310e-01 -2.50009179e-01 5.89055717e-01 2.19056681e-02
-1.96018651e-01 6.17262661e-01 -4.31743152e-02 2.40129277e-01
3.75846438e-02 -7.23244429e-01 -8.13894987e-01 -6.34243786e-01
3.42539877e-01 4.74840790e-01 3.70526463e-01 3.54436547e-01
5.38523355e-03 -2.15644985e-01 -1.66720140e+00 5.78988194e-01
8.79131317e-01 1.31670785e+00 3.38648677e-01 -2.09249556e-01
-1.75736785e-01 8.70017350e-01 -6.89222366e-02 -5.41300714e-01
3.25899482e-01 -2.51265198e-01 -7.66860604e-01 2.54346460e-01
7.74578333e-01 -9.80198756e-02 -2.14614928e-01 1.34108913e+00
3.41971993e-01 5.61991781e-02 4.02822733e-01 -1.44247210e+00
-7.08272755e-01 5.93672752e-01 -7.72494972e-01 -3.24575417e-02
5.85286729e-02 8.32024589e-02 1.05506372e+00 -1.19961011e+00
6.71056509e-01 1.14307284e+00 4.03431982e-01 4.50505376e-01
-1.11463523e+00 -1.18610930e+00 3.51738542e-01 3.71111631e-01
-1.29906273e+00 -2.57401317e-01 7.31292903e-01 -3.09568554e-01
6.14579797e-01 -2.31619794e-02 2.25799873e-01 8.65804613e-01
-3.34423155e-01 1.00343037e+00 1.27169740e+00 -2.69335866e-01
5.52529842e-02 3.02819014e-01 3.66148442e-01 7.72396266e-01
1.78998306e-01 1.15346625e-01 -5.25627255e-01 1.85238030e-02
5.02474308e-01 4.87761617e-01 -1.68272763e-01 -5.65109193e-01
-1.35762358e+00 8.05072904e-01 7.02917218e-01 1.44032285e-01
-2.15378523e-01 -1.45664483e-01 4.17934448e-01 1.86839983e-01
5.92407405e-01 1.54588640e-01 -5.01635730e-01 3.12498212e-01
-8.36972892e-01 5.42784017e-03 4.33260590e-01 1.39802408e+00
1.11471438e+00 -3.15959394e-01 -1.70605272e-01 1.18644953e+00
2.91151971e-01 6.17567122e-01 4.59049761e-01 -1.03292835e+00
6.29959822e-01 1.05102229e+00 -2.38807321e-01 -9.43728685e-01
-1.50444388e-01 -6.83414280e-01 -9.17461991e-01 -1.00443482e-01
1.89090878e-01 2.52603263e-01 -9.18515801e-01 1.85623813e+00
5.09767413e-01 6.68576419e-01 2.38209330e-02 9.70152795e-01
1.12844348e+00 5.05471468e-01 1.35005072e-01 -1.75075069e-01
1.26252317e+00 -1.48703611e+00 -4.99369830e-01 -3.93851310e-01
6.82922959e-01 -7.64326692e-01 1.15708804e+00 2.20863968e-02
-1.00089180e+00 -7.32266426e-01 -1.03503454e+00 4.65139113e-02
-1.67543769e-01 2.25789234e-01 4.05746013e-01 1.01874277e-01
-5.69889307e-01 5.68591654e-01 -7.15417981e-01 -1.62415311e-01
7.04418361e-01 9.43560824e-02 -4.50489491e-01 -5.46018660e-01
-1.11572349e+00 5.49576819e-01 5.53120494e-01 -6.45720437e-02
-7.32601643e-01 -8.13880444e-01 -9.57356155e-01 -1.33502811e-01
5.69944024e-01 -5.22940576e-01 1.16331196e+00 -1.19323313e+00
-8.01168323e-01 1.41848385e+00 -2.89524138e-01 -2.90248930e-01
4.09778118e-01 -2.13698164e-01 -5.85040271e-01 3.88855003e-02
5.58798492e-01 8.23099434e-01 6.81192696e-01 -1.30166936e+00
-8.15375447e-01 -5.49772382e-01 -1.50042772e-01 3.11881274e-01
-3.70165408e-01 -1.23340115e-01 -7.59079516e-01 -6.53150678e-01
4.00448591e-01 -9.89785135e-01 -5.85684960e-04 1.37718827e-01
-3.62691492e-01 -3.31300855e-01 5.92487693e-01 -3.74898672e-01
9.85874116e-01 -2.26739478e+00 -1.26438383e-02 1.59719825e-01
1.88122258e-01 5.44275761e-01 -3.23994309e-01 1.95957586e-01
-2.51617998e-01 -1.78366706e-01 -3.28499585e-01 -5.16819000e-01
-9.81361121e-02 2.11078525e-01 -3.79728049e-01 3.06366444e-01
1.09449588e-02 8.87023091e-01 -1.02459586e+00 -5.90686202e-01
1.32944077e-01 2.33389258e-01 -3.43098015e-01 3.64629537e-01
-1.56596541e-01 3.29102963e-01 -3.92292470e-01 6.88073099e-01
8.13032806e-01 -7.70644128e-01 1.48871884e-01 -4.62761402e-01
3.89695019e-01 2.31108785e-01 -1.36484075e+00 1.84846628e+00
-3.53610516e-01 1.68527961e-01 -3.05766523e-01 -1.17071760e+00
1.13444149e+00 4.39226143e-02 4.21688497e-01 -1.00240409e+00
-1.65341049e-01 3.25236171e-01 -4.21576321e-01 -3.45056295e-01
8.51209611e-02 -6.40370920e-02 1.97050378e-01 4.73159492e-01
1.56540543e-01 2.27169558e-01 1.54036269e-01 3.11058581e-01
6.07862413e-01 2.13133112e-01 1.52211785e-01 -1.84149370e-01
6.16132021e-01 1.34261087e-01 9.88483965e-01 5.01917422e-01
-1.83885068e-01 7.03531921e-01 1.05954096e-01 -4.54551488e-01
-8.14693272e-01 -1.14886689e+00 -2.67369419e-01 1.13930666e+00
7.20655560e-01 -3.20150822e-01 -6.87206328e-01 -9.06771302e-01
7.63973817e-02 5.56195438e-01 -5.56202292e-01 -3.43077034e-01
-4.15669441e-01 -4.92354363e-01 1.74440324e-01 6.70876741e-01
8.74929011e-01 -1.05893469e+00 4.52267118e-02 2.15667021e-02
-2.89948642e-01 -1.32414758e+00 -6.06861651e-01 -7.93522075e-02
-9.24493194e-01 -1.31280553e+00 -4.97360498e-01 -1.14872813e+00
9.80329633e-01 5.29522359e-01 1.23801970e+00 3.61833274e-01
-1.50929376e-01 1.64204389e-02 -4.67160285e-01 -1.49647787e-01
-1.56461656e-01 -2.01686029e-03 7.15848617e-03 2.07035348e-01
6.19268835e-01 -5.24114251e-01 -6.76999927e-01 7.28292882e-01
-7.93222427e-01 3.64749819e-01 5.29618979e-01 9.79972363e-01
1.07330739e+00 -2.41715029e-01 7.32939839e-01 -1.17672813e+00
-1.72440559e-01 -5.85093379e-01 -5.18078804e-01 3.64272326e-01
-9.87058878e-01 -4.15921099e-02 5.74465752e-01 -4.10989374e-01
-1.04405463e+00 1.74741164e-01 1.74856633e-01 -7.15499222e-01
-1.89315096e-01 3.60092998e-01 -3.72765273e-01 2.85929859e-01
3.74551445e-01 3.75200897e-01 3.43873166e-02 -4.76380169e-01
2.92533755e-01 7.31519818e-01 5.77861726e-01 -5.73097050e-01
7.58830607e-01 7.27069497e-01 -4.22428519e-01 -2.05580831e-01
-1.53028214e+00 -6.56655192e-01 -5.19032180e-01 -2.31353678e-02
6.12024426e-01 -1.25246418e+00 -2.29200870e-01 5.94867885e-01
-7.75553942e-01 -3.49641800e-01 -2.25888357e-01 4.40036476e-01
-3.56705278e-01 3.77509981e-01 -4.64068085e-01 -1.76507816e-01
-4.15997207e-01 -1.08762753e+00 1.08574855e+00 4.40930307e-01
6.07268848e-02 -9.09247041e-01 -3.96427289e-02 8.30951810e-01
1.22005410e-01 3.88039625e-03 6.31403923e-01 -1.03638363e+00
-6.08543038e-01 -1.32653654e-01 -5.74315548e-01 5.63792348e-01
2.36586064e-01 -2.23889068e-01 -8.88840675e-01 -4.40288842e-01
-1.63067639e-01 -6.48330808e-01 9.52807784e-01 7.46359536e-03
1.12115383e+00 -2.36031726e-01 -4.60028917e-01 7.23855495e-01
1.37040102e+00 -4.59307767e-02 3.66102457e-01 3.04321289e-01
8.11734557e-01 6.79931521e-01 1.02467871e+00 3.07537526e-01
5.87325513e-01 6.44348264e-01 4.36876982e-01 -1.34064198e-01
-3.62680584e-01 -4.90714729e-01 9.69122350e-02 7.48899579e-01
2.31580302e-01 1.82390153e-01 -8.20658803e-01 5.92105269e-01
-2.09425592e+00 -1.00619483e+00 -1.82010502e-01 2.26709104e+00
1.00867987e+00 1.61766306e-01 -6.55665174e-02 -2.36695841e-01
1.18022609e+00 1.26891315e-01 -8.24073732e-01 3.18340123e-01
-2.13526458e-01 5.32316864e-02 2.79297173e-01 2.92602539e-01
-1.21132636e+00 9.15316343e-01 5.06153584e+00 9.80036438e-01
-9.56798732e-01 2.46556282e-01 8.27976167e-01 -2.33333245e-01
-2.38258928e-01 4.82464693e-02 -9.89905119e-01 7.60622859e-01
3.61256957e-01 -4.22982387e-02 2.17210010e-01 9.82252359e-01
-1.57400593e-01 -3.01850238e-03 -1.20955968e+00 1.05997419e+00
1.80513293e-01 -1.36307943e+00 1.61926508e-01 -1.79498509e-01
1.03461838e+00 2.03235537e-01 5.18929884e-02 3.39495391e-01
3.30673426e-01 -8.93767893e-01 5.51464796e-01 3.67706656e-01
7.30934083e-01 -6.42919421e-01 7.39506006e-01 4.95215476e-01
-1.38528419e+00 8.20993111e-02 -3.47991407e-01 1.51872084e-01
-4.54488695e-02 7.82609761e-01 -4.17192668e-01 6.98872387e-01
8.46024275e-01 1.16849959e+00 -6.24569595e-01 8.07223082e-01
-4.14774626e-01 6.51780725e-01 -1.72201931e-01 3.18786412e-01
2.22191215e-01 -3.19850892e-01 1.21448517e-01 9.20484483e-01
-8.84625874e-03 7.30345473e-02 4.98051554e-01 8.88322353e-01
-1.71874195e-01 1.72825102e-02 -1.96135446e-01 3.95067066e-01
8.56356084e-01 1.31256378e+00 -5.11115551e-01 -6.16342604e-01
-6.23908401e-01 8.83153081e-01 6.05243206e-01 3.18520844e-01
-1.04809964e+00 -2.01344594e-01 5.29369295e-01 1.06683383e-02
2.69795567e-01 3.19221318e-01 -2.81582803e-01 -1.35970592e+00
3.75542730e-01 -7.56178379e-01 7.79539168e-01 -6.98617816e-01
-1.61219084e+00 5.71592867e-01 -1.10235333e-01 -1.42809236e+00
-3.80341187e-02 -3.36064249e-01 -5.60731292e-01 6.77214503e-01
-1.80007994e+00 -1.34229708e+00 -6.46830976e-01 7.46977627e-01
4.95479047e-01 -1.91849828e-01 6.53327405e-01 4.19385821e-01
-7.28692412e-01 8.46990049e-01 2.78590228e-02 2.90977448e-01
1.05301070e+00 -1.05971444e+00 1.60101697e-01 7.00105548e-01
3.17226559e-01 4.97669131e-01 3.26084912e-01 -6.12303078e-01
-9.41712856e-01 -1.45263755e+00 8.83528829e-01 -3.54712188e-01
5.45836866e-01 -1.40766993e-01 -1.20383048e+00 7.44341373e-01
5.92628084e-02 5.17386794e-01 6.95407629e-01 -1.25714451e-01
-8.53994846e-01 -4.52052325e-01 -1.09858608e+00 3.47070009e-01
1.38991380e+00 -6.67238295e-01 -6.62016153e-01 5.23211539e-01
7.04274893e-01 -3.76644224e-01 -9.21033919e-01 7.17402756e-01
2.09249616e-01 -1.15035307e+00 9.21339571e-01 -5.50023496e-01
5.43313980e-01 -4.50788945e-01 -2.60532081e-01 -9.83987689e-01
-3.01976144e-01 -1.15767911e-01 -9.04560536e-02 1.46005189e+00
4.77755278e-01 -7.11454928e-01 7.72374988e-01 5.61135650e-01
-1.70998976e-01 -1.06508160e+00 -6.48593009e-01 -1.01596737e+00
-5.35915792e-02 -2.14948684e-01 5.59266150e-01 1.32673657e+00
-1.50478020e-01 2.74967283e-01 -2.99330503e-01 2.80276805e-01
9.50751245e-01 7.83698738e-01 6.52155042e-01 -1.01404881e+00
-2.28725582e-01 -1.98658749e-01 -3.51213425e-01 -1.00769866e+00
4.51459914e-01 -1.32329679e+00 -1.82192728e-01 -1.39279914e+00
7.40849078e-01 -7.18437433e-01 -5.39356887e-01 7.59240627e-01
-5.58995962e-01 4.00579780e-01 8.79572183e-02 5.78611732e-01
-9.20849621e-01 5.90366364e-01 1.13044643e+00 -1.25928670e-01
1.37053534e-01 -1.48001388e-01 -6.07580066e-01 7.17314422e-01
8.32403779e-01 -5.58575928e-01 -5.13746798e-01 -4.57027942e-01
-6.08033687e-02 -2.88026363e-01 4.36842889e-01 -8.81538868e-01
4.12208259e-01 -1.84842169e-01 1.84663400e-01 -5.13186276e-01
1.23609319e-01 -8.19240391e-01 1.49816245e-01 3.40286672e-01
-6.44832373e-01 -2.03620955e-01 -1.65718779e-01 6.44426823e-01
-5.42714357e-01 -6.58025891e-02 1.00971401e+00 -2.09969655e-03
-9.71317410e-01 6.75270617e-01 6.46257639e-01 3.98491412e-01
1.30457103e+00 -1.67417720e-01 -4.41109210e-01 -2.08571497e-02
-6.22801363e-01 6.40169859e-01 5.64935803e-01 6.44232273e-01
6.16814673e-01 -1.65233064e+00 -7.25420177e-01 1.64229050e-01
5.95972598e-01 3.08482647e-01 3.82672518e-01 7.13190496e-01
6.13778830e-02 1.39385507e-01 -8.62597227e-02 -7.67093837e-01
-1.23665404e+00 7.21021354e-01 2.78484464e-01 -2.50100881e-01
-4.89494056e-01 8.28014970e-01 4.30350810e-01 -7.47374117e-01
3.90959769e-01 9.56106335e-02 7.11341724e-02 -5.23577966e-02
6.69146001e-01 2.74645269e-01 -4.32418622e-02 -6.56411111e-01
-5.60180247e-01 6.21243179e-01 -4.43428725e-01 3.78376037e-01
1.24925756e+00 -1.23550884e-01 -1.40174314e-01 4.02790666e-01
1.36061001e+00 -3.62657964e-01 -1.43955338e+00 -9.31632936e-01
8.26870799e-02 -5.28332472e-01 -2.14231476e-01 -9.14852440e-01
-1.51255620e+00 6.69785798e-01 4.88455951e-01 -2.26244822e-01
1.10017574e+00 3.32458198e-01 8.90527248e-01 2.97934234e-01
3.09400171e-01 -1.08022523e+00 2.07996815e-01 2.33034566e-01
5.67568779e-01 -1.63456929e+00 -1.29387844e-02 -7.49117076e-01
-8.46386552e-01 6.05911434e-01 1.08285630e+00 -2.65964687e-01
4.77546692e-01 5.29958829e-02 1.67576253e-01 -1.91524267e-01
-8.56748223e-01 -1.74220175e-01 4.83982056e-01 3.67300034e-01
3.85509163e-01 3.07216905e-02 -1.01351030e-01 7.51560271e-01
2.57560670e-01 7.02228099e-02 -1.02131821e-01 8.63438070e-01
-4.04881418e-01 -1.03377771e+00 -1.36779770e-02 5.33511519e-01
-1.50331989e-01 -1.60653904e-01 -2.01472118e-01 4.94033545e-01
2.45025247e-01 8.27011287e-01 8.83873329e-02 -3.99456620e-01
2.91366726e-01 4.94789481e-02 2.17735752e-01 -7.74794579e-01
-4.28751379e-01 -1.26989797e-01 7.21644051e-03 -8.06243181e-01
-6.30357623e-01 -5.90578020e-01 -1.58376443e+00 -5.62436553e-03
-4.25750822e-01 4.30610403e-02 2.82440215e-01 9.45050478e-01
7.05806732e-01 1.82223350e-01 9.47875142e-01 -3.15815717e-01
-6.91609502e-01 -7.30955601e-01 -5.09416819e-01 9.28497732e-01
-1.05628625e-01 -6.91720247e-01 -3.93752694e-01 1.74644113e-01] | [9.589088439941406, 2.994605302810669] |
c90032a0-d31f-427b-a5dc-5e693f1bbf73 | draft-and-revise-effective-image-generation | 2206.04452 | null | https://arxiv.org/abs/2206.04452v1 | https://arxiv.org/pdf/2206.04452v1.pdf | Draft-and-Revise: Effective Image Generation with Contextual RQ-Transformer | Although autoregressive models have achieved promising results on image generation, their unidirectional generation process prevents the resultant images from fully reflecting global contexts. To address the issue, we propose an effective image generation framework of Draft-and-Revise with Contextual RQ-transformer to consider global contexts during the generation process. As a generalized VQ-VAE, RQ-VAE first represents a high-resolution image as a sequence of discrete code stacks. After code stacks in the sequence are randomly masked, Contextual RQ-Transformer is trained to infill the masked code stacks based on the unmasked contexts of the image. Then, Contextual RQ-Transformer uses our two-phase decoding, Draft-and-Revise, and generates an image, while exploiting the global contexts of the image during the generation process. Specifically. in the draft phase, our model first focuses on generating diverse images despite rather low quality. Then, in the revise phase, the model iteratively improves the quality of images, while preserving the global contexts of generated images. In experiments, our method achieves state-of-the-art results on conditional image generation. We also validate that the Draft-and-Revise decoding can achieve high performance by effectively controlling the quality-diversity trade-off in image generation. | ['Wook-Shin Han', 'Minsu Cho', 'Saehoon Kim', 'Chiheon Kim', 'Doyup Lee'] | 2022-06-09 | null | null | null | null | ['conditional-image-generation'] | ['computer-vision'] | [ 7.32743263e-01 -5.05440943e-02 9.51179564e-02 -1.38650864e-01
-9.28995013e-01 -4.70895529e-01 7.05524683e-01 -3.38922054e-01
-1.70255918e-02 6.02122068e-01 1.87525183e-01 -3.68809849e-01
2.38220185e-01 -1.00903845e+00 -8.61040950e-01 -9.27807689e-01
2.83289850e-01 -9.93895754e-02 8.50327462e-02 -3.42292562e-02
3.34453642e-01 2.68354192e-02 -1.62107527e+00 5.52215636e-01
1.13495266e+00 8.81664932e-01 7.56414831e-01 1.16154253e+00
-1.19835302e-01 8.67002845e-01 -6.86184168e-01 -3.18832427e-01
3.44072968e-01 -9.50820684e-01 -3.14816117e-01 4.04201925e-01
3.58446509e-01 -4.75537688e-01 -2.10007727e-01 1.03052139e+00
4.28563356e-01 -2.19302818e-01 5.54466307e-01 -1.06182981e+00
-1.09291804e+00 5.61544299e-01 -6.50972664e-01 -1.88459337e-01
2.88383216e-01 4.44580317e-01 7.30414450e-01 -1.14143920e+00
5.82370043e-01 9.52382147e-01 2.39131570e-01 6.03041291e-01
-1.37629342e+00 -7.78361440e-01 1.31771788e-01 -7.05375224e-02
-1.42049575e+00 -5.69080472e-01 6.05763197e-01 -5.46365142e-01
5.41238487e-01 2.85932541e-01 6.38616920e-01 8.30446422e-01
4.06523645e-01 5.48012137e-01 1.06036782e+00 -5.78108191e-01
2.59785980e-01 -2.04327598e-01 -3.49356830e-01 7.17152834e-01
4.36165556e-03 2.04456821e-01 -6.18112564e-01 1.06567085e-01
8.67842674e-01 -2.84628659e-01 -4.14104283e-01 -1.69335902e-01
-1.30323315e+00 5.69318652e-01 3.12754065e-01 1.53164417e-01
-6.01195574e-01 3.24434459e-01 -1.39354467e-01 6.83572590e-02
3.38476837e-01 3.42729419e-01 7.36668631e-02 5.02196327e-02
-1.46173072e+00 2.45535865e-01 6.42885044e-02 1.13239503e+00
9.46039200e-01 2.00828865e-01 -7.28400588e-01 9.39336658e-01
2.14732200e-01 6.25257373e-01 2.44687378e-01 -9.95697737e-01
4.76688445e-01 2.62254655e-01 7.04691559e-02 -7.68706203e-01
2.96784848e-01 -5.47513127e-01 -1.11811447e+00 2.35773981e-01
-1.02515697e-01 -2.24167436e-01 -1.41866159e+00 1.73691440e+00
4.31858115e-02 2.49842644e-01 2.43145730e-02 5.93105197e-01
6.55865550e-01 1.12711000e+00 -6.50432482e-02 -4.58243072e-01
1.22916675e+00 -1.14103413e+00 -6.86175525e-01 -1.92919329e-01
1.56806260e-01 -9.57607627e-01 9.30770874e-01 3.62509400e-01
-1.46410990e+00 -8.42212319e-01 -1.28943706e+00 4.23928984e-02
1.88728496e-01 3.99155378e-01 6.51778430e-02 7.06001043e-01
-1.43833315e+00 1.21566400e-01 -2.61382639e-01 1.80979133e-01
3.71436507e-01 -1.12570688e-01 -1.09529071e-01 -2.22840145e-01
-9.61741447e-01 3.53842556e-01 2.16408297e-01 1.12775117e-01
-1.22270489e+00 -6.57945991e-01 -9.36759710e-01 7.16451854e-02
2.08387792e-01 -8.58616054e-01 1.17563641e+00 -9.18099105e-01
-1.52977085e+00 6.42948270e-01 -5.41240156e-01 -3.05200517e-01
4.77626324e-01 7.12125748e-02 -4.01338995e-01 1.08474776e-01
3.44874531e-01 9.66363907e-01 1.28639829e+00 -1.75728202e+00
-7.89416134e-01 3.77685353e-02 -7.75845796e-02 8.66229385e-02
2.37910785e-02 -2.32788116e-01 -9.39676106e-01 -8.98013949e-01
1.33033141e-01 -9.72685218e-01 -1.74389541e-01 -2.55543470e-01
-5.34238219e-01 5.05085111e-01 7.13991523e-01 -5.55994630e-01
1.65262580e+00 -2.35593247e+00 1.09923683e-01 1.77702338e-01
1.10189423e-01 2.39075482e-01 -6.35334969e-01 2.27904186e-01
-6.95893839e-02 3.74263436e-01 -5.67241073e-01 -5.06426454e-01
-3.17494184e-01 1.77394733e-01 -6.38700724e-01 -5.13160303e-02
4.65481997e-01 9.81688380e-01 -8.04722905e-01 -5.06185651e-01
1.42353147e-01 4.55963731e-01 -7.31867313e-01 4.27998543e-01
-3.44260991e-01 5.46225905e-01 -1.84441850e-01 4.61717993e-01
1.02034581e+00 -3.42031568e-01 1.78340733e-01 -1.32585719e-01
-2.54780114e-01 -6.36717826e-02 -8.36592197e-01 1.58124781e+00
-6.05095625e-01 7.45189846e-01 -1.18266106e-01 -4.75411862e-01
8.68147850e-01 1.86892688e-01 1.51532054e-01 -9.50801015e-01
-2.70950645e-01 1.95270866e-01 -8.40745717e-02 -2.05685586e-01
1.01605630e+00 6.08601496e-02 -8.74135494e-02 3.90329182e-01
-2.91101426e-01 -3.55343968e-01 3.85839641e-01 4.93672788e-01
8.47107828e-01 2.75520116e-01 9.22299176e-02 9.97457132e-02
6.00318432e-01 -2.97683060e-01 3.77110332e-01 9.39400673e-01
1.45342782e-01 1.23928404e+00 5.26950717e-01 -1.51321925e-02
-1.35920691e+00 -1.33711290e+00 8.54128823e-02 8.40251923e-01
2.36044049e-01 -5.75060964e-01 -9.62913871e-01 -4.15112346e-01
-6.12922490e-01 8.92412543e-01 -7.01952338e-01 -1.11526325e-01
-6.07705832e-01 -7.48287797e-01 4.29415166e-01 1.64590269e-01
8.73574972e-01 -1.09635949e+00 -7.11206317e-01 2.89712489e-01
-4.88322318e-01 -1.00060511e+00 -8.62689197e-01 -2.71056473e-01
-4.94966358e-01 -7.60154963e-01 -9.27884102e-01 -8.04398775e-01
9.09096301e-01 4.04926449e-01 1.17619324e+00 4.04777020e-01
-1.42025590e-01 2.29714904e-02 -5.42770088e-01 9.05803144e-02
-7.53944933e-01 -2.85023749e-01 -5.27507663e-01 2.55011797e-01
-5.00016928e-01 -5.73372126e-01 -8.58924389e-01 2.90927410e-01
-1.32331157e+00 6.62088692e-01 7.35538483e-01 9.17579532e-01
9.07109737e-01 6.41368851e-02 2.78936237e-01 -6.60678983e-01
6.76186442e-01 -2.28847116e-01 -5.54634809e-01 3.75438571e-01
-5.43972850e-01 1.95788741e-01 5.15102267e-01 -4.19307739e-01
-1.17707658e+00 -6.26343787e-02 -1.02517083e-01 -2.89269984e-01
3.50265682e-01 2.89758623e-01 -2.35440388e-01 1.75794482e-01
5.07624269e-01 8.78429294e-01 -2.48197943e-01 -2.75784135e-02
5.49970150e-01 6.24219179e-01 8.12747240e-01 -4.59755242e-01
9.73681390e-01 2.94899166e-01 -1.65684879e-01 -5.44037819e-01
-3.77889335e-01 1.73522174e-01 -2.50069618e-01 -4.03835326e-01
1.07770944e+00 -1.11067510e+00 -8.88029709e-02 8.10989738e-01
-1.17927420e+00 -5.73432505e-01 -2.57732600e-01 -6.75175339e-02
-5.62329173e-01 2.93775916e-01 -5.17852664e-01 -7.35916734e-01
-2.52707422e-01 -1.53171635e+00 1.31119323e+00 2.50210404e-01
1.01519316e-01 -3.22413355e-01 -5.80791309e-02 1.69895187e-01
5.30896962e-01 8.75015631e-02 1.11572409e+00 3.43575656e-01
-1.14010191e+00 2.64869243e-01 -4.10890192e-01 3.20801556e-01
2.49674112e-01 2.50179797e-01 -7.83446848e-01 -1.45770401e-01
-2.99263835e-01 -3.51277441e-02 8.78374755e-01 4.36463565e-01
1.22415781e+00 -2.63772696e-01 -1.11111373e-01 6.86945319e-01
1.43637955e+00 4.45428312e-01 1.28677058e+00 2.14220792e-01
5.69788635e-01 3.11260819e-01 5.66616297e-01 5.36063731e-01
3.83775830e-01 7.47856915e-01 3.41256082e-01 -1.39701307e-01
-5.59381068e-01 -5.68949521e-01 5.49661577e-01 7.93965638e-01
1.87347114e-01 -5.18578291e-01 -6.24923944e-01 5.88165045e-01
-1.73169422e+00 -1.01332521e+00 -1.15046024e-01 2.23556542e+00
9.35672164e-01 -4.57581356e-02 -3.45919818e-01 -6.02753311e-02
9.81811404e-01 6.07419491e-01 -2.83450603e-01 -3.84686172e-01
-3.38629901e-01 1.73728079e-01 3.11343968e-01 5.98106980e-01
-7.16981947e-01 9.60503340e-01 6.76813459e+00 1.01056254e+00
-1.10349298e+00 2.39118803e-02 1.00643623e+00 -1.41648322e-01
-7.33047664e-01 1.30815372e-01 -5.64629972e-01 8.70489001e-01
6.71078086e-01 -6.78511932e-02 5.23533642e-01 4.50478613e-01
3.07227135e-01 -3.81844282e-01 -7.60837734e-01 9.09150243e-01
1.31826431e-01 -1.55200136e+00 3.46975535e-01 2.10460395e-01
1.11870742e+00 -4.87496138e-01 4.75526363e-01 2.88285971e-01
1.67540237e-01 -9.61323380e-01 1.16818821e+00 5.51880360e-01
1.26225817e+00 -7.68226743e-01 2.52713442e-01 2.74117261e-01
-1.43302083e+00 -7.01860636e-02 -1.58960730e-01 2.87118495e-01
4.47790593e-01 9.60059881e-01 -5.64912796e-01 5.07317960e-01
4.43826914e-01 3.15025508e-01 -6.57101631e-01 6.49501383e-01
-5.58481157e-01 3.99389237e-01 2.62539864e-01 4.33999956e-01
1.34860098e-01 -4.02316153e-01 5.62760174e-01 1.18129063e+00
8.33976507e-01 1.49418592e-01 -3.70414034e-02 1.16286802e+00
-1.75941154e-01 -1.96069062e-01 -4.06200498e-01 4.19688597e-02
5.85893452e-01 1.01120961e+00 -4.98045504e-01 -5.66700935e-01
-5.17558232e-02 1.21579266e+00 -2.08257698e-04 5.62956572e-01
-1.06208384e+00 -2.69202501e-01 5.32173753e-01 1.67169809e-01
6.16482675e-01 -2.49386638e-01 -2.71721601e-01 -1.03884149e+00
2.23667264e-01 -1.09589291e+00 -8.74228030e-02 -1.27045894e+00
-6.28744602e-01 8.67122114e-01 -7.93235749e-02 -1.37306654e+00
-4.48464215e-01 6.24126457e-02 -5.92207730e-01 9.34482634e-01
-1.49167848e+00 -1.21965432e+00 -4.22397852e-01 5.28454721e-01
8.06511760e-01 -1.13281697e-01 4.88607705e-01 1.73458993e-01
-4.68186677e-01 6.12438262e-01 1.29158227e-02 -5.52860238e-02
6.20195270e-01 -1.01739359e+00 6.53365254e-01 1.42298508e+00
1.91094786e-01 5.46494484e-01 5.59111357e-01 -8.73912454e-01
-1.17366755e+00 -1.27056766e+00 6.26241446e-01 -6.30230382e-02
9.06654373e-02 -4.74359721e-01 -7.24931300e-01 2.72393644e-01
4.30698872e-01 -1.41198784e-01 2.72537649e-01 -6.65088058e-01
-4.23253298e-01 -2.83251423e-02 -9.08908844e-01 9.16256905e-01
1.04506004e+00 -5.00626683e-01 -1.88995421e-01 -2.01783717e-01
9.96332943e-01 -4.99676824e-01 -4.13044512e-01 3.09428513e-01
4.94085044e-01 -1.37000191e+00 8.60242188e-01 2.84457862e-01
1.12143564e+00 -6.88875079e-01 -2.12488711e-01 -1.34994948e+00
-4.61129427e-01 -9.00835276e-01 1.29342198e-01 1.36445320e+00
5.27111351e-01 -4.17390674e-01 6.23396754e-01 2.25136787e-01
-1.08979069e-01 -7.20303833e-01 -7.09403455e-01 -4.88761485e-01
-1.47304580e-01 -4.47749048e-01 8.82847488e-01 3.20654213e-01
-4.96126652e-01 1.97358895e-02 -7.54189193e-01 1.54100284e-01
6.26833916e-01 3.84313941e-01 9.29082811e-01 -4.19349372e-01
-4.93580550e-01 -3.01815093e-01 6.57283142e-02 -1.41696048e+00
-1.62806660e-01 -5.49714863e-01 5.45314431e-01 -1.57196498e+00
4.20588881e-01 -3.92953396e-01 -2.67284177e-03 1.99818403e-01
-6.16350591e-01 4.87055987e-01 5.57849169e-01 3.28652084e-01
-4.06902969e-01 5.26929379e-01 1.49171853e+00 -2.19736233e-01
-1.74971759e-01 -4.34223264e-01 -8.22435200e-01 1.17343239e-01
5.58596849e-01 -3.27242792e-01 -6.62680745e-01 -5.18890917e-01
2.69430637e-01 1.91971049e-01 2.63996899e-01 -9.96578813e-01
1.33498788e-01 -1.15112647e-01 4.63428199e-01 -7.24603415e-01
3.23962986e-01 -4.04460847e-01 5.51206708e-01 3.91748071e-01
-3.33049029e-01 1.19810969e-01 -1.18238106e-02 5.37044644e-01
-3.51189792e-01 -1.41298398e-01 9.01289761e-01 1.15648704e-02
-5.06023407e-01 3.67099464e-01 -5.82444668e-01 -8.59400705e-02
9.28629041e-01 -5.03837705e-01 -3.52790922e-01 -5.18963575e-01
-3.92886162e-01 5.31231612e-02 7.72664249e-01 3.71110499e-01
9.98991489e-01 -1.49461067e+00 -9.29726720e-01 5.72622895e-01
1.53879210e-01 9.87356231e-02 5.81787884e-01 4.53709930e-01
-4.29906964e-01 8.98616090e-02 -5.38292415e-02 -7.57827997e-01
-1.31424212e+00 5.13338566e-01 5.39643988e-02 -5.29183030e-01
-3.34136546e-01 8.01136136e-01 7.47541308e-01 1.73769929e-02
-2.34718531e-01 -1.45931959e-01 2.42241547e-02 -1.14701092e-01
7.24222720e-01 -1.27592042e-01 -1.28629923e-01 -6.98559403e-01
5.16202450e-02 7.29956508e-01 -9.53970179e-02 -6.02237761e-01
1.08470678e+00 -4.22475487e-01 -1.90815315e-01 -2.50123050e-02
1.09233403e+00 3.90029073e-01 -1.52901936e+00 4.97307479e-02
-5.40085018e-01 -7.68157721e-01 2.56803930e-02 -7.93268740e-01
-1.22448146e+00 6.90803707e-01 4.87035364e-01 -9.98836942e-03
1.63002527e+00 -3.97383690e-01 8.58679652e-01 -3.56129438e-01
5.09417951e-01 -8.45765531e-01 4.50655758e-01 4.50873852e-01
1.01873577e+00 -7.48196006e-01 1.61793679e-02 -5.37852407e-01
-7.52466440e-01 7.74607062e-01 4.61509347e-01 1.60020351e-01
2.08322734e-01 3.76187623e-01 -3.55786979e-02 1.41996369e-01
-9.08206701e-01 2.36959495e-02 1.87758520e-01 6.56659901e-01
2.82421440e-01 6.41212612e-02 -1.72800303e-01 2.72984803e-01
-2.91010201e-01 3.75642926e-02 7.36708105e-01 7.91067243e-01
-4.03170139e-01 -1.30843306e+00 -6.54431343e-01 1.78077728e-01
-1.92533702e-01 -6.39550567e-01 -1.17890269e-01 3.47835600e-01
3.34309012e-01 1.17440891e+00 1.47762418e-01 -5.33720970e-01
1.60855561e-01 -1.84770674e-01 4.67172086e-01 -4.21677530e-01
-4.74205256e-01 3.26019287e-01 -6.47610500e-02 -4.99979317e-01
-2.13011503e-01 -3.05665374e-01 -9.97481763e-01 -4.31998596e-02
-1.98432565e-01 -2.50641592e-02 6.61109388e-01 5.40901840e-01
6.61767960e-01 8.71646523e-01 9.69971776e-01 -7.97438979e-01
6.89776149e-03 -6.55023932e-01 -3.60103786e-01 2.48751014e-01
4.58756924e-01 -2.52708554e-01 -2.49934942e-01 4.64554697e-01] | [11.278823852539062, -0.6749674081802368] |
b679c3ff-8db7-4e4e-b464-f5b0f87724c3 | rethinking-action-spaces-for-reinforcement | 1902.08858 | null | http://arxiv.org/abs/1902.08858v2 | http://arxiv.org/pdf/1902.08858v2.pdf | Rethinking Action Spaces for Reinforcement Learning in End-to-end Dialog Agents with Latent Variable Models | Defining action spaces for conversational agents and optimizing their
decision-making process with reinforcement learning is an enduring challenge.
Common practice has been to use handcrafted dialog acts, or the output
vocabulary, e.g. in neural encoder decoders, as the action spaces. Both have
their own limitations. This paper proposes a novel latent action framework that
treats the action spaces of an end-to-end dialog agent as latent variables and
develops unsupervised methods in order to induce its own action space from the
data. Comprehensive experiments are conducted examining both continuous and
discrete action types and two different optimization methods based on
stochastic variational inference. Results show that the proposed latent actions
achieve superior empirical performance improvement over previous word-level
policy gradient methods on both DealOrNoDeal and MultiWoz dialogs. Our detailed
analysis also provides insights about various latent variable approaches for
policy learning and can serve as a foundation for developing better latent
actions in future research. | ['Tiancheng Zhao', 'Kaige Xie', 'Maxine Eskenazi'] | 2019-02-23 | rethinking-action-spaces-for-reinforcement-1 | https://aclanthology.org/N19-1123 | https://aclanthology.org/N19-1123.pdf | naacl-2019-6 | ['goal-oriented-dialogue-systems', 'dialogue-management'] | ['natural-language-processing', 'natural-language-processing'] | [-2.61065946e-03 4.72502679e-01 -7.34268010e-01 -4.40552235e-01
-9.44838643e-01 -6.55596554e-01 1.03991985e+00 -5.00013947e-01
-5.48693478e-01 1.12528062e+00 8.65687370e-01 -4.88050997e-01
1.41994104e-01 -4.92834836e-01 -1.10577606e-01 -8.00606728e-01
2.18026847e-01 9.76365626e-01 -2.11372644e-01 -3.16166788e-01
3.76658082e-01 7.32013658e-02 -9.68168318e-01 6.81131780e-02
9.74025667e-01 4.30503994e-01 3.55671942e-01 8.68992269e-01
-5.93857646e-01 1.18331265e+00 -7.06786811e-01 -4.32057232e-01
1.93116173e-01 -7.89808571e-01 -1.00097346e+00 3.76038462e-01
-7.10136667e-02 -7.39578247e-01 -3.39107960e-01 1.00930333e+00
4.03543770e-01 5.29965580e-01 7.42889881e-01 -1.19903386e+00
-7.14080334e-01 1.05247664e+00 1.18183829e-01 -1.00017101e-01
2.22299695e-01 7.39860773e-01 1.45329595e+00 -4.41371322e-01
5.63399732e-01 1.88729060e+00 1.33130983e-01 1.15925288e+00
-1.31967354e+00 -3.68830025e-01 4.24346685e-01 9.69875604e-02
-4.67157781e-01 -5.85617244e-01 8.29694867e-01 -3.90212923e-01
1.08499384e+00 -7.07043186e-02 4.26202476e-01 1.59350848e+00
1.99454531e-01 1.30159676e+00 1.18583107e+00 -4.84248072e-01
3.67839575e-01 2.65856385e-01 -9.89590809e-02 8.86944532e-01
-5.14503658e-01 1.47962853e-01 -5.19064724e-01 -3.81798446e-01
8.30653250e-01 -1.12773776e-01 1.53275169e-02 -3.87076288e-01
-9.29530263e-01 1.63380742e+00 -1.29904017e-01 1.33849517e-01
-4.77941394e-01 1.51388705e-01 2.91137010e-01 3.83144170e-01
5.88598669e-01 6.83285713e-01 -5.95439970e-01 -7.92042255e-01
-3.92003983e-01 5.18227935e-01 1.27833009e+00 7.62036920e-01
5.44780493e-01 3.50562066e-01 -6.71524286e-01 1.17420816e+00
7.44553924e-01 2.64390349e-01 5.70481956e-01 -1.50212419e+00
6.57539070e-01 4.81240928e-01 5.10858953e-01 -1.71348616e-01
-8.87447298e-02 2.91355491e-01 -2.06764102e-01 4.67878461e-01
6.53145015e-01 -7.51806855e-01 -7.02853978e-01 2.01973224e+00
2.32378855e-01 -4.51763235e-02 5.08972228e-01 7.50527263e-01
3.69983196e-01 8.10786128e-01 2.97004580e-01 -5.21397173e-01
1.07166326e+00 -1.40101087e+00 -1.23532867e+00 -3.34751278e-01
4.55718338e-01 -5.17326176e-01 1.41405308e+00 2.04705909e-01
-1.21527100e+00 -3.00298542e-01 -7.35017896e-01 -8.75516459e-02
-1.34436220e-01 6.11413345e-02 7.15057075e-01 6.08871698e-01
-1.12561095e+00 5.15104949e-01 -1.00885153e+00 -1.45566434e-01
7.39562288e-02 4.02786255e-01 3.84655476e-01 5.53630888e-01
-1.48842800e+00 1.22269559e+00 2.24595144e-01 -3.63534510e-01
-1.33726811e+00 -2.26329401e-01 -9.15606380e-01 1.43439457e-01
7.25310445e-01 -6.08526587e-01 2.11366963e+00 -8.24634612e-01
-2.79066372e+00 4.39413756e-01 -1.73862517e-01 -7.64888704e-01
5.80197453e-01 -3.22130352e-01 1.23211674e-01 -1.14147186e-01
-2.11750176e-02 8.25261891e-01 9.50283945e-01 -1.04016554e+00
-7.67109990e-01 1.03046035e-03 5.32500267e-01 6.89953506e-01
-4.66704577e-01 -1.29529238e-01 -3.44657451e-01 -5.76436341e-01
-4.10843939e-01 -9.80570376e-01 -5.84416747e-01 -3.51810366e-01
-2.66170233e-01 -8.42273295e-01 6.78573787e-01 -6.10703945e-01
1.09489441e+00 -1.64989305e+00 7.86140263e-01 -3.58605742e-01
2.21545652e-01 1.24511749e-01 -1.46017790e-01 5.65087318e-01
5.16664386e-01 6.18813299e-02 -1.36252508e-01 -8.32933486e-01
4.91300285e-01 5.77955186e-01 -4.57872927e-01 2.16695994e-01
-3.70045789e-02 9.14262950e-01 -9.62805569e-01 -4.63319659e-01
5.07842183e-01 1.43012375e-01 -7.70743668e-01 7.01206446e-01
-9.59410787e-01 7.85765469e-01 -8.09538186e-01 2.97725976e-01
-8.39902610e-02 -1.27177864e-01 4.93368506e-01 4.60551977e-01
-2.52401143e-01 8.41742337e-01 -6.66515589e-01 1.76162100e+00
-6.90547049e-01 5.20466506e-01 3.81861776e-01 -7.21846044e-01
8.40866268e-01 7.03633189e-01 6.29160464e-01 -3.06847095e-01
1.42197967e-01 -2.35255629e-01 7.89517015e-02 -5.42052448e-01
6.90481603e-01 -4.38894480e-01 -2.36891732e-01 6.56473398e-01
3.38471353e-01 -1.97937578e-01 2.01571897e-01 1.98955163e-01
7.10424721e-01 2.20959231e-01 3.31569731e-01 -4.66978289e-02
5.59603453e-01 -8.19165260e-03 5.98851323e-01 9.27126110e-01
-5.08593023e-01 -1.95837304e-01 8.61060262e-01 -7.85337761e-02
-1.02812314e+00 -1.04012251e+00 1.70785114e-01 1.56069887e+00
-1.57927737e-01 -3.24864984e-01 -8.40403616e-01 -7.93170035e-01
4.16282285e-03 1.16609073e+00 -5.34859002e-01 -1.23231106e-01
-5.10853410e-01 -4.86565083e-01 3.94432545e-01 5.17124772e-01
5.37184596e-01 -1.40793693e+00 -3.16176325e-01 4.31444466e-01
-2.33805314e-01 -1.01479197e+00 -5.21954298e-01 1.44398764e-01
-8.43622804e-01 -6.28306925e-01 -6.21891320e-01 -5.18587470e-01
2.13315025e-01 -1.73136726e-01 9.94448245e-01 -3.39642167e-01
4.25959080e-01 6.50969565e-01 -3.45212907e-01 -2.86331743e-01
-9.55159187e-01 1.91494331e-01 2.74667978e-01 -2.30648890e-01
5.33830345e-01 -2.68680781e-01 -2.89282322e-01 2.02806100e-01
-4.90912855e-01 1.22069441e-01 4.41840529e-01 1.17785323e+00
-5.72382025e-02 -3.76919091e-01 5.40828288e-01 -8.47401977e-01
1.42301679e+00 -3.69126678e-01 -8.58652174e-01 2.55075812e-01
-7.48266518e-01 7.18382657e-01 5.90502083e-01 -4.94094282e-01
-1.74557948e+00 -1.39383465e-01 -2.39813015e-01 -2.13582277e-01
-1.29087716e-01 1.85725704e-01 -1.06065869e-01 5.35635471e-01
4.59819019e-01 1.36478409e-01 4.38226104e-01 -4.84233558e-01
7.88808525e-01 8.49328876e-01 2.71583032e-02 -8.44295919e-01
5.31888247e-01 1.95926711e-01 -7.06927538e-01 -7.73442626e-01
-7.26396203e-01 -4.06677008e-01 -4.62798178e-01 -2.17963293e-01
1.38264346e+00 -7.56884038e-01 -9.98401880e-01 2.85473168e-01
-1.18281388e+00 -1.08831775e+00 -1.38327748e-01 5.09698987e-01
-9.25497830e-01 2.69009054e-01 -9.93316352e-01 -1.24516010e+00
-2.44842708e-01 -1.65005755e+00 9.92092788e-01 2.78397024e-01
-3.97695303e-01 -1.50847900e+00 3.82492185e-01 6.70586765e-01
2.27025017e-01 -4.26901877e-01 7.60607183e-01 -6.93768680e-01
-5.38850844e-01 4.37893242e-01 4.17129427e-01 2.89517790e-01
4.39241618e-01 -1.81314304e-01 -8.27852428e-01 -2.89034694e-01
2.57587343e-01 -7.78191447e-01 6.20360017e-01 5.21317840e-01
7.43856847e-01 -7.25108266e-01 -7.26590529e-02 2.26717651e-01
9.09629643e-01 5.56305110e-01 1.62498236e-01 1.21356778e-01
4.16559011e-01 6.83915079e-01 6.03506029e-01 6.40582860e-01
3.48574549e-01 8.79008830e-01 2.74211228e-01 1.85201198e-01
3.32204014e-01 -2.71982044e-01 1.06138802e+00 9.19034779e-01
4.14536968e-02 -4.75767344e-01 -6.43559813e-01 2.66508400e-01
-2.14052796e+00 -1.07281876e+00 5.30283272e-01 1.60559249e+00
1.22360528e+00 7.06087947e-02 2.73664325e-01 -6.03174031e-01
5.51332593e-01 5.83609045e-01 -8.47496212e-01 -5.81451833e-01
2.18789771e-01 6.28361776e-02 3.64316583e-01 1.23944938e+00
-1.05088484e+00 1.57031238e+00 6.35408163e+00 5.50621212e-01
-8.02521169e-01 2.71828473e-01 4.94966149e-01 -7.47583434e-02
-2.42149293e-01 1.08672038e-01 -1.11128712e+00 2.44518965e-01
9.47419345e-01 -7.58118704e-02 7.92806029e-01 9.61575031e-01
7.03082204e-01 6.59269765e-02 -8.91420722e-01 6.29610181e-01
-2.23233044e-01 -1.25206232e+00 -1.72756910e-01 1.90147668e-01
8.43781173e-01 -1.24040745e-01 1.01820320e-01 8.73207748e-01
1.40732348e+00 -8.69373918e-01 2.95066804e-01 3.38092208e-01
4.27516699e-01 -4.60502982e-01 4.18088138e-01 5.86261928e-01
-7.82969773e-01 -2.45201543e-01 -3.31863314e-01 -2.60682523e-01
3.87912422e-01 -4.92016315e-01 -1.29812717e+00 -8.04148540e-02
9.64087322e-02 6.17838204e-01 1.07273638e-01 1.42324582e-01
-6.19456947e-01 1.01618063e+00 3.78339030e-02 -5.22302985e-01
8.43428195e-01 -6.40764534e-01 7.39042461e-01 9.38139915e-01
-2.88867921e-01 9.43596289e-02 5.27140617e-01 9.05430555e-01
-7.45900422e-02 4.25096564e-02 -5.78087687e-01 -4.78210598e-01
3.56318653e-01 9.91835654e-01 -4.14730847e-01 -4.74479675e-01
-5.03801525e-01 9.94353771e-01 5.21517336e-01 6.47945762e-01
-8.45804155e-01 2.48528302e-01 1.18595743e+00 -5.39917409e-01
1.84428334e-01 -4.82617170e-01 -3.20483208e-01 -1.37673008e+00
-4.93340313e-01 -1.20776463e+00 2.55597293e-01 -2.81532764e-01
-1.20211840e+00 2.84210622e-01 1.96756482e-01 -8.47780168e-01
-1.08697712e+00 -7.96593904e-01 -6.65894389e-01 8.59657586e-01
-9.91973221e-01 -7.33756840e-01 2.52619833e-01 4.68017012e-01
1.34670472e+00 -6.68389440e-01 9.89128947e-01 -2.69746810e-01
-5.18415630e-01 4.52355236e-01 4.25709754e-01 2.84162581e-01
4.75899339e-01 -1.61803377e+00 1.87355325e-01 5.00684083e-01
2.70180911e-01 6.56457484e-01 9.35326278e-01 -6.82998240e-01
-1.34642708e+00 -5.14850557e-01 4.87070560e-01 -5.57870865e-01
7.98226357e-01 -4.62894440e-01 -6.33957028e-01 9.12320375e-01
8.77774119e-01 -1.11743951e+00 5.41676223e-01 3.50730568e-01
2.85823733e-01 4.43340153e-01 -7.71959364e-01 1.20098960e+00
7.82949388e-01 -6.93098783e-01 -8.48937988e-01 4.36833858e-01
9.67259765e-01 -4.89659965e-01 -6.33418858e-01 -2.33263746e-02
2.69562721e-01 -6.87718034e-01 8.25251222e-01 -1.03496635e+00
3.56319278e-01 3.52204353e-01 1.57705080e-02 -1.57747793e+00
-7.18427747e-02 -1.13416910e+00 -4.13158089e-01 1.00875080e+00
4.53992993e-01 -5.46271205e-01 1.01037073e+00 8.85182142e-01
-1.51104137e-01 -7.48906374e-01 -7.69080162e-01 -4.68109101e-01
4.13720399e-01 -2.02086002e-01 2.66862214e-01 8.70312631e-01
1.80131674e-01 7.85283208e-01 -7.89436340e-01 -2.11895034e-01
3.49445909e-01 3.14808823e-02 1.03375745e+00 -6.87747777e-01
-6.18504405e-01 -6.46948338e-01 3.61206949e-01 -1.77442575e+00
7.52059102e-01 -5.84095955e-01 2.99190283e-01 -1.48547184e+00
-2.35696673e-01 -1.95201188e-01 -2.50406768e-02 3.90036285e-01
-3.68581086e-01 -6.87225342e-01 3.05885226e-01 1.19367518e-01
-5.45543790e-01 1.30837452e+00 1.39637351e+00 -2.69326061e-01
-6.57129407e-01 3.37877631e-01 -4.83822078e-01 7.71039784e-01
1.14997959e+00 -3.13381553e-01 -8.50031734e-01 -1.67298913e-01
-3.13395292e-01 5.93541682e-01 -6.31660819e-02 -4.06586021e-01
1.01619698e-01 -7.65347898e-01 -2.59583831e-01 -3.14090133e-01
8.29812586e-01 -3.60623509e-01 -5.69865108e-01 4.50183272e-01
-1.05939221e+00 1.05759904e-01 -1.27796140e-02 6.01034939e-01
-1.48483947e-01 -4.78066981e-01 7.60317266e-01 -4.03334439e-01
-7.63804793e-01 1.14942864e-01 -9.55116332e-01 2.67571718e-01
7.60664284e-01 7.14711770e-02 8.44813325e-03 -1.10434890e+00
-8.80681813e-01 6.40717924e-01 8.97036344e-02 5.50106704e-01
4.31941837e-01 -1.08944786e+00 -6.60095930e-01 -2.03168020e-01
-2.82671064e-01 -4.24797028e-01 -1.27044857e-01 4.42223698e-01
-7.68785104e-02 7.21802056e-01 -2.63011485e-01 -3.86539459e-01
-1.26825523e+00 2.11089537e-01 4.33052659e-01 -8.53456557e-01
-3.59465241e-01 8.62514913e-01 1.52631700e-01 -7.54421651e-01
5.78182459e-01 -2.02726424e-01 -4.22671467e-01 1.22011974e-01
7.79070854e-02 1.78380162e-01 -7.72492290e-01 -2.61334151e-01
3.65928739e-01 -2.62987643e-01 -2.71585345e-01 -1.00234067e+00
1.19774771e+00 -1.63955674e-01 2.68635362e-01 7.21297562e-01
8.90580773e-01 -1.25481203e-01 -1.96001613e+00 -4.70306933e-01
1.43327072e-01 -2.54061550e-01 -1.14637844e-01 -6.81852043e-01
-5.77624619e-01 8.91858041e-01 4.34013724e-01 2.64634073e-01
4.34488982e-01 1.65700130e-02 5.84537864e-01 6.26756310e-01
1.72083586e-01 -1.54280508e+00 5.80276310e-01 9.73500431e-01
8.33240330e-01 -1.40748727e+00 -2.35715628e-01 1.09906361e-01
-1.23586977e+00 1.02985620e+00 9.35888886e-01 8.57992768e-02
3.32363576e-01 1.02353401e-01 4.43440855e-01 -8.57022479e-02
-1.27249384e+00 -3.41602087e-01 -2.62903683e-02 3.37508589e-01
6.52492583e-01 1.07480727e-01 -5.23535073e-01 1.01989746e-01
-1.86265007e-01 -3.19073528e-01 2.28933826e-01 9.06706154e-01
-4.40172195e-01 -1.78847682e+00 -2.69297570e-01 4.19273376e-01
-4.32641447e-01 -4.91256826e-02 -5.72565734e-01 6.01758122e-01
-4.28149134e-01 1.20769513e+00 -1.43085077e-01 -1.62314862e-01
-1.37445852e-01 5.87352037e-01 2.27969512e-01 -7.27431118e-01
-6.68153346e-01 2.06044182e-01 3.23654324e-01 -3.79466474e-01
-5.21246672e-01 -7.82724440e-01 -1.31067717e+00 -7.33376667e-02
-3.05247456e-01 4.96825069e-01 4.18940693e-01 1.13096344e+00
-8.11229087e-03 5.35844862e-01 6.25962734e-01 -6.66616738e-01
-1.40518093e+00 -1.46643567e+00 -1.84569746e-01 4.14404392e-01
2.31631950e-01 -9.46629047e-01 -2.93336421e-01 2.24753562e-02] | [12.946786880493164, 8.068065643310547] |
754e51e8-c4f3-4a95-83e8-bf8d4b231284 | robust-portfolio-design-and-stock-price | 2204.01850 | null | https://arxiv.org/abs/2204.01850v1 | https://arxiv.org/pdf/2204.01850v1.pdf | Robust Portfolio Design and Stock Price Prediction Using an Optimized LSTM Model | Accurate prediction of future prices of stocks is a difficult task to perform. Even more challenging is to design an optimized portfolio with weights allocated to the stocks in a way that optimizes its return and the risk. This paper presents a systematic approach towards building two types of portfolios, optimum risk, and eigen, for four critical economic sectors of India. The prices of the stocks are extracted from the web from Jan 1, 2016, to Dec 31, 2020. Sector-wise portfolios are built based on their ten most significant stocks. An LSTM model is also designed for predicting future stock prices. Six months after the construction of the portfolios, i.e., on Jul 1, 2021, the actual returns and the LSTM-predicted returns for the portfolios are computed. A comparison of the predicted and the actual returns indicate a high accuracy level of the LSTM model. | ['Gourab Nath', 'Saikat Mondal', 'Jaydip Sen'] | 2022-03-02 | null | null | null | null | ['stock-price-prediction'] | ['time-series'] | [-5.17773092e-01 2.88376268e-02 -1.05190545e-01 -1.42560273e-01
-5.20875990e-01 -7.63714135e-01 6.58910573e-01 -2.36403704e-01
-3.15320373e-01 6.54921949e-01 5.37522554e-01 -6.84560657e-01
-4.61482853e-01 -1.13964021e+00 -3.55745941e-01 -4.78960276e-01
-3.36216360e-01 3.03816408e-01 -1.16729230e-01 -7.95076862e-02
7.78934658e-01 4.68048453e-01 -1.14986205e+00 2.84892678e-01
3.42095226e-01 1.71023953e+00 2.96164989e-01 4.49890733e-01
-4.68444735e-01 7.09339380e-01 -4.79780167e-01 -8.65635335e-01
8.28245342e-01 3.58260684e-02 -1.62396491e-01 -4.97051954e-01
-2.91040242e-01 -6.39085591e-01 -8.63640383e-02 1.00909662e+00
1.14904597e-01 -7.91155621e-02 5.05975366e-01 -6.61165237e-01
-4.10731941e-01 1.07154179e+00 -3.15616965e-01 3.65654737e-01
-4.51999128e-01 -6.18788740e-03 1.37154794e+00 -1.00308645e+00
1.51144087e-01 6.13511801e-01 7.59570360e-01 -5.06536625e-02
-7.22681880e-01 -7.88100421e-01 -5.56762516e-02 3.73766050e-02
-8.14420462e-01 -1.91811875e-01 6.32174432e-01 -7.86640763e-01
1.11317050e+00 2.21377403e-01 9.51308668e-01 4.67106998e-01
1.03973687e+00 1.83003142e-01 1.12074614e+00 -2.17139900e-01
1.96283162e-01 2.87015826e-01 -3.38582665e-01 -1.18297294e-01
5.83900154e-01 4.82387751e-01 -3.55414689e-01 -1.79465175e-01
4.25614536e-01 2.22317085e-01 2.61670262e-01 2.12055683e-01
-1.07633328e+00 6.84784472e-01 2.08344713e-01 6.52502298e-01
-1.08945286e+00 1.47524387e-01 1.56199306e-01 4.21583444e-01
6.79397762e-01 5.23290396e-01 -7.33650506e-01 -1.10213675e-01
-1.39268386e+00 1.81348935e-01 8.02423060e-01 2.68116713e-01
2.62713581e-01 6.24053240e-01 -1.75274372e-01 3.22553843e-01
3.42291415e-01 6.75234437e-01 5.98095894e-01 -8.67774904e-01
6.42275155e-01 3.77116740e-01 4.79566246e-01 -1.23297012e+00
-1.14811063e-01 -9.39674318e-01 -5.63128471e-01 3.08406830e-01
2.92355210e-01 -4.71257001e-01 -5.65179229e-01 1.16103899e+00
-4.91512746e-01 1.65794343e-01 3.37828070e-01 1.78132564e-01
7.67830014e-02 1.09922242e+00 -3.81202064e-02 -2.64865249e-01
8.53273094e-01 -6.18761480e-01 -6.45845592e-01 -4.18619484e-01
6.57919794e-02 -7.04127908e-01 1.31854445e-01 5.29161513e-01
-1.15004992e+00 -2.80077308e-01 -8.16174448e-01 9.15387928e-01
-4.85430837e-01 1.11778699e-01 4.15131301e-01 3.91149402e-01
-9.69204485e-01 1.17927682e+00 -5.31270027e-01 4.93592352e-01
-5.88657893e-03 3.46859813e-01 1.91650525e-01 9.73444879e-01
-1.33562136e+00 1.39960134e+00 7.92276859e-01 3.63791943e-01
-6.87594593e-01 -5.57423949e-01 -2.23038718e-01 5.37211299e-01
-1.13945484e-01 1.50177709e-03 1.26902223e+00 -1.02371442e+00
-1.30166018e+00 2.41896763e-01 4.58782911e-01 -1.09601057e+00
5.66506624e-01 -1.72041774e-01 -4.44813043e-01 -4.75426793e-01
-2.24285558e-01 -6.33597746e-02 6.05101287e-01 -7.79592812e-01
-1.23554003e+00 -1.64358944e-01 -2.45989799e-01 -7.83467963e-02
-7.82428265e-01 1.23993650e-01 1.58880785e-01 -9.83438969e-01
-3.69123630e-02 -6.36494696e-01 -2.82636017e-01 -8.13186884e-01
-1.55058935e-01 3.28999221e-01 1.31861195e-01 -1.48404813e+00
1.48224115e+00 -1.75382316e+00 -4.74195659e-01 6.10901177e-01
-1.52738869e-01 1.37969926e-01 1.86630502e-01 4.54616666e-01
-4.08359081e-01 2.97490388e-01 2.20226292e-02 5.17395362e-02
3.67500097e-01 4.09722663e-02 -1.06236911e+00 7.98730776e-02
1.14530802e-01 1.06506848e+00 -3.71777296e-01 1.98153570e-01
5.80754802e-02 -1.50339352e-03 2.15819683e-02 2.79661030e-01
-1.59441248e-01 -4.95362468e-02 -2.81605244e-01 4.37256753e-01
4.95075732e-01 4.90361191e-02 2.33927578e-01 8.95866752e-02
-7.80729294e-01 5.17205596e-01 -8.08470309e-01 5.24027348e-01
-4.13217932e-01 4.57465053e-01 -4.85883981e-01 -6.77092195e-01
1.30528021e+00 5.28155148e-01 6.34811699e-01 -8.97586823e-01
-2.17278209e-02 6.98182404e-01 1.97212085e-01 1.88723411e-02
7.23412991e-01 -5.35948694e-01 -1.45439148e-01 6.96403503e-01
-2.14610189e-01 2.47763902e-01 1.65331185e-01 -4.64148521e-01
6.65619910e-01 1.60026804e-01 2.37085849e-01 -3.53276938e-01
2.33323008e-01 -2.10352689e-01 7.44268894e-01 3.85944307e-01
2.51573622e-01 9.54710990e-02 5.96684754e-01 -7.07178473e-01
-1.03841364e+00 -9.07043099e-01 9.84835401e-02 6.92326307e-01
-8.68037105e-01 4.01632011e-01 -4.58545923e-01 -2.81311631e-01
2.63960242e-01 1.22231269e+00 -6.52042925e-01 1.12085760e-01
-3.99682730e-01 -9.21483159e-01 2.19004586e-01 3.35917681e-01
4.48386580e-01 -1.26525092e+00 -9.37658608e-01 7.23322809e-01
1.14673831e-01 -5.06462634e-01 -1.74811646e-01 1.92914635e-01
-1.07095921e+00 -6.55936778e-01 -1.07288694e+00 -2.02488422e-01
2.39095792e-01 -3.51944000e-01 1.14721537e+00 -4.06393975e-01
5.36540210e-01 -2.56246924e-01 1.66678458e-01 -1.16028190e+00
-6.11273199e-03 -7.18474910e-02 -1.26933813e-01 2.22584337e-01
7.26362243e-02 -5.38582504e-01 -4.18538541e-01 -6.55747950e-02
-5.14494359e-01 -2.03892127e-01 8.58346701e-01 4.57758665e-01
4.78480101e-01 5.01279891e-01 7.12590456e-01 -3.46874744e-01
6.95783257e-01 -6.38578355e-01 -1.15250182e+00 4.64809477e-01
-1.08199072e+00 6.43351153e-02 4.03899968e-01 -1.73738033e-01
-1.23188174e+00 -2.07696229e-01 -6.73406664e-03 -2.04233930e-01
6.93766475e-01 1.02794707e+00 3.94608676e-01 2.12788120e-01
-1.43139064e-01 4.33109641e-01 -3.68098319e-01 -7.98839331e-01
-1.15019716e-01 4.90063816e-01 6.52406871e-01 -3.07854980e-01
9.69552577e-01 6.90987781e-02 -2.56692827e-01 -1.48684964e-01
-8.60330164e-01 1.80768952e-01 -5.63306868e-01 -6.44176662e-01
3.59900624e-01 -7.05090463e-01 -3.06369722e-01 7.24611938e-01
-8.60999167e-01 -3.50253701e-01 -5.85032582e-01 6.73281312e-01
-3.98719043e-01 -3.02241147e-01 -4.59730029e-01 -1.29022741e+00
-9.87279713e-01 -7.47960150e-01 1.51015788e-01 2.28504553e-01
-3.39592010e-01 -1.19018543e+00 1.32178187e-01 -1.25751600e-01
9.73984361e-01 6.20165825e-01 9.12536442e-01 -9.78815138e-01
-5.05699337e-01 -5.82513094e-01 -2.34678015e-01 7.27588654e-01
2.50265360e-01 2.27121562e-01 -5.89382648e-01 -1.39756978e-01
4.27279890e-01 4.93150651e-01 6.91878796e-01 6.47285879e-01
5.27800322e-01 -8.06874394e-01 2.65667558e-01 3.92291754e-01
1.63530719e+00 9.22964513e-01 6.05859816e-01 9.83893871e-01
-7.08435150e-03 8.80467653e-01 6.65775537e-01 7.93060839e-01
3.25948238e-01 1.98260933e-01 3.40932995e-01 3.50870669e-01
4.79606509e-01 -2.23841429e-01 6.39086366e-01 1.26444054e+00
-2.14383230e-01 1.45190194e-01 -1.19295466e+00 6.74780846e-01
-1.45476818e+00 -1.01601779e+00 1.78208217e-01 2.42278528e+00
4.88390207e-01 8.88149619e-01 -1.12293819e-02 1.98447511e-01
3.94670874e-01 3.81656289e-01 -5.05490899e-01 -5.66203058e-01
-6.53334782e-02 1.45869076e-01 1.16633654e+00 3.55398268e-01
-8.94647658e-01 6.12065017e-01 6.93456602e+00 4.51891035e-01
-1.26750410e+00 -1.57030314e-01 1.14084148e+00 -2.71700829e-01
-7.27438629e-01 1.97379306e-01 -9.18543696e-01 9.62153137e-01
1.56588793e+00 -8.86900306e-01 2.39087418e-01 8.29413474e-01
4.85606223e-01 -1.02278173e-01 -4.02583331e-01 3.72194886e-01
-3.54073524e-01 -1.60150385e+00 1.08019255e-01 5.00530601e-01
7.53072441e-01 1.04001485e-01 5.49559534e-01 1.23404272e-01
4.65860486e-01 -7.39672661e-01 1.30727386e+00 1.47910285e+00
2.40444377e-01 -1.22240603e+00 1.17041314e+00 3.41392338e-01
-1.23353970e+00 -5.54045439e-01 -3.81288469e-01 -1.67803213e-01
2.86949366e-01 8.84489775e-01 -3.61519098e-01 4.29629833e-01
9.16164696e-01 5.52028954e-01 -4.21608865e-01 9.05269325e-01
3.92234921e-02 5.16355038e-01 -1.55211851e-01 1.18593173e-03
6.54545605e-01 -5.77255845e-01 2.96819180e-01 9.15732026e-01
1.17531633e+00 -2.43747309e-01 -5.29607892e-01 7.69270480e-01
8.06697980e-02 -1.01528466e-01 -6.34549081e-01 -3.64805758e-01
3.85882199e-01 8.95043850e-01 -3.25595886e-01 -2.15608418e-01
-3.40525299e-01 1.19588733e-01 -1.69029325e-01 2.48043552e-01
-4.91293192e-01 -4.89394784e-01 5.18533230e-01 1.17909640e-01
4.08275008e-01 -2.53724009e-01 -8.23760629e-01 -7.62257695e-01
3.29133123e-02 -7.05472708e-01 2.00329453e-01 -5.98131657e-01
-1.13499808e+00 5.98510802e-01 -3.97817269e-02 -1.32211316e+00
-7.08600640e-01 -6.29188478e-01 -1.07937825e+00 1.28149927e+00
-1.57496536e+00 -5.04680574e-01 4.85037088e-01 -1.64000824e-01
6.43016547e-02 -7.98146248e-01 4.81762558e-01 2.90718339e-02
-5.04335105e-01 3.96950878e-02 8.67204249e-01 1.31656602e-01
-7.13480636e-02 -1.31642365e+00 1.12570453e+00 8.70663762e-01
-7.76817929e-03 3.84743601e-01 6.21707141e-01 -9.55688655e-01
-8.51553023e-01 -1.10542130e+00 1.48245239e+00 -1.42574981e-02
1.20586920e+00 2.39756778e-01 -6.99038982e-01 9.29075420e-01
2.40432173e-01 -7.92999983e-01 5.60990334e-01 -5.29286444e-01
2.39241436e-01 -6.23130739e-01 -1.05937779e+00 5.11397064e-01
8.30946639e-02 -1.62856758e-01 -7.38065720e-01 -1.23518862e-01
2.60103077e-01 1.04428373e-01 -1.31958020e+00 3.63805145e-01
9.91789997e-01 -1.17756701e+00 9.55669582e-01 -2.57567376e-01
1.25798419e-01 2.98249722e-01 -2.80158579e-01 -1.39579606e+00
-4.58571583e-01 -5.41055739e-01 -2.10775897e-01 1.32153964e+00
7.28160083e-01 -1.23688519e+00 8.83973658e-01 9.56401050e-01
2.30801344e-01 -9.19613600e-01 -1.19464660e+00 -8.67247343e-01
2.78389663e-01 -4.03819382e-01 1.20914459e+00 4.64175135e-01
-4.94845212e-01 -5.64078212e-01 -5.32231450e-01 -6.27532974e-02
8.86102736e-01 3.41518253e-01 3.67549062e-01 -1.31309950e+00
-3.32379341e-02 -9.69610631e-01 1.29359946e-01 -1.87904358e-01
-3.99445295e-02 -5.17024040e-01 -3.73035580e-01 -1.58719683e+00
-2.32203677e-01 -2.21009329e-01 -9.94508028e-01 3.61320645e-01
3.13606679e-01 -1.88873351e-01 6.13073587e-01 4.87004429e-01
3.96683991e-01 4.41165328e-01 7.12927461e-01 -6.78843558e-02
-5.63911104e-04 3.87021601e-01 -7.78794944e-01 9.71345782e-01
1.38887584e+00 -5.71543992e-01 7.18615502e-02 -3.45773399e-01
5.15544713e-01 3.15016270e-01 -5.45386821e-02 -1.00652885e+00
-2.26393491e-02 -5.83162963e-01 5.63834310e-01 -1.08621919e+00
2.69768238e-01 -6.96728230e-01 9.99489069e-01 9.17843461e-01
-2.85567075e-01 7.44021595e-01 2.27897599e-01 1.13501489e-01
-3.92471671e-01 -6.53263569e-01 5.63042521e-01 -4.43022817e-01
-4.29591715e-01 1.27981365e-01 -5.04776180e-01 -3.12589794e-01
9.55777109e-01 -2.58585274e-01 2.36474618e-01 -4.26812112e-01
-8.16409588e-01 2.31338501e-01 -1.56526286e-02 3.06505978e-01
5.95431745e-01 -1.40497935e+00 -8.02295864e-01 -4.03742008e-02
-4.50267047e-01 -8.47819626e-01 -2.01388732e-01 3.17520320e-01
-6.30357027e-01 8.90255868e-01 -4.60571766e-01 4.41229254e-01
-5.41360915e-01 1.02887340e-01 5.96190035e-01 -1.10027838e+00
-3.36028039e-01 6.43330038e-01 -2.88046710e-02 1.60163250e-02
1.18530057e-01 -4.70472693e-01 -6.40952349e-01 7.93581188e-01
7.14554071e-01 4.43473905e-01 1.75228879e-01 -6.48305953e-01
-5.77435791e-02 5.50674260e-01 1.55568093e-01 -5.36417067e-01
1.97691178e+00 -9.35374852e-03 -3.75956833e-01 7.32931554e-01
7.95946538e-01 -2.15554014e-01 -1.30029488e+00 -2.34894663e-01
1.07319069e+00 -5.32664239e-01 1.48165405e-01 -8.52488279e-01
-1.61715317e+00 7.67052770e-01 4.73632723e-01 2.82508701e-01
1.03426325e+00 -8.06387424e-01 1.09893453e+00 3.26743424e-01
4.31861222e-01 -1.35060918e+00 -3.32578510e-01 5.95638394e-01
1.20037150e+00 -4.64921772e-01 8.48074257e-03 5.93087494e-01
-5.38568974e-01 1.25666392e+00 -8.93709436e-02 -2.14378834e-01
9.82469201e-01 1.99535564e-01 2.01048739e-02 6.77002594e-02
-8.90481830e-01 2.10696936e-01 4.60799485e-01 1.47135958e-01
1.27883703e-01 2.82705039e-01 -1.00291535e-01 9.89297271e-01
-5.75664699e-01 -7.19507486e-02 4.68458265e-01 6.84412360e-01
-6.15057230e-01 -8.98111224e-01 -5.19341946e-01 9.83874321e-01
-1.15814757e+00 -3.04976285e-01 -1.72761127e-01 4.97339785e-01
-2.28230879e-01 5.85634828e-01 4.00605410e-01 -4.77275014e-01
5.95950961e-01 2.09837064e-01 -4.07995373e-01 -2.01629594e-01
-1.01041222e+00 8.45245644e-02 1.06757030e-01 -1.11251041e-01
2.10402280e-01 -8.32046926e-01 -9.21719670e-01 -4.91195470e-01
-1.57115102e-01 2.83313543e-01 7.49057412e-01 7.97984660e-01
3.99890468e-02 4.40425545e-01 1.12834644e+00 -1.10204911e+00
-1.17613053e+00 -8.24865580e-01 -9.33104694e-01 -2.98627079e-01
1.58125475e-01 -3.31783086e-01 -5.68844020e-01 -2.51652211e-01] | [4.530671119689941, 4.164480686187744] |
48525220-87bc-40a1-a3c7-0955a98b9c8d | distributionally-robust-survival-analysis-a | 2211.10508 | null | https://arxiv.org/abs/2211.10508v1 | https://arxiv.org/pdf/2211.10508v1.pdf | Distributionally Robust Survival Analysis: A Novel Fairness Loss Without Demographics | We propose a general approach for training survival analysis models that minimizes a worst-case error across all subpopulations that are large enough (occurring with at least a user-specified minimum probability). This approach uses a training loss function that does not know any demographic information to treat as sensitive. Despite this, we demonstrate that our proposed approach often scores better on recently established fairness metrics (without a significant drop in prediction accuracy) compared to various baselines, including ones which directly use sensitive demographic information in their training loss. Our code is available at: https://github.com/discovershu/DRO_COX | ['George H. Chen', 'Shu Hu'] | 2022-11-18 | null | null | null | null | ['survival-analysis'] | ['miscellaneous'] | [-1.32168621e-01 2.56396574e-03 -6.90371752e-01 -7.22128272e-01
-1.33302915e+00 -3.46713513e-01 4.11068618e-01 7.98146307e-01
-7.81778991e-01 1.30886793e+00 2.29933426e-01 -3.87411147e-01
-1.65954500e-01 -1.01277423e+00 -4.86640811e-01 -7.40390718e-01
-2.30591819e-01 6.23867035e-01 3.40216421e-02 -6.84912596e-03
2.13594884e-01 3.16472530e-01 -1.13748515e+00 -2.01034099e-01
1.00272727e+00 6.94361031e-01 -7.38660574e-01 5.32337606e-01
6.43032253e-01 8.62902284e-01 -4.79685456e-01 -8.19103003e-01
1.17268860e-01 -1.98460445e-01 -7.51995981e-01 -7.70940125e-01
4.48612571e-01 -3.72479886e-01 -3.36917400e-01 7.54879773e-01
7.24143803e-01 1.40229151e-01 1.02562094e+00 -1.23007214e+00
-3.19838226e-01 5.73146164e-01 -5.46758413e-01 5.20031452e-01
1.23324424e-01 1.58759411e-02 1.04759848e+00 -5.89884341e-01
4.62088317e-01 9.03102398e-01 1.13821912e+00 7.62187481e-01
-1.55941868e+00 -9.68968332e-01 -3.92225415e-01 -8.21134001e-02
-1.72805226e+00 -7.52404630e-01 2.63624221e-01 -4.49492365e-01
7.29559779e-01 5.38440466e-01 9.22456756e-02 1.10359633e+00
4.61348861e-01 4.10401553e-01 1.04022419e+00 -2.06488743e-01
3.24254721e-01 1.47297546e-01 5.39211094e-01 6.20567501e-01
4.51331884e-01 6.24181569e-01 -2.51257569e-01 -8.97033811e-01
2.99932659e-01 1.35601545e-02 -2.04600483e-01 -3.38398367e-01
-9.46830392e-01 1.13008845e+00 1.56309023e-01 -5.20496294e-02
-1.10087045e-01 3.28009754e-01 5.17372727e-01 2.67920315e-01
7.26883471e-01 1.67680040e-01 -6.37506604e-01 -6.99354112e-02
-1.13266122e+00 1.49309114e-01 6.90389574e-01 5.96564889e-01
4.48116332e-01 -3.79879266e-01 -4.85524595e-01 5.94391048e-01
-1.64193474e-02 4.40351695e-01 8.96435827e-02 -1.10761631e+00
2.23042756e-01 3.17422777e-01 6.50163069e-02 -5.69354117e-01
-6.95926726e-01 -6.74283564e-01 -9.92469788e-01 9.09794718e-02
6.83173180e-01 -3.78704727e-01 -4.89053518e-01 2.34869385e+00
3.89980316e-01 3.61481309e-01 -1.75847393e-02 5.09837270e-01
5.27680635e-01 2.86128968e-01 5.23099720e-01 -3.34132046e-01
1.13470995e+00 -4.26917166e-01 -3.22631299e-01 1.64652959e-01
9.99234140e-01 -3.75192702e-01 9.97153759e-01 1.44945413e-01
-1.09118545e+00 1.43903151e-01 -8.48718941e-01 -1.02369770e-01
-2.07928076e-01 -6.59609139e-02 6.52858555e-01 1.16202581e+00
-1.16012692e+00 7.57835984e-01 -7.24011481e-01 -6.66114688e-01
6.73255622e-01 4.91531640e-01 -2.77125835e-01 9.39780548e-02
-1.30323350e+00 8.26281965e-01 2.83254176e-01 -5.28054059e-01
-8.94073486e-01 -1.27786815e+00 -5.83951294e-01 5.18330000e-02
1.53388977e-01 -1.29513347e+00 1.00609374e+00 -6.55943871e-01
-9.30077374e-01 9.26094532e-01 7.19435560e-03 -7.59489179e-01
8.71752024e-01 -6.85433522e-02 -1.76044703e-01 -1.88772932e-01
-7.57293776e-02 2.88335472e-01 1.58318117e-01 -9.72332776e-01
-7.23719656e-01 -4.20817167e-01 1.80316027e-02 -3.09166666e-02
-5.50394773e-01 2.32962862e-01 2.76242960e-02 -7.81935215e-01
-7.29452550e-01 -7.29021013e-01 -4.51478869e-01 2.94146352e-02
-5.67704082e-01 -9.39140022e-02 1.69183146e-02 -5.66229939e-01
1.58062983e+00 -1.79643786e+00 -2.16877803e-01 3.03310901e-01
2.16359273e-01 -1.80548951e-01 9.86681134e-02 4.59281683e-01
6.53661191e-02 4.40224320e-01 -6.45692050e-01 -5.39882660e-01
-1.96257636e-01 -1.44677863e-01 -1.36585031e-02 7.62618601e-01
-3.42499584e-01 4.30865049e-01 -9.11853373e-01 -6.12466335e-01
3.94084789e-02 5.57627976e-01 -6.84050262e-01 5.97301237e-02
3.43995988e-01 3.53952587e-01 -3.21401507e-01 6.22745514e-01
5.50880075e-01 -3.25614363e-01 7.11341277e-02 2.53931582e-01
1.09418191e-01 3.13382857e-02 -6.09378874e-01 1.16578162e+00
-3.26061904e-01 1.55571595e-01 -2.91552663e-01 -9.91591513e-01
6.60073519e-01 2.24760070e-01 6.71523988e-01 -6.69268668e-02
2.72964656e-01 2.35578939e-01 -2.35807598e-01 1.05308555e-01
-2.05805562e-02 -4.29100275e-01 -3.00581425e-01 3.34191382e-01
-5.82224615e-02 3.96927178e-01 -9.68231484e-02 3.34954321e-01
1.40305531e+00 -4.67437953e-01 6.92023575e-01 -6.56492472e-01
5.73818088e-01 4.34055701e-02 9.88258660e-01 1.01892602e+00
-4.78602946e-01 7.07642555e-01 6.91045880e-01 -2.45280832e-01
-9.74194705e-01 -1.21793091e+00 -7.60550678e-01 9.46453393e-01
-2.39572972e-01 -3.88129950e-01 -5.74246824e-01 -9.10677016e-01
2.56704837e-01 9.34913039e-01 -9.36954439e-01 -3.93562108e-01
-1.08807743e-01 -1.50896442e+00 9.54984009e-01 3.19886535e-01
3.98271419e-02 -4.63543922e-01 -3.36122900e-01 -2.77938217e-01
-5.74142598e-02 -3.90839756e-01 -5.86076260e-01 1.11337252e-01
-1.05272746e+00 -1.11100698e+00 -8.47185194e-01 -3.17193687e-01
6.98984206e-01 -4.84611452e-01 1.31096053e+00 2.61019260e-01
-2.70598263e-01 2.25977153e-01 -2.20578730e-01 -3.91724050e-01
-3.83394331e-01 4.29916501e-01 1.71217576e-01 -3.52891475e-01
5.10124445e-01 -5.26850283e-01 -9.42788005e-01 1.63678631e-01
-5.46526611e-01 -3.54877889e-01 2.61823565e-01 9.50788736e-01
4.83951569e-01 -2.77154148e-02 9.21338379e-01 -1.35412467e+00
3.75937879e-01 -9.59011257e-01 -5.10986388e-01 3.24851990e-01
-1.08972335e+00 -1.25239357e-01 6.27406418e-01 -6.35143816e-02
-7.77702928e-01 -9.13045779e-02 -8.96294340e-02 -9.82145220e-02
-2.09182635e-01 4.80822325e-01 3.67839970e-02 -2.95445696e-02
8.72375786e-01 -1.31882310e-01 9.27717909e-02 -4.64232743e-01
-1.35234594e-01 4.60836530e-01 2.80124694e-01 -5.36536932e-01
7.96929002e-01 5.13274133e-01 3.59176517e-01 -2.09205598e-01
-7.78898597e-01 -3.34851593e-01 -5.22289395e-01 5.87003604e-02
4.68937278e-01 -7.99129069e-01 -7.18004107e-01 5.14526963e-01
-4.76731390e-01 -5.38632631e-01 -2.81568468e-01 5.30185044e-01
-4.94263232e-01 3.95387053e-01 -6.80711925e-01 -7.78603256e-01
-6.26567304e-01 -5.40450335e-01 6.90372884e-01 9.96714830e-02
-3.36692631e-01 -1.36034060e+00 3.70937884e-01 3.52043539e-01
4.91941750e-01 6.54437006e-01 1.03718543e+00 -9.15285766e-01
-7.78943673e-02 -4.08253312e-01 -4.59662378e-02 2.60058399e-02
4.87118699e-02 2.21804574e-01 -9.78070021e-01 -7.35065579e-01
-3.69760782e-01 -3.77876252e-01 1.05324399e+00 8.57251048e-01
1.35985398e+00 -4.29046929e-01 -5.96098065e-01 8.57618988e-01
1.55051827e+00 -1.04179308e-01 6.29786372e-01 2.18851507e-01
2.50944406e-01 6.16113186e-01 4.88677144e-01 7.90599048e-01
6.60037577e-01 6.14049554e-01 9.38941240e-02 -2.95781285e-01
1.83664083e-01 -1.39429867e-01 2.47477695e-01 2.53240943e-01
-1.56168506e-01 -3.25389564e-01 -1.19340873e+00 5.47975838e-01
-1.91473508e+00 -1.08718014e+00 -1.45562798e-01 2.91495371e+00
1.09560561e+00 8.74869451e-02 5.04185438e-01 -1.29386902e-01
6.49055481e-01 1.62906554e-02 -6.44160390e-01 -4.17634547e-01
-1.85901225e-01 2.87569374e-01 1.00961399e+00 7.35312104e-01
-1.33070862e+00 5.45881152e-01 7.31803370e+00 1.01975477e+00
-7.38333225e-01 3.50099117e-01 1.35447693e+00 -4.57123518e-01
-3.28030854e-01 8.11826736e-02 -6.75848603e-01 5.88753581e-01
1.43840945e+00 -6.85264587e-01 9.21970755e-02 5.72584867e-01
3.41365308e-01 -1.38391227e-01 -1.09206128e+00 4.49492812e-01
-2.41005272e-01 -1.09560275e+00 -9.22217220e-02 1.68738231e-01
5.70336521e-01 -9.21321586e-02 2.27201864e-01 3.50218236e-01
6.80203736e-01 -1.09928477e+00 3.85306418e-01 7.07998991e-01
1.03319299e+00 -1.14590991e+00 9.11222219e-01 2.58238345e-01
-8.23301256e-01 -4.05729674e-02 -2.90259600e-01 5.35371117e-02
9.16291177e-02 1.12296987e+00 -6.16262496e-01 5.70210993e-01
6.08524799e-01 6.10620797e-01 -6.43655717e-01 1.40628755e+00
2.57601231e-01 1.00087833e+00 -3.38923961e-01 3.23652983e-01
-3.62552971e-01 2.15908706e-01 5.83701491e-01 1.09093273e+00
4.74408060e-01 2.81331480e-01 3.47132012e-02 5.45285761e-01
-2.47503921e-01 3.97023231e-01 -4.05966133e-01 5.12432396e-01
7.27734506e-01 8.81954730e-01 -4.41306174e-01 -1.44568950e-01
-4.13301677e-01 5.80558121e-01 2.60572881e-01 1.21987790e-01
-8.73355806e-01 -3.91357958e-01 8.05344820e-01 3.95248055e-01
-2.61091202e-01 3.89717281e-01 -5.39451301e-01 -1.06143045e+00
-4.63831037e-01 -4.12989914e-01 1.31510973e+00 -1.84153125e-01
-1.89031374e+00 2.33499855e-01 2.38619432e-01 -1.10148478e+00
-1.28159359e-01 -2.03607529e-01 -6.24575615e-01 8.14131737e-01
-1.43161654e+00 -9.51567531e-01 -1.53211206e-02 5.79573214e-01
-6.74659908e-02 -4.54663783e-02 8.46397340e-01 5.60546517e-01
-9.19268966e-01 1.41985846e+00 3.07953417e-01 1.36655243e-02
1.12113833e+00 -1.28855419e+00 1.67210922e-01 5.87556899e-01
-5.41788280e-01 4.63391989e-01 7.11535513e-01 -5.30518055e-01
-6.51657283e-01 -1.19540536e+00 1.23019850e+00 -8.05120230e-01
4.06921506e-01 -2.42942169e-01 -7.00607836e-01 7.49906003e-01
-1.99920818e-01 -2.02216655e-02 1.14536786e+00 4.75807369e-01
-2.73324102e-01 -2.92356014e-01 -1.61710310e+00 3.14780653e-01
1.04963911e+00 -1.48925602e-01 -1.51455954e-01 3.16675693e-01
3.11092466e-01 -7.86546990e-02 -1.29196894e+00 6.59612477e-01
6.04711652e-01 -1.06049526e+00 1.13090169e+00 -7.99310565e-01
3.90148580e-01 4.63216193e-02 -2.68408228e-02 -1.07993078e+00
-4.40520078e-01 -4.96405244e-01 1.98529754e-02 1.28290868e+00
7.87432313e-01 -1.07082260e+00 8.98637593e-01 9.27633882e-01
2.70559758e-01 -8.42487395e-01 -1.14866519e+00 -8.85998905e-01
7.66206145e-01 5.12039028e-02 5.67357242e-01 1.29165578e+00
8.99565294e-02 -8.26499909e-02 -5.27556419e-01 1.17428817e-01
1.16556299e+00 -1.86882600e-01 4.47254360e-01 -1.60532260e+00
5.55508621e-02 -4.41679746e-01 -4.07153994e-01 -8.10869634e-02
3.34374666e-01 -1.05796862e+00 -4.07243460e-01 -1.18392229e+00
9.20980275e-01 -8.52245808e-01 -8.97241890e-01 6.70074046e-01
-3.22257489e-01 3.67167085e-01 -3.19135815e-01 3.10595125e-01
-3.47967327e-01 4.43168759e-01 5.21241128e-01 1.20673813e-01
-2.38034036e-02 3.16681266e-01 -9.99118447e-01 3.71057212e-01
1.18368351e+00 -8.57925951e-01 -2.64653981e-01 1.09956443e-01
3.78715433e-02 3.11317563e-01 4.76650476e-01 -1.15267551e+00
9.09253210e-02 -4.35716003e-01 3.92708063e-01 -1.58107877e-01
-5.54734543e-02 -4.87901926e-01 5.46081245e-01 6.78442836e-01
-6.93974793e-01 -1.04735948e-01 8.86473134e-02 4.58627492e-01
1.56865910e-01 -1.13635369e-01 9.77835715e-01 2.59079039e-01
-1.81281805e-01 5.64796567e-01 -2.97229916e-01 2.68355906e-01
1.17449057e+00 1.33271843e-01 -7.49608219e-01 -5.31859636e-01
-6.21481299e-01 3.90044749e-01 9.43786383e-01 -1.49099991e-01
1.47244915e-01 -1.25171471e+00 -1.07271552e+00 -3.54640633e-01
2.89165288e-01 -4.90794718e-01 3.11708838e-01 1.15653908e+00
-3.23163718e-01 2.72327393e-01 -1.46525308e-01 -1.23581491e-01
-1.21870971e+00 5.92048824e-01 6.79643035e-01 -4.69630778e-01
-3.25951606e-01 8.84635806e-01 2.01329902e-01 -6.13792181e-01
1.05314292e-01 1.07905671e-01 8.11411068e-03 -1.40438573e-02
4.25790697e-01 7.12015569e-01 -5.83037436e-02 -6.03930891e-01
-6.82261169e-01 1.86414137e-01 1.17737122e-01 4.47210297e-02
1.10215223e+00 -9.74583998e-02 1.13806561e-01 3.96727860e-01
1.26192117e+00 1.29555240e-01 -9.43624675e-01 -9.87297222e-02
-2.17038617e-02 -5.78800499e-01 3.07072997e-01 -1.04762459e+00
-1.19923365e+00 4.96322453e-01 7.53342092e-01 -1.43379211e-01
1.20196867e+00 -1.21262498e-01 7.61367083e-01 1.32151013e-02
4.00107592e-01 -7.62727141e-01 -5.79401314e-01 1.96847051e-01
3.12351167e-01 -1.38154387e+00 3.11628431e-01 -3.57806683e-01
-5.67653894e-01 6.39363050e-01 4.50714588e-01 -5.80047965e-02
1.10675049e+00 4.27066565e-01 5.45470081e-02 -2.95716394e-02
-9.49254334e-01 -1.21181965e-01 8.73410329e-02 7.61673987e-01
5.40362418e-01 1.71449736e-01 -7.88505137e-01 7.56973982e-01
-1.55405030e-01 1.56250641e-01 4.88515764e-01 5.73370457e-01
-1.92889765e-01 -1.24136400e+00 -7.44607747e-02 1.09114099e+00
-9.27449107e-01 -2.74294049e-01 -2.58710057e-01 6.06241286e-01
-2.49235556e-01 7.87020385e-01 1.64037645e-01 -2.91395515e-01
1.38648078e-01 4.23665792e-02 1.49408504e-01 -3.88598800e-01
-6.96066260e-01 -4.57648158e-01 3.54311973e-01 -5.46602130e-01
-4.59463328e-01 -9.58488464e-01 -9.82141018e-01 -1.01313663e+00
-2.43681949e-02 1.83712095e-01 -3.10431644e-02 4.43214029e-01
3.50756913e-01 4.94218431e-02 7.96082914e-01 -2.00682506e-01
-7.20406294e-01 -5.41156709e-01 -6.63261592e-01 3.48336160e-01
4.33062941e-01 -6.64078712e-01 -6.73873901e-01 -9.94108021e-02] | [7.830067157745361, 5.511399745941162] |
4eacc4dd-9a84-43cd-8033-8f4f9a9a9374 | recycle-gan-unsupervised-video-retargeting | 1808.05174 | null | http://arxiv.org/abs/1808.05174v1 | http://arxiv.org/pdf/1808.05174v1.pdf | Recycle-GAN: Unsupervised Video Retargeting | We introduce a data-driven approach for unsupervised video retargeting that
translates content from one domain to another while preserving the style native
to a domain, i.e., if contents of John Oliver's speech were to be transferred
to Stephen Colbert, then the generated content/speech should be in Stephen
Colbert's style. Our approach combines both spatial and temporal information
along with adversarial losses for content translation and style preservation.
In this work, we first study the advantages of using spatiotemporal constraints
over spatial constraints for effective retargeting. We then demonstrate the
proposed approach for the problems where information in both space and time
matters such as face-to-face translation, flower-to-flower, wind and cloud
synthesis, sunrise and sunset. | ['Deva Ramanan', 'Aayush Bansal', 'Yaser Sheikh', 'Shugao Ma'] | 2018-08-15 | recycle-gan-unsupervised-video-retargeting-1 | http://openaccess.thecvf.com/content_ECCV_2018/html/Aayush_Bansal_Recycle-GAN_Unsupervised_Video_ECCV_2018_paper.html | http://openaccess.thecvf.com/content_ECCV_2018/papers/Aayush_Bansal_Recycle-GAN_Unsupervised_Video_ECCV_2018_paper.pdf | eccv-2018-9 | ['face-to-face-translation'] | ['computer-vision'] | [ 5.71660817e-01 -7.65516758e-02 -4.38527949e-02 -2.89563447e-01
-3.47403109e-01 -1.11359799e+00 6.30127430e-01 -3.54561090e-01
-4.74608764e-02 9.71934676e-01 1.33461326e-01 -1.43206969e-01
4.12949212e-02 -6.46043777e-01 -7.79576480e-01 -7.07939684e-01
2.39789039e-01 7.86802620e-02 3.09194952e-01 -2.69517004e-01
2.17928275e-01 1.04098105e+00 -1.37678826e+00 9.24452171e-02
7.61420429e-01 7.57512808e-01 -3.34454812e-02 1.10725582e+00
-9.70373582e-03 6.72828794e-01 -8.23997915e-01 -3.87386233e-01
4.56551373e-01 -8.02042365e-01 -7.82655060e-01 4.34107989e-01
4.26572561e-01 -2.67136753e-01 -4.95169282e-01 9.72333670e-01
4.33549494e-01 2.50312716e-01 5.75179935e-01 -1.45851409e+00
-7.57431507e-01 1.19389594e-01 -6.61846757e-01 1.76312611e-01
4.85428393e-01 6.03934601e-02 2.96108723e-01 -5.92722178e-01
1.05417264e+00 1.06959295e+00 3.45415562e-01 5.90272665e-01
-1.27189052e+00 -5.21628320e-01 9.06517580e-02 -1.82620153e-01
-1.36681700e+00 -7.44877696e-01 7.48140454e-01 -3.92395705e-01
4.81383353e-01 5.40654182e-01 4.57330078e-01 1.13258278e+00
3.16375941e-01 3.41267020e-01 9.48713422e-01 -4.66682345e-01
2.34015569e-01 2.44053334e-01 -8.18562806e-01 2.98284203e-01
-1.90574542e-01 4.69984651e-01 -8.95455360e-01 -6.38167933e-02
7.09077179e-01 -5.35940051e-01 -1.01495966e-01 -4.17121589e-01
-1.14705670e+00 4.95856255e-01 1.26543701e-01 -4.81239147e-02
-6.37224764e-02 6.38399869e-02 4.32582259e-01 4.48593378e-01
8.13108265e-01 4.58104461e-01 -3.89587253e-01 -1.04534134e-01
-1.31224954e+00 3.94796610e-01 4.52505589e-01 1.40942967e+00
6.26029730e-01 3.73402417e-01 -2.38064051e-01 5.45861781e-01
5.34133427e-02 7.70379961e-01 2.02162683e-01 -1.09992230e+00
2.60206908e-01 -1.61038369e-01 2.99263388e-01 -1.14149439e+00
1.09040454e-01 1.46781325e-01 -5.40481985e-01 3.75195324e-01
2.01266199e-01 -3.95616770e-01 -9.74331200e-01 1.75700736e+00
5.55996954e-01 3.31183702e-01 1.85837999e-01 1.00714147e+00
5.47237396e-01 7.78063655e-01 -2.59894699e-01 -5.30819058e-01
8.93796980e-01 -7.08018541e-01 -9.23126340e-01 -8.70871320e-02
5.99003769e-02 -1.21156371e+00 8.54872108e-01 -4.24568728e-02
-1.19874609e+00 -3.97321165e-01 -9.09421086e-01 -2.53382120e-02
-4.58881229e-01 -9.25425887e-02 5.71483895e-02 7.18824446e-01
-1.32092381e+00 5.08908868e-01 -3.57467830e-01 -6.96766973e-01
1.64640144e-01 3.37416261e-01 -4.84616786e-01 2.17456132e-01
-1.12661994e+00 6.64734423e-01 2.26206332e-01 -3.83825630e-01
-9.46663678e-01 -5.90510130e-01 -6.31478310e-01 -2.79683590e-01
2.59387076e-01 -4.72349316e-01 1.13306487e+00 -1.71556878e+00
-1.92673826e+00 8.63806129e-01 -2.41729289e-01 -3.12344909e-01
6.74873590e-01 8.26012045e-02 -7.51057565e-01 3.93031061e-01
1.63193598e-01 7.91309536e-01 1.48624516e+00 -1.31054199e+00
-6.69900358e-01 -1.19020879e-01 2.33607069e-01 4.81472641e-01
-2.69739866e-01 3.72677982e-01 -5.79236031e-01 -1.18323457e+00
-1.85546190e-01 -1.19897985e+00 2.94051856e-01 3.78038436e-01
-5.33008993e-01 5.98150015e-01 1.57971358e+00 -8.84655476e-01
8.96795928e-01 -2.28450990e+00 4.84157860e-01 1.75541714e-01
-1.72899827e-01 2.34108433e-01 -2.35482141e-01 4.72265005e-01
-2.91999340e-01 2.33393639e-01 -3.94477665e-01 -2.76165754e-01
-4.51510251e-01 2.54100591e-01 -4.50453043e-01 5.44371307e-01
2.90671080e-01 5.57752550e-01 -8.10709357e-01 -4.96455759e-01
2.56827213e-02 4.42662269e-01 -6.11661911e-01 3.27096879e-01
-2.89731264e-01 9.54590440e-01 -1.62108559e-02 7.09110081e-01
9.29846346e-01 5.67134738e-01 1.36897609e-01 3.79904583e-02
-3.79192293e-01 -9.71663818e-02 -9.20244753e-01 1.54738033e+00
-3.46278608e-01 9.50835586e-01 3.16927165e-01 -4.84882057e-01
7.36333191e-01 3.29998791e-01 4.60918486e-01 -6.32669866e-01
-1.13973372e-01 7.74833336e-02 -4.68654424e-01 -3.22915703e-01
6.63397849e-01 -2.34561726e-01 -3.89283560e-02 3.93977433e-01
-2.53644317e-01 -6.02957785e-01 3.48065495e-02 2.21880764e-01
7.14790881e-01 5.47611713e-01 -2.22345982e-02 -5.32053351e-01
2.51667768e-01 6.52698725e-02 5.88013232e-01 3.76968920e-01
-1.44387618e-01 9.81266797e-01 5.68063974e-01 -8.29562619e-02
-1.43209374e+00 -1.04034615e+00 7.93503150e-02 1.02651286e+00
1.67504475e-01 -1.98077392e-02 -1.09410858e+00 -6.33596420e-01
-2.29887113e-01 7.74538696e-01 -7.01069772e-01 -4.96573120e-01
-5.02121150e-01 -1.46372899e-01 8.66952777e-01 2.87816793e-01
6.89438105e-01 -6.98768735e-01 -6.31329477e-01 -1.97406203e-01
-1.58523589e-01 -9.50751066e-01 -1.12425792e+00 -8.66123736e-02
-6.75655007e-01 -7.19230354e-01 -8.91676664e-01 -5.90099514e-01
7.27227688e-01 2.30295673e-01 9.10471022e-01 -2.22079843e-01
-2.88364261e-01 4.72574145e-01 -5.97751677e-01 -3.09913099e-01
-5.91556966e-01 -1.88625142e-01 2.76593834e-01 2.60344088e-01
-4.14433688e-01 -5.10333419e-01 -3.05168688e-01 5.16170204e-01
-1.34185529e+00 -9.27118957e-03 -1.77012339e-01 6.08099937e-01
4.30921853e-01 2.53914475e-01 2.18804970e-01 -8.81016254e-01
3.86796206e-01 -2.40068406e-01 -6.78543508e-01 2.23086461e-01
-2.56398410e-01 -1.45939708e-01 6.51048779e-01 -6.18257999e-01
-1.29441988e+00 2.85212457e-01 2.05553502e-01 -7.51700401e-01
-1.13656916e-01 4.31208871e-02 -4.09226745e-01 -3.38143557e-01
7.42979944e-01 2.29255512e-01 -2.21878476e-02 -2.31464043e-01
5.44841409e-01 6.29713178e-01 5.80976248e-01 -5.76414168e-01
1.21738017e+00 7.34046638e-01 1.43757847e-03 -1.06947792e+00
-2.89053142e-01 1.32351499e-02 -9.16486681e-01 -3.33836585e-01
1.00513649e+00 -7.00477004e-01 -2.98310012e-01 7.21035182e-01
-1.25355625e+00 -3.72167736e-01 -6.18660033e-01 5.10538630e-02
-7.07166791e-01 3.05715561e-01 -1.01331353e-01 -6.75867379e-01
5.34801856e-02 -1.03312707e+00 1.09454775e+00 3.42667282e-01
-1.25257030e-01 -1.02606666e+00 -2.54940949e-02 -5.23519590e-02
5.04878044e-01 3.80850792e-01 7.66060531e-01 -1.51339099e-01
-4.66573656e-01 1.18101977e-01 -1.42205358e-01 3.04242462e-01
5.66094577e-01 5.83295047e-01 -8.85267675e-01 -4.80653614e-01
-1.38059273e-01 1.56455822e-02 3.15170437e-01 2.14816123e-01
7.99140573e-01 -6.75092578e-01 -5.63578047e-02 9.68032241e-01
1.23565638e+00 5.78806281e-01 8.90170753e-01 1.95993364e-01
7.06310689e-01 7.00308681e-01 7.74922013e-01 5.99706709e-01
2.93165743e-02 9.63673115e-01 3.15511793e-01 -2.31468394e-01
-3.02382141e-01 -3.38294595e-01 5.06900370e-01 1.33410573e-01
2.12583855e-01 -5.92813432e-01 -5.49153328e-01 7.26417065e-01
-1.53089714e+00 -9.56942022e-01 1.50235921e-01 2.30485344e+00
7.37818122e-01 -4.08077568e-01 1.53450862e-01 -1.50951296e-01
1.05692911e+00 2.71692127e-01 -5.31161010e-01 -7.27362931e-01
-3.42601836e-01 8.37400369e-03 8.21658432e-01 7.05330133e-01
-1.08287930e+00 1.17890596e+00 7.14732456e+00 6.80135489e-01
-1.41016650e+00 1.09456725e-01 6.01527035e-01 -2.76478410e-01
-6.08933389e-01 1.93007708e-01 -3.43146980e-01 6.00765765e-01
7.79092133e-01 -3.28043312e-01 8.86034548e-01 4.40724880e-01
5.56829810e-01 -1.39693066e-01 -8.60138834e-01 7.41099894e-01
2.82187790e-01 -1.26814008e+00 5.36520360e-03 -1.90196991e-01
1.02048457e+00 -6.17206573e-01 3.07642907e-01 -3.36713195e-01
2.29278862e-01 -8.57091486e-01 1.19967687e+00 5.37660778e-01
1.20785224e+00 -8.96173716e-01 1.21831842e-01 1.18415542e-01
-1.13201475e+00 2.81730860e-01 -2.38060489e-01 3.39382082e-01
6.23185076e-02 3.59275907e-01 -6.49737537e-01 6.00684166e-01
6.00540459e-01 4.96394902e-01 -3.24902356e-01 5.92548251e-01
-5.52260689e-02 2.40456507e-01 -3.22350174e-01 4.51397538e-01
2.33485145e-04 -3.64301652e-01 1.10227883e+00 1.07086027e+00
5.61230242e-01 8.89961347e-02 -3.50089669e-02 7.46321380e-01
-8.35018419e-03 -3.02147251e-02 -1.12169075e+00 -2.02861562e-01
4.91768271e-01 7.11854637e-01 -7.01683700e-01 -1.55816212e-01
-9.48803872e-02 1.67861509e+00 -2.38428012e-01 6.40063822e-01
-1.11933064e+00 -5.67828119e-01 9.81985450e-01 4.50346231e-01
2.49454692e-01 -2.18683586e-01 -3.98105145e-01 -9.85631287e-01
3.70707512e-02 -8.13127577e-01 2.04337761e-01 -1.04047978e+00
-7.78931916e-01 6.48436785e-01 1.54013976e-01 -1.21405172e+00
-7.33002424e-02 -2.11661294e-01 -6.50373399e-01 8.93559158e-01
-1.34664178e+00 -1.28189898e+00 -6.95973262e-02 8.30998540e-01
6.51622891e-01 -4.09894168e-01 5.70573270e-01 1.86906323e-01
-4.13507611e-01 6.89889014e-01 3.42061490e-01 -9.80407223e-02
9.65021491e-01 -8.84824514e-01 5.24152219e-01 1.49496543e+00
-9.21511948e-02 3.06912065e-01 9.25456583e-01 -8.63938332e-01
-1.30131829e+00 -1.49344695e+00 8.20135474e-01 -2.86818296e-01
3.15923840e-01 -2.87121445e-01 -4.52663213e-01 6.52845740e-01
5.27485073e-01 -1.65377006e-01 2.44641155e-01 -5.54662466e-01
-3.37112874e-01 -1.69130430e-01 -1.66442776e+00 8.63797128e-01
8.89523804e-01 -6.39141500e-01 -6.34248927e-02 5.12842298e-01
8.76137853e-01 -7.37043262e-01 -5.88549972e-01 1.36400029e-01
3.29533279e-01 -7.72250950e-01 8.09399664e-01 -7.30975807e-01
3.43348712e-01 -6.21966541e-01 -4.13030922e-01 -1.51374602e+00
-2.05651730e-01 -1.22245586e+00 3.71134251e-01 1.56356955e+00
3.04057062e-01 -3.80633324e-01 5.93466043e-01 5.69456935e-01
9.49402452e-02 1.53512567e-01 -1.11758935e+00 -9.03432965e-01
1.83231398e-01 -8.05597976e-02 7.10085154e-01 1.05958557e+00
-3.04581851e-01 -1.00966454e-01 -8.71865273e-01 4.33976501e-01
3.28737795e-01 -8.49544480e-02 6.10000014e-01 -5.67317188e-01
-4.40901108e-02 2.58953199e-02 -3.03650647e-01 -6.81310892e-01
1.90661967e-01 -5.31072021e-01 1.49970472e-01 -1.12684190e+00
-1.92316622e-01 -2.43275672e-01 4.42116380e-01 2.16880083e-01
2.00160474e-01 4.27339613e-01 4.66226131e-01 1.50763974e-01
3.02505190e-03 4.34302986e-01 1.29670620e+00 -1.01369537e-01
-2.60029674e-01 -6.78612292e-02 -5.00461459e-01 3.83454442e-01
8.47099483e-01 -5.15789449e-01 -6.16115689e-01 -6.73122108e-01
-9.59534850e-03 2.54770130e-01 3.19762498e-01 -6.89279675e-01
2.26520430e-02 -7.11437881e-01 2.52573907e-01 -1.04549155e-01
4.13600236e-01 -1.08961666e+00 4.64046687e-01 3.59992087e-02
-3.27340573e-01 3.22590351e-01 5.14490962e-01 6.54324651e-01
-1.78569958e-01 -3.74130346e-02 1.36329627e+00 3.36422995e-02
-5.25923431e-01 2.91810334e-01 -6.52150691e-01 3.74822021e-02
1.45756567e+00 -2.37887278e-01 -2.73327112e-01 -7.44400263e-01
-8.36747110e-01 -2.19822958e-01 1.01653421e+00 5.51529288e-01
5.82548201e-01 -1.43911767e+00 -7.75352359e-01 4.36566621e-01
-6.82721958e-02 -3.01219732e-01 2.07135156e-01 6.40740454e-01
-1.00172675e+00 6.11478724e-02 -3.44940484e-01 -4.31957453e-01
-1.48032689e+00 6.94626510e-01 3.58461618e-01 3.69194657e-01
-3.81658912e-01 9.87401545e-01 3.01273495e-01 -3.25828999e-01
-1.81476977e-02 5.36834523e-02 3.00346196e-01 -2.29295716e-03
3.86356562e-01 3.34521115e-01 -3.15634087e-02 -1.08550096e+00
-4.99033749e-01 6.53687894e-01 1.80767909e-01 -6.61430240e-01
8.09690595e-01 -4.65804279e-01 -2.24629909e-01 1.00767650e-01
1.15167952e+00 3.63780975e-01 -1.39517283e+00 -2.27031969e-02
-4.21273887e-01 -9.44749296e-01 -2.31658649e-02 -8.75346839e-01
-1.26789725e+00 6.49603605e-01 7.53316045e-01 6.53537586e-02
1.58281481e+00 -2.80809402e-01 6.49856091e-01 -8.50644857e-02
2.18990490e-01 -1.27066469e+00 1.00300573e-02 4.35337514e-01
9.53946054e-01 -7.17698514e-01 -9.96763632e-02 -4.03933495e-01
-1.15888524e+00 9.47465479e-01 2.80839711e-01 7.97764957e-02
6.02762878e-01 5.31331003e-01 2.10724801e-01 2.77021021e-01
-5.20765483e-01 1.31698586e-02 2.01496869e-01 9.89727199e-01
1.18272416e-01 -2.78568566e-02 1.04793668e-01 -2.33175203e-01
-6.53421059e-02 -2.33743414e-02 7.23502517e-01 1.12389612e+00
-1.46825522e-01 -1.17142177e+00 -7.02856719e-01 5.92058077e-02
-3.98356944e-01 -2.14762270e-01 -8.58130455e-01 4.57293689e-01
2.53254414e-01 1.09598768e+00 1.45589963e-01 -3.37310046e-01
3.36348861e-01 8.23495612e-02 3.74760449e-01 -5.25846720e-01
-5.20217121e-01 3.76191467e-01 7.37899542e-02 -4.15201962e-01
-5.03413200e-01 -7.87849307e-01 -9.13476825e-01 -6.97911978e-01
3.60633545e-02 -1.55783460e-01 6.09283626e-01 7.90958405e-01
5.65521181e-01 4.62232351e-01 1.03335452e+00 -7.57532656e-01
1.24833830e-01 -3.97064149e-01 -6.80375993e-01 2.62970001e-01
5.84601998e-01 -3.69920254e-01 -2.91642278e-01 7.92198956e-01] | [11.662042617797852, -0.47789284586906433] |
9006d575-16cb-4bf0-97d9-a6d34aa0570b | skill-structured-knowledge-infusion-for-large | 2205.08184 | null | https://arxiv.org/abs/2205.08184v1 | https://arxiv.org/pdf/2205.08184v1.pdf | SKILL: Structured Knowledge Infusion for Large Language Models | Large language models (LLMs) have demonstrated human-level performance on a vast spectrum of natural language tasks. However, it is largely unexplored whether they can better internalize knowledge from a structured data, such as a knowledge graph, or from text. In this work, we propose a method to infuse structured knowledge into LLMs, by directly training T5 models on factual triples of knowledge graphs (KGs). We show that models pre-trained on Wikidata KG with our method outperform the T5 baselines on FreebaseQA and WikiHop, as well as the Wikidata-answerable subset of TriviaQA and NaturalQuestions. The models pre-trained on factual triples compare competitively with the ones on natural language sentences that contain the same knowledge. Trained on a smaller size KG, WikiMovies, we saw 3x improvement of exact match score on MetaQA task compared to T5 baseline. The proposed method has an advantage that no alignment between the knowledge graph and text corpus is required in curating training data. This makes our method particularly useful when working with industry-scale knowledge graphs. | ['Martin Jaggi', 'Enrique Alfonseca', 'Zhe Dong', 'Fedor Moiseev'] | 2022-05-17 | null | https://aclanthology.org/2022.naacl-main.113 | https://aclanthology.org/2022.naacl-main.113.pdf | naacl-2022-7 | ['triviaqa'] | ['miscellaneous'] | [-2.69941360e-01 9.65668321e-01 -2.31821835e-01 -3.18451315e-01
-1.14060640e+00 -8.49268377e-01 6.27085567e-01 2.93464124e-01
-4.92721796e-01 8.78289878e-01 2.64488429e-01 -4.76591438e-01
-1.69586226e-01 -1.26841271e+00 -1.36579335e+00 8.42572190e-03
6.28770366e-02 9.92186189e-01 6.01846814e-01 -7.00733423e-01
-4.24408205e-02 -3.47868174e-01 -1.00581658e+00 8.59834969e-01
1.32974875e+00 6.93949044e-01 6.26612306e-02 3.96823853e-01
-6.75202727e-01 1.31863129e+00 -5.38877487e-01 -1.28116512e+00
-7.68419281e-02 -1.32567912e-01 -1.47905755e+00 -5.55226624e-01
1.29614627e+00 1.44861624e-01 -2.51174569e-01 9.57755685e-01
1.49884284e-01 -2.93089062e-01 5.33599555e-01 -1.10829258e+00
-1.00621712e+00 1.42337656e+00 -1.45899430e-01 1.40387326e-01
3.52383524e-01 -8.53562430e-02 1.57415104e+00 -7.86090612e-01
1.04017615e+00 1.42969954e+00 7.61084735e-01 4.92197156e-01
-8.06109369e-01 -3.72436464e-01 -1.98481176e-02 7.22843170e-01
-1.24378669e+00 -3.07443291e-01 3.07086051e-01 -2.01534450e-01
1.59595382e+00 7.60094002e-02 3.11084926e-01 7.65447378e-01
1.98281094e-01 9.00467217e-01 7.96956301e-01 -6.62865281e-01
-2.28365824e-01 3.03881347e-01 4.41659212e-01 1.19407499e+00
5.08793414e-01 -4.96377677e-01 -7.61935174e-01 -1.78209484e-01
2.85047144e-01 -8.42567861e-01 -3.86244774e-01 -4.95687425e-01
-1.17216873e+00 7.73443520e-01 5.15849948e-01 2.81686485e-01
-1.87076911e-01 4.48484749e-01 5.55692077e-01 6.94123507e-01
3.36184829e-01 8.10875475e-01 -9.72502947e-01 -4.88974229e-02
-5.34848630e-01 1.80158854e-01 1.29834819e+00 1.20171106e+00
9.77513373e-01 -2.68821239e-01 -6.22357195e-03 6.11261845e-01
2.62939423e-01 1.04226029e+00 3.26559633e-01 -8.64095211e-01
1.17346823e+00 1.01946604e+00 9.46116820e-02 -9.02584732e-01
-3.32789689e-01 -4.16647315e-01 -2.45864943e-01 -5.36412239e-01
6.91004455e-01 -5.17156199e-02 -9.63866949e-01 1.63669753e+00
2.13802114e-01 -1.19788200e-01 6.79252326e-01 6.05458200e-01
1.39222276e+00 6.32045746e-01 1.82908326e-01 7.76162893e-02
1.41371870e+00 -1.03349340e+00 -6.02615774e-01 -4.72624689e-01
1.25302267e+00 -4.68024969e-01 1.20590985e+00 2.70612776e-01
-1.01048684e+00 -2.22146884e-01 -8.91140103e-01 -4.40195173e-01
-7.99923241e-01 -3.87220010e-02 6.98686600e-01 6.73596561e-01
-1.26823068e+00 2.88263410e-01 -5.32365322e-01 -4.91421223e-01
2.42749780e-01 2.03335583e-01 -5.06738603e-01 -2.84880877e-01
-1.77127469e+00 1.30611825e+00 8.33759606e-01 -9.33361650e-02
-9.44123626e-01 -8.88882458e-01 -1.05040550e+00 5.09764291e-02
9.24626291e-01 -1.08870697e+00 1.37829351e+00 -8.36926579e-01
-1.25139022e+00 1.04090583e+00 -6.48592412e-02 -8.30254853e-01
2.28127211e-01 -4.82275456e-01 -4.46712852e-01 -2.17598882e-02
2.02298671e-01 3.37124765e-01 5.55528164e-01 -1.07662725e+00
-4.72848862e-01 -3.07050943e-01 6.40725195e-01 1.60588697e-01
-3.83314788e-01 -2.02525109e-01 -5.22924840e-01 -2.91242879e-02
-3.49612385e-01 -6.79298103e-01 1.40839264e-01 -6.79870129e-01
-4.72673357e-01 -5.50834239e-01 4.42185879e-01 -7.53368318e-01
1.21477020e+00 -1.60568821e+00 2.60417432e-01 2.86414385e-01
3.20176750e-01 6.25101089e-01 -3.40662092e-01 8.41387510e-01
2.84764558e-01 2.83912808e-01 8.11277106e-02 2.39363864e-01
2.42924631e-01 5.43037355e-01 -6.43473506e-01 -2.65204370e-01
1.59320772e-01 1.59014201e+00 -1.33540070e+00 -6.86736226e-01
-3.50509912e-01 -8.59648734e-02 -3.16408008e-01 -3.26303281e-02
-9.80755389e-01 -3.09927046e-01 -5.37276089e-01 4.41945076e-01
2.64798492e-01 -7.16966331e-01 4.97904330e-01 -3.34320992e-01
3.94787431e-01 6.54339433e-01 -7.99716353e-01 1.91864955e+00
-6.11610830e-01 3.46510589e-01 -2.52170593e-01 -7.55695760e-01
6.84829950e-01 3.96317840e-01 -1.63183138e-01 -7.93853998e-01
-3.09502780e-01 4.28335071e-01 -2.05596760e-02 -6.64569557e-01
5.57888925e-01 -2.28258327e-01 -1.93169028e-01 4.33372080e-01
6.99826658e-01 -4.22276020e-01 6.03205085e-01 9.12533581e-01
1.36316788e+00 3.43577303e-02 3.07655245e-01 -3.11931223e-01
6.26197994e-01 5.47897220e-01 1.28387690e-01 8.61867368e-01
3.84319454e-01 -7.73288086e-02 5.58408201e-01 -4.13799822e-01
-6.23472512e-01 -1.13824272e+00 2.30532005e-01 1.16238356e+00
1.63609385e-02 -1.03665626e+00 -6.04926109e-01 -1.29562283e+00
1.92719758e-01 1.01242745e+00 -6.08117819e-01 -1.06204100e-01
-7.42286921e-01 -3.71609539e-01 9.45396841e-01 4.62796807e-01
5.38587987e-01 -1.07061708e+00 -7.95420855e-02 2.61539519e-01
-4.31463748e-01 -1.34022331e+00 -1.93501279e-01 -1.74892932e-01
-7.51971185e-01 -1.45052528e+00 -2.07086533e-01 -7.98469663e-01
3.84005100e-01 1.78269800e-02 1.91939783e+00 8.82661939e-02
1.09034881e-01 8.21283758e-01 -5.83740771e-01 -6.58161819e-01
-5.66001832e-01 4.01415050e-01 -2.86860287e-01 -4.55149680e-01
7.12799072e-01 -3.05524468e-01 -1.91061944e-02 9.64110345e-02
-8.03415477e-01 1.97124444e-02 7.47668028e-01 5.87746739e-01
2.75327235e-01 -4.96991873e-02 9.45772409e-01 -1.50370562e+00
6.92440987e-01 -5.47400534e-01 -5.62961459e-01 1.01136863e+00
-5.97105443e-01 5.10299265e-01 6.94105744e-01 7.24083260e-02
-1.17202246e+00 -3.39745909e-01 1.52106062e-01 1.49098589e-04
4.40216810e-01 1.15287185e+00 -9.96709764e-02 -9.97608900e-02
1.10993409e+00 1.40431337e-02 -2.59729803e-01 -2.95963585e-01
1.09344208e+00 3.67485404e-01 4.70098138e-01 -1.06966043e+00
9.09813106e-01 1.99191734e-01 -2.45903075e-01 -6.41618609e-01
-1.41672111e+00 -6.68874621e-01 -6.45251334e-01 1.05285861e-01
5.65017819e-01 -9.86454725e-01 -4.14998382e-01 8.25300887e-02
-1.36037040e+00 -6.13308072e-01 -2.55079895e-01 1.36985853e-01
-4.55457568e-01 4.44235414e-01 -4.60393578e-01 -2.52089977e-01
-7.43974388e-01 -3.20970386e-01 9.75393593e-01 -6.88952655e-02
-9.31951851e-02 -1.48756194e+00 4.39071596e-01 9.56948698e-01
3.02886426e-01 -6.43086731e-02 1.43316281e+00 -7.68353045e-01
-7.93170571e-01 -5.08022644e-02 -3.13432097e-01 5.02000690e-01
7.77212977e-02 -2.29477257e-01 -7.35150516e-01 -8.10121000e-02
-6.07053161e-01 -1.07412410e+00 8.64221275e-01 -2.08626062e-01
5.04049182e-01 -5.47092676e-01 -2.47356921e-01 -3.77030745e-02
1.40071940e+00 -3.81425261e-01 7.62948036e-01 4.30174351e-01
8.53041708e-01 4.87191379e-01 6.26317739e-01 -3.11117619e-01
9.77552474e-01 4.41892773e-01 3.71652603e-01 2.71272391e-01
-4.53080535e-01 -5.93006968e-01 5.41568935e-01 1.31349468e+00
-1.63347065e-01 -4.20595497e-01 -1.30395329e+00 9.74985957e-01
-2.06103802e+00 -8.23081672e-01 -3.53752851e-01 1.86329436e+00
1.37716663e+00 9.76880640e-02 -2.31578857e-01 -4.82148975e-01
2.76580900e-01 -8.91365036e-02 -3.61855537e-01 -2.88414478e-01
-4.62089509e-01 3.97793084e-01 3.33044767e-01 8.28969777e-01
-8.07667494e-01 1.39254594e+00 5.60552073e+00 9.88123178e-01
-6.24078333e-01 2.57027656e-01 -1.98931202e-01 1.79672807e-01
-5.81727445e-01 2.19755620e-01 -9.54023182e-01 -5.62886074e-02
1.06781781e+00 -5.25305152e-01 1.90597579e-01 7.29421437e-01
-3.99228424e-01 1.04154386e-01 -1.15376008e+00 6.23607099e-01
3.53429794e-01 -1.59138548e+00 6.55782580e-01 -2.60688007e-01
8.69210541e-01 3.97633821e-01 -2.75117993e-01 9.31989133e-01
7.72594273e-01 -1.15642369e+00 5.01300216e-01 5.22990823e-01
4.65307772e-01 -3.63680094e-01 1.09643662e+00 4.96843010e-01
-9.87664104e-01 2.32034713e-01 -4.93892401e-01 8.19949359e-02
1.10262744e-01 5.22713184e-01 -1.45109749e+00 1.19123149e+00
4.89701062e-01 5.29400229e-01 -9.24420297e-01 6.84038341e-01
-7.16928065e-01 8.50678205e-01 -2.09745258e-01 -2.64318287e-01
3.28005970e-01 9.72699225e-02 2.60919631e-01 1.24278462e+00
-3.18185426e-02 3.82859707e-02 3.45616788e-02 6.87116742e-01
-5.59513807e-01 4.14440274e-01 -6.95052922e-01 -3.99757028e-01
1.89776316e-01 1.21079564e+00 3.51461768e-02 -6.94375515e-01
-7.49242067e-01 5.54816782e-01 1.01055479e+00 3.44406128e-01
-4.68628943e-01 -5.28494418e-01 -1.16539232e-01 2.53535122e-01
2.08996713e-01 -1.19056940e-01 2.29162663e-01 -1.28648317e+00
3.78773510e-01 -9.82005835e-01 9.96244311e-01 -1.00929570e+00
-1.50763857e+00 4.86990124e-01 1.47926852e-01 -5.33307254e-01
-3.94766062e-01 -8.70914400e-01 -2.83429772e-01 5.86164653e-01
-1.77719915e+00 -1.54950953e+00 -2.33969435e-01 7.51566172e-01
4.13037054e-02 -1.08097620e-01 8.57811272e-01 2.81505793e-01
1.03066072e-01 5.52209795e-01 -1.69708744e-01 4.98289764e-01
9.30302441e-01 -1.63679266e+00 4.44785416e-01 5.99387646e-01
7.12966144e-01 9.32942212e-01 5.67288637e-01 -9.72847342e-01
-1.71184909e+00 -1.21467686e+00 1.27462959e+00 -1.22152436e+00
1.23869073e+00 -1.56054765e-01 -1.19781446e+00 1.17747486e+00
5.80139577e-01 -2.43963525e-01 6.33275270e-01 7.17144072e-01
-8.97331715e-01 -1.50425896e-01 -6.96191072e-01 3.07324529e-01
1.19941127e+00 -7.07075894e-01 -1.34225583e+00 7.44913340e-01
9.33344901e-01 -4.54270393e-01 -1.18777680e+00 4.77296919e-01
1.52037352e-01 -3.39368165e-01 7.44547844e-01 -1.11794567e+00
3.90064955e-01 -3.23658824e-01 -3.52804251e-02 -1.64094925e+00
4.08914648e-02 -6.05155170e-01 -5.01860082e-01 1.13754427e+00
1.08504140e+00 -6.65252686e-01 7.45194256e-01 4.49885458e-01
-3.61410141e-01 -3.59426677e-01 -1.00354314e+00 -1.12241423e+00
3.65077049e-01 -2.74822265e-01 2.47918680e-01 1.06220520e+00
6.34406805e-01 9.25583839e-01 -4.50716987e-02 1.98635578e-01
5.74356556e-01 3.45157921e-01 1.09015894e+00 -1.04333138e+00
-2.48782501e-01 4.05462123e-02 -3.65949631e-01 -8.40983093e-01
4.76823628e-01 -1.44883263e+00 -2.23483026e-01 -1.96388841e+00
3.11482787e-01 -4.86465096e-01 -2.17090244e-03 9.96182561e-01
-3.09224874e-01 -1.71006337e-01 -1.14232793e-01 -1.21228606e-01
-1.24056077e+00 4.64174747e-01 1.33712006e+00 -4.03503627e-01
1.59927011e-01 -5.55849552e-01 -6.88242555e-01 5.62034011e-01
6.90826654e-01 -4.22093928e-01 -8.20217788e-01 -9.37242806e-01
1.05075181e+00 -1.23456046e-01 2.93046683e-01 -6.17534041e-01
5.61661899e-01 -9.19699445e-02 -4.21643615e-01 -1.40936524e-01
1.83206111e-01 -4.10475105e-01 -6.76821470e-02 2.87606567e-01
-5.59375510e-02 -7.89487958e-02 4.51953590e-01 5.53300858e-01
-5.05999506e-01 -3.24837536e-01 8.44386071e-02 -5.04262388e-01
-1.06483781e+00 2.39441656e-02 8.69364068e-02 9.19899106e-01
5.44749379e-01 3.04630578e-01 -1.20864844e+00 -4.84865248e-01
-5.29997051e-01 5.74761629e-01 2.14308485e-01 4.83649939e-01
4.97809142e-01 -1.09534800e+00 -9.19767737e-01 -4.45042729e-01
5.72950542e-01 2.77213268e-02 9.30035487e-02 7.17529058e-01
-5.68454623e-01 1.02277553e+00 7.54183978e-02 -2.92215079e-01
-9.59770381e-01 4.63848174e-01 3.77732843e-01 -6.81529284e-01
-3.55393678e-01 8.96820068e-01 7.70379901e-02 -9.62185025e-01
-1.15108296e-01 -5.29981256e-01 -1.73097134e-01 -7.22989962e-02
4.54726130e-01 1.18541501e-01 4.58983988e-01 -3.57693911e-01
-3.87933820e-01 5.38890123e-01 -3.22190225e-01 1.50797799e-01
1.09464169e+00 1.81676596e-01 -5.01651824e-01 3.60619277e-01
8.14585328e-01 3.21079940e-01 -3.23272347e-01 -6.61116481e-01
4.72342551e-01 -8.36783424e-02 -2.29478851e-01 -1.28859806e+00
-8.12333465e-01 6.61396742e-01 -2.12284282e-01 1.82316333e-01
5.07677674e-01 5.06194174e-01 9.14286494e-01 1.19263768e+00
8.50692272e-01 -9.30347383e-01 2.59976685e-01 9.03159857e-01
1.09109855e+00 -1.21153104e+00 -8.57345983e-02 -7.52948284e-01
-6.54364049e-01 1.03212202e+00 8.02388668e-01 1.82600588e-01
2.56949186e-01 -2.28844181e-01 3.06776524e-01 -8.64327729e-01
-1.11522365e+00 -4.68813777e-01 6.15373671e-01 5.74577093e-01
3.26884538e-01 2.81704962e-02 -6.51999041e-02 7.08875537e-01
-4.47070867e-01 8.66962895e-02 4.92544025e-01 7.69210637e-01
-5.45779645e-01 -1.10628474e+00 9.67717618e-02 5.82000792e-01
-3.50244552e-01 -6.09387875e-01 -7.65206575e-01 1.11861730e+00
-1.38184384e-01 9.91556346e-01 -5.04765987e-01 -1.58898935e-01
5.59242487e-01 3.94581109e-01 8.57346296e-01 -1.10840845e+00
-6.46133602e-01 -9.08707142e-01 1.05614924e+00 -5.05545616e-01
-4.40957904e-01 -1.90815955e-01 -1.33614266e+00 -1.79685116e-01
-5.18646419e-01 7.03272820e-01 2.76334763e-01 1.08467162e+00
4.11228269e-01 9.77540761e-02 -3.28095764e-01 2.79900342e-01
-4.10209179e-01 -9.69040394e-01 -3.19607347e-01 4.70796704e-01
-1.26409411e-01 -4.75797504e-01 -1.66287661e-01 1.29519194e-01] | [10.445562362670898, 7.939227104187012] |
64bf6bbb-1ba0-44d8-8691-52fd443f2bb1 | understanding-worldwide-private-information | 2102.12869 | null | https://arxiv.org/abs/2102.12869v1 | https://arxiv.org/pdf/2102.12869v1.pdf | Understanding Worldwide Private Information Collection on Android | Mobile phones enable the collection of a wealth of private information, from unique identifiers (e.g., email addresses), to a user's location, to their text messages. This information can be harvested by apps and sent to third parties, which can use it for a variety of purposes. In this paper we perform the largest study of private information collection (PIC) on Android to date. Leveraging an anonymized dataset collected from the customers of a popular mobile security product, we analyze the flows of sensitive information generated by 2.1M unique apps installed by 17.3M users over a period of 21 months between 2018 and 2019. We find that 87.2% of all devices send private information to at least five different domains, and that actors active in different regions (e.g., Asia compared to Europe) are interested in collecting different types of information. The United States (62% of the total) and China (7% of total flows) are the countries that collect most private information. Our findings raise issues regarding data regulation, and would encourage policymakers to further regulate how private information is used by and shared among the companies and how accountability can be truly guaranteed. | ['Gianluca Stringhini', 'Pierre-Antoine Vervier', 'Yun Shen'] | 2021-02-25 | null | null | null | null | ['mobile-security'] | ['miscellaneous'] | [ 5.23714609e-02 2.67410666e-01 -1.04014337e+00 -9.75032821e-02
-7.78454304e-01 -1.32722533e+00 6.07953191e-01 5.33475839e-02
-1.82649493e-01 7.45242953e-01 6.36630535e-01 -6.95564806e-01
2.09832683e-01 -7.79706180e-01 -4.38390911e-01 -1.26096264e-01
3.98040026e-01 6.28654659e-02 2.09894441e-02 1.59834817e-01
2.92033374e-01 3.30324829e-01 -8.34720492e-01 3.44085932e-01
7.27191806e-01 1.10866416e+00 -2.65897393e-01 1.24700658e-01
-2.18808025e-01 4.88590688e-01 -7.34373748e-01 -9.31304038e-01
2.86748230e-01 8.02741349e-02 -7.04172850e-01 -4.70255315e-01
1.56776428e-01 -6.83264256e-01 1.35957375e-02 9.69876707e-01
8.51737410e-02 -5.29340208e-01 2.40325361e-01 -1.23430014e+00
-7.47958958e-01 7.20399380e-01 -7.92702138e-01 5.01056276e-02
4.42067057e-01 6.33212253e-02 5.76668084e-01 -3.08742076e-01
9.59706128e-01 6.02757812e-01 9.60156739e-01 3.49747509e-01
-9.53119874e-01 -1.22116745e+00 -3.43145043e-01 -7.98850715e-01
-1.04645729e+00 -7.66164899e-01 5.02218664e-01 -5.82215011e-01
3.74496341e-01 6.24969602e-01 2.86476403e-01 1.44350684e+00
2.63672888e-01 5.99407032e-02 1.14497781e+00 1.50012895e-01
1.57339483e-01 8.25283885e-01 2.79918462e-01 -6.80146217e-02
1.19051600e+00 -5.79599500e-01 -5.61255693e-01 -9.13328826e-01
3.43020409e-01 3.77642781e-01 -8.55750889e-02 -2.44366694e-02
-8.23708892e-01 7.34266341e-01 -1.49186403e-01 4.17145103e-01
-5.23584366e-01 -7.58649558e-02 4.03926104e-01 2.21004523e-02
5.75635374e-01 3.18973869e-01 -8.73387575e-01 -8.63540530e-01
-5.62730730e-01 -2.48798624e-01 1.05335355e+00 9.96424198e-01
8.44401538e-01 -5.89697063e-01 4.87283587e-01 5.49875975e-01
4.41501081e-01 7.29101241e-01 5.59368014e-01 -1.17790627e+00
1.02790189e+00 7.88556516e-01 5.10684311e-01 -1.38480687e+00
-4.54790965e-02 -1.23173162e-01 -6.29919350e-01 -4.92327720e-01
6.53883934e-01 -7.54806161e-01 6.84409589e-02 1.44688058e+00
1.44626856e-01 -1.99686378e-01 -3.04642409e-01 3.30232441e-01
3.38867307e-01 3.59340906e-01 -7.21654389e-03 -2.21241280e-01
1.44434249e+00 -3.39618586e-02 -5.89146256e-01 -7.54506141e-02
6.65603101e-01 -6.92438126e-01 7.98915505e-01 2.62195647e-01
-9.10785913e-01 -5.73256165e-02 -4.52562183e-01 1.54068306e-01
-8.70056868e-01 -2.64811635e-01 6.82852030e-01 1.45674515e+00
-7.29149580e-01 3.18183631e-01 -7.20902324e-01 -5.47529042e-01
1.16489255e+00 4.84750688e-01 -5.64967096e-01 2.39324749e-01
-9.46378350e-01 1.51342139e-01 -1.90775245e-01 -3.70779037e-01
-2.84613341e-01 -9.76105034e-01 -4.60885376e-01 -2.65226990e-01
2.06280500e-01 -3.41544509e-01 1.17489111e+00 -8.65562558e-01
-1.03528929e+00 8.83001149e-01 -1.63666487e-01 -3.91238630e-01
2.05120325e-01 -2.17518225e-01 -7.16144919e-01 7.36555383e-02
5.68355203e-01 -1.48124043e-02 6.12176061e-01 -9.89961565e-01
-8.22504640e-01 -7.32021153e-01 6.31765500e-02 -5.75964749e-01
-9.67422962e-01 4.30084050e-01 -2.07147107e-01 -3.64982933e-01
-1.93596080e-01 -1.14995682e+00 3.30779314e-01 -6.98570013e-01
-5.10160744e-01 2.97377914e-01 1.41945171e+00 -1.26705897e+00
1.53753507e+00 -2.16391373e+00 -7.16199815e-01 2.63675004e-01
4.27977681e-01 2.59251297e-01 5.71566999e-01 6.44558787e-01
2.20489234e-01 1.25260055e+00 -1.87428415e-01 -5.33249199e-01
-4.77004424e-02 -4.08966355e-02 -5.54576755e-01 5.61771154e-01
-4.38318849e-01 1.17845321e+00 -8.70258093e-01 -5.96176051e-02
-4.89531346e-02 3.63192201e-01 -1.96360365e-01 -3.66107136e-01
2.06103623e-01 7.24388063e-01 -6.95930302e-01 8.21168959e-01
1.07505441e+00 -4.04847264e-01 5.16576767e-01 3.82478893e-01
-1.76473603e-01 7.57235050e-01 -5.74950516e-01 1.14582205e+00
-6.70021355e-01 7.42177606e-01 5.38572729e-01 2.20942441e-02
4.72658187e-01 8.89209136e-02 7.36509979e-01 -3.94513994e-01
1.73875064e-01 2.06572801e-01 -3.75673771e-01 -1.90405726e-01
7.16647506e-01 5.34075499e-01 -5.97447395e-01 1.05403972e+00
-7.22092032e-01 4.73711193e-01 -4.22994196e-01 2.56349415e-01
1.30741978e+00 -4.39953834e-01 4.34569240e-01 -1.15173161e-01
2.44305328e-01 -1.69561207e-01 6.54735565e-01 5.16840935e-01
-4.61915791e-01 1.04284488e-01 9.14272130e-01 -3.88515368e-02
-6.30128682e-01 -8.01198602e-01 -1.42526150e-01 7.32604504e-01
1.28960414e-02 -6.44604802e-01 -8.36699784e-01 -1.03816342e+00
4.70543712e-01 3.98888409e-01 -3.25022906e-01 3.38412002e-02
-5.53767979e-01 -3.33191574e-01 5.80564678e-01 1.56842902e-01
8.99696767e-01 -6.88282549e-01 -5.30444682e-01 -1.22729316e-01
-4.32423413e-01 -1.33766127e+00 -7.33379841e-01 -6.68329835e-01
-7.93364584e-01 -1.07890046e+00 -5.12567163e-01 -7.99445659e-02
5.33685267e-01 4.38240379e-01 8.27352524e-01 -3.51315767e-01
3.33669096e-01 6.49548352e-01 -1.61924288e-01 -6.08143628e-01
-4.58228320e-01 7.15772986e-01 5.71937785e-02 5.37433326e-01
6.92035437e-01 -5.98490357e-01 -4.55489874e-01 6.66300952e-01
-5.75542808e-01 -6.90605521e-01 1.17738433e-01 -2.27712676e-01
1.53749719e-01 1.91564187e-01 8.37254286e-01 -1.59531140e+00
7.15159178e-01 -1.35442615e+00 -3.91880691e-01 -2.85709411e-01
-5.72941005e-01 -1.00553155e+00 4.20137823e-01 -2.83884436e-01
-1.18864334e+00 -1.62962362e-01 2.44990453e-01 3.59756887e-01
-3.35427821e-01 1.46204725e-01 -6.38158679e-01 -1.88808903e-01
3.38572055e-01 -4.27754730e-01 2.46947389e-02 -5.14329970e-01
2.17595130e-01 1.40946734e+00 1.58681780e-01 4.11654683e-03
8.34593236e-01 8.99373233e-01 -5.87903082e-01 -1.04375374e+00
-2.61842310e-01 -6.12957835e-01 -2.22508967e-01 -7.63076469e-02
7.23721802e-01 -8.06977451e-01 -1.16021502e+00 7.30074883e-01
-9.07948911e-01 -7.21193254e-02 -2.06060126e-01 2.17183113e-01
1.24771424e-01 3.99568409e-01 -6.91553175e-01 -8.91516149e-01
-1.40867516e-01 -5.45934677e-01 1.10008514e+00 2.09507376e-01
-9.57319677e-01 -9.13547695e-01 -1.89443275e-01 9.04226422e-01
8.00589502e-01 3.44230175e-01 4.01552737e-01 -6.32590353e-01
-7.11701810e-01 -5.19846201e-01 -2.52211899e-01 9.00660828e-02
7.85406291e-01 -1.20615222e-01 -1.07644200e+00 -1.14735074e-01
4.30466026e-01 3.66149336e-01 2.18901068e-01 4.34848875e-01
1.24112833e+00 -9.58833098e-01 -9.46549773e-01 3.89439285e-01
1.23797357e+00 3.93485993e-01 9.32365358e-01 2.00691938e-01
8.26085389e-01 5.35417795e-01 4.30913478e-01 6.79455280e-01
5.56531072e-01 5.03165960e-01 5.13613522e-01 3.86515677e-01
3.56558889e-01 -7.10764825e-01 5.06578982e-01 6.17972672e-01
1.99934691e-01 -1.49411082e-01 -7.86943972e-01 5.78083456e-01
-1.42555046e+00 -7.36824453e-01 -3.78837675e-01 2.50325561e+00
6.13762259e-01 2.73820937e-01 6.02662265e-01 -1.39837325e-01
8.81018221e-01 1.78332984e-01 -4.01747018e-01 -3.99758488e-01
1.70622870e-01 1.17381951e-02 1.42177045e+00 3.20952505e-01
-7.85792708e-01 4.76286829e-01 5.98200512e+00 4.82797563e-01
-9.49382544e-01 5.71796536e-01 9.92631555e-01 -1.52858779e-01
-6.34995222e-01 1.78706720e-01 -9.48426127e-01 1.35773456e+00
1.56555259e+00 -9.85064581e-02 3.19427669e-01 7.47199237e-01
3.68037790e-01 -3.56739372e-01 -3.41613919e-01 9.47613895e-01
-3.47412556e-01 -1.38697779e+00 -5.15443504e-01 1.08847690e+00
1.05292821e+00 -1.35542408e-01 4.60479170e-01 -2.60137916e-01
6.99422136e-02 -6.43199921e-01 3.02791685e-01 7.90330529e-01
8.90797317e-01 -8.67239773e-01 6.09273374e-01 4.18238938e-01
-8.21824193e-01 -4.30661887e-01 1.98079288e-01 -2.67182797e-01
4.36456323e-01 1.25356328e+00 -5.81301212e-01 3.52176487e-01
8.92897010e-01 8.53114963e-01 -5.89804709e-01 2.71694660e-01
3.54658179e-02 1.00816476e+00 -2.24717990e-01 3.31251621e-02
-3.08007210e-01 -5.28015256e-01 4.63467836e-01 6.77825511e-01
3.37909937e-01 -3.34227413e-01 -8.22803259e-01 6.70007169e-01
-6.72101438e-01 -3.02570522e-01 -1.23987961e+00 -6.30753040e-01
9.37914670e-01 1.36638749e+00 -6.51909351e-01 -4.81922552e-02
-5.20810425e-01 8.88252437e-01 -1.98740661e-01 2.88417011e-01
-5.99700153e-01 -5.55972636e-01 1.14353836e+00 8.93339694e-01
1.03684198e-02 -3.78560007e-01 -7.36921787e-01 -7.87024856e-01
3.71929705e-01 -8.64418626e-01 7.58694019e-03 -4.50310796e-01
-1.11841297e+00 -1.54266790e-01 -3.00575197e-01 -8.20146322e-01
-3.52135897e-01 -2.53092855e-01 -3.28586280e-01 8.80019784e-01
-1.05389750e+00 -8.02264035e-01 -6.59697428e-02 3.21714878e-01
-4.64034557e-01 -2.50896722e-01 5.65394998e-01 7.32567668e-01
-3.85468900e-01 6.50679111e-01 2.19115421e-01 2.50012189e-01
5.96220672e-01 -6.71140969e-01 9.62646723e-01 6.15554988e-01
-1.50340974e-01 1.31811953e+00 -1.71018243e-01 -1.21463776e+00
-1.44116247e+00 -1.09213567e+00 1.31354368e+00 -1.51644814e+00
5.33945084e-01 -8.09567332e-01 -3.49363387e-01 1.08827746e+00
1.11718185e-01 -2.45238632e-01 1.11075675e+00 -6.09552152e-02
-2.56139059e-02 -3.79563987e-01 -1.62108684e+00 2.74833769e-01
1.09468281e+00 -1.05245578e+00 1.26374438e-01 8.98683071e-02
9.61557865e-01 -5.45775332e-02 -1.18712533e+00 -2.80102760e-01
8.14356804e-01 -1.20853150e+00 5.86863637e-01 -4.03007090e-01
2.02691764e-01 1.50784284e-01 -9.14057419e-02 -7.00774968e-01
1.32460788e-01 -1.43909717e+00 -5.33907175e-01 2.14672399e+00
5.34400344e-01 -1.58917868e+00 1.06841719e+00 1.17345643e+00
3.88238668e-01 -5.99842608e-01 -8.63113940e-01 -5.20697415e-01
-1.71117097e-01 -5.48793554e-01 1.27287257e+00 1.18201625e+00
-1.31238893e-01 -8.08311775e-02 -2.19582066e-01 -1.00616375e-02
2.33686730e-01 -1.92621842e-01 9.81526077e-01 -1.25439656e+00
1.03491306e-01 6.36738632e-03 -6.33640662e-02 -8.52372766e-01
6.25730082e-02 -5.49294949e-01 -1.09264338e+00 -1.10954976e+00
1.82363093e-01 -7.21137166e-01 2.49435052e-01 3.56138408e-01
6.32786214e-01 2.54455149e-01 3.41038331e-02 4.20479149e-01
-2.40800217e-01 -2.37964675e-01 7.27680862e-01 2.02224553e-01
-4.25193518e-01 6.09807909e-01 -1.87501276e+00 6.84311748e-01
1.13586581e+00 -4.95152414e-01 -1.78726017e-01 -1.32573843e-01
6.92039967e-01 -1.82561547e-01 4.30838346e-01 -6.37025654e-01
-1.93477005e-01 -9.75463912e-02 2.62770772e-01 -5.06195664e-01
6.37681186e-02 -1.29793668e+00 5.45265913e-01 2.16156662e-01
2.17506155e-01 1.53372079e-01 2.10654587e-02 5.27600586e-01
1.63279936e-01 -1.16336644e-01 1.81199521e-01 -3.00288685e-02
-1.52813029e-02 2.25065827e-01 -4.39913332e-01 2.41671816e-01
1.20654619e+00 -5.91643512e-01 -7.56691992e-01 -6.83022797e-01
-2.24100500e-01 -1.69511318e-01 9.82244790e-01 4.54906762e-01
-8.86113271e-02 -1.37957942e+00 4.69493195e-02 1.22177027e-01
-2.00547233e-01 -6.12593889e-01 1.47269983e-02 8.25914800e-01
5.15011437e-02 7.47555852e-01 -1.75324982e-04 -1.23129562e-01
-8.71220410e-01 2.56208450e-01 -9.74471308e-03 3.00111677e-02
-6.72944337e-02 1.18131585e-01 -3.52845818e-01 -2.49225065e-01
-1.83153614e-01 -4.43922967e-01 1.09496273e-01 4.68755156e-01
7.97737062e-01 1.06551623e+00 2.61768326e-02 -9.19034123e-01
-7.19526529e-01 4.41864640e-01 8.47160593e-02 -9.21064094e-02
8.90770257e-01 -5.20593107e-01 -4.09558207e-01 3.65253836e-01
1.85329890e+00 1.05602801e+00 -8.63723040e-01 4.60963815e-01
-1.59152418e-01 -9.07547712e-01 -5.57056546e-01 -5.95743537e-01
-1.37272680e+00 2.62386978e-01 2.55817562e-01 9.05001223e-01
9.49009478e-01 1.95694193e-01 1.35872793e+00 -6.87563062e-01
9.23875213e-01 -1.05523729e+00 -4.88043457e-01 -1.57047268e-02
2.54054070e-01 -1.05269146e+00 -1.82480678e-01 -6.61956489e-01
-3.96185845e-01 4.69819546e-01 1.78579763e-01 4.94338304e-01
9.69634533e-01 7.97476247e-02 -1.10283025e-01 -8.65068063e-02
-1.07630178e-01 5.37018359e-01 -2.86025345e-01 1.02027059e+00
1.94161445e-01 3.00809652e-01 -4.97283697e-01 1.26482165e+00
-2.64643610e-01 6.15948811e-03 8.99853528e-01 8.26101422e-01
-3.89218368e-02 -1.38866985e+00 -2.38526657e-01 1.09370017e+00
-1.13831222e+00 1.38870642e-01 -8.64368320e-01 7.24195838e-01
4.06735390e-01 1.27997935e+00 1.00755945e-01 -4.64336932e-01
4.04792689e-02 -1.12503439e-01 -3.90360922e-01 -5.94658136e-01
-7.08362997e-01 -7.82441199e-01 3.65681171e-01 -6.55171096e-01
-2.36756563e-01 -1.12954557e+00 -9.54109609e-01 -7.85264492e-01
-6.20156191e-02 -9.20343921e-02 9.91444767e-01 4.12660688e-01
1.12930620e+00 -2.49312595e-01 8.81755829e-01 -1.81796983e-01
-1.57636538e-01 -6.07270300e-01 -6.52627051e-01 2.56092608e-01
3.38245571e-01 -2.83445358e-01 -5.30759752e-01 -1.60670996e-01] | [6.311082363128662, 6.976050853729248] |
837034b6-10fb-443d-82e5-0ba0188c7556 | spectral-geometric-verification-re-ranking | 2210.04432 | null | https://arxiv.org/abs/2210.04432v2 | https://arxiv.org/pdf/2210.04432v2.pdf | Spectral Geometric Verification: Re-Ranking Point Cloud Retrieval for Metric Localization | In large-scale metric localization, an incorrect result during retrieval will lead to an incorrect pose estimate or loop closure. Re-ranking methods propose to take into account all the top retrieval candidates and re-order them to increase the likelihood of the top candidate being correct. However, state-of-the-art re-ranking methods are inefficient when re-ranking many potential candidates due to their need for resource intensive point cloud registration between the query and each candidate. In this work, we propose an efficient spectral method for geometric verification (named SpectralGV) that does not require registration. We demonstrate how the optimal inter-cluster score of the correspondence compatibility graph of two point clouds represents a robust fitness score measuring their spatial consistency. This score takes into account the subtle geometric differences between structurally similar point clouds and therefore can be used to identify the correct candidate among potential matches retrieved by global similarity search. SpectralGV is deterministic, robust to outlier correspondences, and can be computed in parallel for all potential candidates. We conduct extensive experiments on 5 large-scale datasets to demonstrate that SpectralGV outperforms other state-of-the-art re-ranking methods and show that it consistently improves the recall and pose estimation of 3 state-of-the-art metric localization architectures while having a negligible effect on their runtime. The open-source implementation and trained models are available at: https://github.com/csiro-robotics/SpectralGV. | ['Clinton Fookes', 'Sridha Sridharan', 'Peyman Moghadam', 'Kavisha Vidanapathirana'] | 2022-10-10 | null | null | null | null | ['point-cloud-registration'] | ['computer-vision'] | [-1.96732774e-01 -4.89427030e-01 3.90632637e-03 -1.52606696e-01
-1.47377241e+00 -7.94209003e-01 5.65154612e-01 5.70818126e-01
-3.15565556e-01 1.93054333e-01 -2.09806770e-01 2.53271237e-02
-4.24378812e-01 -6.41614795e-01 -1.08480883e+00 -4.04412746e-01
-2.00897008e-01 9.57938552e-01 5.75784922e-01 -1.30853266e-01
6.00711286e-01 8.62549245e-01 -1.78091407e+00 -1.35463953e-01
7.83925951e-01 9.42726791e-01 4.56481457e-01 3.46167147e-01
1.37608692e-01 -1.68777123e-01 -5.73503375e-01 -3.48335914e-02
5.39885938e-01 7.29661062e-03 -6.80101037e-01 -4.28075403e-01
1.17749107e+00 -1.08634487e-01 1.52734034e-02 1.12610173e+00
5.92884421e-01 2.00271100e-01 4.01958883e-01 -1.16527212e+00
-3.37327570e-01 4.23083715e-02 -5.65953195e-01 -9.57289711e-02
5.99890649e-01 -2.00708866e-01 1.10367763e+00 -1.51036501e+00
6.73130512e-01 1.02477920e+00 1.00908530e+00 4.78748754e-02
-1.16341102e+00 -8.74898374e-01 -1.65701157e-03 2.44523093e-01
-1.95359397e+00 -3.38272572e-01 6.92099392e-01 -3.46395820e-01
8.53905082e-01 3.69414568e-01 5.85516751e-01 2.65306711e-01
5.69965737e-03 2.27590680e-01 6.60966992e-01 -2.95355737e-01
2.27018416e-01 -4.95816410e-01 -9.96502638e-02 7.48835444e-01
5.22203863e-01 2.45451197e-01 -7.29575455e-01 -6.12841249e-01
5.65498412e-01 6.20782524e-02 -4.76113632e-02 -1.00961852e+00
-1.58909953e+00 5.34970760e-01 9.34970856e-01 1.56468511e-01
-3.42997909e-01 5.35374284e-01 -6.19309805e-02 1.66907296e-01
8.56541619e-02 7.86922634e-01 -3.44568431e-01 2.83692982e-02
-9.38653171e-01 3.24134082e-01 3.23483974e-01 1.01096320e+00
1.11586487e+00 -5.42882562e-01 2.96359807e-01 7.73078620e-01
3.75174195e-01 8.37681532e-01 1.95038468e-01 -1.01402676e+00
3.46146405e-01 7.33382106e-01 2.48788342e-01 -1.45084476e+00
-4.64041561e-01 -4.52628314e-01 -2.65211225e-01 1.80219337e-01
1.10192150e-01 4.75412399e-01 -7.99398243e-01 1.41779971e+00
5.32233417e-01 5.18708527e-01 -1.29288211e-01 9.88110125e-01
8.13680112e-01 3.52032214e-01 -3.80254358e-01 2.26434454e-01
9.63631690e-01 -6.35841548e-01 -1.20531702e-02 -3.08717728e-01
6.89775467e-01 -1.22071064e+00 6.86461270e-01 1.86391369e-01
-7.46397376e-01 -5.12375891e-01 -1.31519413e+00 1.15014859e-01
-3.41597646e-01 3.62764567e-01 3.67692471e-01 6.07207976e-02
-1.17136359e+00 8.10352504e-01 -9.44903135e-01 -6.08481884e-01
-6.67116493e-02 6.93124354e-01 -5.87258160e-01 -1.75003558e-01
-7.60589182e-01 1.04041219e+00 2.39736944e-01 1.73663795e-01
-5.11526048e-01 -6.96675599e-01 -6.35331094e-01 -2.45107502e-01
2.31514350e-01 -5.94607770e-01 1.10093808e+00 -3.31413805e-01
-1.00477326e+00 8.13659966e-01 -3.81543338e-01 -1.55972550e-02
2.64277011e-01 -3.18171382e-01 -1.99043512e-01 2.00634420e-01
4.66217458e-01 7.54362702e-01 4.41563159e-01 -1.43906498e+00
-6.82117224e-01 -4.03652698e-01 -3.34667534e-01 2.71560311e-01
1.25129089e-01 -7.53060728e-02 -7.24714398e-01 -3.20586562e-01
1.01289964e+00 -1.37261355e+00 -2.43404210e-01 1.19135469e-01
-9.17971879e-02 -1.32283494e-01 6.96998775e-01 -2.76683241e-01
9.04608488e-01 -2.11146235e+00 1.03753069e-02 6.61901534e-01
-5.93737885e-02 -1.44321406e-02 -3.40880245e-01 6.38575912e-01
2.54838299e-02 -5.64857423e-02 9.06766485e-03 -2.90683776e-01
-1.60399362e-01 -4.78962287e-02 -1.93038881e-01 9.23136294e-01
-1.58980250e-01 6.25972748e-01 -1.09411573e+00 -3.61961216e-01
4.58676994e-01 4.00091827e-01 -3.27733099e-01 -1.49110764e-01
7.13808164e-02 3.55865270e-01 -4.46101815e-01 7.76701748e-01
9.43012953e-01 -3.59373242e-01 -1.06933728e-01 -5.09275973e-01
-1.64955556e-01 3.65614414e-01 -1.53146124e+00 2.00309443e+00
-2.23139107e-01 4.31999832e-01 -1.86103508e-01 -4.79250371e-01
1.09217298e+00 -2.16631934e-01 7.83082426e-01 -5.39324820e-01
-1.35395706e-01 8.70720565e-01 -2.52124667e-01 3.19971919e-01
8.76892507e-01 4.32517231e-01 -2.21872658e-01 2.30418056e-01
-2.31018066e-01 -4.45161074e-01 -1.98117197e-02 2.11095005e-01
1.15114188e+00 2.05679506e-01 2.61213422e-01 -1.73941359e-01
4.07500625e-01 3.97096306e-01 4.19376522e-01 6.85627699e-01
1.11616805e-01 1.06295443e+00 -3.42101634e-01 -2.90912062e-01
-7.59178758e-01 -1.23795736e+00 -1.32147551e-01 7.69426584e-01
9.11585391e-01 -6.77942157e-01 -2.66356200e-01 -2.80164570e-01
4.37857151e-01 2.14916795e-01 -2.03874364e-01 -1.67773530e-01
-5.86284935e-01 -2.07201153e-01 3.57029945e-01 4.22604889e-01
1.77377269e-01 -5.76561868e-01 -6.48140609e-01 7.54923224e-02
-3.43281239e-01 -9.03707564e-01 -4.09886122e-01 4.56785187e-02
-1.03145885e+00 -1.29326892e+00 -3.77718389e-01 -8.65191102e-01
1.04482639e+00 8.77816021e-01 1.10755682e+00 4.96731997e-01
-6.29621670e-02 5.17782569e-01 -4.93332237e-01 -4.01980728e-02
-1.27236962e-01 1.14089675e-01 3.82078975e-01 -4.86885548e-01
2.51303852e-01 -3.79482210e-01 -6.35810494e-01 8.42606306e-01
-4.35070246e-01 -3.94670308e-01 4.88561064e-01 5.95347583e-01
1.26557791e+00 -3.55226666e-01 1.99478138e-02 -2.00997859e-01
1.68645456e-01 -1.00495651e-01 -1.03643775e+00 3.80807132e-01
-5.72538733e-01 1.79047048e-01 1.78468987e-01 -2.74126321e-01
-1.36501372e-01 5.60952008e-01 3.12010378e-01 -7.41058886e-01
2.53820866e-01 6.55608475e-01 2.12740749e-01 -8.01633835e-01
7.03210056e-01 -8.44634976e-03 -1.74667388e-01 -5.35084903e-01
3.12775046e-01 4.22884673e-01 6.81535125e-01 -6.66278839e-01
1.28525555e+00 5.14455438e-01 3.64691675e-01 -5.17201245e-01
-4.73316163e-01 -1.01014268e+00 -8.24596286e-01 -2.03416139e-01
2.50771463e-01 -9.75711346e-01 -5.55581272e-01 2.41557494e-01
-1.28133368e+00 2.10856780e-01 7.80033916e-02 5.60692608e-01
-4.36748117e-01 3.83820534e-01 -7.86413625e-03 -3.76157969e-01
-2.99368709e-01 -1.25574911e+00 1.50073552e+00 -8.73107016e-02
-2.83933252e-01 -4.62842435e-01 3.48042548e-01 5.69097437e-02
1.83470726e-01 2.01327845e-01 3.73130232e-01 -5.51394880e-01
-1.04864907e+00 -5.84739923e-01 -6.32886291e-02 -2.35886142e-01
1.25027657e-01 1.18744820e-01 -6.28316343e-01 -6.22106135e-01
-6.09980822e-01 8.71480778e-02 6.63332820e-01 1.14532419e-01
7.23033845e-01 1.83249593e-01 -7.37582862e-01 6.10014856e-01
1.52315402e+00 -2.15005428e-01 4.11830038e-01 7.00505674e-01
7.47226357e-01 3.11254472e-01 1.31535029e+00 2.11188540e-01
3.90321732e-01 1.11159611e+00 7.28209615e-01 -6.15967531e-03
2.67970785e-02 -3.61589938e-01 1.93467215e-01 8.37323546e-01
7.64459297e-02 1.46010533e-01 -1.26010668e+00 6.19845629e-01
-1.99180996e+00 -6.90023780e-01 -4.16478902e-01 2.75476670e+00
4.41669226e-01 -1.26365259e-01 -1.55625001e-01 -1.51170954e-01
8.66110802e-01 -8.46236348e-02 -3.24507922e-01 2.24354103e-01
3.34474705e-02 1.30533770e-01 9.42885399e-01 5.92092812e-01
-1.19652569e+00 1.05643773e+00 5.71741915e+00 6.40417576e-01
-1.13672006e+00 -5.10335453e-02 -7.84989297e-02 -1.49926111e-01
-2.52513085e-02 2.59837776e-01 -8.53528440e-01 1.70306325e-01
6.02586985e-01 -9.75193009e-02 3.34525049e-01 1.05023420e+00
-1.44408599e-01 -3.07635546e-01 -1.20763910e+00 1.14852405e+00
8.00272301e-02 -1.35795188e+00 -1.28921673e-01 8.29988047e-02
7.00982571e-01 7.62489915e-01 -6.16820380e-02 -2.17950150e-01
1.80997387e-01 -6.76298916e-01 9.65182781e-01 2.62015492e-01
8.34858954e-01 -6.76091135e-01 7.14866877e-01 1.13609828e-01
-1.66991389e+00 2.11458936e-01 -5.73238373e-01 3.48960459e-01
-4.18956503e-02 4.50541675e-01 -8.79921734e-01 6.89840913e-01
1.03119016e+00 5.95401764e-01 -9.13812459e-01 1.60264862e+00
-7.23178685e-02 -3.58443037e-02 -9.18756545e-01 3.62958200e-02
4.59168516e-02 6.27732463e-03 6.91614509e-01 7.90989757e-01
8.93554389e-01 -1.18139409e-01 3.98201764e-01 4.83483464e-01
1.77405730e-01 1.18420422e-01 -4.81123239e-01 4.69336003e-01
1.04142642e+00 1.12464476e+00 -8.33045959e-01 -1.59143973e-02
-1.27213895e-01 8.31444621e-01 3.18604946e-01 -5.94240129e-02
-7.19225347e-01 -3.54860723e-01 7.51821995e-01 2.07892820e-01
3.45026284e-01 -5.27274728e-01 -2.29018882e-01 -8.19261193e-01
3.45540643e-01 -5.53762734e-01 3.01891714e-01 -8.85497093e-01
-1.07329774e+00 5.62463880e-01 2.16796156e-03 -1.95761049e+00
-2.80646563e-01 -3.81215811e-01 -2.53837019e-01 7.29010880e-01
-1.40122950e+00 -1.17082572e+00 -6.35115862e-01 4.29728746e-01
-1.23567972e-02 1.68063700e-01 8.30817044e-01 3.29478413e-01
1.03732102e-01 5.64990878e-01 2.90546238e-01 -1.55525833e-01
1.01915157e+00 -1.01332247e+00 5.58889031e-01 8.99204969e-01
4.18960303e-01 9.45135057e-01 7.34861076e-01 -8.41532290e-01
-1.54169154e+00 -9.69611645e-01 9.31688607e-01 -6.88130796e-01
5.99905729e-01 -2.34527647e-01 -7.48135448e-01 3.38791221e-01
-5.75155735e-01 2.68334389e-01 2.54950106e-01 1.01096638e-01
-4.80551094e-01 -3.27176273e-01 -1.01263642e+00 4.21040654e-01
1.19888008e+00 -4.71494675e-01 -4.37965989e-01 3.75516534e-01
5.66812873e-01 -7.51356244e-01 -8.52149725e-01 7.35937476e-01
5.73252559e-01 -8.12849164e-01 1.33783638e+00 2.41205469e-01
-2.35754818e-01 -9.65816081e-01 -3.47515047e-01 -1.21196222e+00
-4.11286294e-01 -4.47684139e-01 1.42030761e-01 9.08987403e-01
4.10788357e-01 -6.53787732e-01 8.25254083e-01 2.64807403e-01
-3.23499948e-01 -4.86239314e-01 -1.14822912e+00 -1.15744579e+00
-3.01955104e-01 -3.20120245e-01 8.11596692e-01 8.57819378e-01
-2.26485074e-01 -7.37973079e-02 -1.55072808e-02 8.68407488e-01
6.36779308e-01 5.19286096e-01 1.06382263e+00 -1.69105434e+00
4.13213708e-02 -2.94628382e-01 -8.70730340e-01 -1.10144269e+00
-6.84054941e-02 -8.82366538e-01 5.19369006e-01 -1.70657647e+00
-7.36662745e-02 -8.86201501e-01 -2.96580046e-01 5.98859251e-01
2.28861421e-02 5.60915291e-01 2.39404693e-01 8.46233904e-01
-1.02144527e+00 3.69318426e-01 6.64804757e-01 -1.40636414e-02
-2.38500029e-01 -3.67001556e-02 -3.78664911e-01 5.29879987e-01
7.58106828e-01 -7.40656793e-01 1.04071768e-02 -5.11260748e-01
3.61885279e-01 -2.66399771e-01 6.47984982e-01 -1.46965981e+00
6.14855826e-01 -1.08841322e-01 2.20633984e-01 -1.05525768e+00
5.43241978e-01 -1.18929672e+00 5.68190694e-01 4.74362195e-01
7.48107359e-02 5.05001605e-01 2.58272529e-01 4.90622461e-01
-2.32148722e-01 -1.57939076e-01 5.76191425e-01 6.53421283e-02
-8.53837073e-01 3.00069273e-01 2.65870929e-01 -3.84455502e-01
9.25816417e-01 -4.55493361e-01 -3.88874710e-01 -2.36670345e-01
-3.29639524e-01 2.51731485e-01 1.12385368e+00 5.69526672e-01
8.14808965e-01 -1.54514313e+00 -5.30797362e-01 -3.88187356e-02
6.74223483e-01 1.68631896e-01 -2.14570597e-01 7.33467996e-01
-7.96430171e-01 3.33777338e-01 1.42334580e-01 -1.06838465e+00
-1.64022958e+00 2.09232226e-01 2.86187977e-01 -4.23516659e-03
-3.86114091e-01 8.63317251e-01 -2.39829496e-01 -7.95106649e-01
1.83872327e-01 -3.15305084e-01 3.38251561e-01 -2.43013948e-01
1.77543566e-01 4.31127936e-01 6.29855037e-01 -1.00485420e+00
-1.06231892e+00 1.39819288e+00 2.12023497e-01 -2.82696523e-02
1.21966422e+00 -1.43017443e-02 -2.57394046e-01 9.94335115e-02
1.25306797e+00 3.21423143e-01 -8.77809167e-01 -2.41539448e-01
2.83070743e-01 -7.64013708e-01 6.62893951e-02 -6.12975121e-01
-9.17922437e-01 3.79004091e-01 8.04932594e-01 -3.13900322e-01
8.24796081e-01 3.43621522e-01 6.20809019e-01 7.50617802e-01
1.04172659e+00 -9.48180497e-01 9.47445780e-02 6.14471257e-01
1.00855267e+00 -1.21651363e+00 4.84529734e-01 -6.09843254e-01
-1.56589642e-01 9.87539351e-01 6.00792587e-01 -3.98117334e-01
4.82673377e-01 -4.43823449e-02 1.05554886e-01 -4.99048263e-01
-1.88588515e-01 -1.96796447e-01 6.73639417e-01 5.15396416e-01
1.57336161e-01 2.56979093e-02 -3.37053090e-02 -8.50879326e-02
-5.17560720e-01 -3.46199214e-01 1.61674377e-02 1.05946505e+00
-5.34860373e-01 -1.16223300e+00 -7.61196911e-01 2.60419607e-01
4.71399650e-02 -5.91133498e-02 -6.45265877e-01 5.89849234e-01
-1.06131941e-01 8.72981548e-01 1.44225195e-01 -7.68572748e-01
4.86328900e-01 -4.03403908e-01 5.57171226e-01 -6.04927540e-01
-5.09898245e-01 4.26367857e-02 -1.03345662e-01 -1.03029990e+00
-4.01894182e-01 -8.90576899e-01 -1.52906001e+00 -2.75134325e-01
-7.03656614e-01 2.49364704e-01 9.15071666e-01 4.83923852e-01
8.85030270e-01 -1.47490889e-01 5.06784558e-01 -1.20906818e+00
-4.63629305e-01 -6.23914301e-01 -1.17005974e-01 2.49849260e-01
2.48064891e-01 -1.06361389e+00 -4.51918185e-01 -4.67270166e-01] | [7.590928077697754, -2.337731122970581] |
1c1db442-1304-4664-93b6-e8663b87275a | paper-abstract-writing-through-editing | 1805.06064 | null | http://arxiv.org/abs/1805.06064v1 | http://arxiv.org/pdf/1805.06064v1.pdf | Paper Abstract Writing through Editing Mechanism | We present a paper abstract writing system based on an attentive neural
sequence-to-sequence model that can take a title as input and automatically
generate an abstract. We design a novel Writing-editing Network that can attend
to both the title and the previously generated abstract drafts and then
iteratively revise and polish the abstract. With two series of Turing tests,
where the human judges are asked to distinguish the system-generated abstracts
from human-written ones, our system passes Turing tests by junior domain
experts at a rate up to 30% and by non-expert at a rate up to 80%. | ['Zhi-Hao Zhou', 'Lifu Huang', 'Heng Ji', 'Boliang Zhang', 'Qingyun Wang', 'Kevin Knight', 'Spencer Whitehead'] | 2018-05-15 | paper-abstract-writing-through-editing-1 | https://aclanthology.org/P18-2042 | https://aclanthology.org/P18-2042.pdf | acl-2018-7 | ['paper-generation'] | ['natural-language-processing'] | [ 4.86570239e-01 6.36236906e-01 5.83551824e-02 -4.45363492e-01
-8.01724315e-01 -9.33776379e-01 5.38581908e-01 2.05696017e-01
-4.80306447e-01 9.27752674e-01 9.44949239e-02 -5.95953465e-01
1.61709458e-01 -5.37696421e-01 -6.87190175e-01 1.28386602e-01
6.04153872e-01 6.24303401e-01 2.18409836e-01 -1.82903409e-01
6.58290267e-01 3.78670335e-01 -1.18695724e+00 6.38438404e-01
1.13342190e+00 2.02148378e-01 3.87886137e-01 1.40134656e+00
-1.82363212e-01 7.41688311e-01 -1.26315355e+00 -6.89948082e-01
-4.79214564e-02 -6.14189446e-01 -1.24117827e+00 -1.40188932e-01
3.67957741e-01 -3.20722461e-01 -2.03319356e-01 1.08325112e+00
3.19091141e-01 -2.99610179e-02 7.75036156e-01 -8.21246982e-01
-1.35323799e+00 1.18613827e+00 -4.97168377e-02 1.84230089e-01
9.24475312e-01 4.49395716e-01 7.16830432e-01 -1.08585167e+00
8.11581612e-01 1.15573871e+00 1.83949366e-01 9.17080462e-01
-1.00144708e+00 -4.19334531e-01 2.95196056e-01 -1.43726885e-01
-1.08349478e+00 -3.32901835e-01 4.43799675e-01 -3.82459581e-01
1.17735565e+00 3.06329966e-01 4.55498934e-01 1.19593930e+00
3.30815852e-01 6.55290484e-01 5.73887110e-01 -6.60063565e-01
2.95322746e-01 2.41475284e-01 3.95902485e-01 5.56799889e-01
1.74260557e-01 -2.79547572e-01 -5.64792871e-01 2.10040286e-02
5.28216541e-01 -5.31870246e-01 -5.71670532e-01 5.06255507e-01
-1.19327211e+00 2.74691790e-01 -1.15035623e-01 3.96020524e-02
-6.95439696e-01 -2.39644572e-01 2.30674922e-01 5.25183499e-01
-1.45409763e-01 1.15989316e+00 -3.59627515e-01 -1.44531399e-01
-1.18912756e+00 3.93319190e-01 1.24055326e+00 1.38585603e+00
2.22424120e-01 3.93631570e-02 -8.74083698e-01 6.64941609e-01
9.34195742e-02 6.25781596e-01 4.76946294e-01 -8.01346600e-01
5.87610424e-01 6.35217130e-01 5.96487999e-01 -7.64666021e-01
2.85692453e-01 -1.67779893e-01 -7.33811200e-01 2.98032612e-01
1.48771524e-01 -2.47885421e-01 -8.78136396e-01 1.50104070e+00
-2.36998945e-01 -6.68084264e-01 3.01127106e-01 7.83866048e-01
1.13316154e+00 9.74252641e-01 7.02790990e-02 -6.85803235e-01
1.45335114e+00 -1.02202642e+00 -1.02759695e+00 -1.06308311e-01
2.74960041e-01 -7.87062347e-01 1.49337566e+00 8.28163862e-01
-1.73489857e+00 -8.37013066e-01 -1.40376318e+00 -1.50291190e-01
-2.91509956e-01 4.77785349e-01 -2.16557831e-01 2.95667231e-01
-1.10164332e+00 8.12966406e-01 -3.30770433e-01 -3.50055069e-01
2.27795690e-01 2.02698857e-01 -1.00372337e-01 8.36891532e-02
-1.38193214e+00 9.77841318e-01 5.76110482e-01 2.64241010e-01
-7.92599380e-01 -9.05815363e-01 -4.41321850e-01 2.99092770e-01
2.53903598e-01 -9.47254539e-01 2.02804565e+00 -7.88891375e-01
-2.09154773e+00 9.21781301e-01 -1.60513476e-01 -1.12302005e-01
7.64531136e-01 -4.52845931e-01 -6.04049206e-01 2.72474170e-01
5.03689013e-02 5.20892620e-01 6.13590181e-01 -9.41640377e-01
-4.73485887e-01 5.17454110e-02 7.97362700e-02 1.29197419e-01
-1.68392509e-01 2.24934518e-01 -6.37697816e-01 -7.74770439e-01
-4.41202790e-01 -7.82039046e-01 1.28606707e-01 5.57819474e-03
-5.66803157e-01 -5.78747332e-01 3.57394636e-01 -8.39244127e-01
1.49899971e+00 -1.64105630e+00 3.70256603e-01 -8.10791645e-03
3.34140956e-01 5.81136644e-01 -4.82124031e-01 6.54033303e-01
-4.40879241e-02 3.13808411e-01 -3.02672595e-01 -1.59613773e-01
2.83654667e-02 -2.27874607e-01 -6.07545614e-01 -3.20765376e-01
4.26458567e-01 9.71814632e-01 -1.22614443e+00 -3.06804061e-01
-3.78074437e-01 -1.03152648e-01 -2.89017409e-01 8.55768621e-01
-4.77471471e-01 -2.09779695e-01 -1.54465914e-01 1.47380099e-01
5.53130209e-01 -4.11787629e-01 2.38597870e-01 2.07300454e-01
-6.83186576e-02 4.02487576e-01 -9.26425159e-01 1.55845225e+00
-3.89563978e-01 8.76032352e-01 -1.09929442e-01 -4.46080506e-01
1.11200821e+00 5.63417315e-01 -9.32564914e-01 -6.55469954e-01
-1.86190143e-01 1.81605697e-01 -3.53718437e-02 -7.17790246e-01
7.81370401e-01 8.07616487e-02 -8.27909112e-02 1.03061855e+00
-1.76463351e-02 -4.76230562e-01 5.49110830e-01 6.86885953e-01
9.36420560e-01 1.11216694e-01 2.86182702e-01 -1.86703756e-01
6.99290633e-01 -4.50325236e-02 2.31903106e-01 1.53144777e+00
1.64707348e-01 7.43419886e-01 8.93007159e-01 -6.77867711e-01
-1.20274115e+00 -1.03803039e+00 7.13958979e-01 1.01144087e+00
-2.18446493e-01 -4.00298685e-01 -1.10645413e+00 -5.83820879e-01
-4.63465810e-01 1.33390749e+00 -5.49321592e-01 -4.98264074e-01
-4.06592220e-01 1.42680183e-01 6.88995361e-01 7.23003566e-01
3.27801049e-01 -1.75083721e+00 -8.04196060e-01 1.55000225e-01
-1.42629027e-01 -9.25229907e-01 -8.32393765e-01 -7.54251629e-02
-5.48068702e-01 -5.76822221e-01 -1.14372313e+00 -9.71438646e-01
7.02261925e-01 -4.14549947e-01 1.15184617e+00 4.40278172e-01
-1.23649046e-01 -3.56075056e-02 -1.54004782e-01 -6.82707906e-01
-1.05417311e+00 3.77418905e-01 1.14125073e-01 -3.72815818e-01
2.78690189e-01 -1.58464044e-01 -5.56516387e-02 -3.54963429e-02
-1.05019665e+00 4.83081341e-01 7.85380006e-01 4.27028298e-01
-9.24452916e-02 -4.39906657e-01 7.74485707e-01 -9.85703945e-01
1.43861973e+00 1.13976918e-01 -5.89989364e-01 7.19380975e-01
-6.69998169e-01 3.81571531e-01 1.06954384e+00 -7.56844640e-01
-1.16179311e+00 -2.33739674e-01 2.87668616e-01 -2.63487756e-01
-1.47503749e-01 5.86450040e-01 -1.13517269e-01 5.78117609e-01
9.80422556e-01 4.86977309e-01 -9.89382938e-02 -1.86621562e-01
1.65369242e-01 8.23265135e-01 9.96295869e-01 -5.44068873e-01
9.33689475e-01 -5.90358615e-01 -6.09933794e-01 -6.66080177e-01
-9.03769433e-01 3.02902609e-01 -7.94746459e-01 -2.52605855e-01
6.50522768e-01 -5.51696122e-01 -9.00761604e-01 3.06758851e-01
-1.98221684e+00 -4.43709105e-01 -7.81500116e-02 2.77940392e-01
-2.81539202e-01 2.99506694e-01 -6.70778215e-01 -6.27805114e-01
-7.60679245e-01 -1.14042354e+00 6.83813274e-01 6.81971073e-01
-1.16514730e+00 -4.32112902e-01 -1.34656549e-01 1.99972484e-02
2.89541334e-01 -1.72155187e-01 1.18775880e+00 -7.52024055e-01
-2.81466424e-01 -3.61201644e-01 -4.01408255e-01 2.65792459e-01
-5.32304980e-02 4.77053046e-01 -6.41444206e-01 8.49527791e-02
-3.76692593e-01 -3.15569788e-01 4.57352072e-01 -8.73872116e-02
1.09620428e+00 -6.89547181e-01 -2.09548041e-01 1.49229364e-02
6.48617566e-01 5.81124425e-01 7.84392953e-01 -6.78559989e-02
3.79723549e-01 4.28176373e-01 1.09000728e-01 1.33576304e-01
1.48969635e-01 3.15366268e-01 -3.18773478e-01 1.57975212e-01
6.77535543e-03 -5.74128449e-01 1.18346207e-01 9.67855036e-01
-4.27621901e-02 -7.15261638e-01 -9.48436260e-01 6.51935816e-01
-1.71460378e+00 -9.42642868e-01 -1.21216469e-01 2.04472685e+00
1.44115055e+00 6.34817958e-01 -9.96285304e-02 -1.13618365e-02
7.96707213e-01 -1.14641272e-01 -4.71286058e-01 -1.20921481e+00
1.74268529e-01 3.80058140e-01 -1.62399635e-01 7.43913651e-01
-4.18818653e-01 1.20883405e+00 7.52649546e+00 3.29935133e-01
-1.04207242e+00 -7.32252896e-01 1.82286501e-01 -1.11145101e-01
-4.25928771e-01 -9.89523157e-02 -5.37715793e-01 4.36111122e-01
1.18294871e+00 -6.67922497e-01 4.22025949e-01 7.69370437e-01
2.89592087e-01 8.04866329e-02 -1.47094953e+00 7.68363893e-01
2.21879959e-01 -1.51282215e+00 9.28732276e-01 -3.98965865e-01
4.95306522e-01 -7.64797628e-01 -2.17535377e-01 3.39270353e-01
4.82754797e-01 -1.35991812e+00 1.02900553e+00 9.42584574e-01
8.88875425e-01 -6.82839632e-01 6.79209054e-01 4.96755898e-01
-5.82551539e-01 1.60052627e-01 -1.86119109e-01 -4.86514926e-01
1.56511478e-02 3.99546653e-01 -1.07646298e+00 4.67691720e-01
3.90312195e-01 2.57113963e-01 -5.33737898e-01 7.47033000e-01
-1.05191565e+00 5.17170787e-01 3.49212103e-02 -9.09711719e-01
-3.75379138e-02 2.84546912e-01 4.18786407e-01 1.69043756e+00
1.04534097e-01 7.58148372e-01 2.29874589e-02 1.36161435e+00
-5.65536380e-01 -6.68098256e-02 -3.02748621e-01 -5.73196292e-01
4.49096233e-01 1.08518183e+00 -5.89885771e-01 -9.37270880e-01
2.38316506e-01 1.35080814e+00 3.43830645e-01 6.71650708e-01
-4.92758423e-01 -1.54962420e+00 2.71612015e-02 -2.01606497e-01
2.84863472e-01 -1.16513968e-01 -5.86433351e-01 -9.59261417e-01
3.66898417e-01 -1.09565544e+00 -6.41362928e-03 -1.67420995e+00
-8.43604326e-01 1.10796535e+00 -1.91446647e-01 -8.01059425e-01
-1.02789424e-01 -5.01131952e-01 -9.99611199e-01 1.35609245e+00
-9.20346081e-01 -4.53911662e-01 -3.55157375e-01 -2.93643177e-01
8.94567490e-01 -2.46468350e-01 7.95617819e-01 -2.04162419e-01
-5.25088489e-01 7.42039979e-01 -5.58781624e-01 4.63733196e-01
5.17061949e-01 -1.10720336e+00 7.95365632e-01 1.00931835e+00
-1.85905576e-01 1.24976110e+00 1.06688797e+00 -9.76276875e-01
-1.06054950e+00 -7.54450500e-01 1.59216309e+00 -5.72295427e-01
2.35708624e-01 -4.46441710e-01 -1.41807652e+00 3.38748515e-01
7.89224029e-01 -6.16170287e-01 4.25088376e-01 -2.50152528e-01
-4.16187257e-01 1.95995212e-01 -8.49678457e-01 1.06191659e+00
9.26440299e-01 -6.55349493e-01 -1.51185954e+00 2.44427472e-01
9.70234156e-01 -5.66599965e-01 -5.20061135e-01 1.05595306e-01
8.02098989e-01 -3.35692465e-01 3.57870430e-01 -1.15172124e+00
1.28199887e+00 -5.49337745e-01 5.13029158e-01 -1.40804398e+00
-5.16505182e-01 -1.05213320e+00 1.50948077e-01 1.02159095e+00
8.60667348e-01 -1.81145847e-01 1.09618604e-01 8.95538390e-01
2.45035738e-02 -3.49596888e-01 -3.78565103e-01 -5.77658236e-01
-9.64411870e-02 -2.11315267e-02 4.91928667e-01 6.19751632e-01
6.27651632e-01 8.55304778e-01 -8.10243562e-02 1.60458162e-01
2.05560163e-01 2.42679641e-02 5.90677738e-01 -1.37619472e+00
-1.00417351e-02 -7.58228242e-01 4.22554523e-01 -1.08644927e+00
1.75922453e-01 -7.19587326e-01 4.31652039e-01 -1.78130376e+00
1.72614232e-02 5.14069915e-01 4.19286862e-02 6.00882888e-01
-4.85727042e-01 -1.46018818e-01 8.39127600e-02 5.47151342e-02
-6.73669636e-01 3.55381429e-01 1.00413990e+00 -1.82417095e-01
-4.24425960e-01 2.45709885e-02 -1.03465927e+00 3.83029550e-01
3.63996774e-01 -4.92930144e-01 -2.15867132e-01 -5.78116715e-01
4.61822867e-01 3.12183678e-01 1.37309253e-01 -8.08764875e-01
5.55828810e-01 -4.61296409e-01 7.17282653e-01 -6.23986304e-01
-4.31545347e-01 -6.65883766e-03 -2.24924833e-01 5.24837315e-01
-1.29052532e+00 5.18407166e-01 4.64259565e-01 1.97771519e-01
9.45630819e-02 -4.41028744e-01 6.39204562e-01 -5.04172087e-01
-2.56066889e-01 -1.64898813e-01 -9.96691406e-01 4.40068170e-02
5.68958700e-01 -2.80716509e-01 -3.05646688e-01 -5.29303551e-01
-5.33477306e-01 3.59219134e-01 3.24293464e-01 7.49874175e-01
6.92772269e-01 -1.00082481e+00 -8.13203752e-01 1.02602303e-01
1.93821058e-01 -4.63520661e-02 -1.30142793e-01 2.11460456e-01
-6.97589219e-01 5.96165001e-01 -2.06961125e-01 -1.63422674e-01
-1.09219074e+00 6.41308248e-01 7.05757439e-02 -9.00901780e-02
-5.76227725e-01 9.13511813e-01 -4.20140363e-02 -3.72170448e-01
4.65237767e-01 -6.01741314e-01 -1.71869159e-01 -1.64655268e-01
1.24544585e+00 6.01212718e-02 1.13166712e-01 2.75659531e-01
-2.63201654e-01 5.01229405e-01 -5.87497175e-01 -5.57689428e-01
1.00480807e+00 2.75341243e-01 7.70994946e-02 6.28499448e-01
6.84348166e-01 -2.43822765e-02 -7.42522061e-01 -1.99944988e-01
2.00551659e-01 1.58082128e-01 -3.17390770e-01 -1.46756136e+00
-2.20335528e-01 8.81164730e-01 -2.71644801e-01 2.18587160e-01
7.06611216e-01 -3.53502035e-01 5.23274958e-01 9.35722113e-01
-3.09953749e-01 -1.39423847e+00 3.32898021e-01 8.45910668e-01
1.32487333e+00 -8.16824317e-01 5.38992509e-02 9.91824456e-03
-8.63577247e-01 1.53095722e+00 7.21088886e-01 1.57392342e-02
-8.41232613e-02 2.07384408e-01 2.56460458e-01 -3.68156619e-02
-1.22528946e+00 5.91346920e-01 4.17984962e-01 4.94573116e-01
6.17416143e-01 -1.30720451e-01 -3.17614883e-01 8.09324503e-01
-4.57367539e-01 6.53651714e-01 1.18315423e+00 8.31989169e-01
-3.33604932e-01 -7.61505246e-01 -3.24461937e-01 1.91134930e-01
-8.20562243e-02 -2.26866305e-01 -9.87822235e-01 2.00442985e-01
-6.30091786e-01 8.54452908e-01 1.45306259e-01 -1.38593361e-01
7.40567565e-01 4.55422193e-01 2.25330845e-01 -9.04396296e-01
-9.33822930e-01 -5.04956722e-01 3.34921367e-02 1.11018009e-01
1.48167223e-01 -2.50987232e-01 -1.34009564e+00 -1.01670571e-01
1.87822923e-01 5.86861610e-01 3.71645212e-01 1.06740236e+00
6.05177045e-01 7.27318108e-01 3.41273487e-01 -5.74533045e-01
-5.99584937e-01 -1.30689180e+00 2.12740824e-02 2.04851136e-01
4.92121518e-01 1.34031311e-01 -2.50260681e-01 5.49374223e-01] | [11.978239059448242, 9.034523963928223] |
354f1a05-d5d1-4862-b777-715861162457 | exploiting-an-external-microphone-for | 2307.04460 | null | https://arxiv.org/abs/2307.04460v1 | https://arxiv.org/pdf/2307.04460v1.pdf | Exploiting an External Microphone for Binaural RTF-Vector-Based Direction of Arrival Estimation for Multiple Speakers | In hearing aid applications, an important objective is to accurately estimate the direction of arrival (DOA) of multiple speakers in noisy and reverberant environments. Recently, we proposed a binaural DOA estimation method, where the DOAs of the speakers are estimated by selecting the directions for which the so-called Hermitian angle spectrum between the estimated relative transfer function (RTF) vector and a database of prototype anechoic RTF vectors is maximized. The RTF vector is estimated using the covariance whitening (CW) method, which requires a computationally complex generalized eigenvalue decomposition. The spatial spectrum is obtained by only considering frequencies where it is likely that one speaker dominates over the other speakers, noise and reverberation. In this contribution, we exploit the availability of an external microphone that is spatially separated from the hearing aid microphones and consider a low-complexity RTF vector estimation method that assumes a low spatial coherence between the undesired components in the external microphone and the hearing aid microphones. Using recordings of two speakers and diffuse-like babble noise in acoustic environments with mild reverberation and low signal-to-noise ratio, simulation results show that the proposed method yields a comparable DOA estimation performance as the CW method at a lower computational complexity. | ['Simon Doclo', 'Daniel Fejgin'] | 2023-07-10 | null | null | null | null | ['direction-of-arrival-estimation'] | ['audio'] | [-2.77692806e-02 -3.28220516e-01 9.81960237e-01 2.69621968e-01
-9.52952385e-01 -4.68845814e-01 1.03693038e-01 -1.12082675e-01
-2.23403379e-01 4.71672088e-01 7.24496901e-01 -2.91647255e-01
-3.20874482e-01 -3.07917178e-01 -3.09822142e-01 -1.05067790e+00
-2.73371518e-01 -2.69566596e-01 9.33985971e-03 -6.74966797e-02
-3.39350224e-01 3.00225258e-01 -1.79613292e+00 -1.04096808e-01
6.68076754e-01 1.08583772e+00 3.42568278e-01 1.08457196e+00
3.99111301e-01 5.69354713e-01 -1.12824595e+00 6.77570924e-02
3.83154243e-01 -6.27821207e-01 2.66725242e-01 -1.73109666e-01
2.19186142e-01 -1.46389723e-01 -1.40665278e-01 1.08710659e+00
1.10933340e+00 3.27262700e-01 6.39297903e-01 -6.28238082e-01
2.95549989e-01 2.59640545e-01 -2.76228249e-01 2.61002481e-01
4.84341145e-01 -1.15660012e-01 7.01731026e-01 -1.08660603e+00
7.82574415e-02 9.59642410e-01 7.28188336e-01 1.59603637e-02
-9.63902056e-01 -6.19568348e-01 -3.81330192e-01 1.52384654e-01
-1.71314156e+00 -6.69885635e-01 1.07204282e+00 -4.02580053e-01
6.69749916e-01 4.70332146e-01 7.00642407e-01 4.75203663e-01
1.48589402e-01 1.81822196e-01 1.09493351e+00 -9.28287208e-01
5.17166257e-01 6.21109158e-02 7.13584572e-02 3.31174165e-01
9.60744247e-02 3.80623162e-01 -5.90280771e-01 -6.28112555e-01
2.37238511e-01 -5.39244533e-01 -9.56395447e-01 -3.57356936e-01
-1.21512687e+00 4.60925728e-01 1.24716744e-01 6.41091764e-01
-6.88351870e-01 -5.30506410e-02 -2.02995390e-01 2.88282007e-01
3.40713888e-01 2.30299518e-01 6.58486262e-02 -5.41337617e-02
-6.85990453e-01 -3.53754424e-02 1.20103228e+00 5.85411429e-01
3.94056171e-01 5.11346519e-01 1.17489278e-01 1.14077497e+00
8.06361198e-01 1.11591470e+00 2.27768674e-01 -5.10217726e-01
4.37896699e-01 -7.54369080e-01 4.88175184e-01 -1.23190725e+00
-2.59661704e-01 -1.16265440e+00 -6.33764684e-01 3.29844654e-01
5.80016613e-01 -6.20773792e-01 -6.20998859e-01 1.63382757e+00
7.06590652e-01 4.50667113e-01 2.23493308e-01 1.03511488e+00
2.46588737e-01 6.70077860e-01 -5.52412033e-01 -7.19641984e-01
1.19293177e+00 -4.60631251e-01 -1.14549673e+00 -2.64664084e-01
5.69559447e-02 -1.51346040e+00 6.23645723e-01 8.26892614e-01
-8.84486973e-01 -5.11417270e-01 -1.21763289e+00 6.19635403e-01
1.53490350e-01 -5.44010475e-02 -1.38619497e-01 1.23752868e+00
-8.73803496e-01 -1.39234051e-01 -6.98884666e-01 3.01631182e-01
-5.40685356e-01 -2.89064497e-01 -3.44083905e-02 -1.64793000e-01
-9.84951079e-01 6.29331708e-01 -5.40564716e-01 5.33630610e-01
-7.99918950e-01 -8.04532826e-01 -6.15890026e-01 7.13540101e-03
-7.41793588e-02 -4.59034830e-01 1.21495771e+00 -7.08768129e-01
-1.58926463e+00 9.71543640e-02 -5.57107329e-01 -6.67862073e-02
3.57427210e-01 -4.89896625e-01 -1.04438150e+00 8.76512155e-02
-9.78109539e-02 -7.60753214e-01 1.16439891e+00 -1.36392379e+00
-3.81553292e-01 -2.93398827e-01 -6.90671682e-01 2.88504571e-01
-1.03902921e-01 -1.55445457e-01 1.03563800e-01 -7.49351203e-01
7.99017847e-01 -6.32622898e-01 -1.08324125e-01 -2.23528579e-01
-5.03000319e-01 2.97799110e-01 1.46038920e-01 -9.03170586e-01
1.38005388e+00 -2.58954763e+00 -3.00593913e-01 6.52222097e-01
-4.13448876e-03 1.84150487e-01 6.96342587e-02 6.60280406e-01
-2.04602569e-01 -8.41435254e-01 7.53059834e-02 -2.29797766e-01
-1.87276140e-01 -3.64180386e-01 -4.08617824e-01 8.85877848e-01
-5.36630630e-01 -3.10546488e-01 -8.40301096e-01 -1.00408494e-01
1.91197231e-01 9.04451489e-01 -5.07906079e-01 6.57003522e-01
6.47334516e-01 3.72760803e-01 -1.84970930e-01 2.68992931e-01
9.85118568e-01 5.92874646e-01 -6.92429617e-02 -3.07624638e-01
-3.40132117e-01 3.15543771e-01 -2.03347969e+00 9.05893564e-01
-1.05553842e+00 8.11403751e-01 8.18005443e-01 -7.24208593e-01
1.11942244e+00 6.05940700e-01 2.01166019e-01 -3.27931851e-01
-7.15203360e-02 7.54703939e-01 3.35795969e-01 -5.56464016e-01
-2.30363682e-01 -4.20020640e-01 4.31899071e-01 3.17851424e-01
1.10745899e-01 -1.72212571e-01 -2.91577190e-01 -2.34278038e-01
1.11184442e+00 -5.94994605e-01 6.21587396e-01 -3.76799703e-01
6.46933377e-01 -8.17638814e-01 4.61222678e-01 6.95845127e-01
-1.29281148e-01 5.75280428e-01 -2.71292031e-01 1.76568434e-01
-6.13129258e-01 -1.34057140e+00 -1.35281727e-01 8.08693290e-01
-1.32931381e-01 -1.18194319e-01 -6.81373417e-01 3.76709133e-01
-2.14937434e-01 9.13617909e-01 6.33819476e-02 8.75381678e-02
-7.17378497e-01 -2.19407633e-01 2.39415467e-01 -1.08866051e-01
3.37484092e-01 -1.19166009e-01 -3.73924136e-01 5.47062576e-01
-4.01698023e-01 -8.67885947e-01 -5.31663179e-01 9.16972309e-02
-4.31752533e-01 -8.43414485e-01 -1.12200177e+00 -7.34114110e-01
5.22386193e-01 5.42213857e-01 6.06725633e-01 -6.58156633e-01
-3.06747139e-01 8.34546924e-01 -1.33291215e-01 -8.25213492e-01
-3.04728985e-01 -1.02780867e+00 3.68798941e-01 6.37677014e-01
-3.42634358e-02 -9.84586954e-01 -9.51017737e-01 6.70440257e-01
-3.93075436e-01 -3.55484784e-01 1.22430153e-01 7.24918067e-01
2.08199963e-01 5.28945327e-01 4.85933959e-01 1.51344374e-01
8.07558358e-01 -3.69623125e-01 -6.44465506e-01 -1.22175731e-01
-1.08049132e-01 -3.41872275e-01 7.15633035e-01 -6.28474355e-01
-1.35959053e+00 -7.79589564e-02 -3.01059335e-01 -6.57877550e-02
-1.91786617e-01 3.22307974e-01 -4.32277411e-01 1.02324128e-01
9.89572287e-01 3.81041169e-01 -2.56572902e-01 -7.57925630e-01
3.74506712e-02 1.00223982e+00 5.39064109e-01 1.40772415e-02
1.02474558e+00 3.58840168e-01 5.33905029e-02 -1.51194334e+00
-2.98701674e-01 -9.36355948e-01 -1.09349601e-02 -4.31781828e-01
4.24425572e-01 -8.67875814e-01 -5.63641310e-01 7.00903833e-01
-1.13078368e+00 2.47963611e-02 5.99001795e-02 1.62709129e+00
-3.02680194e-01 4.04183805e-01 -1.42898895e-02 -1.64559460e+00
-1.60781071e-01 -8.17377627e-01 4.11664099e-01 -1.05498940e-01
-2.20689252e-01 -7.15985894e-01 3.03004026e-01 9.25997049e-02
7.28630662e-01 9.47101638e-02 6.69625580e-01 -3.34039122e-01
-3.08977008e-01 -4.21386838e-01 6.05605960e-01 5.59139073e-01
4.57505286e-01 -5.75197816e-01 -1.25080264e+00 -2.68523455e-01
7.81815767e-01 4.69079256e-01 8.46473798e-02 1.11020494e+00
2.17043623e-01 -3.79793644e-01 -1.86340779e-01 5.08548677e-01
1.44099152e+00 5.18378198e-01 3.80789459e-01 -2.00622410e-01
1.87050059e-01 6.79182768e-01 4.90549415e-01 5.85164845e-01
-1.30137041e-01 6.22807086e-01 1.54346138e-01 -1.59046993e-01
-4.46899056e-01 -8.35561454e-02 1.90220132e-01 1.15944219e+00
3.23566012e-02 -4.89080042e-01 -7.36132443e-01 7.09300041e-01
-1.19848156e+00 -6.57001257e-01 -2.54927039e-01 2.82281423e+00
4.81054902e-01 -2.23153263e-01 -3.56512964e-02 6.23162568e-01
6.93311453e-01 5.26222549e-02 -3.30678701e-01 -8.78235400e-02
-4.23176885e-02 3.93990070e-01 3.67919058e-01 1.18062580e+00
-5.99116564e-01 -9.59187001e-02 5.89862156e+00 6.49512529e-01
-1.05162680e+00 1.31916225e-01 -1.03087788e-02 -2.47207228e-02
-2.96586096e-01 -4.79471982e-01 -4.72680509e-01 2.21213818e-01
8.71322989e-01 -2.55151302e-01 4.76383209e-01 6.35122836e-01
5.18357158e-01 -2.88272113e-01 -8.15841317e-01 1.17406631e+00
1.77581459e-01 -5.07878840e-01 -7.39642382e-01 5.15809730e-02
3.96969438e-01 -1.18933298e-01 2.34035805e-01 -2.12479442e-01
-9.88188386e-02 -6.78795576e-01 6.90661490e-01 5.31662107e-01
2.63494670e-01 -6.43187165e-01 6.16505980e-01 4.07868385e-01
-1.27292323e+00 -2.78269142e-01 -2.34837055e-01 -5.10388985e-02
3.12853009e-01 1.41072321e+00 -1.18314707e+00 2.18047678e-01
5.58841348e-01 -1.84480697e-01 4.00833637e-01 1.70223260e+00
-3.85263592e-01 1.10405946e+00 -7.35630989e-01 -1.94488704e-01
-6.89247176e-02 -3.85119170e-02 1.29354942e+00 1.16536200e+00
9.46481168e-01 3.02056372e-01 -2.61953473e-01 4.03517246e-01
3.93048316e-01 4.15921092e-01 -5.71047127e-01 5.86627364e-01
6.46717548e-01 6.66913390e-01 -2.73436606e-01 1.71005890e-01
-1.93601698e-01 5.14106691e-01 -8.52042973e-01 1.04504800e+00
-3.87464792e-01 -8.64136159e-01 8.54104102e-01 3.15156579e-01
4.47942287e-01 -4.85501170e-01 6.70235455e-02 -6.79361880e-01
3.56367171e-01 -7.67507672e-01 5.65185398e-02 -8.42590988e-01
-9.34127569e-01 7.38238931e-01 -2.96466023e-01 -1.76631331e+00
-4.95024830e-01 -4.52882499e-01 -5.69908917e-01 1.47819614e+00
-1.26691830e+00 -4.73637223e-01 1.46496817e-01 7.64617085e-01
1.76233873e-01 -1.98540226e-01 9.34044182e-01 4.78022635e-01
1.53397530e-01 6.05013788e-01 7.74236023e-01 -1.73442066e-01
6.01615906e-01 -1.06642175e+00 -3.58463138e-01 8.77951443e-01
-4.81331758e-02 8.59788775e-01 1.48694015e+00 -2.13357583e-01
-1.27226257e+00 -6.31874681e-01 1.06773984e+00 2.25450084e-01
4.95994210e-01 -5.24066567e-01 -5.50251901e-01 6.63377196e-02
7.47399479e-02 1.06197104e-01 1.15475631e+00 -6.69731796e-02
-1.78226858e-01 -6.25502586e-01 -9.44607794e-01 4.64365542e-01
5.61477900e-01 -5.57592392e-01 -5.72045505e-01 3.09200853e-01
2.18994319e-01 -1.89745352e-01 -4.82910961e-01 -2.40032468e-02
6.35382175e-01 -9.81491864e-01 1.15036452e+00 2.51244158e-01
-3.65733594e-01 -6.64242029e-01 -6.44821644e-01 -1.75126016e+00
-1.29240468e-01 -1.22064042e+00 4.83982041e-02 1.09135234e+00
3.05919796e-01 -1.07152998e+00 1.43399000e-01 1.01534829e-01
2.38656420e-02 -2.05929160e-01 -1.32889032e+00 -9.62069392e-01
-3.12523544e-01 -6.86753690e-01 1.93727031e-01 4.55169201e-01
8.15690681e-02 3.36687028e-01 -5.65294027e-01 9.36576724e-01
1.05409443e+00 -1.55236766e-01 6.37206972e-01 -9.75673437e-01
-7.16586530e-01 3.29073668e-02 -3.80239874e-01 -1.36841524e+00
-3.92312378e-01 -2.52129912e-01 5.64851522e-01 -1.42085290e+00
-8.08447838e-01 -2.81197727e-01 -3.82033944e-01 -4.79960471e-01
-1.52874410e-01 -9.40387845e-02 -1.22125939e-01 -1.97547957e-01
3.65886480e-01 4.08037245e-01 9.10374582e-01 2.13926192e-02
-5.04139781e-01 9.49288368e-01 -2.40350649e-01 9.59343970e-01
2.03637406e-01 -5.75565159e-01 -5.47278345e-01 -2.34681919e-01
1.94483809e-02 5.40864229e-01 7.94515237e-02 -1.39888585e+00
4.29370642e-01 3.31321478e-01 -6.44285008e-02 -5.68890035e-01
6.28299773e-01 -1.03043830e+00 4.40669328e-01 5.86707175e-01
-1.24796890e-01 -5.89355052e-01 1.00371741e-01 9.30674136e-01
-3.63471895e-01 -8.37436840e-02 8.56258154e-01 1.87364087e-01
1.05921715e-01 -3.87863100e-01 -1.03106034e+00 -4.42814201e-01
5.43969274e-01 -3.52863260e-02 1.26583785e-01 -1.27519703e+00
-7.21240878e-01 -5.83824575e-01 -5.46518803e-01 -1.18464917e-01
5.99789202e-01 -1.07699442e+00 -8.97624493e-01 2.26375341e-01
-1.63923919e-01 -4.64466304e-01 5.37208259e-01 6.96274579e-01
-2.29806080e-01 4.77433950e-01 2.95613199e-01 -6.93581581e-01
-1.51633179e+00 2.52776265e-01 7.23179102e-01 9.96374711e-02
-3.75271827e-01 1.08086908e+00 4.75404173e-01 -9.00541991e-02
3.65158647e-01 -2.49751106e-01 -3.09559256e-01 -1.47424862e-01
9.12656724e-01 7.05376625e-01 3.88707191e-01 -7.23732412e-01
-2.71876574e-01 8.44535291e-01 6.50668085e-01 -8.18564355e-01
1.03260791e+00 -5.72556674e-01 9.27986428e-02 4.65640634e-01
1.57981336e+00 1.17365563e+00 -6.17518842e-01 -4.89854634e-01
-3.66792321e-01 -7.48137236e-01 3.45649600e-01 -8.57160330e-01
-6.23197615e-01 1.06206834e+00 1.31971538e+00 2.97210574e-01
1.48739076e+00 -2.24696428e-01 5.23433864e-01 3.51048619e-01
5.23627579e-01 -8.08782041e-01 2.07199678e-01 8.31660926e-02
9.33613539e-01 -4.95101929e-01 -3.59173328e-01 -5.21798074e-01
-7.21810609e-02 1.01845384e+00 -1.93497222e-02 9.43889469e-02
1.19754827e+00 2.69372463e-01 7.75457084e-01 2.34576896e-01
-1.14539109e-01 -1.45390779e-01 2.77100265e-01 9.96940434e-01
4.04252738e-01 2.13973999e-01 -2.56262004e-01 4.70268488e-01
-6.81950629e-01 -5.45674384e-01 5.21770298e-01 6.68100297e-01
-7.03111768e-01 -6.06242776e-01 -1.23076546e+00 -3.52174416e-02
-6.21611595e-01 -2.68241763e-01 1.92797646e-01 1.05125187e-02
-7.55734518e-02 1.69277430e+00 2.68561067e-03 -1.46837756e-01
9.37623024e-01 1.38556316e-01 6.68474883e-02 -2.12625965e-01
-2.66529411e-01 9.65762258e-01 2.82380342e-01 -1.94070652e-01
-4.89097089e-02 -5.17855942e-01 -7.60665357e-01 1.62866190e-01
-6.17288768e-01 5.34501255e-01 1.05215752e+00 5.82766294e-01
1.46940410e-01 5.99573374e-01 1.17383957e+00 -4.74965543e-01
-5.51914692e-01 -9.73865986e-01 -1.15155220e+00 -8.84684771e-02
1.01473069e+00 -3.74326825e-01 -9.35535073e-01 -6.23690076e-02] | [15.157198905944824, 5.7487382888793945] |
b9975938-9fdf-479f-8ecc-cfb03798692b | the-circor-digiscope-dataset-from-murmur | 2108.00813 | null | https://arxiv.org/abs/2108.00813v2 | https://arxiv.org/pdf/2108.00813v2.pdf | The CirCor DigiScope Dataset: From Murmur Detection to Murmur Classification | Cardiac auscultation is one of the most cost-effective techniques used to detect and identify many heart conditions. Computer-assisted decision systems based on auscultation can support physicians in their decisions. Unfortunately, the application of such systems in clinical trials is still minimal since most of them only aim to detect the presence of extra or abnormal waves in the phonocardiogram signal, i.e., only a binary ground truth variable (normal vs abnormal) is provided. This is mainly due to the lack of large publicly available datasets, where a more detailed description of such abnormal waves (e.g., cardiac murmurs) exists. To pave the way to more effective research on healthcare recommendation systems based on auscultation, our team has prepared the currently largest pediatric heart sound dataset. A total of 5282 recordings have been collected from the four main auscultation locations of 1568 patients, in the process, 215780 heart sounds have been manually annotated. Furthermore, and for the first time, each cardiac murmur has been manually annotated by an expert annotator according to its timing, shape, pitch, grading, and quality. In addition, the auscultation locations where the murmur is present were identified as well as the auscultation location where the murmur is detected more intensively. Such detailed description for a relatively large number of heart sounds may pave the way for new machine learning algorithms with a real-world application for the detection and analysis of murmur waves for diagnostic purposes. | ['Miguel T. Coimbra', 'Gari D Clifford', 'Reza Sameni', 'Ali Bahrami Rad', 'Andoni Elola', 'Thiago Tavares', 'Thamine Hatem', 'Sandra Mattos', 'Alipio Jorge', 'Carlos Ferreira', 'Cristina Oliveira', 'Marcelo Nogueira', 'Paulo Dias Costa', 'Francesco Renna', 'Jorge Oliveira'] | 2021-08-02 | null | null | null | null | ['predict-clinical-outcome', 'classify-murmurs'] | ['time-series', 'time-series'] | [ 1.97945535e-01 1.00068405e-01 2.10783333e-02 -7.05313236e-02
-5.85247576e-01 -7.72553563e-01 -4.05401617e-01 5.86251259e-01
-1.10377856e-01 6.06060028e-01 8.34415760e-03 -5.38268089e-01
-2.95624703e-01 -5.27643502e-01 -9.94127840e-02 -7.80857325e-01
-2.66823202e-01 6.06119514e-01 2.20296741e-01 2.18523249e-01
2.29642894e-02 3.21741134e-01 -1.33285189e+00 -3.16831544e-02
8.21464658e-01 8.78182232e-01 3.39374542e-01 1.00838447e+00
2.17135742e-01 5.42794168e-01 -9.21519399e-01 -3.43464762e-01
-1.74600884e-01 -9.44896460e-01 -4.43862468e-01 -1.35341331e-01
2.38711149e-01 -3.35264266e-01 1.36460423e-01 8.18855643e-01
1.08629203e+00 -4.04713094e-01 4.29889262e-01 -6.27419829e-01
2.01998185e-02 7.14320421e-01 -1.05087809e-01 5.86362481e-01
2.74790615e-01 -6.52057081e-02 8.95344913e-01 -4.78373945e-01
4.26593512e-01 2.59967953e-01 8.49148095e-01 3.32880795e-01
-8.94363344e-01 -4.37424749e-01 -6.01093113e-01 1.65762857e-01
-1.12229967e+00 -3.48952264e-02 8.13430965e-01 -6.39736593e-01
3.48307878e-01 6.06772423e-01 1.00433969e+00 3.96762371e-01
-9.40837711e-02 2.72883028e-01 9.96751845e-01 -5.73371291e-01
2.46707782e-01 6.77611306e-02 2.23485883e-02 7.36105025e-01
5.36648214e-01 -1.94707453e-01 -7.88928494e-02 -3.13347310e-01
6.42648041e-01 6.82361647e-02 -6.93037033e-01 1.28956875e-02
-1.29498398e+00 6.01278067e-01 -1.01842947e-01 7.83779979e-01
-4.54758227e-01 -3.30690235e-01 5.32735407e-01 1.74451888e-01
1.70013681e-01 8.32079887e-01 -6.38139367e-01 -6.63053513e-01
-1.02196944e+00 -1.28018195e-02 8.81069183e-01 3.86761189e-01
2.38364071e-01 -3.63516957e-02 -1.26691446e-01 1.04725230e+00
6.73157489e-03 6.37050152e-01 5.70104003e-01 -1.07151580e+00
2.91985244e-01 3.55759203e-01 7.19808191e-02 -1.23328042e+00
-6.19009972e-01 -5.61448038e-01 -8.72578979e-01 -1.48281649e-01
7.14676142e-01 -4.41970974e-01 -2.68584251e-01 1.16595614e+00
3.86729807e-01 2.10069954e-01 -1.46071807e-01 1.13014781e+00
1.08492076e+00 1.13249242e-01 -2.12015674e-01 -4.96838272e-01
1.55824554e+00 -4.25662875e-01 -9.95757759e-01 3.20672512e-01
8.15915942e-01 -9.21026886e-01 6.28666401e-01 7.69229352e-01
-1.05265188e+00 -4.59602833e-01 -9.03145194e-01 7.52109766e-01
-1.18052274e-01 4.05092984e-01 3.42814475e-01 9.80705857e-01
-5.38408697e-01 8.45928848e-01 -8.68124545e-01 -7.64357764e-03
2.35178456e-01 4.03909534e-02 -1.56055808e-01 2.60979772e-01
-1.34948552e+00 7.22387493e-01 1.21201813e-01 2.72008210e-01
-2.40269199e-01 -6.41779661e-01 -7.43468642e-01 -1.13286927e-01
2.38568172e-01 -3.25463563e-01 1.09013855e+00 -4.88897264e-01
-1.22181356e+00 8.94811988e-01 5.84117956e-02 -3.02184641e-01
4.00600016e-01 1.21760756e-01 -6.23535097e-01 5.59490085e-01
-2.38538850e-02 -2.08569095e-02 7.14517474e-01 -6.11264288e-01
-6.48995221e-01 -2.77623415e-01 -2.78744698e-01 -7.06388503e-02
-2.10881338e-01 1.42489105e-01 -1.59613833e-01 -8.46976876e-01
4.00316268e-01 -7.11654127e-01 1.62254199e-01 -1.54966980e-01
-3.08865637e-01 -1.50582805e-01 5.17432213e-01 -1.04331696e+00
1.74918437e+00 -2.27194095e+00 -2.34458044e-01 1.04739934e-01
6.93520725e-01 7.42405951e-01 3.93410295e-01 2.62828141e-01
-1.14650279e-01 2.78539985e-01 -2.24043652e-01 1.96137875e-01
-2.89218843e-01 1.20638475e-01 -8.44981372e-02 5.40015936e-01
-1.34609431e-01 5.78288138e-01 -1.01207089e+00 -4.81285512e-01
5.00788152e-01 3.98690909e-01 -2.76454031e-01 4.35708761e-01
2.94527292e-01 7.63921201e-01 -2.62485117e-01 4.94346052e-01
3.73387069e-01 -2.07861945e-01 2.41083309e-01 -1.27069235e-01
-2.10362568e-01 4.24612433e-01 -1.06360626e+00 1.32591546e+00
-2.28582218e-01 7.28898466e-01 -7.47772828e-02 -9.60337579e-01
1.00520611e+00 1.01584196e+00 8.26010168e-01 -9.28399861e-02
7.35113993e-02 5.14705181e-01 7.50458241e-01 -1.06083500e+00
-2.27892369e-01 -1.21088885e-01 3.21063340e-01 5.61853707e-01
-3.36490154e-01 -2.25066960e-01 3.07610184e-01 -2.82645017e-01
1.16598320e+00 -2.45301113e-01 4.19486612e-01 -1.46227181e-02
4.98052478e-01 -1.49202973e-01 6.86683118e-01 7.06854165e-01
-4.30873871e-01 1.13596404e+00 4.46594387e-01 -5.99553347e-01
-6.83691204e-01 -4.74956781e-01 -6.31012082e-01 5.15802026e-01
-3.27607989e-01 -3.57748955e-01 -7.55427599e-01 -3.72614503e-01
-2.32080281e-01 1.95409670e-01 -3.77775341e-01 9.34614539e-02
-5.90304375e-01 -6.94240510e-01 8.18543792e-01 4.16514724e-01
2.56451279e-01 -1.25230372e+00 -1.28225541e+00 4.69808400e-01
-4.48906183e-01 -7.98400998e-01 -2.32138813e-01 -2.97579132e-02
-8.27679634e-01 -1.41703856e+00 -1.14944923e+00 -6.54630661e-01
2.49876723e-01 -1.49279088e-01 1.00736380e+00 3.36643845e-01
-7.90357947e-01 1.36954382e-01 -6.11778915e-01 -6.48572087e-01
-6.38253152e-01 -1.04361720e-01 -1.45407930e-01 -1.42814413e-01
-6.17382228e-02 -5.81955075e-01 -8.51033270e-01 3.02711815e-01
-5.11313915e-01 -1.91948503e-01 2.96498239e-01 5.58026969e-01
4.37430769e-01 2.45296091e-01 8.29011321e-01 -9.42052245e-01
6.42638981e-01 -3.72730047e-01 -3.39284182e-01 1.18703730e-02
-6.31228507e-01 -5.77006042e-01 6.09693110e-01 -2.89804935e-01
-4.35966223e-01 -3.36823881e-01 -3.56405914e-01 -3.49501014e-01
-6.56286120e-01 5.87868929e-01 3.68752033e-01 2.75001794e-01
6.77015543e-01 9.16931704e-02 -7.44464202e-03 -5.69566190e-01
-2.49504313e-01 8.71205032e-01 6.11002982e-01 -1.24297544e-01
3.94657314e-01 4.88818511e-02 7.85134360e-02 -8.03970098e-01
-7.28482723e-01 -6.27561927e-01 -5.55950105e-01 -2.79543906e-01
1.15101182e+00 -3.59584779e-01 -6.39078438e-01 4.74443287e-01
-9.60776627e-01 2.21159831e-02 -3.16693902e-01 8.44210863e-01
-2.14068979e-01 6.38659239e-01 -5.06776154e-01 -1.12395370e+00
-5.49198627e-01 -9.19268012e-01 6.17140830e-01 1.70680046e-01
-6.87497973e-01 -1.02130651e+00 2.24537879e-01 5.48097908e-01
3.84155273e-01 5.53181589e-01 1.09102488e+00 -7.85977900e-01
1.16901211e-01 -5.89745522e-01 2.23263770e-01 4.07673717e-01
5.58077633e-01 1.66602656e-01 -9.08775330e-01 -1.18897244e-01
1.35567352e-01 9.21068564e-02 2.77047783e-01 6.37886047e-01
1.24265182e+00 1.00345790e-01 7.99955055e-02 2.63286591e-01
9.10426319e-01 7.51367271e-01 3.25042009e-01 -2.21137807e-01
5.92717230e-01 6.53945565e-01 5.95939875e-01 4.35860246e-01
1.94226712e-01 4.60023522e-01 3.14335108e-01 -4.25251633e-01
-1.04671285e-01 1.46168536e-02 -2.08825856e-01 1.14103508e+00
-5.09731293e-01 7.63234198e-02 -1.08281386e+00 4.67474103e-01
-1.47267389e+00 -8.30565631e-01 -6.41397119e-01 2.57611346e+00
8.30078244e-01 -5.40082436e-03 2.86700517e-01 7.93382764e-01
9.69311535e-01 -6.35852367e-02 -3.59212458e-01 -3.38218927e-01
2.69286763e-02 4.08912539e-01 7.80636221e-02 9.73491669e-02
-1.03319371e+00 -6.04491495e-02 6.13988924e+00 2.62002021e-01
-1.52718186e+00 -3.99018787e-02 4.94021118e-01 4.14500117e-01
1.44934505e-01 -3.01351666e-01 -3.20736796e-01 8.18417847e-01
6.18940890e-01 3.00219089e-01 1.28567830e-01 5.62447608e-01
4.61712956e-01 -2.04541117e-01 -9.45343614e-01 1.22808373e+00
-3.71115692e-02 -1.33770752e+00 -9.25889194e-01 -1.81967035e-01
3.41361582e-01 -2.91547567e-01 -2.84062356e-01 -8.35495368e-02
-6.94341838e-01 -7.32687533e-01 3.57214600e-01 5.14698267e-01
1.05709600e+00 -3.81865591e-01 1.07499194e+00 4.50871408e-01
-9.09506083e-01 1.09009512e-01 -4.74507129e-03 -7.53233433e-02
6.15806261e-04 7.87978590e-01 -1.06424320e+00 4.42374706e-01
5.94305515e-01 6.33189023e-01 -2.45688722e-01 1.46960604e+00
-4.22036290e-01 1.30910027e+00 -2.50386417e-01 4.28079925e-02
-4.49138999e-01 -2.30719745e-02 7.01868713e-01 9.73814905e-01
3.81523967e-01 3.33111167e-01 -5.79749700e-03 7.36626565e-01
3.59666407e-01 3.64262879e-01 -5.30331612e-01 -1.87067352e-02
6.05160594e-01 1.26001859e+00 -8.82143319e-01 -4.05721933e-01
-3.83462250e-01 2.75849700e-01 -3.22535932e-01 1.54964715e-01
-5.87673128e-01 -7.01058686e-01 9.31286290e-02 5.40864646e-01
6.08088709e-02 3.74479741e-01 -4.86473322e-01 -8.60152245e-01
-3.30316392e-03 -1.04894269e+00 5.60595810e-01 -7.06922114e-01
-1.00480759e+00 6.53962493e-01 -1.82349727e-01 -1.26262689e+00
-6.31130040e-01 -3.27407956e-01 -8.63434911e-01 1.03808939e+00
-1.08292234e+00 -1.55768022e-01 -2.91980684e-01 5.91727830e-02
3.22922975e-01 -6.19547740e-02 1.26987922e+00 5.80330670e-01
-5.43791711e-01 3.69444489e-01 -4.37904477e-01 3.72395962e-01
6.50094271e-01 -1.44700694e+00 -2.13266686e-01 4.94656414e-01
4.92202081e-02 5.96591055e-01 7.33654976e-01 -4.74561483e-01
-8.36608589e-01 -7.10073173e-01 1.01837599e+00 -2.96919197e-01
4.47185874e-01 2.45606869e-01 -9.87343132e-01 4.15330258e-04
-1.32999107e-01 9.61317718e-02 9.51951802e-01 -1.50362030e-01
2.71877974e-01 -1.55885994e-01 -8.90373230e-01 4.66489382e-02
3.37819189e-01 -2.69362122e-01 -6.22296631e-01 2.73244172e-01
1.53340220e-01 -6.47376180e-01 -1.29803813e+00 6.49373353e-01
6.35360479e-01 -9.19873476e-01 5.92005372e-01 -2.12677732e-01
4.31346953e-01 -2.89932966e-01 3.85913163e-01 -1.09741449e+00
-1.86379299e-01 -9.44662035e-01 -2.75032427e-02 1.06835592e+00
3.37778926e-01 -8.82177949e-01 5.55987418e-01 5.07534921e-01
-1.30663961e-01 -1.12459135e+00 -7.01913178e-01 -2.17305422e-01
-3.44173104e-01 -4.18446183e-01 4.92891997e-01 1.06430078e+00
1.74543157e-01 7.84260482e-02 -2.64109939e-01 -1.59079451e-02
3.13796282e-01 4.39003110e-01 2.72732556e-01 -1.41230357e+00
-5.76557100e-01 -4.49425071e-01 -4.67980862e-01 -4.76842076e-01
-4.84213412e-01 -6.27366841e-01 -9.93320495e-02 -1.37950516e+00
3.95866781e-02 -8.29227567e-01 -3.25043619e-01 4.36490208e-01
-3.63508791e-01 4.34837669e-01 -2.83207279e-02 1.36345446e-01
-3.54718827e-02 -3.37789208e-01 1.51120365e+00 1.54062137e-01
-3.21015209e-01 6.71580255e-01 -4.14659113e-01 9.75938022e-01
8.59152079e-01 -6.54974043e-01 -3.75335842e-01 7.95455277e-02
1.51981920e-01 7.31570005e-01 2.20549107e-01 -7.44353771e-01
-8.67154077e-02 1.42516285e-01 1.07082896e-01 -6.39942229e-01
6.22381754e-02 -7.21369863e-01 1.81852058e-01 7.55470932e-01
-2.42526546e-01 -5.98797388e-03 -2.23369554e-01 1.67863712e-01
-3.83384943e-01 -6.46368623e-01 4.96212631e-01 -1.40585497e-01
-5.96594028e-02 1.58367246e-01 -6.98550045e-01 3.63600940e-01
7.58835196e-01 -3.34931016e-01 -9.97230560e-02 -4.11693007e-01
-1.16220999e+00 -1.81562185e-01 -2.86880173e-02 5.06604165e-02
5.80540478e-01 -8.83019626e-01 -8.42376411e-01 2.17524216e-01
1.64438486e-02 9.72606614e-02 3.23935181e-01 1.43727231e+00
-8.23036194e-01 4.22237009e-01 4.24286090e-02 -7.45526373e-01
-1.49301243e+00 8.54452252e-02 5.08989751e-01 -1.56593248e-01
-8.45605850e-01 5.61108887e-01 -2.06377804e-01 1.01508565e-01
4.90405530e-01 -5.77280462e-01 -6.59684420e-01 1.42569035e-01
4.95600373e-01 5.87152421e-01 3.14524531e-01 -3.97569805e-01
-2.55812645e-01 5.22342741e-01 4.43689257e-01 1.98531300e-01
1.07437563e+00 3.01418863e-02 -2.00692102e-01 5.54278255e-01
7.20546007e-01 2.31648907e-01 -6.31844044e-01 1.73238188e-01
-2.67038852e-01 -4.47275817e-01 -3.21514904e-02 -8.65031958e-01
-1.14583433e+00 1.13256848e+00 7.24064469e-01 7.95519233e-01
1.23126805e+00 -1.74579933e-01 7.33414114e-01 8.42794925e-02
2.69172758e-01 -7.92690098e-01 -1.03209838e-01 -2.46417493e-01
5.78060448e-01 -1.04782438e+00 -2.80786693e-01 -4.97461140e-01
-6.00225449e-01 1.24146664e+00 4.13159057e-02 1.08495831e-01
7.77106941e-01 2.93968111e-01 4.66449082e-01 -2.91587293e-01
-3.67345035e-01 6.67108893e-02 3.51011038e-01 6.26682699e-01
9.06089127e-01 4.53470588e-01 -8.04954827e-01 7.19443381e-01
-3.02983075e-01 1.79521216e-03 5.01723588e-01 6.47351861e-01
-4.30385947e-01 -1.07210350e+00 -4.27001655e-01 7.35521376e-01
-1.11124122e+00 -8.75599906e-02 -1.28465980e-01 3.51354539e-01
5.53132772e-01 1.27454484e+00 -2.41471589e-01 -1.42364157e-02
4.09088641e-01 1.37880519e-01 4.26298141e-01 -7.89020836e-01
-7.66313076e-01 3.33715558e-01 5.26758097e-02 2.19909493e-02
-2.89629310e-01 -6.86434865e-01 -1.27916634e+00 2.91136533e-01
-4.41782862e-01 4.89722580e-01 5.01217604e-01 7.30965495e-01
3.57661918e-02 6.72403812e-01 6.14784300e-01 -2.14575902e-01
-5.20569384e-01 -1.05589843e+00 -8.62784386e-01 4.36285794e-01
4.17310357e-01 -3.24057102e-01 -4.52130377e-01 1.79954484e-01] | [14.299718856811523, 3.313336133956909] |
bcf97bf1-8371-4d61-a02e-adb2e4299289 | multiple-object-tracking-in-recent-times-a | 2209.04796 | null | https://arxiv.org/abs/2209.04796v1 | https://arxiv.org/pdf/2209.04796v1.pdf | Multiple Object Tracking in Recent Times: A Literature Review | Multiple object tracking gained a lot of interest from researchers in recent years, and it has become one of the trending problems in computer vision, especially with the recent advancement of autonomous driving. MOT is one of the critical vision tasks for different issues like occlusion in crowded scenes, similar appearance, small object detection difficulty, ID switching, etc. To tackle these challenges, as researchers tried to utilize the attention mechanism of transformer, interrelation of tracklets with graph convolutional neural network, appearance similarity of objects in different frames with the siamese network, they also tried simple IOU matching based CNN network, motion prediction with LSTM. To take these scattered techniques under an umbrella, we have studied more than a hundred papers published over the last three years and have tried to extract the techniques that are more focused on by researchers in recent times to solve the problems of MOT. We have enlisted numerous applications, possibilities, and how MOT can be related to real life. Our review has tried to show the different perspectives of techniques that researchers used overtimes and give some future direction for the potential researchers. Moreover, we have included popular benchmark datasets and metrics in this review. | ['Md. Hasanul Kabir', 'A. B. M. Ashikur Rahman', 'Md. Bakhtiar Hasan', 'Kashifa Kawaakib Hussain', 'Samia Islam', 'Mk Bashar'] | 2022-09-11 | null | null | null | null | ['small-object-detection'] | ['computer-vision'] | [-4.54544090e-02 -5.16370952e-01 -1.64815173e-01 -4.40515503e-02
1.19621843e-01 -1.66732833e-01 5.30565023e-01 -1.05393521e-01
-6.84515476e-01 6.95409000e-01 -1.52262151e-01 1.52502000e-01
-4.01267320e-01 -4.77599859e-01 -7.22317517e-01 -6.61050081e-01
-1.09320223e-01 3.19557041e-01 9.32831109e-01 -4.90932286e-01
4.90225434e-01 5.58971763e-01 -2.15361166e+00 -3.67650948e-02
4.80692357e-01 1.05609369e+00 3.62127751e-01 5.20220101e-01
-2.11765274e-01 1.31348789e+00 -5.39811313e-01 -5.24125159e-01
3.05976331e-01 -1.97616056e-01 -4.20426458e-01 -1.94203168e-01
8.40058088e-01 9.88037139e-02 -7.59904861e-01 1.13172162e+00
5.04834116e-01 4.97199029e-01 4.59303916e-01 -1.76176405e+00
-6.38559401e-01 1.90534562e-01 -8.29690039e-01 1.01663136e+00
2.67188232e-02 3.34864944e-01 5.58826327e-01 -4.36504930e-01
5.91355205e-01 1.42286730e+00 6.72025144e-01 6.38181925e-01
-3.01533490e-01 -7.72566617e-01 2.38944203e-01 1.26721191e+00
-1.13736618e+00 -2.57852465e-01 5.46176434e-01 -4.17434275e-01
9.32277977e-01 1.82368636e-01 9.21969771e-01 8.60591948e-01
7.10072041e-01 9.79022861e-01 6.77120686e-01 -8.19388628e-02
-3.44832420e-01 1.04503505e-01 3.74507844e-01 8.20961237e-01
5.03572464e-01 7.32839331e-02 -4.71072555e-01 2.54949093e-01
3.09023887e-01 2.37388358e-01 -1.41433686e-01 -3.37188452e-01
-1.44357884e+00 6.58858180e-01 5.79878092e-01 5.71067274e-01
8.52068812e-02 3.75998259e-01 5.03337979e-01 8.41325819e-02
2.33745016e-02 7.26066604e-02 -1.02871776e-01 1.60036907e-01
-6.37167037e-01 3.23403716e-01 3.60737294e-01 1.09827340e+00
5.96833944e-01 2.62422770e-01 -2.62994796e-01 4.02662545e-01
2.60877341e-01 8.34734142e-01 8.18744302e-01 -9.47494864e-01
4.02877390e-01 5.04988909e-01 8.67404044e-03 -1.44533408e+00
-4.91050392e-01 -3.44692171e-01 -9.46403801e-01 4.31447595e-01
3.75583321e-01 -9.13203284e-02 -7.16079354e-01 1.44381344e+00
3.66130650e-01 4.92508978e-01 -1.59645230e-01 8.63773286e-01
1.22656631e+00 6.33599401e-01 6.70846691e-03 -1.37496963e-01
1.41942024e+00 -1.47744358e+00 -1.02659488e+00 -2.51406670e-01
3.05158019e-01 -8.87494445e-01 2.93399960e-01 2.88901448e-01
-1.03373539e+00 -1.08456922e+00 -1.28616476e+00 -5.00997342e-02
-8.27126980e-01 -1.29630521e-01 6.44268930e-01 3.14760745e-01
-1.02386510e+00 7.95066714e-01 -8.79076838e-01 -8.79597187e-01
5.33500314e-01 5.73569179e-01 -2.29260951e-01 -5.08313514e-02
-1.09424353e+00 1.29021072e+00 4.87779826e-01 3.40534836e-01
-7.05475569e-01 1.64330639e-02 -5.51952422e-01 -2.99809486e-01
3.92148018e-01 -1.04821849e+00 8.04711282e-01 -9.64641988e-01
-9.99606609e-01 7.15994895e-01 -2.04338282e-01 -6.33973837e-01
6.02470100e-01 -9.90881622e-02 -8.92365098e-01 -3.39033693e-01
1.92323402e-01 6.95741653e-01 7.26603389e-01 -6.14185035e-01
-1.19176352e+00 -3.40964168e-01 4.67415899e-03 3.65959615e-01
-1.94372088e-01 2.89117813e-01 -4.22669679e-01 -2.17298806e-01
-1.92241132e-01 -1.01464748e+00 -1.98808223e-01 9.54912230e-02
-6.91238269e-02 -7.14368105e-01 1.31345475e+00 -1.12407416e-01
1.05190682e+00 -2.11841345e+00 2.83268809e-01 -5.86550951e-01
3.05636257e-01 6.42034650e-01 -1.55579597e-02 1.72470674e-01
2.92235702e-01 -3.37751687e-01 3.74004105e-03 -2.27965772e-01
-1.58612475e-01 2.38291204e-01 -8.58043805e-02 7.73434341e-01
8.81537143e-03 1.03060079e+00 -9.00406778e-01 -8.20874333e-01
5.64427495e-01 4.07965273e-01 1.55621856e-01 -6.72167093e-02
-8.82944837e-02 4.74135250e-01 -4.75265771e-01 5.25453150e-01
7.66106844e-01 -2.42341176e-01 -5.27072012e-01 -3.82793516e-01
-4.10632759e-01 -5.15626311e-01 -1.46729851e+00 1.41505814e+00
2.88625747e-01 1.32812369e+00 -1.66662976e-01 -1.18387377e+00
7.46931374e-01 1.38571793e-02 5.85697472e-01 -9.06740725e-01
4.13167387e-01 3.42040479e-01 2.93231547e-01 -1.02774799e+00
6.13775313e-01 3.32489550e-01 5.17534733e-01 -1.34288386e-01
-1.41330659e-01 2.87991822e-01 5.87406099e-01 -1.53740957e-01
8.10370684e-01 1.82543337e-01 2.88465977e-01 -8.77244696e-02
9.31975961e-01 4.63514179e-01 5.63463688e-01 6.31516218e-01
-8.41111720e-01 3.34988326e-01 -2.05479845e-01 -9.18804765e-01
-7.69770920e-01 -5.25394022e-01 -1.98654104e-02 8.45142305e-01
7.46333122e-01 1.81408525e-01 -1.31883889e-01 -3.85453314e-01
-2.68364791e-02 1.26846373e-01 -6.30774260e-01 6.08766917e-03
-1.16647732e+00 -7.56871402e-01 5.61685145e-01 3.94424826e-01
1.04146051e+00 -1.36845601e+00 -9.23127592e-01 3.19117159e-01
-9.63400230e-02 -1.22204626e+00 -2.73819476e-01 -8.08342826e-03
-7.30263293e-01 -1.23516619e+00 -1.02795935e+00 -1.08359039e+00
1.00411266e-01 9.15404201e-01 9.83998358e-01 3.90518248e-01
-3.71301293e-01 1.32859468e-01 -2.25354671e-01 -7.66839802e-01
2.48379678e-01 7.76536344e-03 1.77096054e-01 9.57974512e-03
6.69037938e-01 -3.46721292e-01 -6.69429481e-01 5.04266262e-01
-8.33285332e-01 -2.76487648e-01 5.03746927e-01 4.33710426e-01
2.71545291e-01 -1.74140781e-01 2.37351134e-01 -3.17872167e-01
1.45524710e-01 -4.83104467e-01 -7.57161558e-01 2.77374983e-01
-2.31784821e-01 -7.93993101e-02 5.38946688e-01 -4.88770306e-01
-7.17810631e-01 -8.09877552e-03 2.50359415e-03 -5.07972240e-01
-8.01156983e-02 -8.35918337e-02 1.24145247e-01 -6.83477581e-01
4.20196652e-01 3.35560024e-01 -1.41893893e-01 -6.43001646e-02
2.58304536e-01 5.98886371e-01 4.47032720e-01 9.63150486e-02
7.77427793e-01 8.39703560e-01 3.87181193e-01 -6.75916076e-01
-8.82682145e-01 -7.41877198e-01 -4.96450901e-01 -6.14773154e-01
1.25493634e+00 -7.14561522e-01 -9.88191366e-01 5.88381350e-01
-1.40990794e+00 1.75556362e-01 7.84499720e-02 7.39851952e-01
-2.97873229e-01 5.47070980e-01 -2.67548382e-01 -6.54381633e-01
-2.88990647e-01 -1.38137853e+00 5.56461632e-01 8.69531631e-01
2.76011020e-01 -1.07737446e+00 8.39019418e-02 2.78558284e-01
6.04325771e-01 3.89361292e-01 4.01839733e-01 -5.32579243e-01
-1.13222373e+00 3.86369675e-02 -3.12004626e-01 -1.00724041e-01
1.67887490e-02 1.51780501e-01 -8.65064323e-01 -3.82160008e-01
5.12393825e-02 2.92175282e-02 1.13392425e+00 7.26169646e-01
7.28109181e-01 1.26092359e-01 -9.69256639e-01 6.63411319e-01
1.50917459e+00 6.91374183e-01 5.06085038e-01 7.35666096e-01
7.27059662e-01 4.61949468e-01 8.20684373e-01 -1.58880979e-01
4.19167638e-01 7.83918560e-01 6.95563734e-01 1.21143267e-01
-5.46183288e-01 2.30603069e-01 2.59936333e-01 1.02464414e+00
-4.38448817e-01 -6.00725710e-01 -5.08903265e-01 5.69814444e-01
-2.13266039e+00 -1.24550080e+00 -6.80276632e-01 1.90934300e+00
-2.89915204e-01 3.14806104e-01 1.80975810e-01 -2.48555318e-01
1.07015574e+00 4.51778382e-01 -7.18697488e-01 1.45290769e-03
-4.66275513e-01 -4.09188688e-01 8.26140642e-01 1.89979330e-01
-1.19799113e+00 7.99796402e-01 5.91738033e+00 8.74578178e-01
-1.24156535e+00 1.59372583e-01 2.10744545e-01 -3.51232551e-02
3.56136918e-01 -1.51098073e-01 -1.28352821e+00 5.74659705e-01
6.42942429e-01 -2.72028923e-01 2.20273331e-01 6.81345522e-01
1.32559380e-02 -1.27267331e-01 -8.14629078e-01 1.17062259e+00
4.44303870e-01 -1.35432208e+00 -1.66296959e-03 -1.57303914e-01
7.76974082e-01 5.04732966e-01 2.38296479e-01 3.38954270e-01
1.45470481e-02 -9.04988945e-01 5.55171072e-01 7.67297924e-01
1.28196448e-01 -3.96052837e-01 1.04145801e+00 3.50126296e-01
-1.72449040e+00 -2.75754780e-01 -8.29519928e-01 -2.99659967e-01
2.59527057e-01 7.95344636e-02 -2.08212867e-01 8.78023982e-01
1.05252433e+00 1.17905557e+00 -7.45746076e-01 1.85361028e+00
4.28476691e-01 -5.86381219e-02 -2.95006901e-01 -7.12769330e-01
5.27508676e-01 -1.72345728e-01 7.95366764e-01 9.54888284e-01
3.09198946e-01 -2.52589345e-01 6.40906915e-02 5.24526775e-01
2.93539345e-01 1.43926293e-01 -8.43014598e-01 2.80156761e-01
1.36024594e-01 1.34417558e+00 -9.14134800e-01 -6.10407948e-01
-5.36321700e-01 5.02315342e-01 1.04492351e-01 1.58460170e-01
-1.14984798e+00 -5.25132477e-01 4.77789581e-01 -1.73119843e-01
3.61862630e-01 -2.80710220e-01 1.48847163e-01 -9.10298645e-01
-3.14385109e-02 -5.12894154e-01 5.05857527e-01 -9.52229440e-01
-1.08073640e+00 8.31336498e-01 2.44249068e-02 -1.65744317e+00
2.40769908e-01 -7.47835040e-01 -6.49649441e-01 6.04723394e-01
-1.75342703e+00 -9.03409362e-01 -6.80454969e-01 6.12826169e-01
9.86781120e-01 -4.16739881e-01 1.18705921e-01 7.10965812e-01
-8.30035090e-01 2.85653055e-01 2.96462119e-01 8.79454762e-02
7.53217041e-01 -6.87876999e-01 4.33020860e-01 8.56872022e-01
1.89063475e-01 4.51598883e-01 8.40446830e-01 -5.12313724e-01
-1.26260340e+00 -1.17267299e+00 7.90229499e-01 -5.70041418e-01
6.65028751e-01 1.96901754e-01 -6.20130777e-01 8.47829878e-01
7.51686811e-01 2.11506769e-01 -7.36889765e-02 -6.51560724e-01
2.73613662e-01 -3.70800257e-01 -8.51250887e-01 5.33215642e-01
1.18030286e+00 3.21955800e-01 -4.71381068e-01 4.98210669e-01
7.38506317e-01 -5.69323957e-01 -3.75815362e-01 4.67744231e-01
5.35480380e-01 -1.22524929e+00 9.32705581e-01 -5.85680544e-01
-1.68499485e-01 -7.82662153e-01 2.31673028e-02 -8.62589240e-01
-5.44811785e-01 -4.58566070e-01 -8.60356688e-02 9.00226355e-01
2.46264059e-02 -7.30751157e-01 8.34711134e-01 -5.65856658e-02
-2.81245440e-01 -5.83483756e-01 -9.69634235e-01 -8.79182518e-01
-1.74191117e-01 5.14023425e-03 4.15265292e-01 6.50603890e-01
-5.31269848e-01 2.50142902e-01 -5.80287635e-01 3.26148234e-02
7.83839464e-01 -8.21169242e-02 8.02962780e-01 -1.36501396e+00
5.06547466e-02 -6.06116176e-01 -1.04790604e+00 -1.30350101e+00
-1.94232315e-01 -6.79871619e-01 -1.73680529e-01 -1.78276527e+00
2.00085983e-01 -9.63477790e-02 -5.26906550e-01 3.78566906e-02
-1.13901898e-01 3.30850720e-01 4.15029407e-01 3.61056805e-01
-1.24067092e+00 4.28053379e-01 1.63592422e+00 -6.29748762e-01
1.70783713e-01 2.66029418e-01 -2.38659829e-01 7.58189201e-01
5.60968399e-01 -7.63708711e-01 -2.38666654e-01 -6.64823651e-01
2.30629712e-01 -1.24503829e-01 4.08018231e-01 -1.85136759e+00
9.04310763e-01 -2.30742380e-01 3.05913270e-01 -9.85679150e-01
5.50509572e-01 -9.64941382e-01 4.52157617e-01 8.91364992e-01
9.34078991e-02 6.57836258e-01 3.49931538e-01 6.82691514e-01
-2.97193438e-01 -3.84338260e-01 7.92373836e-01 -4.51429933e-01
-1.40955031e+00 5.95997751e-01 -3.75238955e-01 9.96408388e-02
1.41997182e+00 -6.08468413e-01 -4.84309584e-01 3.70971160e-03
-2.64479816e-01 3.72619897e-01 -7.34144896e-02 7.05042839e-01
3.73106927e-01 -1.28790355e+00 -5.65663636e-01 -2.86665559e-01
9.05483291e-02 -2.72360295e-01 4.12953526e-01 1.16416013e+00
-6.30966485e-01 7.65010357e-01 -7.29314744e-01 -8.50840867e-01
-1.55504346e+00 9.39440429e-01 5.64812362e-01 -7.03357952e-03
-7.64135420e-01 8.61753047e-01 -2.40322156e-03 1.70260921e-01
4.42290902e-01 -2.47090146e-01 -8.73428643e-01 -2.38985587e-02
5.66275895e-01 7.14169502e-01 -7.44203106e-02 -9.16135013e-01
-5.34777045e-01 1.19345832e+00 4.66577485e-02 4.07742023e-01
1.07621002e+00 -3.34639251e-01 4.19376716e-02 6.69479311e-01
1.05604172e+00 -3.52126300e-01 -1.00981736e+00 -1.78088069e-01
-1.26667425e-01 -3.71382773e-01 -3.10769916e-01 -2.72435069e-01
-1.31599498e+00 9.48814571e-01 1.14182842e+00 5.16459048e-01
6.92901075e-01 -2.52217591e-01 9.75141585e-01 6.34093225e-01
4.49508607e-01 -8.64264011e-01 9.14341137e-02 5.84780395e-01
4.39371794e-01 -1.50270259e+00 -3.44323926e-02 -4.51557375e-02
-3.90883833e-01 1.27687466e+00 8.47786963e-01 -3.07727873e-01
6.05604589e-01 -6.83971867e-02 7.70959258e-02 -3.49654078e-01
-4.92699593e-01 -5.13198137e-01 2.96897382e-01 7.74795890e-01
2.08954275e-01 -3.99262637e-01 -3.27643931e-01 -2.90425926e-01
4.06359695e-02 7.39868358e-02 4.15387928e-01 8.07909369e-01
-8.66764128e-01 -8.29391539e-01 -5.33836484e-01 4.74225789e-01
-2.79006630e-01 2.48803571e-01 1.07571073e-01 1.11635232e+00
8.52476716e-01 1.07412267e+00 3.08606513e-02 -3.68001312e-01
5.15352190e-01 -4.81573164e-01 4.91502285e-01 4.21971753e-02
-4.58312035e-01 -6.03953838e-01 -2.94286281e-01 -4.25119728e-01
-9.88146126e-01 -6.27305984e-01 -9.45635140e-01 -4.84497368e-01
-4.26650614e-01 1.35905355e-01 4.97634828e-01 1.11674440e+00
1.09903842e-01 8.37741375e-01 5.45562878e-02 -8.61199558e-01
-5.13408110e-02 -1.00015032e+00 -3.81688386e-01 3.50426912e-01
5.42175233e-01 -1.10153747e+00 -3.29701513e-01 -1.54264301e-01] | [6.508151531219482, -1.9875929355621338] |
bd1c2949-23b1-437c-8135-760911be7e09 | accurate-and-versatile-3d-segmentation-of | null | null | https://elifesciences.org/articles/57613 | https://elifesciences.org/articles/57613 | Accurate and versatile 3D segmentation of plant tissues at cellular resolution | Quantitative analysis of plant and animal morphogenesis requires accurate segmentation of individual cells in volumetric images of growing organs. In the last years, deep learning has provided robust automated algorithms that approach human performance, with applications to bio-image analysis now starting to emerge. Here, we present PlantSeg, a pipeline for volumetric segmentation of plant tissues into cells. PlantSeg employs a convolutional neural network to predict cell boundaries and graph partitioning to segment cells based on the neural network predictions. PlantSeg was trained on fixed and live plant organs imaged with confocal and light sheet microscopes. PlantSeg delivers accurate results and generalizes well across different tissues, scales, acquisition settings even on non plant samples. We present results of PlantSeg applications in diverse developmental contexts. PlantSeg is free and open-source, with both a command line and a user-friendly graphical interface. | ['Anna Kreshuk', 'Alexis Maizel', 'Kay Schneitz', 'Fred A Hamprecht', 'Miltos Tsiantis', 'Jan U Lohmann', 'George W Bassel', 'Salva Duran-Nebreda', 'Alberto Bailoni', 'Constantin Pape', 'Susanne S Steigleder', 'Rena Lymbouridou', 'David Wilson-Sánchez', 'Sören Strauss', 'Christian Wenzl', 'Marion Louveaux', 'Amaya Vilches Barro', 'Rachele Tofanelli', 'Athul Vijayan', 'Lorenzo Cerrone', 'Adrian Wolny'] | 2020-07-29 | null | null | null | elife-2020-7 | ['graph-partitioning'] | ['graphs'] | [ 1.14044823e-01 5.50090522e-02 2.15954229e-01 -6.96863830e-02
-1.45764917e-01 -1.00784540e+00 8.87412354e-02 5.12840390e-01
-3.57069559e-02 4.36089426e-01 -7.26833701e-01 -4.66389626e-01
1.71195865e-01 -9.03786898e-01 -3.81681621e-01 -5.89462399e-01
-3.33055049e-01 7.93056309e-01 3.48250180e-01 1.53950512e-01
1.96124107e-01 1.11623359e+00 -9.13427889e-01 1.65443867e-01
4.84147251e-01 1.12465751e+00 7.01039791e-01 1.30612755e+00
-3.53822678e-01 3.96519095e-01 -3.06998879e-01 1.35685638e-01
2.96017379e-01 -1.15302727e-01 -1.16274261e+00 4.03804809e-01
1.54709563e-01 -3.03432465e-01 2.50409335e-01 3.91499817e-01
3.79273593e-01 -6.64176464e-01 6.61453724e-01 -1.10528862e+00
-9.90803719e-01 3.94471973e-01 -6.63227081e-01 1.31199315e-01
-2.34309807e-01 4.87786144e-01 9.84642982e-01 -9.80979979e-01
8.86263251e-01 8.49567115e-01 9.24520791e-01 3.35628361e-01
-1.63551795e+00 -5.56309447e-02 -2.08605438e-01 -3.95978123e-01
-1.15745413e+00 -1.95042938e-01 1.25706330e-01 -9.11006749e-01
9.69859004e-01 4.14041355e-02 9.11811054e-01 3.11441690e-01
6.16728902e-01 5.62492788e-01 7.21273959e-01 -1.44691810e-01
3.14356923e-01 -5.28176427e-01 -1.36144012e-01 1.00532711e+00
2.06728429e-01 -3.86336118e-01 1.74569726e-01 1.32873729e-01
1.24279952e+00 -3.29882115e-01 -1.44002244e-01 -4.10919011e-01
-1.22590792e+00 4.01782334e-01 3.89642030e-01 5.96750319e-01
-4.08363611e-01 2.63713837e-01 4.99590993e-01 -4.41949427e-01
4.97092783e-01 5.46496034e-01 -1.07369602e+00 3.83466363e-01
-8.57437015e-01 2.53996193e-01 1.01038492e+00 9.40835059e-01
6.15524113e-01 1.58193246e-01 -5.74670434e-02 8.28042924e-01
2.65599221e-01 1.23745620e-01 8.92158374e-02 -1.36085796e+00
-5.45078337e-01 8.53749990e-01 -1.40337408e-01 -7.08124995e-01
-7.65763044e-01 -1.69014454e-01 -9.11440134e-01 5.81813693e-01
9.95451629e-01 -1.37350723e-01 -9.96616900e-01 1.25851977e+00
4.14181322e-01 -1.15750238e-01 -3.61595750e-01 6.48966730e-01
1.09990215e+00 6.50973558e-01 2.68851727e-01 3.92960720e-02
1.18983579e+00 -4.46319848e-01 -2.57971585e-01 -9.51968282e-02
8.66471887e-01 -7.22472966e-01 9.11319613e-01 4.68439996e-01
-1.35984957e+00 -2.47127146e-01 -7.92877734e-01 -2.88513631e-01
-8.26575279e-01 4.99681383e-01 7.98467696e-01 2.28559196e-01
-1.17281747e+00 1.06582940e+00 -9.57161784e-01 -7.07231879e-01
1.27434909e+00 5.30553102e-01 -5.87932169e-01 3.25276583e-01
-1.52880490e-01 7.17753053e-01 4.08659935e-01 1.18750930e-01
-1.01981115e+00 -1.02588844e+00 -6.37821615e-01 4.22883421e-01
-1.00631535e-01 -8.31501186e-01 1.26052570e+00 -5.94063580e-01
-1.70122790e+00 1.68023932e+00 1.66930586e-01 -3.93878192e-01
1.30181968e-01 3.16496462e-01 4.97415274e-01 1.54568134e-02
8.59534442e-02 1.04861093e+00 9.40355286e-02 -1.26787949e+00
-4.86853421e-01 -7.11381316e-01 -4.01801974e-01 -4.02468026e-01
4.16547433e-02 -2.29657680e-01 -2.40066871e-01 -4.64893132e-02
3.39378566e-01 -7.73142874e-01 -3.53976339e-01 8.10247302e-01
-6.74208879e-01 9.23849344e-02 1.24145186e+00 -7.38758504e-01
2.62094796e-01 -1.53218448e+00 1.68385535e-01 -1.24223247e-01
4.88493323e-01 3.53939682e-01 -1.98562220e-01 2.78492063e-01
-9.11538005e-02 3.99404407e-01 -5.14211416e-01 -2.25084022e-01
-3.84071290e-01 1.44238338e-01 1.88675880e-01 6.63978875e-01
3.62620503e-01 1.14415514e+00 -6.33785188e-01 -7.57701874e-01
5.64412475e-01 4.04049039e-01 -1.71614990e-01 8.24651048e-02
-4.48175997e-01 8.22254777e-01 -2.13519081e-01 1.10418212e+00
1.10173905e+00 -6.81552529e-01 1.73272267e-01 -9.22856629e-02
-4.22722399e-01 -3.32931757e-01 -5.36097348e-01 1.65652037e+00
-2.07235709e-01 7.52269328e-01 8.70615125e-01 -1.03401589e+00
8.70186269e-01 1.28728181e-01 9.77098107e-01 1.57692924e-01
3.85179281e-01 3.91469300e-01 -1.65310860e-01 1.03117723e-03
-8.87651443e-02 1.41523436e-01 3.16800714e-01 -6.81492267e-03
6.75204098e-01 -9.29350019e-01 5.99908590e-01 7.60203525e-02
1.12798858e+00 5.51206350e-01 4.61371601e-01 -5.76476753e-01
5.35713494e-01 3.27236950e-01 4.86224115e-01 1.78094938e-01
-5.12330711e-01 7.16590524e-01 5.63246965e-01 -5.85305870e-01
-1.36553741e+00 -1.00925076e+00 -3.03397328e-01 1.02137530e+00
-6.66626394e-02 -2.04378203e-01 -9.71844554e-01 -6.13369979e-02
1.13433510e-01 1.82608232e-01 -5.71404159e-01 3.71422708e-01
-2.65339732e-01 -7.03896224e-01 6.48766339e-01 4.41858947e-01
2.92741597e-01 -1.42837310e+00 -7.99181700e-01 5.28815746e-01
2.90527850e-01 -1.33751607e+00 3.07623476e-01 6.21204376e-01
-8.02467823e-01 -1.14228654e+00 -9.63335752e-01 -1.07015526e+00
4.65906799e-01 8.39091167e-02 1.29911196e+00 3.87027711e-01
-1.12186921e+00 2.32582912e-01 -5.88251390e-02 -7.73615837e-01
-3.50787044e-01 4.58169967e-01 -6.84871674e-01 -8.72907758e-01
-2.22478658e-01 -7.46653557e-01 -6.50426447e-01 1.03267312e-01
-6.65171683e-01 2.86046505e-01 5.62858023e-02 8.57256174e-01
9.59638596e-01 -3.56549293e-01 2.84276754e-01 -9.59308624e-01
1.51459977e-01 -3.62553537e-01 -1.27734363e+00 1.63954228e-01
-3.33731353e-01 -5.43044388e-01 1.04812753e+00 2.76033282e-01
-6.19383752e-01 5.52290261e-01 -1.41607121e-01 7.11366609e-02
-6.97849095e-01 5.31657755e-01 -1.61599398e-01 -2.93586224e-01
6.39058828e-01 2.80746706e-02 8.32458809e-02 -2.81178802e-02
2.97449052e-01 7.29626194e-02 6.91663027e-01 -5.91727495e-01
4.38873291e-01 7.13969231e-01 6.03244305e-01 -9.83904958e-01
-5.98970473e-01 -3.44553739e-01 -1.20217574e+00 -4.44638282e-01
9.07283425e-01 -2.75761068e-01 -1.32071137e+00 8.70259583e-01
-1.09869528e+00 -9.03045654e-01 -3.67094874e-01 -1.05230883e-01
-8.21929932e-01 1.88641682e-01 -1.14932823e+00 -4.14397240e-01
-6.86355770e-01 -9.69654500e-01 1.31882441e+00 6.32452905e-01
-5.59607185e-02 -1.23050582e+00 -2.78836321e-02 -1.22227922e-01
3.19668472e-01 8.14906299e-01 8.89891326e-01 -1.83482632e-01
-7.21865833e-01 -2.45038137e-01 -4.97653216e-01 1.03394300e-01
4.25791711e-01 1.21571708e+00 -1.11671674e+00 -6.52642548e-02
-8.53238285e-01 -4.96636122e-01 7.49194741e-01 1.23387575e+00
1.77740765e+00 1.73505068e-01 -8.05179298e-01 1.00641823e+00
1.60050726e+00 6.60098791e-02 4.88245934e-01 2.34480634e-01
8.35895061e-01 8.74983549e-01 2.19148353e-01 6.80546999e-01
-2.45536282e-03 -7.12241083e-02 9.49160039e-01 -6.26584768e-01
1.47402927e-01 3.94967347e-01 -4.29960251e-01 2.14818463e-01
-3.71358693e-01 -6.51930571e-01 -1.24686360e+00 6.81348920e-01
-1.27422082e+00 -7.92730570e-01 -5.84621072e-01 1.68802977e+00
6.71020508e-01 -8.91871378e-02 1.43198576e-02 -6.27885712e-03
6.60660982e-01 -2.25513071e-01 -9.27271008e-01 -4.07233953e-01
-4.45720106e-01 5.85127652e-01 8.23831141e-01 2.57140875e-01
-1.16527796e+00 1.46000671e+00 7.04842997e+00 4.21284914e-01
-1.32985735e+00 -3.26273769e-01 1.24615979e+00 5.29323220e-01
1.17358483e-01 3.07730228e-01 -4.40377951e-01 -1.58782423e-01
7.27504730e-01 -1.91289410e-01 2.89315253e-01 7.24589944e-01
3.31917286e-01 -1.38455763e-01 -9.09652889e-01 3.01324815e-01
-8.59626174e-01 -1.90892792e+00 -2.90921062e-01 3.46151173e-01
3.98256749e-01 4.12927449e-01 -1.95035949e-01 -1.37674674e-01
6.88398838e-01 -1.22963607e+00 4.23906058e-01 2.79134393e-01
7.98064470e-01 -4.35336143e-01 5.73804855e-01 4.35385436e-01
-1.45068157e+00 8.68681818e-02 -5.75810373e-01 2.76441038e-01
7.14291707e-02 8.25344443e-01 -1.35030425e+00 2.82692194e-01
7.47453868e-01 6.67775273e-01 -6.63954079e-01 9.69877601e-01
4.78583962e-01 7.63296425e-01 -3.38743538e-01 5.06452382e-01
1.89322345e-02 -4.29483443e-01 1.97628498e-01 1.47940195e+00
3.97782862e-01 -2.50090137e-02 2.15628788e-01 1.48213136e+00
-1.51185049e-02 2.41809592e-01 -6.30514443e-01 -6.71197474e-01
1.77715749e-01 2.16579366e+00 -1.70785737e+00 -5.51449321e-02
-1.23366058e-01 9.84861851e-01 3.97268802e-01 -2.49796417e-02
-4.22618628e-01 -3.67435932e-01 3.05782527e-01 2.57925272e-01
2.21977055e-01 -6.10736132e-01 -6.84914470e-01 -4.74426717e-01
-9.70810592e-01 -1.06999017e-01 1.39746249e-01 -1.20980632e+00
-1.08959937e+00 6.15669750e-02 -4.98612493e-01 -5.01971245e-01
1.44358322e-01 -1.33775270e+00 -7.75256574e-01 8.26550722e-01
-9.62036252e-01 -1.52841365e+00 -7.77888179e-01 -6.92767426e-02
5.14592171e-01 1.24209926e-01 1.42414069e+00 -2.05304503e-01
-7.47846007e-01 -2.76468992e-01 1.05936691e-01 9.41595361e-02
2.74762064e-01 -1.67061758e+00 7.01952815e-01 4.90047485e-01
-2.09096581e-01 3.19499135e-01 6.15836620e-01 -4.47246134e-01
-9.76717889e-01 -1.17903972e+00 4.58746910e-01 -1.56829908e-01
3.19547057e-01 -2.05343142e-01 -9.71267223e-01 8.11118066e-01
4.34819996e-01 6.31323278e-01 8.16484451e-01 -3.13326508e-01
2.17075005e-01 3.03896040e-01 -1.38396442e+00 5.33465147e-01
7.58522093e-01 -5.19740582e-02 5.25059104e-01 6.86224580e-01
3.91746014e-01 -6.94906056e-01 -1.43006957e+00 4.31605875e-01
6.99148715e-01 -9.69642043e-01 9.08444405e-01 -3.12826157e-01
6.17706120e-01 -3.92307758e-01 -5.88721922e-03 -1.15210247e+00
-7.95176864e-01 -3.49381983e-01 1.18462130e-01 1.31155634e+00
4.31856066e-01 -2.33507887e-01 8.93860221e-01 2.92621017e-01
-4.81812418e-01 -8.13541055e-01 -4.37909186e-01 -3.80373806e-01
5.41142046e-01 1.93354592e-01 5.09921372e-01 7.79851615e-01
-1.63093597e-01 8.94428790e-02 4.55412567e-01 3.32013071e-02
3.86290342e-01 2.14033097e-01 7.39664137e-01 -1.85027349e+00
1.64776236e-01 -6.10342503e-01 -2.85426170e-01 -5.89511693e-01
2.97520727e-01 -8.60759854e-01 1.37485385e-01 -1.89807582e+00
-1.55747971e-02 -1.70438349e-01 2.36408263e-01 6.15858555e-01
1.77718908e-01 2.27169067e-01 -8.11412781e-02 1.95965379e-01
-9.39230919e-02 3.17684077e-02 1.60603499e+00 -3.11062366e-01
-9.49025676e-02 -3.10254917e-02 -4.88686740e-01 8.30290854e-01
1.18975472e+00 -8.63904506e-02 -1.00792922e-01 -1.49353892e-01
-2.07114398e-01 -2.49232188e-01 4.33019608e-01 -1.12063742e+00
-2.13067636e-01 -3.56047034e-01 1.00949442e+00 -8.22090626e-01
2.16595218e-01 -6.35433733e-01 3.91190760e-02 6.03515446e-01
-2.25379765e-01 -1.63350776e-01 6.44686818e-01 5.64952269e-02
3.24403346e-01 -2.42273092e-01 1.17360163e+00 -7.67209232e-01
-4.77577895e-01 6.89009070e-01 -9.00910676e-01 -1.97369337e-01
1.16332030e+00 -6.49621189e-01 -4.43780482e-01 -9.15751159e-02
-9.16417778e-01 1.10637531e-01 6.35161281e-01 -3.45313519e-01
3.86996925e-01 -7.53215015e-01 -6.14193678e-01 -1.44759059e-01
1.38349235e-01 5.33241689e-01 -4.10587564e-02 6.88659668e-01
-1.61575770e+00 4.18161690e-01 -6.92784548e-01 -1.11408150e+00
-1.10195649e+00 2.35213920e-01 5.80119014e-01 -2.22126350e-01
-3.51502508e-01 9.75072920e-01 4.06643063e-01 -9.27747011e-01
-4.37130868e-01 -8.50199938e-01 -3.20167989e-01 -3.51862639e-01
-1.70188565e-02 3.00119966e-01 -6.92457557e-02 -6.55284524e-01
-1.83924973e-01 4.84843969e-01 2.50594318e-01 2.57848114e-01
1.61879838e+00 -1.79400042e-01 -3.14249456e-01 6.96344316e-01
7.29566574e-01 -4.12988007e-01 -1.41028941e+00 4.31220561e-01
9.86309424e-02 -1.44033894e-01 1.28046438e-01 -7.59050906e-01
-1.30856490e+00 9.72746015e-01 6.09637678e-01 6.00986838e-01
1.02092075e+00 -1.30081758e-01 7.08730400e-01 6.58496097e-02
2.69410551e-01 -8.91045928e-01 -2.65249431e-01 6.98696077e-01
7.16941595e-01 -1.24121881e+00 1.09742917e-01 -8.67207170e-01
-9.40883439e-03 1.46854830e+00 8.84148657e-01 -1.59940571e-01
8.53021204e-01 6.75507069e-01 8.36170837e-02 -4.45955187e-01
-7.46056437e-01 -1.74872160e-01 -2.38308817e-01 1.03628540e+00
1.21101141e+00 1.56354066e-02 -4.88620112e-03 2.92556345e-01
-1.39226029e-02 3.48631591e-01 6.34665430e-01 9.69552100e-01
-5.92525065e-01 -8.98218989e-01 -2.16231436e-01 4.01413769e-01
-6.63912773e-01 9.09835622e-02 -8.32457781e-01 7.63465106e-01
2.70234615e-01 4.15649652e-01 1.86627880e-01 2.07734197e-01
-3.76724498e-03 -1.88260481e-01 7.94570267e-01 -8.09873343e-01
-5.87456048e-01 1.58575028e-01 -3.18540275e-01 -5.23408294e-01
-1.81443542e-01 -5.25905132e-01 -1.83639908e+00 -4.03530151e-01
-5.88504791e-01 -5.34828603e-01 8.46513391e-01 7.56931484e-01
5.42023242e-01 7.47701406e-01 1.28086567e-01 -1.44413459e+00
4.23758924e-01 -9.24359381e-01 -1.04896355e+00 -1.93058327e-01
-1.59739003e-01 -3.25647622e-01 -1.01465970e-01 7.31256604e-01] | [14.305516242980957, -3.1190946102142334] |
2b8fdff9-e4c1-4434-8960-506fec310f86 | automated-pii-extraction-from-social-media | 2111.09415 | null | https://arxiv.org/abs/2111.09415v1 | https://arxiv.org/pdf/2111.09415v1.pdf | Automated PII Extraction from Social Media for Raising Privacy Awareness: A Deep Transfer Learning Approach | Internet users have been exposing an increasing amount of Personally Identifiable Information (PII) on social media. Such exposed PII can cause severe losses to the users, and informing users of their PII exposure is crucial to raise their privacy awareness and encourage them to take protective measures. To this end, advanced automatic techniques are needed. While Information Extraction (IE) techniques can be used to extract the PII automatically, Deep Learning (DL)-based IE models alleviate the need for feature engineering and further improve the efficiency. However, DL-based IE models often require large-scale labeled data for training, but PII-labeled social media posts are difficult to obtain due to privacy concerns. Also, these models rely heavily on pre-trained word embeddings, while PII in social media often varies in forms and thus has no fixed representations in pre-trained word embeddings. In this study, we propose the Deep Transfer Learning for PII Extraction (DTL-PIIE) framework to address these two limitations. DTL-PIIE transfers knowledge learned from publicly available PII data to social media to address the problem of rare PII-labeled data. Moreover, our framework leverages Graph Convolutional Networks (GCNs) to incorporate syntactic patterns to guide PIIE without relying on pre-trained word embeddings. Evaluation against benchmark IE models indicates that our approach outperforms state-of-the-art DL-based IE models. Our framework can facilitate various applications, such as PII misuse prediction and privacy risk assessment, protecting the privacy of internet users. | ['Hsinchun Chen', 'Weifeng Li', 'MohammadReza Ebrahimi', 'Fang Yu Lin', 'Yizhi Liu'] | 2021-11-11 | null | null | null | null | ['embeddings-evaluation'] | ['natural-language-processing'] | [ 3.09541374e-01 1.40943080e-01 -6.11391902e-01 -1.88049391e-01
-4.33352739e-01 -7.56859601e-01 5.05791187e-01 4.09239411e-01
-3.47197175e-01 3.93065393e-01 3.51521969e-01 -7.22661912e-01
-7.05943257e-02 -1.24137974e+00 -5.11996746e-01 -1.21376358e-01
-2.34089047e-01 2.93101110e-02 -6.26624897e-02 -2.24469304e-02
7.72304684e-02 5.80523849e-01 -7.38690317e-01 2.02244565e-01
9.16971743e-01 1.03130102e+00 -5.12479842e-01 2.82141119e-01
-3.12984705e-01 3.45347136e-01 -3.58241558e-01 -8.74980807e-01
3.79396021e-01 -4.04520817e-02 -6.39975011e-01 -4.95099396e-01
2.25754693e-01 -5.98127306e-01 -7.29612589e-01 1.24790323e+00
3.44718695e-01 -1.04835972e-01 7.01964676e-01 -1.47351956e+00
-1.08030868e+00 3.29868168e-01 -7.98393667e-01 5.17393090e-02
5.29173672e-01 -1.28249666e-02 1.13177037e+00 -5.19708991e-01
5.08689761e-01 1.01492739e+00 8.76488030e-01 5.85150421e-01
-1.05667639e+00 -1.20945895e+00 2.17916016e-02 1.40060574e-01
-1.46488166e+00 -1.91336825e-01 8.70330513e-01 -2.76591241e-01
9.87408578e-01 1.89564824e-01 4.23762977e-01 1.48159385e+00
-1.25771970e-01 9.93761122e-01 5.92628717e-01 -7.26150051e-02
-9.13974345e-02 6.02637589e-01 4.28373009e-01 6.89037502e-01
7.59838045e-01 7.17689842e-02 -3.80532503e-01 -6.28714442e-01
3.74236137e-01 6.10589743e-01 -2.54637927e-01 -1.14205971e-01
-4.87291783e-01 1.00418842e+00 3.77682775e-01 3.79181206e-01
-3.76075581e-02 -2.98899829e-01 6.16154850e-01 8.46415088e-02
6.01887703e-01 3.55933756e-01 -3.74295622e-01 1.53418913e-01
-5.74555874e-01 1.00421503e-01 1.09431458e+00 1.20965207e+00
1.09576106e+00 -4.69595730e-01 -1.31246084e-02 7.21551538e-01
4.67431575e-01 4.00255293e-01 3.74544203e-01 -2.01535717e-01
9.13571656e-01 1.08915901e+00 -3.56935784e-02 -1.65719211e+00
-2.21697435e-01 -2.53533870e-01 -9.32703972e-01 -1.23482175e-01
2.43355140e-01 -2.11143419e-01 -6.25483572e-01 1.47361267e+00
5.93521036e-02 2.34613538e-01 -1.91231847e-01 3.66955787e-01
6.28188968e-01 6.69427514e-01 3.61051232e-01 2.35266998e-01
1.29501319e+00 -3.79706442e-01 -6.25000954e-01 -3.70722950e-01
8.32241535e-01 -1.78561613e-01 9.16527689e-01 7.62393624e-02
-4.20334339e-01 -6.98624477e-02 -1.06966484e+00 -9.84198526e-02
-8.79883885e-01 -1.28352180e-01 7.12224364e-01 1.16563201e+00
-8.17920208e-01 6.04855239e-01 -4.92447197e-01 -4.78060842e-01
1.21213675e+00 5.60069263e-01 -6.56476080e-01 -1.05110727e-01
-1.54050052e+00 4.72367704e-01 3.39715868e-01 -2.64163576e-02
-5.34188509e-01 -1.05261648e+00 -1.09964085e+00 4.15255904e-01
5.57744503e-01 -2.80371398e-01 6.91345692e-01 -5.79230428e-01
-1.06992722e+00 8.40193748e-01 2.47568041e-02 -4.64884937e-01
1.84348166e-01 -2.79006660e-01 -8.05898905e-01 4.48822603e-02
-1.01237901e-01 2.19499812e-01 9.89288211e-01 -1.04546177e+00
-5.51493347e-01 -5.75060844e-01 5.79663664e-02 -1.84950083e-01
-1.20115113e+00 2.28344843e-01 -5.08978426e-01 -3.83519530e-01
-6.70732200e-01 -8.83865952e-01 -6.71220049e-02 4.34305072e-02
-5.51012278e-01 -3.01961899e-01 1.22290289e+00 -1.03827369e+00
1.86238408e+00 -2.27989316e+00 -5.08553088e-01 5.96564591e-01
7.35230744e-01 1.16050744e+00 -1.81905583e-01 5.53581536e-01
4.98030111e-02 9.34657156e-01 -2.33266100e-01 -3.26922953e-01
1.49735346e-01 5.40657975e-02 -2.46250734e-01 2.19019011e-01
1.96043149e-01 1.04883695e+00 -9.46602046e-01 -3.86202008e-01
1.36159435e-01 4.87867206e-01 -8.06392252e-01 4.45036411e-01
-1.76061573e-03 2.46032402e-01 -6.90305293e-01 5.85808277e-01
9.26761806e-01 -2.69380361e-01 2.42830589e-01 -1.89882472e-01
3.50442261e-01 4.43506002e-01 -6.72193825e-01 1.25772810e+00
-5.41988492e-01 2.42629364e-01 4.26508160e-03 -8.35518479e-01
7.62491167e-01 3.12087268e-01 5.85152268e-01 -4.41946030e-01
2.77919620e-01 -6.51873052e-02 -2.04539374e-01 -4.99734253e-01
1.17640167e-01 3.12377095e-01 -2.14086115e-01 7.15729058e-01
2.67469846e-02 7.36432910e-01 -6.44931570e-02 2.97210127e-01
1.24692965e+00 -2.69623160e-01 4.74348903e-01 9.05222073e-02
8.13760877e-01 -5.61219931e-01 8.35482776e-01 6.31986737e-01
-4.63765353e-01 2.59070247e-01 4.91181314e-01 -4.17949885e-01
-6.59270763e-01 -8.33971739e-01 1.13328723e-02 8.84356201e-01
-4.59259413e-02 -6.84912443e-01 -7.51571596e-01 -1.34964371e+00
2.46105000e-01 7.33902395e-01 -5.63744485e-01 -2.91131377e-01
-5.31559885e-01 -7.15184808e-01 6.65793896e-01 4.44021761e-01
6.24548674e-01 -9.68568861e-01 1.79483727e-01 2.72579640e-01
-1.13053225e-01 -9.85370755e-01 -8.66828144e-01 -4.40601557e-01
-3.74322385e-01 -1.22464991e+00 -2.49697834e-01 -5.40076733e-01
7.01155186e-01 4.20615226e-01 7.22744465e-01 1.62610963e-01
-3.65175307e-01 5.22526264e-01 -3.21608722e-01 -6.99597597e-01
-1.66686252e-01 5.29933929e-01 -1.39809221e-01 4.08456326e-01
1.24296379e+00 -8.27886522e-01 -7.59872913e-01 3.46086830e-01
-1.33333898e+00 -4.03866708e-01 2.19856754e-01 4.55782413e-01
1.00953579e-01 1.03598610e-01 5.95199585e-01 -1.53776395e+00
8.51826370e-01 -8.85656714e-01 -3.98969322e-01 4.45954144e-01
-7.51351416e-01 -1.84778243e-01 6.79233670e-01 -5.35694838e-01
-1.14900017e+00 -4.00840372e-01 -3.70401382e-01 -1.83067963e-01
-5.30584976e-02 4.68186498e-01 -8.43437195e-01 -3.77362967e-01
4.40039217e-01 -9.25495178e-02 -4.85412069e-02 -4.29206550e-01
3.75734329e-01 1.04937232e+00 -8.32073539e-02 -1.92517251e-01
1.06992912e+00 2.75905609e-01 -3.98922592e-01 -7.87589133e-01
-8.87301922e-01 -6.34746671e-01 -3.59892696e-01 2.64620334e-01
7.99853623e-01 -5.34102201e-01 -8.87320697e-01 4.75530863e-01
-1.01371622e+00 9.04402509e-02 2.24648356e-01 8.89852419e-02
-2.89067272e-02 8.44179392e-01 -6.08670533e-01 -8.13383222e-01
-7.53043473e-01 -7.47794390e-01 6.43203437e-01 1.12755798e-01
-4.80345607e-01 -1.32313204e+00 1.28531605e-01 5.27802408e-01
5.86709857e-01 1.31146371e-01 1.01307487e+00 -1.16436660e+00
-5.09723961e-01 -8.05156589e-01 -7.05308795e-01 4.53562617e-01
5.32928765e-01 -3.36534709e-01 -1.06340384e+00 -3.17590505e-01
-1.10103741e-01 -1.06760144e-01 5.63350677e-01 -2.58266509e-01
1.66746271e+00 -9.64137435e-01 -5.81520081e-01 9.64423358e-01
1.40851092e+00 2.33226091e-01 8.70710492e-01 1.95594057e-01
1.21143413e+00 4.68826056e-01 4.95986007e-02 5.32735765e-01
4.58818108e-01 3.02251130e-01 3.09707552e-01 8.90282094e-02
1.80009618e-01 -9.28063989e-01 1.03909649e-01 2.61556387e-01
1.85271144e-01 -4.76633847e-01 -8.71436298e-01 5.26206970e-01
-1.62286603e+00 -8.08806777e-01 2.29438171e-01 2.26805663e+00
7.94631839e-01 -3.69181409e-02 6.88906461e-02 2.01022830e-02
4.78068531e-01 3.47356111e-01 -5.68687856e-01 -3.64331126e-01
4.61269081e-01 2.86239833e-01 9.83516037e-01 4.44595575e-01
-1.22606838e+00 8.56192887e-01 4.97279406e+00 7.60586739e-01
-9.38310981e-01 1.70363262e-01 4.93769825e-01 2.98106432e-01
-6.29624665e-01 -1.73239142e-01 -1.35521340e+00 6.36385679e-01
1.17484748e+00 -3.26606542e-01 7.06758380e-01 8.38810086e-01
-1.95481583e-01 5.74518561e-01 -1.02335143e+00 1.14612281e+00
9.04058814e-02 -1.16171694e+00 6.72976822e-02 5.97865283e-01
2.21866608e-01 -8.73699039e-02 2.92488992e-01 5.92512667e-01
3.32540989e-01 -9.63476837e-01 -3.06175679e-01 2.10108310e-01
9.47636127e-01 -9.78002191e-01 6.99070573e-01 1.89077944e-01
-1.03992665e+00 -2.53175825e-01 -4.91724491e-01 3.17162901e-01
-1.13292597e-01 5.14995873e-01 -1.01515985e+00 5.19203305e-01
5.28163612e-01 9.88876760e-01 -5.75078011e-01 6.59448206e-01
-4.09798652e-01 8.29818130e-01 -4.84634846e-01 -1.85062245e-01
2.45516896e-01 -4.63696271e-01 3.11667830e-01 1.33546913e+00
2.26577654e-01 7.82456994e-02 5.58730774e-02 8.35781276e-01
-5.63291192e-01 2.07661092e-01 -1.21054482e+00 -7.01891422e-01
3.86456460e-01 1.31316733e+00 4.04443443e-02 -9.88816693e-02
-1.07309914e+00 1.13941646e+00 6.54595494e-01 4.55446154e-01
-5.06384552e-01 -5.62502265e-01 1.14447427e+00 4.90615457e-01
8.05195123e-02 -1.58135191e-01 -1.48456275e-01 -1.40438080e+00
-2.92716939e-02 -8.41128409e-01 8.01009119e-01 -9.08647105e-02
-1.72350442e+00 4.36299354e-01 -2.03355715e-01 -1.20279014e+00
-1.57143816e-01 -6.91785932e-01 -7.60786772e-01 8.55435669e-01
-1.96221375e+00 -1.37311566e+00 -1.42931910e-02 6.92649245e-01
-3.02349478e-01 -1.63541809e-01 1.01218569e+00 6.43513978e-01
-9.13318336e-01 1.25712073e+00 -1.40894324e-01 8.04294348e-01
5.43785334e-01 -9.30953979e-01 5.96916854e-01 7.23237276e-01
7.41003975e-02 1.14370203e+00 -6.95059970e-02 -9.01768982e-01
-1.37882912e+00 -1.52964413e+00 1.23294258e+00 -6.85440361e-01
8.50161433e-01 -6.02089822e-01 -1.22474456e+00 1.01114380e+00
2.98427064e-02 4.81398143e-02 1.59128845e+00 2.38298610e-01
-6.93733871e-01 -2.16558799e-01 -1.46504748e+00 7.21664131e-01
1.38644671e+00 -8.67573082e-01 -1.64420620e-01 4.31240767e-01
5.98500967e-01 2.43352056e-01 -1.09048724e+00 2.80455481e-02
3.78269643e-01 -6.31976187e-01 1.10028863e+00 -8.20328712e-01
8.93864557e-02 7.45123327e-02 3.21879417e-01 -9.44550157e-01
-4.04100120e-01 -7.37223923e-01 -3.49145174e-01 1.58218956e+00
6.27198517e-01 -1.03882539e+00 1.07895827e+00 1.47327614e+00
4.36260998e-01 -4.79542166e-01 -5.39008439e-01 -5.81940114e-01
-1.88642129e-01 -6.02330208e-01 8.92231762e-01 1.01540399e+00
2.70191491e-01 1.26073420e-01 -6.03467822e-01 4.84648734e-01
7.48638272e-01 -3.23919296e-01 8.29946280e-01 -1.18634439e+00
-6.80036172e-02 -2.17268780e-01 -5.44707298e-01 -6.55115724e-01
4.34431523e-01 -1.24999297e+00 -7.51847088e-01 -1.33626580e+00
8.99428800e-02 -5.76838851e-01 -4.89225775e-01 8.45477760e-01
-3.27510804e-01 5.22176437e-02 -5.08913659e-02 4.23089899e-02
-4.06030864e-01 4.32977676e-01 7.52491295e-01 -3.93339694e-01
-3.84867847e-01 2.40269542e-01 -1.26753604e+00 7.38117397e-01
9.71634388e-01 -5.22507310e-01 -6.47183418e-01 -2.12896794e-01
5.08986771e-01 -3.12628120e-01 5.17670512e-02 -7.74653435e-01
8.47444311e-02 -2.58803498e-02 2.51954347e-01 -3.14072788e-01
-6.42412379e-02 -1.23217523e+00 -2.74888963e-01 -1.51755875e-02
-2.68753886e-01 -4.97768998e-01 2.23524332e-01 8.75811160e-01
-4.79460359e-02 -3.47088337e-01 4.58496153e-01 -6.34688139e-02
-5.37540317e-01 1.25475538e+00 -3.60476881e-01 1.42226785e-01
8.66709471e-01 -4.14603800e-02 -5.37944101e-02 -3.78973663e-01
-3.10846835e-01 2.94069409e-01 3.96091610e-01 4.81700689e-01
7.30016291e-01 -1.33121109e+00 -3.71679157e-01 5.47075272e-01
5.36258399e-01 -3.76777798e-01 5.04102446e-02 2.16412514e-01
-2.38742977e-01 3.24257284e-01 -1.54488254e-02 8.57731104e-02
-1.17385125e+00 1.05509162e+00 -9.24348012e-02 -6.02904141e-01
-6.29326165e-01 6.83518648e-01 3.06718171e-01 -7.00136960e-01
4.06932682e-01 -7.07598701e-02 -3.24474454e-01 -5.10781668e-02
1.11041510e+00 2.05232516e-01 -8.96945372e-02 -2.84602195e-01
-3.66675407e-01 1.51928395e-01 -6.20315671e-01 4.21962947e-01
1.32540953e+00 -6.22378178e-02 -6.24678060e-02 -3.56468946e-01
2.04628587e+00 1.85882062e-01 -9.04846609e-01 -6.13041222e-01
2.25125626e-01 -8.93176854e-01 -1.45020917e-01 -5.30222356e-01
-1.05616450e+00 1.09578478e+00 2.38971874e-01 7.73967728e-02
9.95267451e-01 -4.53237981e-01 1.61908913e+00 4.31328714e-01
4.74135876e-01 -8.81190181e-01 -1.92063108e-01 1.12739317e-01
5.70683181e-01 -1.40037978e+00 -1.26922905e-01 -6.53654337e-01
-6.68625295e-01 9.26696956e-01 4.91051197e-01 8.15880299e-02
1.22786283e+00 3.35215032e-02 -1.94822311e-01 -7.10347369e-02
4.37862203e-02 1.04741685e-01 4.24255162e-01 9.82034326e-01
1.27260476e-01 3.03039737e-02 -1.68077365e-01 9.17201877e-01
2.09267691e-01 3.48030287e-03 1.35662228e-01 8.84527743e-01
-2.70001423e-02 -1.61212802e+00 2.72935271e-01 8.02732229e-01
-6.92988515e-01 -1.60992995e-01 -4.62566584e-01 4.76834416e-01
-1.06209680e-01 9.08490300e-01 -4.32980418e-01 -6.07247949e-01
1.87310904e-01 2.15276822e-01 -7.63355419e-02 -7.35069573e-01
-7.40329087e-01 -6.42208695e-01 2.51017064e-01 -5.85566640e-01
4.64033410e-02 -3.61847728e-01 -8.25813293e-01 -5.80402017e-01
-4.34506051e-02 -1.74956709e-01 5.58558106e-01 7.07214355e-01
7.19196975e-01 -1.17912620e-01 7.65425861e-01 -4.06221777e-01
-3.45562816e-01 -6.40315771e-01 -4.55655038e-01 8.68673265e-01
1.74877420e-01 -2.41110995e-01 -3.64871770e-01 -3.95262361e-01] | [6.05755615234375, 7.040432929992676] |
58cf7e6d-222f-4f61-8f4a-138f2d74d185 | unsupervised-person-re-identification-by-soft | 1903.06325 | null | http://arxiv.org/abs/1903.06325v2 | http://arxiv.org/pdf/1903.06325v2.pdf | Unsupervised Person Re-identification by Soft Multilabel Learning | Although unsupervised person re-identification (RE-ID) has drawn increasing
research attentions due to its potential to address the scalability problem of
supervised RE-ID models, it is very challenging to learn discriminative
information in the absence of pairwise labels across disjoint camera views. To
overcome this problem, we propose a deep model for the soft multilabel learning
for unsupervised RE-ID. The idea is to learn a soft multilabel (real-valued
label likelihood vector) for each unlabeled person by comparing (and
representing) the unlabeled person with a set of known reference persons from
an auxiliary domain. We propose the soft multilabel-guided hard negative mining
to learn a discriminative embedding for the unlabeled target domain by
exploring the similarity consistency of the visual features and the soft
multilabels of unlabeled target pairs. Since most target pairs are cross-view
pairs, we develop the cross-view consistent soft multilabel learning to achieve
the learning goal that the soft multilabels are consistently good across
different camera views. To enable effecient soft multilabel learning, we
introduce the reference agent learning to represent each reference person by a
reference agent in a joint embedding. We evaluate our unified deep model on
Market-1501 and DukeMTMC-reID. Our model outperforms the state-of-the-art
unsupervised RE-ID methods by clear margins. Code is available at
https://github.com/KovenYu/MAR. | ['An-Cong Wu', 'Jian-Huang Lai', 'Wei-Shi Zheng', 'Hong-Xing Yu', 'Xiaowei Guo', 'Shaogang Gong'] | 2019-03-15 | unsupervised-person-re-identification-by-soft-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Yu_Unsupervised_Person_Re-Identification_by_Soft_Multilabel_Learning_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Yu_Unsupervised_Person_Re-Identification_by_Soft_Multilabel_Learning_CVPR_2019_paper.pdf | cvpr-2019-6 | ['unsupervised-person-re-identification'] | ['computer-vision'] | [-3.7110221e-01 -9.1752447e-02 -3.9248070e-01 -8.4366363e-01
-9.0923136e-01 -7.4645746e-01 9.9625582e-01 -1.0825861e-01
-5.9853351e-01 6.5087545e-01 4.3969771e-01 4.2832989e-01
1.5311673e-01 -1.3549976e-01 -5.5397326e-01 -6.5478885e-01
2.7639100e-01 9.3421388e-01 -3.3200413e-01 3.3690891e-01
-3.3639687e-01 3.2743615e-01 -1.3239814e+00 -9.9326849e-02
6.6915202e-01 4.3098396e-01 -4.2635489e-02 4.1882756e-01
4.0008286e-01 6.4945650e-01 -3.5764861e-01 -9.0895021e-01
6.5908688e-01 -2.9170087e-01 -7.8408056e-01 2.2546807e-01
1.1901920e+00 -5.0342351e-01 -5.3206730e-01 1.3530308e+00
6.7760271e-01 2.2218785e-01 1.0067345e+00 -1.5716175e+00
-1.0193630e+00 2.8226703e-01 -8.6179954e-01 2.1358799e-02
4.5383146e-01 1.8799338e-02 1.0938610e+00 -1.0348115e+00
7.2414052e-01 1.4866122e+00 5.2823418e-01 8.9168781e-01
-1.4286857e+00 -9.7910577e-01 3.4513959e-01 3.2754764e-01
-1.6700957e+00 -4.9616739e-01 4.7369459e-01 -6.7414242e-01
5.2392137e-01 -1.2703307e-01 1.9099019e-01 1.2137835e+00
-2.1953665e-01 8.0419433e-01 1.1766645e+00 -2.4063703e-01
-7.2219968e-02 3.7554628e-01 5.2003586e-01 7.1236950e-01
2.0588863e-01 2.7678445e-01 -4.2708305e-01 -1.2985817e-01
3.9372647e-01 4.0983814e-01 -1.7877136e-03 -6.4481109e-01
-1.1957188e+00 7.3562443e-01 4.6796441e-01 3.3338215e-02
8.9877151e-02 6.0565490e-03 3.1775683e-01 2.1896005e-01
3.5022277e-01 9.4808586e-02 -1.4054793e-01 3.8588467e-01
-6.8955296e-01 2.3974317e-01 5.9717375e-01 9.6724498e-01
9.9145830e-01 -2.1425986e-01 -1.6907535e-01 1.0591186e+00
6.1796600e-01 8.8183105e-01 2.9979208e-01 -9.2651582e-01
4.1245624e-01 5.6641191e-01 2.2800331e-01 -7.5342947e-01
-1.9508061e-01 -2.9923332e-01 -1.0130090e+00 3.0741298e-01
5.1768863e-01 -4.7529873e-02 -9.4274145e-01 2.0484152e+00
4.0182376e-01 4.0462640e-01 7.9971917e-02 1.0988321e+00
9.1314316e-01 4.6083185e-01 3.9198038e-01 1.7566200e-01
1.3224825e+00 -1.1418378e+00 -5.0113827e-01 -4.0046218e-01
7.9987264e-01 -5.2475369e-01 6.8603975e-01 -6.0283117e-02
-5.9085912e-01 -7.4563813e-01 -1.0792810e+00 -2.3545048e-01
-3.9627787e-01 4.4176731e-01 3.5351470e-02 5.8677208e-01
-1.0145510e+00 1.0627956e-02 -4.1299599e-01 -5.8353829e-01
1.8243313e-01 5.8160859e-01 -1.0694108e+00 -5.2275079e-01
-1.1566504e+00 9.0349150e-01 2.3918833e-01 -3.8958713e-02
-1.1313039e+00 -4.3271151e-01 -1.0311412e+00 -1.5837643e-01
2.5989309e-01 -6.3731211e-01 1.0368677e+00 -1.1385741e+00
-1.0641109e+00 1.5128752e+00 -3.0936748e-01 -1.2284532e-01
5.1787490e-01 -1.9222228e-01 -4.8657805e-01 -9.0679832e-02
5.8117855e-01 8.9946151e-01 7.6212150e-01 -1.6358628e+00
-6.1756784e-01 -5.5782795e-01 9.8042682e-02 4.9895969e-01
-3.7467700e-01 1.8649893e-01 -6.8181252e-01 -7.0624411e-01
-2.6436710e-01 -1.3715519e+00 -2.0144226e-02 -1.9740063e-01
-5.0457412e-01 -5.3657222e-01 6.9217092e-01 -7.6959872e-01
6.8967652e-01 -2.2553124e+00 2.0878614e-01 1.5096939e-01
4.6727166e-01 2.4476162e-01 -4.0137494e-01 7.6052551e-03
-3.6384740e-01 -4.9398858e-02 3.4547407e-02 -1.0346712e+00
2.6285440e-01 1.9086511e-01 -6.1428741e-02 8.9174777e-01
-1.2853703e-01 9.3576211e-01 -9.3339300e-01 -5.8703589e-01
2.8643689e-01 4.2679706e-01 -9.6090980e-02 5.6454188e-01
3.5707906e-01 4.6744967e-01 -1.1043547e-01 7.2563839e-01
8.3098483e-01 -3.5338655e-01 3.7790811e-01 -4.4761056e-01
2.3415750e-01 -3.1040949e-01 -1.4459205e+00 1.3912202e+00
-6.3873935e-03 3.8667834e-01 1.0197021e-01 -8.4969962e-01
6.5484548e-01 1.3905901e-01 3.1375948e-01 -6.9094265e-01
-6.0829122e-02 -4.1603319e-02 -4.0681776e-01 -9.4921574e-02
3.6858949e-01 -1.5231189e-01 -3.3304599e-01 6.8951440e-01
3.3662599e-01 5.6532729e-01 1.6246809e-01 4.6765891e-01
5.2234119e-01 -2.3466416e-02 3.2211342e-01 -1.8078014e-01
6.3022387e-01 -3.9093894e-01 8.3160031e-01 7.5205332e-01
-4.7527960e-01 6.2546563e-01 -1.8363552e-02 -4.7797459e-01
-1.0434921e+00 -1.4854829e+00 -8.2265861e-02 1.2576990e+00
5.7978326e-01 -3.4076318e-01 -5.1059926e-01 -1.1319201e+00
2.2496761e-01 2.1989733e-01 -6.8245023e-01 1.3155915e-01
-4.3527025e-01 -5.8200997e-01 4.2236745e-01 5.7000959e-01
5.6283402e-01 -5.5016756e-01 4.0983424e-01 -2.9678488e-01
-2.9782423e-01 -1.1432337e+00 -1.0552748e+00 6.5041415e-02
-1.6636440e-01 -1.1324961e+00 -1.0270656e+00 -1.1330053e+00
9.7763038e-01 5.2624524e-01 1.0352170e+00 -2.0545982e-01
-8.1795350e-02 7.7914864e-01 -1.6535544e-01 8.7619640e-02
-2.2080040e-01 -1.4958221e-01 7.9600096e-01 1.7674841e-01
7.9138255e-01 -8.9517543e-03 -4.2237630e-01 6.8639618e-01
-7.0473480e-01 2.1326274e-02 3.6825737e-01 1.0036021e+00
8.3946788e-01 -3.1067154e-01 5.5417240e-01 -6.4648521e-01
3.9916211e-01 -5.0443316e-01 -5.0366247e-01 6.4882165e-01
-6.0269666e-01 1.2695311e-01 4.6808559e-01 -6.1287636e-01
-1.0039897e+00 2.5558719e-01 3.0180082e-01 -4.8090622e-01
-2.2401609e-01 1.6150558e-01 -5.2591366e-01 -1.7924054e-02
3.9158094e-01 -1.5819651e-01 -1.9577698e-01 -4.3723017e-01
6.0515827e-01 8.0252403e-01 7.8108370e-01 -6.3237256e-01
1.0615383e+00 4.4118226e-01 -2.6836941e-01 -4.2058900e-01
-9.2647201e-01 -8.7102360e-01 -7.9497725e-01 -2.6164699e-01
1.1410763e+00 -1.5264146e+00 -5.8610243e-01 5.2283067e-01
-9.6547729e-01 -2.2424901e-01 -1.1128021e-01 5.5580437e-01
-1.8640535e-01 7.6575226e-01 -5.2093285e-01 -4.9897602e-01
-1.6647916e-01 -1.1778668e+00 1.0020093e+00 4.1502926e-01
-3.3569553e-01 -1.0223564e+00 4.3088001e-01 5.6302965e-01
-3.5688916e-01 -5.4822754e-02 3.8978860e-01 -9.3017066e-01
-4.0403536e-01 -4.2649868e-01 -5.4439050e-01 6.0474008e-01
1.3978010e-01 -4.3715957e-01 -1.0205915e+00 -7.3876798e-01
-4.0095377e-01 -6.9906378e-01 7.7354020e-01 2.0388681e-01
6.6842496e-01 -1.7985585e-01 -5.7754749e-01 7.0447350e-01
1.2217817e+00 -2.5216594e-01 1.2242120e-01 3.4191537e-01
1.2269931e+00 6.3950372e-01 5.7176518e-01 3.2109600e-01
9.1598713e-01 8.2888478e-01 1.3026640e-01 -3.1965569e-01
-6.5604761e-02 -3.9528701e-01 5.5116749e-01 4.3682975e-01
1.7918283e-02 -3.6423168e-01 -8.3653635e-01 6.4116049e-01
-2.0287073e+00 -1.0285560e+00 1.6956457e-01 2.2952085e+00
5.8735925e-01 -3.9541858e-01 2.9274046e-01 -5.5356830e-01
1.1978955e+00 8.1551298e-02 -6.5678847e-01 3.0810518e-02
-3.2509330e-01 -5.0233215e-01 6.4665085e-01 6.3683367e-01
-1.4073309e+00 8.6714309e-01 5.4351945e+00 6.5201217e-01
-6.3193846e-01 3.1583047e-01 7.1689743e-01 -5.4359805e-02
-8.6151153e-02 -1.6050591e-01 -1.3017503e+00 4.8152032e-01
6.6909307e-01 -1.0011133e-01 3.4506443e-01 8.2876885e-01
-1.1710448e-01 1.2144009e-02 -1.5545079e+00 1.5305495e+00
5.3061372e-01 -9.1526896e-01 1.0940383e-02 2.3247562e-01
1.1942143e+00 4.7418069e-02 2.3875879e-01 3.9401287e-01
8.8764924e-01 -1.0329012e+00 5.5541110e-01 4.2508367e-01
1.0028783e+00 -7.0182568e-01 8.8885742e-01 2.0733172e-01
-1.2702690e+00 -1.3364953e-01 -3.5895318e-01 3.2503879e-01
1.9323546e-01 3.5823017e-01 -6.5741462e-01 3.9815730e-01
8.8493961e-01 1.1531532e+00 -7.0412123e-01 6.4712852e-01
-2.7637154e-01 1.4306280e-01 -2.2199708e-01 5.9446889e-01
1.1166099e-02 -2.3004304e-01 2.5181744e-01 1.3114196e+00
8.5759766e-02 -8.6714208e-02 6.7665142e-01 6.4705652e-01
-4.9132526e-01 -7.6881453e-02 -5.2721709e-01 1.4480524e-01
3.7578571e-01 1.2102264e+00 -1.8283491e-01 -4.9938953e-01
-6.6189879e-01 1.3028511e+00 5.2261311e-01 6.3797247e-01
-7.1321410e-01 2.2149824e-01 8.7894154e-01 -6.4722300e-02
-1.5815966e-01 -8.6731188e-02 4.9284890e-02 -1.4703159e+00
1.9305695e-02 -8.3584088e-01 8.8125092e-01 -6.0815632e-01
-1.9176592e+00 3.9018503e-01 8.4587924e-02 -1.3723747e+00
-4.5433038e-01 -5.2329451e-01 -2.7584317e-01 1.1494230e+00
-1.5383906e+00 -1.5808647e+00 -3.4189469e-01 8.6541015e-01
2.4180450e-01 -5.6578940e-01 8.2061267e-01 3.4238729e-01
-6.7245239e-01 1.1584563e+00 3.4056824e-01 5.1744419e-01
1.5869136e+00 -1.3914809e+00 2.1541916e-01 9.7104168e-01
-1.1160953e-02 7.5159883e-01 4.2905292e-01 -7.3714179e-01
-1.1141168e+00 -1.2578193e+00 9.5670968e-01 -7.4220753e-01
5.3318578e-01 -4.7261882e-01 -6.3955837e-01 1.0464228e+00
9.7889677e-02 2.3983029e-01 1.0046766e+00 1.0962994e-01
-8.7181240e-01 -2.2708096e-01 -1.1656578e+00 4.5201585e-01
9.0939939e-01 -8.3286726e-01 -6.0441244e-01 3.5324201e-01
3.1467655e-01 3.9159648e-02 -8.1064087e-01 2.8386348e-01
5.8342767e-01 -6.9479346e-01 1.3029946e+00 -3.5474479e-01
-2.4947330e-02 -6.6460013e-01 -1.4171317e-01 -1.3148061e+00
-5.5376685e-01 -8.7334793e-03 5.1163312e-02 1.5500445e+00
1.9506181e-02 -6.8334389e-01 6.4959890e-01 9.7391355e-01
2.9994106e-01 -5.9688058e-02 -8.5165888e-01 -8.9449358e-01
5.2567208e-03 1.5207343e-01 3.0655620e-01 1.2663412e+00
-8.3000809e-02 4.2551053e-01 -8.6493784e-01 5.2883023e-01
1.1987399e+00 -6.6076457e-03 9.7757226e-01 -1.3437181e+00
-3.0674052e-01 1.7918764e-02 -5.8941907e-01 -1.1084496e+00
8.4218156e-01 -1.2606864e+00 2.2200478e-02 -1.4403983e+00
9.6037370e-01 -3.4737748e-01 -5.1409942e-01 5.8022720e-01
-3.2583910e-01 4.9243441e-01 3.7309524e-01 5.1179457e-01
-1.1303188e+00 4.2979670e-01 9.0314478e-01 -6.2973309e-01
6.1566275e-02 -3.5529980e-01 -7.4999535e-01 6.6335225e-01
4.8749417e-01 -5.3810829e-01 -2.8129512e-01 -5.8638573e-01
-2.0649903e-02 -3.1171304e-01 6.6373551e-01 -8.4110999e-01
3.9755532e-01 -1.1127023e-02 6.7244858e-01 -4.9775377e-01
2.6325139e-01 -8.1196010e-01 2.4420451e-01 9.2144720e-02
-4.6174404e-01 4.5199145e-02 -1.9518290e-01 6.8181068e-01
-2.3786028e-01 -1.4906858e-01 9.7715998e-01 -6.9586150e-02
-1.1350046e+00 7.1091133e-01 5.7943622e-03 8.7002635e-02
1.0600890e+00 -2.3812026e-01 -4.4096968e-01 -3.5947955e-01
-8.4550655e-01 7.2023576e-01 9.1643268e-01 5.4279041e-01
4.2912650e-01 -1.6119851e+00 -9.2087907e-01 1.1733719e-01
6.6875923e-01 -2.7742410e-01 4.3788379e-01 4.3672484e-01
4.0511142e-02 2.0576580e-01 -3.0463365e-01 -6.4831054e-01
-1.6284326e+00 6.8157703e-01 4.5614317e-01 -3.0287123e-01
-3.3526626e-01 8.7577027e-01 8.7917966e-01 -8.9622152e-01
3.6054519e-01 5.4504257e-01 -1.9453475e-01 -1.4414106e-04
8.6228639e-01 4.1865352e-01 -3.8324538e-01 -1.2699715e+00
-5.3240561e-01 8.1694955e-01 -4.6045107e-01 -7.6403655e-02
1.0305327e+00 -4.8566622e-01 -1.5380681e-01 1.8433356e-01
1.5574375e+00 -1.3216160e-01 -1.4005654e+00 -6.1126101e-01
-2.1920392e-01 -5.1656193e-01 -2.5523302e-01 -7.8587347e-01
-7.8886420e-01 5.6976908e-01 1.1430179e+00 -5.0178516e-01
7.1898139e-01 3.1995592e-01 4.5102477e-01 4.2848116e-01
1.6017880e-01 -1.1826967e+00 2.3894294e-01 4.5633104e-01
5.1659793e-01 -1.9209831e+00 5.5425152e-02 -8.9324944e-02
-8.3404267e-01 6.5978187e-01 8.4486192e-01 2.2266475e-02
5.7113606e-01 6.3348420e-02 3.1967723e-01 6.8646766e-02
-1.5863901e-01 -2.3173258e-01 4.3747067e-01 8.7236667e-01
2.9977059e-01 4.5715815e-01 1.2898651e-01 5.1465899e-01
1.7172423e-01 -3.0757877e-01 2.3399736e-01 4.0828916e-01
-6.7178351e-03 -1.3250588e+00 -5.2221507e-01 2.3811436e-01
-2.1583892e-01 1.3975202e-01 -4.9816734e-01 5.1245457e-01
3.0118483e-01 9.0085703e-01 -4.1078158e-02 -4.7373080e-01
6.0151637e-02 -5.5131819e-03 2.9878694e-01 -7.2187650e-01
-2.4673173e-01 -1.4286935e-01 -7.5918652e-02 -2.6075569e-01
-6.7898250e-01 -7.7077353e-01 -9.3946087e-01 -4.2298311e-01
4.9602180e-03 -6.0248166e-02 1.7606588e-01 9.1785198e-01
1.3593486e-01 -1.4314595e-01 6.2170291e-01 -1.0117711e+00
-4.9223414e-01 -8.2377577e-01 -7.9202598e-01 9.6887589e-01
3.5654110e-01 -8.1345469e-01 -5.1799875e-01 3.1308520e-01] | [14.780006408691406, 1.0562111139297485] |
90e5fabc-7ea4-42cb-9e68-d061eafd8d4a | robust-optimization-and-validation-of-echo | 2103.03174 | null | https://arxiv.org/abs/2103.03174v2 | https://arxiv.org/pdf/2103.03174v2.pdf | Robust Optimization and Validation of Echo State Networks for learning chaotic dynamics | An approach to the time-accurate prediction of chaotic solutions is by learning temporal patterns from data. Echo State Networks (ESNs), which are a class of Reservoir Computing, can accurately predict the chaotic dynamics well beyond the predictability time. Existing studies, however, also showed that small changes in the hyperparameters may markedly affect the network's performance. The aim of this paper is to assess and improve the robustness of Echo State Networks for the time-accurate prediction of chaotic solutions. The goal is three-fold. First, we investigate the robustness of routinely used validation strategies. Second, we propose the Recycle Validation, and the chaotic versions of existing validation strategies, to specifically tackle the forecasting of chaotic systems. Third, we compare Bayesian optimization with the traditional Grid Search for optimal hyperparameter selection. Numerical tests are performed on two prototypical nonlinear systems that have both chaotic and quasiperiodic solutions. Both model-free and model-informed Echo State Networks are analysed. By comparing the network's robustness in learning chaotic versus quasiperiodic solutions, we highlight fundamental challenges in learning chaotic solutions. The proposed validation strategies, which are based on the dynamical systems properties of chaotic time series, are shown to outperform the state-of-the-art validation strategies. Because the strategies are principled-they are based on chaos theory such as the Lyapunov time-they can be applied to other Recurrent Neural Networks architectures with little modification. This work opens up new possibilities for the robust design and application of Echo State Networks, and Recurrent Neural Networks, to the time-accurate prediction of chaotic systems. | ['Luca Magri', 'Alberto Racca'] | 2021-02-09 | null | null | null | null | ['robust-design'] | ['miscellaneous'] | [-2.20212579e-01 -3.90656382e-01 2.81140655e-01 5.09886146e-02
-6.51254654e-02 -5.63822925e-01 7.81792581e-01 -2.98582911e-01
-2.08537340e-01 9.71292794e-01 -2.61738777e-01 -4.43184465e-01
-6.87716722e-01 -5.11003435e-01 -1.41112536e-01 -1.17055237e+00
-7.27269232e-01 5.57037413e-01 1.68912172e-01 -5.68819582e-01
5.96700132e-01 7.89533556e-01 -1.64928293e+00 -3.01674575e-01
7.51464546e-01 1.05882728e+00 -4.21944335e-02 7.16912150e-01
2.79833287e-01 6.35964096e-01 -5.26224196e-01 3.96218419e-01
2.37997159e-01 -5.77037930e-01 -1.47097141e-01 -6.27869785e-01
-3.29098940e-01 4.23017330e-03 -4.54958498e-01 9.68636215e-01
5.90391994e-01 1.38064146e-01 7.48866022e-01 -9.08621013e-01
-3.90131205e-01 6.46258831e-01 4.72551495e-01 5.26412129e-01
-2.30885014e-01 4.27573889e-01 1.71785176e-01 -7.50482678e-01
3.07706416e-01 7.82545805e-01 1.21586037e+00 4.97115165e-01
-1.16915524e+00 -8.07930768e-01 -4.34672743e-01 2.61164218e-01
-1.37708640e+00 -6.39553010e-01 9.02545333e-01 -6.56086028e-01
1.05225611e+00 3.00058633e-01 8.59882653e-01 8.88192058e-01
7.21601784e-01 -2.01501697e-01 1.79293025e+00 -5.41241407e-01
3.61645311e-01 3.25709581e-01 3.07129234e-01 5.30992746e-01
3.22500288e-01 1.16881347e+00 -2.24173158e-01 -2.69707918e-01
6.28426075e-01 -4.07563984e-01 -3.57612491e-01 -1.37368530e-01
-1.08234549e+00 8.72402847e-01 1.59140065e-01 8.39217126e-01
-3.51255506e-01 1.71372995e-01 3.63395631e-01 6.15342021e-01
3.99105161e-01 1.04484117e+00 -3.63610387e-01 -1.99844703e-01
-1.25783312e+00 1.57557413e-01 1.37247348e+00 3.76664400e-01
4.74645138e-01 8.78267586e-01 3.46267894e-02 3.93855661e-01
3.83358777e-01 9.02967989e-01 1.01890707e+00 -7.80521512e-01
-2.56623596e-01 4.07775156e-02 1.81558073e-01 -1.17861819e+00
-7.16551960e-01 -7.03723788e-01 -1.21260238e+00 2.68448025e-01
2.74680793e-01 -3.60278726e-01 -7.71492362e-01 1.45271003e+00
-1.42030060e-01 4.70106840e-01 6.57674074e-01 6.86832190e-01
6.21562302e-01 9.23427761e-01 -4.23595428e-01 -7.49229312e-01
7.71454751e-01 -6.24975920e-01 -8.44668031e-01 2.13578328e-01
3.53948683e-01 -6.04170501e-01 5.21632135e-01 3.33817512e-01
-8.99579644e-01 -4.55952257e-01 -1.15904188e+00 8.74047577e-01
-6.24831557e-01 -9.86563042e-03 2.25076184e-01 7.69405305e-01
-1.24805534e+00 1.22722042e+00 -8.42294395e-01 -1.21129550e-01
-5.53193033e-01 4.47104961e-01 2.12053418e-01 7.02468514e-01
-1.68001986e+00 1.45011401e+00 6.19801700e-01 6.36085570e-01
-7.91875303e-01 -4.62250918e-01 -5.51362932e-01 6.40617609e-02
-2.47052506e-01 -1.05220281e-01 9.92770433e-01 -7.88841605e-01
-2.02718973e+00 1.70639008e-01 1.29620461e-02 -7.55228996e-01
3.39306712e-01 4.85637575e-01 -7.60466754e-01 1.21493533e-01
-2.59489805e-01 -3.88233922e-02 9.52178657e-01 -9.98414338e-01
1.31033868e-01 2.89570898e-01 -1.00777829e+00 -3.23600978e-01
-1.32669136e-01 -3.04511368e-01 4.40532923e-01 -4.85117346e-01
2.27005914e-01 -1.25566554e+00 -4.10736471e-01 -6.89269543e-01
-1.26754567e-01 -1.74678601e-02 9.04195189e-01 -4.79645550e-01
1.39426458e+00 -1.85195875e+00 -1.56081561e-02 8.26700211e-01
-9.32948142e-02 5.56119800e-01 1.26085803e-01 8.36527467e-01
-8.32724795e-02 1.12388901e-01 -4.79355872e-01 -1.19191393e-01
-9.23417732e-02 2.23177597e-01 -8.92057300e-01 4.77490932e-01
1.75348803e-01 9.24195230e-01 -6.40665054e-01 -5.94636463e-02
1.57764152e-01 4.90445971e-01 7.33722448e-02 1.45385712e-01
3.49948257e-02 6.04926109e-01 -2.29467258e-01 2.74419963e-01
3.52857053e-01 -2.76690543e-01 8.82621855e-02 9.01032388e-02
-7.52238691e-01 1.26806512e-01 -9.60166454e-01 5.82593262e-01
-2.52930135e-01 1.10330474e+00 -3.93167913e-01 -1.21060371e+00
1.56922162e+00 4.50647682e-01 4.38222975e-01 -7.80024946e-01
3.46156321e-02 8.76478136e-01 3.84455562e-01 -5.64850986e-01
4.43398207e-01 -2.60342658e-01 2.25969795e-02 6.74039721e-01
-4.42742482e-02 -4.46441650e-01 7.70676881e-02 -4.37389374e-01
5.85216999e-01 -4.64529656e-02 2.00344771e-01 -7.75023401e-01
6.54666781e-01 1.16937317e-01 2.68116981e-01 9.69588876e-01
-6.79201558e-02 1.84757888e-01 4.99061137e-01 -7.39393473e-01
-1.36821723e+00 -5.62689543e-01 -6.47104144e-01 9.21319649e-02
3.74460523e-03 -4.13574278e-02 -3.41550797e-01 2.04052776e-01
-4.75766994e-02 5.45345366e-01 -6.90513372e-01 -3.80248576e-01
-9.37517524e-01 -1.13837397e+00 8.96152616e-01 4.67332192e-02
3.44482571e-01 -1.39278853e+00 -8.39506328e-01 6.35114312e-01
2.49044716e-01 -7.19217718e-01 1.82619393e-01 4.64350224e-01
-1.07278359e+00 -8.37103903e-01 -8.86748672e-01 -5.73059678e-01
2.30177775e-01 -3.84332299e-01 1.00609231e+00 4.06015098e-01
1.05784632e-01 3.68623465e-01 -1.55428812e-01 -1.61842421e-01
-1.00213242e+00 1.34044006e-01 6.58555567e-01 -2.61193246e-01
1.41108140e-01 -9.30933416e-01 -3.80344033e-01 5.80193520e-01
-6.30525291e-01 -3.42075020e-01 4.54748839e-01 1.11986160e+00
4.61815864e-01 1.76775351e-01 9.38949168e-01 -3.65509480e-01
1.02193809e+00 -6.08824790e-01 -1.11952567e+00 4.39900100e-01
-1.48652601e+00 4.36080605e-01 9.19541419e-01 -8.45230460e-01
-4.13225383e-01 -2.52758563e-01 1.68160573e-01 -6.17114305e-01
2.64589459e-01 8.49949956e-01 1.00472069e+00 -6.86800182e-01
8.46652329e-01 8.50593269e-01 4.70821053e-01 -4.01224382e-02
-3.07921082e-01 4.88674998e-01 3.60821217e-01 -4.87435043e-01
7.68357813e-01 3.71643007e-02 4.29725945e-01 -1.14583731e+00
-2.73892172e-02 -1.67720035e-01 -4.28445756e-01 -4.40182090e-01
2.43112460e-01 -5.02831876e-01 -8.59768271e-01 7.33562589e-01
-1.01466942e+00 -5.79340696e-01 -3.81982446e-01 6.23563528e-01
-7.36742318e-01 1.63608253e-01 -5.49134493e-01 -1.22881556e+00
-5.03544748e-01 -1.07673311e+00 2.36117437e-01 4.70705479e-01
1.47900119e-01 -1.27078903e+00 7.37877071e-01 -9.01939034e-01
1.12435806e+00 3.65432531e-01 6.09631836e-01 -7.98892856e-01
-3.97703052e-01 -1.11895226e-01 2.80323148e-01 3.56802702e-01
-2.81783283e-01 4.15209860e-01 -8.10151398e-01 -4.40298557e-01
4.19786513e-01 -1.68399572e-01 8.92157435e-01 5.45281410e-01
2.03854442e-01 -4.88333523e-01 -2.36638710e-01 6.66465282e-01
1.41276610e+00 4.46987748e-01 5.91523945e-01 4.40902174e-01
-5.84565401e-02 7.55652905e-01 6.58579171e-02 3.18432987e-01
-1.90963268e-01 4.17648256e-01 1.36356816e-01 3.99841368e-01
1.93918526e-01 1.05562411e-01 4.71186697e-01 1.61381352e+00
-1.29480168e-01 6.98509142e-02 -1.23339748e+00 2.94083178e-01
-1.75318825e+00 -1.24886072e+00 -1.72795460e-01 2.11285424e+00
6.33771896e-01 9.69334915e-02 5.24980798e-02 2.29921669e-01
5.83440006e-01 3.28702405e-02 -6.15437627e-01 -6.05209589e-01
-4.30705845e-01 8.73547494e-02 3.59825760e-01 5.79383016e-01
-7.64720857e-01 7.56187677e-01 7.20586300e+00 4.99999225e-01
-1.69986451e+00 -1.32482886e-01 4.37053978e-01 2.58260876e-01
-5.27071133e-02 5.85832447e-02 -9.34586167e-01 6.77881718e-01
1.77054501e+00 -3.37522656e-01 6.97791040e-01 5.87362409e-01
3.66582960e-01 6.82342723e-02 -4.94302750e-01 7.22771943e-01
-5.85755594e-02 -1.54768276e+00 -3.63635868e-01 1.62225235e-02
9.47475493e-01 2.38607556e-01 2.48495251e-01 4.99699444e-01
-1.45522639e-01 -1.20838475e+00 6.09311342e-01 1.34291708e+00
6.89308167e-01 -6.58647597e-01 9.33060229e-01 3.70160431e-01
-9.30170774e-01 -2.22174436e-01 -4.68565464e-01 -1.41039297e-01
1.69709086e-01 5.44123888e-01 -5.75007558e-01 1.21148482e-01
7.13035882e-01 6.98448777e-01 -5.63835621e-01 1.21996856e+00
1.37450114e-01 8.01543415e-01 -6.54746175e-01 -8.18378150e-01
4.99333382e-01 -4.90008771e-01 8.96973252e-01 1.21857977e+00
7.30113983e-01 2.03972712e-01 -2.66971201e-01 1.06784797e+00
8.27252507e-01 -3.16308826e-01 -8.57358694e-01 -2.55724490e-01
5.86717308e-01 9.34154451e-01 -7.84655929e-01 -2.80121565e-01
4.20223117e-01 1.07341059e-01 -1.53488621e-01 4.57862347e-01
-4.67492133e-01 -5.78510642e-01 1.41458213e-01 -2.50312775e-01
3.69539291e-01 -5.77524185e-01 -2.03534052e-01 -1.05788815e+00
-2.16872633e-01 -8.39891672e-01 -1.46360919e-01 -6.07686460e-01
-1.23870885e+00 1.23371649e+00 2.32990980e-01 -1.46568251e+00
-1.10395491e+00 -6.76838458e-01 -8.55459690e-01 1.04457641e+00
-1.70793808e+00 -5.21726072e-01 -3.00304648e-02 4.64067668e-01
4.16683927e-02 -6.73002601e-01 9.60303009e-01 -2.93068737e-01
-3.28707516e-01 2.92451501e-01 9.28230762e-01 -2.58355290e-01
9.97697636e-02 -9.26576257e-01 2.12153152e-01 7.01519132e-01
-2.52529413e-01 6.93641782e-01 1.18330133e+00 -6.90172970e-01
-1.06858015e+00 -6.43076897e-01 8.05410862e-01 -2.20199496e-01
8.99006426e-01 -1.61666542e-01 -1.22370219e+00 3.10402978e-02
4.54967730e-02 3.34448218e-02 2.57729858e-01 -3.35230589e-01
5.94940186e-02 6.91560358e-02 -9.24009860e-01 3.54707301e-01
2.27017671e-01 -4.93683934e-01 -6.20420218e-01 2.47166336e-01
2.74584234e-01 -2.25129902e-01 -8.75303447e-01 6.46382093e-01
9.09711599e-01 -1.10642719e+00 7.20164180e-01 -5.73119164e-01
1.55752882e-01 -2.60783523e-01 2.12633163e-02 -1.35105884e+00
-2.61118650e-01 -1.08518195e+00 -4.68840778e-01 6.72833204e-01
3.78259927e-01 -1.26516521e+00 4.35467750e-01 3.42130750e-01
-8.41746032e-02 -8.47498477e-01 -1.32726026e+00 -1.27518749e+00
4.16718245e-01 -5.46243824e-02 4.60609227e-01 9.28034067e-01
5.08606322e-02 -2.74155110e-01 -4.87523884e-01 3.03005368e-01
5.82939327e-01 2.39837989e-01 2.70359367e-01 -1.52495027e+00
-1.74346641e-01 -7.82713771e-01 -2.73334622e-01 -3.63001078e-01
1.84610978e-01 -5.23892701e-01 2.62333721e-01 -6.40658617e-01
-6.44625485e-01 -9.12956536e-01 -3.63588870e-01 -4.50891210e-03
3.88852358e-01 2.61608839e-01 1.32271171e-01 9.41277027e-01
1.02978416e-01 6.39435709e-01 8.72385323e-01 2.86321104e-01
-5.42130470e-01 3.19514424e-01 1.45338908e-01 2.98979729e-01
1.10055745e+00 -7.93196201e-01 -1.55592546e-01 3.62101078e-01
5.09982884e-01 3.57281476e-01 4.99918461e-01 -1.40750742e+00
6.64140522e-01 7.35325888e-02 2.01781064e-01 -5.52081048e-01
2.21068203e-01 -6.58938408e-01 4.40333813e-01 1.08569312e+00
-2.17372283e-01 5.37926912e-01 3.86221349e-01 5.96381664e-01
-2.68962473e-01 -6.90403581e-01 8.73024523e-01 -1.02452867e-01
-4.49338496e-01 -2.99824942e-02 -7.71163106e-01 -1.54362962e-01
8.33552003e-01 -2.81367600e-01 -3.87772262e-01 -4.39052731e-01
-7.17837751e-01 5.11340536e-02 1.71075046e-01 7.18853157e-03
5.06274700e-01 -1.12608862e+00 -6.84757292e-01 5.52085042e-01
-4.20771211e-01 -7.32684135e-01 4.44970392e-02 1.17416894e+00
-5.65207541e-01 9.06638563e-01 -2.82143176e-01 -7.63490021e-01
-6.17430925e-01 4.68679428e-01 1.12876332e+00 -3.70907277e-01
-4.16469157e-01 1.16516016e-01 -7.01433539e-01 -5.50023913e-01
-6.94627464e-02 -3.57537866e-01 -5.32837689e-01 -5.42612467e-03
2.41629273e-01 3.04810226e-01 3.57649364e-02 -6.11160874e-01
-1.63341224e-01 9.04240251e-01 4.47309911e-01 -3.63042444e-01
1.44063735e+00 7.96029419e-02 -3.52975965e-01 8.54127705e-01
9.58101392e-01 -3.99893522e-01 -1.08402061e+00 3.32367374e-03
2.11624861e-01 2.49660268e-01 -2.21403246e-03 -8.54437709e-01
-6.66967690e-01 8.03730190e-01 1.00956786e+00 9.47956860e-01
8.66163194e-01 -5.54788470e-01 4.10898834e-01 7.26518869e-01
2.63679951e-01 -9.44855034e-01 -1.66790277e-01 1.04938352e+00
1.00812888e+00 -1.10087454e+00 -8.78779665e-02 4.54275042e-01
-1.72272936e-01 1.83054101e+00 1.19658783e-01 -6.35467410e-01
1.01226819e+00 2.48656839e-01 6.60805553e-02 -1.80250630e-01
-8.98843884e-01 1.72655731e-01 4.51043695e-01 4.10654634e-01
3.42864096e-02 -1.04423642e-01 -3.71382117e-01 3.90803963e-01
-4.40116853e-01 -9.67580080e-02 6.49208307e-01 5.65079093e-01
-4.58298743e-01 -7.26619184e-01 -5.04399240e-01 2.92294532e-01
-7.61875585e-02 -2.20140427e-01 -1.28806129e-01 1.03646243e+00
-4.72656310e-01 7.67046094e-01 -4.20439281e-02 -5.00455201e-01
5.57019562e-02 2.24048913e-01 2.46936250e-02 1.66125353e-02
-9.35862064e-01 -2.12162629e-01 -1.97780982e-01 -8.67378116e-02
-3.48930836e-01 -6.97227776e-01 -7.77895927e-01 -3.43056232e-01
-5.59112549e-01 7.68192112e-01 9.51748550e-01 8.99103880e-01
3.75237137e-01 2.57614881e-01 9.59474266e-01 -1.21040213e+00
-1.04214156e+00 -1.09307325e+00 -5.82387149e-01 -4.12951171e-01
7.49928832e-01 -5.50242186e-01 -1.10428536e+00 -4.54098523e-01] | [6.6200103759765625, 3.398724317550659] |
19872222-b9b6-4554-bb62-3c8852575b88 | addressing-the-confounds-of-accompaniments-in | 2002.06817 | null | https://arxiv.org/abs/2002.06817v1 | https://arxiv.org/pdf/2002.06817v1.pdf | Addressing the confounds of accompaniments in singer identification | Identifying singers is an important task with many applications. However, the task remains challenging due to many issues. One major issue is related to the confounding factors from the background instrumental music that is mixed with the vocals in music production. A singer identification model may learn to extract non-vocal related features from the instrumental part of the songs, if a singer only sings in certain musical contexts (e.g., genres). The model cannot therefore generalize well when the singer sings in unseen contexts. In this paper, we attempt to address this issue. Specifically, we employ open-unmix, an open source tool with state-of-the-art performance in source separation, to separate the vocal and instrumental tracks of music. We then investigate two means to train a singer identification model: by learning from the separated vocal only, or from an augmented set of data where we "shuffle-and-remix" the separated vocal tracks and instrumental tracks of different songs to artificially make the singers sing in different contexts. We also incorporate melodic features learned from the vocal melody contour for better performance. Evaluation results on a benchmark dataset called the artist20 shows that this data augmentation method greatly improves the accuracy of singer identification. | ['Yi-Hsuan Yang', 'Yu-Ching Yang', 'Zhe-Cheng Fan', 'Kai-Hsiang Cheng', 'Tsung-Han Hsieh'] | 2020-02-17 | null | null | null | null | ['singer-identification'] | ['music'] | [ 3.73036683e-01 -4.79325950e-01 1.03377640e-01 -5.28614745e-02
-9.04888690e-01 -1.10159135e+00 2.37591207e-01 -2.46045321e-01
-1.78463176e-01 4.09137845e-01 1.87789381e-01 1.10464469e-01
-2.02680603e-01 -2.90793002e-01 -4.71298993e-01 -9.14350271e-01
-6.50315592e-03 1.60007268e-01 2.13772934e-02 -1.38837814e-01
1.12964004e-01 1.72563821e-01 -1.84971499e+00 3.97544980e-01
6.51568890e-01 7.39281535e-01 2.86800802e-01 7.80479789e-01
-1.01067767e-01 3.78263295e-01 -9.43253279e-01 -9.67578143e-02
4.99982625e-01 -8.78628671e-01 -5.22786379e-01 -4.75239456e-02
5.90167940e-01 1.29111379e-01 1.33636668e-01 1.14799213e+00
4.06545728e-01 2.07817435e-01 5.13826251e-01 -8.23347807e-01
-1.48670420e-01 7.90081322e-01 -5.45876622e-01 1.69883475e-01
1.90716341e-01 -8.92782733e-02 1.18068552e+00 -7.94464231e-01
1.75878689e-01 1.07842636e+00 7.98745513e-01 6.43273413e-01
-1.50253820e+00 -1.07650912e+00 -3.05207580e-01 1.30198365e-02
-1.37820101e+00 -6.06568694e-01 1.07314563e+00 -5.41993260e-01
1.25817731e-01 7.03302741e-01 4.84755427e-01 9.87755060e-01
-2.57140011e-01 6.81993723e-01 1.12705183e+00 -5.35172880e-01
-1.78675316e-02 1.92653254e-01 1.51855871e-01 1.57825947e-01
-4.02411699e-01 3.07456583e-01 -5.88613272e-01 -2.38220751e-01
6.36662126e-01 -3.51695299e-01 -5.30868888e-01 1.31248698e-01
-1.21550167e+00 7.02923536e-01 1.17807873e-01 6.80090308e-01
-2.11540252e-01 2.95696445e-02 3.47956061e-01 4.67265934e-01
2.57269472e-01 8.80254447e-01 -3.35183769e-01 -2.44596511e-01
-1.25247192e+00 3.41348350e-01 9.80243862e-01 3.18277329e-01
5.79846323e-01 2.52385616e-01 -2.32260764e-01 1.31364059e+00
3.24244751e-03 2.13995755e-01 7.62912035e-01 -9.76379454e-01
1.71672165e-01 2.01585501e-01 -8.08620453e-02 -6.27942622e-01
2.28152387e-02 -8.58116269e-01 -6.63069367e-01 1.04402862e-01
6.91483021e-01 -5.70973642e-02 -6.63869619e-01 1.83163190e+00
2.06449300e-01 7.08380938e-01 -9.17489752e-02 9.15934384e-01
7.73848295e-01 7.30846286e-01 -2.77484328e-01 -3.98198724e-01
1.22632647e+00 -1.11041927e+00 -8.97957206e-01 -9.88080502e-02
-1.83599573e-02 -1.16484129e+00 1.34702992e+00 5.96400261e-01
-9.42418396e-01 -1.06439614e+00 -1.11557448e+00 3.62460554e-01
-2.92091668e-02 3.57623339e-01 1.12293616e-01 6.66697025e-01
-3.28448474e-01 8.40934515e-01 -4.23387468e-01 8.37984756e-02
-3.80050465e-02 3.75297576e-01 -2.40334928e-01 6.22732520e-01
-1.11321831e+00 2.82373786e-01 3.39691222e-01 -1.26877517e-01
-9.00681257e-01 -8.93799841e-01 -4.43952382e-01 4.45136316e-02
4.47712392e-01 -2.95382023e-01 1.50545168e+00 -1.26797962e+00
-1.46865964e+00 6.16527796e-01 7.05901813e-03 -2.62899518e-01
3.15632045e-01 -1.78933918e-01 -6.09836340e-01 -1.48367137e-01
3.11538577e-02 -2.63110213e-02 1.34778953e+00 -1.25567389e+00
-6.49599433e-01 -3.18864465e-01 -3.68195415e-01 1.22310743e-01
-5.61675906e-01 1.81740522e-01 -4.62881327e-01 -1.11344051e+00
-7.54954219e-02 -1.26124811e+00 2.11598262e-01 -6.93357825e-01
-3.97460341e-01 -7.91529790e-02 8.64342391e-01 -6.10869408e-01
1.54180348e+00 -2.72992420e+00 1.40669912e-01 1.97700575e-01
-2.03877673e-01 4.42394853e-01 -2.25353718e-01 2.94166207e-01
-2.83287197e-01 -4.42762338e-02 -1.53018475e-01 -3.85437995e-01
-1.74465477e-01 2.27899849e-01 -6.41792119e-01 8.39420855e-02
-1.72032729e-01 3.77202690e-01 -8.04212093e-01 -1.06311217e-01
-7.43511692e-02 3.50529581e-01 -4.53092963e-01 4.80374008e-01
1.69328731e-02 9.33892250e-01 1.47520844e-02 4.43021297e-01
3.64223540e-01 2.73301184e-01 -1.21490378e-03 -2.91578710e-01
-1.42007887e-01 5.09146690e-01 -1.54466391e+00 1.69729722e+00
-4.46129829e-01 4.98747408e-01 3.77636194e-01 -6.58857346e-01
9.03730273e-01 6.27952456e-01 3.82254750e-01 -8.91009122e-02
1.21780494e-02 5.94218135e-01 6.48004889e-01 -3.02647144e-01
4.70578641e-01 -5.95225453e-01 3.79971787e-02 2.77979076e-01
1.96779564e-01 -2.02403963e-01 2.44675219e-01 -4.85553086e-01
8.58936369e-01 -1.17479831e-01 1.78814799e-01 -1.58792526e-01
6.58629835e-01 -4.21484411e-01 6.14977658e-01 6.00651979e-01
3.90884057e-02 8.82609546e-01 1.69046775e-01 -9.29520577e-02
-7.19257355e-01 -1.06933081e+00 -1.71590075e-01 1.37581170e+00
-2.19272792e-01 -6.26047134e-01 -8.46808195e-01 -5.82490146e-01
2.12154854e-02 6.89348698e-01 -5.01836538e-01 -2.56651580e-01
-7.79244542e-01 -6.17584705e-01 8.38563800e-01 4.34613764e-01
4.51725498e-02 -1.21035361e+00 -1.82587683e-01 4.13799256e-01
-1.79879189e-01 -7.06505299e-01 -1.00968409e+00 4.25271720e-01
-5.87722003e-01 -9.20610487e-01 -6.53457880e-01 -8.47766876e-01
8.04576054e-02 1.37698308e-01 8.52100730e-01 -3.06453198e-01
-1.41981602e-01 5.45583963e-02 -2.64210343e-01 -5.03156662e-01
-7.60949135e-01 1.01094797e-01 2.55773097e-01 5.18251181e-01
2.65348479e-02 -7.12915957e-01 -2.06147730e-01 7.13426352e-01
-9.45464075e-01 -1.69559777e-01 3.34155202e-01 1.02582574e+00
6.97282970e-01 5.49265683e-01 8.95341396e-01 -7.84538925e-01
9.13104653e-01 -2.25349098e-01 -2.89296508e-01 -8.19568187e-02
-1.74506724e-01 2.74680983e-02 8.20109010e-01 -9.70560312e-01
-6.46436453e-01 1.96173653e-01 -3.25553417e-01 -8.76001000e-01
-2.44223952e-01 4.11480576e-01 -3.68383795e-01 9.27994400e-02
6.46361470e-01 2.14903340e-01 -1.41413342e-02 -1.00902760e+00
1.10551178e-01 8.60698581e-01 9.91852522e-01 -6.24654472e-01
9.47618425e-01 3.97114605e-02 -2.53097653e-01 -1.03357840e+00
-8.83768976e-01 -8.12203109e-01 -4.30433333e-01 -1.25659972e-01
5.41882396e-01 -5.96269846e-01 -4.80468303e-01 4.30489719e-01
-9.82934237e-01 -7.46313334e-02 -7.00489998e-01 5.71269274e-01
-5.17161846e-01 2.35744491e-01 -4.21209097e-01 -9.26215947e-01
-3.84710938e-01 -1.04352283e+00 8.58761370e-01 3.27098399e-01
-5.64047933e-01 -6.66023970e-01 5.15856266e-01 2.63759077e-01
1.67511970e-01 6.16064817e-02 7.90224075e-01 -9.83187556e-01
-3.73264030e-02 -2.49176890e-01 3.65497798e-01 8.95815492e-01
7.29383349e-01 -1.43361956e-01 -1.44841909e+00 -3.29444766e-01
3.78121138e-01 -5.84880076e-02 8.21891010e-01 1.52720541e-01
1.11639690e+00 -4.05400604e-01 4.84587662e-02 6.55296445e-01
8.87923837e-01 3.45243633e-01 2.24226564e-01 -4.45412509e-02
7.16838717e-01 8.58695805e-01 5.00272095e-01 1.21278942e-01
-4.79297400e-01 8.58778477e-01 1.41813114e-01 1.49955988e-01
-2.80834347e-01 -4.91890997e-01 4.11016613e-01 1.20123410e+00
-2.03165188e-01 -4.76377904e-02 -6.13992929e-01 6.30144000e-01
-1.64995182e+00 -9.70326245e-01 -5.94275184e-02 2.56875896e+00
1.07993305e+00 -5.19350171e-03 3.53054285e-01 7.77379274e-01
8.41930449e-01 1.21037856e-01 -4.05279249e-01 -1.06489480e-01
-1.36275515e-01 6.26158476e-01 5.77870086e-02 3.94163787e-01
-1.19522119e+00 6.30813360e-01 5.56254625e+00 1.12117529e+00
-1.26789486e+00 6.17102310e-02 1.30056918e-01 -2.97643363e-01
5.27831092e-02 -3.51580158e-02 -7.94414937e-01 5.97359300e-01
9.67080474e-01 2.84693465e-02 8.17383289e-01 5.74227810e-01
7.43472278e-02 3.66743684e-01 -1.32719195e+00 1.11793756e+00
8.90759230e-02 -9.09553051e-01 -1.40407234e-01 1.42212242e-01
6.77301645e-01 -3.71109575e-01 2.61837333e-01 4.60406601e-01
-3.51373643e-01 -1.08367038e+00 7.94883668e-01 3.12970102e-01
6.57488108e-01 -7.77749658e-01 5.28770864e-01 5.32036662e-01
-1.25957680e+00 -9.97900367e-02 6.13985620e-02 9.87256318e-02
-5.41290268e-02 2.77620465e-01 -1.03804207e+00 5.46076357e-01
5.97710252e-01 6.08419955e-01 -4.61703360e-01 1.14662337e+00
-7.16620237e-02 1.08504641e+00 -2.65009761e-01 3.65285188e-01
-6.98609576e-02 -3.12911153e-01 1.05581009e+00 1.08881068e+00
4.64014024e-01 -3.51116598e-01 8.00021663e-02 9.21677947e-01
-2.12636217e-02 3.22988510e-01 -4.47933137e-01 -2.88562298e-01
2.57645637e-01 1.34459186e+00 -3.50072294e-01 -5.96678779e-02
-1.55720413e-01 9.26662803e-01 -1.66055143e-01 2.36582235e-01
-6.01223350e-01 -3.05129111e-01 6.07979178e-01 3.83536220e-01
2.34375387e-01 2.00455338e-02 9.15471017e-02 -9.54076529e-01
-1.33760393e-01 -1.33249307e+00 3.36243391e-01 -4.29220676e-01
-1.45538104e+00 7.03239560e-01 -3.09617758e-01 -1.53540182e+00
-5.35167694e-01 -4.17519867e-01 -8.35552514e-01 1.10330653e+00
-1.16381192e+00 -9.37267840e-01 -6.49291724e-02 5.78730702e-01
7.79796243e-01 -2.89088249e-01 9.82943177e-01 4.93524909e-01
-3.18173289e-01 5.79623103e-01 2.72580773e-01 2.29919344e-01
1.12058401e+00 -1.43526554e+00 1.62738413e-01 4.15410429e-01
8.52439404e-01 7.52808571e-01 7.33076572e-01 -3.90807927e-01
-1.11784852e+00 -9.62924123e-01 6.90603077e-01 -3.45337451e-01
6.57850504e-01 -5.13856649e-01 -1.25798452e+00 3.56656849e-01
5.26666455e-02 -8.88240114e-02 1.18376207e+00 2.29537755e-01
-2.04587936e-01 -3.56890678e-01 -6.18611693e-01 4.65626359e-01
5.83799124e-01 -7.03563213e-01 -9.87345695e-01 -2.38660984e-02
4.57550168e-01 -1.84510589e-01 -6.01918459e-01 5.21794200e-01
8.59462142e-01 -9.20345306e-01 1.01273894e+00 -5.48612595e-01
-4.87885624e-03 -6.89215064e-01 -2.19639570e-01 -1.59984231e+00
-1.81664541e-01 -9.35458839e-01 -5.25503457e-02 1.78970289e+00
2.64501125e-01 -2.59083986e-01 5.08795679e-01 -9.91849452e-02
-7.90211707e-02 -1.81566969e-01 -8.22962284e-01 -1.11077380e+00
-7.08377287e-02 -4.52256918e-01 6.07953131e-01 9.87160027e-01
-2.81035244e-01 7.47057796e-01 -6.60390913e-01 -7.24628121e-02
4.01270926e-01 3.48950863e-01 9.29025531e-01 -1.48709857e+00
-8.34118009e-01 -4.27643865e-01 -7.46822357e-02 -7.54322529e-01
2.47145310e-01 -9.51507449e-01 1.79907352e-01 -6.07284486e-01
-2.50661578e-02 -3.51563275e-01 -6.93506896e-01 3.19296956e-01
-3.30509841e-01 4.55343634e-01 5.08633375e-01 5.02464533e-01
-6.83081150e-02 4.73859459e-01 1.27867723e+00 -1.89635694e-01
-6.31641030e-01 6.47865713e-01 -5.10296464e-01 9.85626221e-01
7.12247729e-01 -5.55991828e-01 -2.77778387e-01 1.41227394e-01
-1.99878395e-01 1.67828843e-01 2.74027735e-01 -1.19582844e+00
5.88384718e-02 2.99665630e-01 8.74780566e-02 -5.33390999e-01
5.68817616e-01 -8.58850479e-01 4.83667225e-01 2.92185843e-01
-4.31072652e-01 -2.84758747e-01 3.99997056e-01 5.01119494e-01
-5.10354877e-01 -6.96399927e-01 8.86739373e-01 7.98273087e-02
-1.13665417e-01 -1.31949395e-01 -1.18385017e-01 7.53536262e-03
5.24065912e-01 -9.69005078e-02 1.87807605e-01 -3.46404195e-01
-8.09823334e-01 -4.40549374e-01 2.08536208e-01 7.47724116e-01
1.07015148e-01 -1.37850213e+00 -8.35924149e-01 6.23269677e-01
8.88166502e-02 -3.78087133e-01 1.20535120e-01 6.53641820e-01
9.59680900e-02 1.23254821e-01 3.04012336e-02 -5.56014180e-01
-1.63493395e+00 5.20695269e-01 2.54051626e-01 -2.10931629e-01
-2.12065443e-01 6.93374574e-01 6.45952106e-01 -5.00440657e-01
3.69412959e-01 -3.31276298e-01 -4.51072931e-01 4.98704821e-01
6.29840136e-01 4.05982465e-01 1.13280572e-01 -8.45687985e-01
-3.41254175e-02 7.13683903e-01 1.09755829e-01 -2.98744619e-01
1.03164864e+00 1.76510528e-01 -5.19231707e-02 1.07235646e+00
1.19374990e+00 8.21515560e-01 -9.45822537e-01 -1.72147125e-01
4.10458893e-02 -5.69207132e-01 1.55198267e-02 -8.07551861e-01
-9.35603499e-01 8.52536440e-01 6.38261080e-01 6.10243082e-01
1.14798892e+00 -2.02992395e-01 7.27983177e-01 -4.68261130e-02
1.73204206e-03 -9.57355380e-01 7.34563842e-02 4.33691621e-01
1.11351514e+00 -9.01979923e-01 -4.45017695e-01 -2.56590486e-01
-5.26453376e-01 1.03862023e+00 2.60417163e-01 -2.01211482e-01
5.92328906e-01 2.66890764e-01 2.49690622e-01 1.59541398e-01
-3.42044830e-01 -1.11232325e-01 8.53062987e-01 1.52312428e-01
6.03730321e-01 1.17039941e-02 -1.82817966e-01 1.24559200e+00
-6.72853291e-01 -4.08771068e-01 1.26283541e-01 4.45227593e-01
-2.04436153e-01 -1.46060348e+00 -9.45049882e-01 2.92883933e-01
-8.13250780e-01 -2.19025254e-01 -7.26644993e-01 4.89543527e-01
4.82724607e-01 1.01917148e+00 -7.65616074e-02 -5.82348228e-01
4.96858567e-01 3.74753892e-01 3.50412637e-01 -7.85018563e-01
-1.11945176e+00 8.54941487e-01 -1.22965336e-01 -2.23949440e-02
-3.61204147e-01 -8.53245497e-01 -1.03270936e+00 2.54651248e-01
-5.99843502e-01 5.65027058e-01 6.65543854e-01 6.57035112e-01
-7.60599375e-02 8.46608341e-01 8.80801022e-01 -8.72315168e-01
-7.16214299e-01 -1.17255294e+00 -1.04603124e+00 5.39102912e-01
5.40536404e-01 -4.51607108e-01 -7.05312371e-01 2.33101636e-01] | [15.717881202697754, 5.527655601501465] |
36139ca9-063f-4cd0-b946-a79b77d94101 | memory-augmented-non-local-attention-for | 2108.11048 | null | https://arxiv.org/abs/2108.11048v1 | https://arxiv.org/pdf/2108.11048v1.pdf | Memory-Augmented Non-Local Attention for Video Super-Resolution | In this paper, we propose a novel video super-resolution method that aims at generating high-fidelity high-resolution (HR) videos from low-resolution (LR) ones. Previous methods predominantly leverage temporal neighbor frames to assist the super-resolution of the current frame. Those methods achieve limited performance as they suffer from the challenge in spatial frame alignment and the lack of useful information from similar LR neighbor frames. In contrast, we devise a cross-frame non-local attention mechanism that allows video super-resolution without frame alignment, leading to be more robust to large motions in the video. In addition, to acquire the information beyond neighbor frames, we design a novel memory-augmented attention module to memorize general video details during the super-resolution training. Experimental results indicate that our method can achieve superior performance on large motion videos comparing to the state-of-the-art methods without aligning frames. Our source code will be released. | ['Tao Mei', 'Liefeng Bo', 'Jingen Liu', 'Jiyang Yu'] | 2021-08-25 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Yu_Memory-Augmented_Non-Local_Attention_for_Video_Super-Resolution_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Yu_Memory-Augmented_Non-Local_Attention_for_Video_Super-Resolution_CVPR_2022_paper.pdf | cvpr-2022-1 | ['video-super-resolution'] | ['computer-vision'] | [ 3.97180289e-01 -1.60457745e-01 -2.66773283e-01 -2.73429483e-01
-1.05494952e+00 -9.89139080e-02 4.57167059e-01 -5.22711277e-01
-2.26225778e-01 8.95152032e-01 5.46503067e-01 2.78142631e-01
1.02822125e-01 -6.78864241e-01 -8.43634844e-01 -6.59322023e-01
1.39850661e-01 -3.28520030e-01 5.82338750e-01 -2.49143660e-01
2.60790616e-01 2.87653148e-01 -1.69960964e+00 6.58980668e-01
7.47516692e-01 7.24466085e-01 7.39947677e-01 5.77876389e-01
9.23479348e-02 1.01455140e+00 -2.97807783e-01 5.04534654e-02
2.23201677e-01 -6.06706262e-01 -7.43548453e-01 4.69477102e-02
8.32966268e-01 -9.30511951e-01 -6.04753792e-01 9.59027767e-01
5.88583529e-01 3.63504916e-01 2.98967306e-02 -4.95692760e-01
-8.94779563e-01 5.45824170e-01 -7.91212082e-01 8.45172107e-01
7.56038785e-01 -7.25113228e-02 7.35550880e-01 -1.07872009e+00
8.33913863e-01 1.34863377e+00 4.21558589e-01 8.23426247e-01
-1.07999337e+00 -7.01910377e-01 3.71224046e-01 4.22263443e-01
-1.44172978e+00 -7.39754498e-01 7.81285346e-01 -1.98276713e-01
7.00718999e-01 2.05012918e-01 3.47800165e-01 1.18177867e+00
6.52629063e-02 3.92641664e-01 8.18450749e-01 -1.61557496e-01
-3.38350944e-02 -4.10767645e-01 -3.31183225e-01 4.77242887e-01
4.43154499e-02 2.75184095e-01 -8.45442355e-01 2.12040424e-01
1.58211040e+00 9.34108347e-02 -7.09824145e-01 -4.77223955e-02
-1.35913765e+00 5.12678027e-01 3.95043701e-01 6.16438329e-01
-6.08938813e-01 -5.40729128e-02 4.56304811e-02 -2.51901727e-02
4.28546548e-01 2.08140999e-01 -1.52211413e-01 -4.99172211e-02
-1.13871264e+00 1.40261278e-01 5.90600856e-02 8.99925709e-01
6.80158377e-01 3.34851682e-01 -4.07526910e-01 6.28891945e-01
-1.28586693e-02 1.13874106e-02 6.28770351e-01 -1.40088832e+00
5.83346248e-01 6.74452772e-03 4.06484276e-01 -1.04013717e+00
-6.73658252e-02 -2.17461914e-01 -9.65765297e-01 1.00900121e-01
6.75799102e-02 7.50838667e-02 -7.98172951e-01 1.72308254e+00
2.50384212e-01 6.48661017e-01 1.28865838e-01 1.34019613e+00
9.70823288e-01 9.02787209e-01 -2.83078011e-02 -7.31652617e-01
1.00713289e+00 -1.12641156e+00 -1.03765917e+00 -6.13528192e-02
-1.97505355e-01 -6.91484690e-01 8.23508978e-01 1.80388927e-01
-1.57699037e+00 -1.11182284e+00 -1.03126514e+00 -3.43242764e-01
2.10527956e-01 -2.40740359e-01 3.01176399e-01 8.14191625e-02
-1.17851794e+00 5.99445999e-01 -7.57154942e-01 -3.19226906e-02
2.73538172e-01 1.92522958e-01 -4.42539096e-01 -3.17299992e-01
-1.26638460e+00 6.03073537e-01 3.98166567e-01 -4.65471065e-03
-8.20928574e-01 -6.71862364e-01 -9.31667686e-01 -1.27593894e-02
4.02167141e-01 -6.78049684e-01 1.04203129e+00 -1.11914062e+00
-1.54906952e+00 3.61477494e-01 -5.58334172e-01 -3.29782814e-01
3.42666864e-01 -4.61070240e-01 -4.24854189e-01 4.78868812e-01
-1.24684963e-02 6.96865678e-01 1.18524957e+00 -1.34497476e+00
-9.04725432e-01 -6.29914328e-02 2.24830583e-01 3.97404969e-01
-1.98599681e-01 9.48662609e-02 -6.85668290e-01 -1.05773056e+00
8.78477767e-02 -4.62771058e-01 -1.98131487e-01 -3.44497919e-01
5.47205172e-02 2.07667902e-01 1.02778363e+00 -7.80034661e-01
1.46539080e+00 -2.08253074e+00 5.23500919e-01 -4.50946689e-01
2.35146895e-01 4.01876837e-01 -1.98167533e-01 -7.89046064e-02
-5.85003607e-02 1.15777329e-01 8.63102898e-02 -2.15256840e-01
-6.15849853e-01 5.04067466e-02 -3.82878751e-01 1.70498818e-01
3.00479800e-01 7.16672242e-01 -9.49827909e-01 -6.02788627e-01
3.82458329e-01 1.12964797e+00 -5.63064277e-01 4.24293905e-01
7.09111914e-02 9.86234963e-01 -4.89523709e-01 4.43325251e-01
7.07389176e-01 -4.31164801e-01 1.81242287e-01 -5.71423650e-01
-3.12460959e-01 8.95793438e-02 -1.12913322e+00 1.92310810e+00
-3.67595464e-01 4.62738693e-01 1.93518147e-01 -5.50085187e-01
7.38367796e-01 3.45866472e-01 5.64748168e-01 -8.69252741e-01
-1.46835163e-01 -5.72248325e-02 -3.26432735e-01 -4.66254532e-01
8.72026622e-01 3.54311652e-02 3.89331967e-01 3.23913284e-02
-1.61765143e-01 4.81570661e-01 1.67042077e-01 -1.94953773e-02
8.27202022e-01 4.43080306e-01 4.05295074e-01 1.18296660e-01
8.10613632e-01 -4.70680118e-01 8.75085115e-01 4.70339566e-01
-8.10402930e-02 1.08859289e+00 -7.45500997e-02 -5.77163398e-01
-1.26472735e+00 -1.07616770e+00 -8.86850655e-02 9.85941350e-01
5.39620340e-01 -5.73683918e-01 -7.75735259e-01 -2.47525588e-01
-5.75733006e-01 3.16840261e-01 -4.58457053e-01 1.50379136e-01
-1.07235956e+00 -5.06344855e-01 -1.46600734e-02 6.71533346e-01
7.16687977e-01 -9.41427708e-01 -6.80772007e-01 3.75828177e-01
-6.59648418e-01 -1.50701940e+00 -8.52804422e-01 -5.98920047e-01
-1.03950667e+00 -8.06987345e-01 -9.89864230e-01 -7.62413204e-01
5.54873645e-01 7.41306365e-01 1.01038909e+00 -2.33910289e-02
-6.20922893e-02 -1.19945058e-03 -5.32813728e-01 4.79023874e-01
-2.37087175e-01 -4.73692045e-02 8.34652707e-02 7.03531504e-02
7.68988207e-02 -5.74657261e-01 -8.26356828e-01 4.57014710e-01
-1.04927099e+00 4.50327128e-01 7.17464864e-01 9.04884160e-01
1.01016366e+00 1.98679611e-01 5.54640114e-01 -5.31234503e-01
1.27160758e-01 -3.69285196e-01 -4.59881872e-01 4.11187746e-02
-2.69899577e-01 3.72318551e-02 7.36886382e-01 -4.82599884e-01
-1.39389884e+00 4.89191711e-02 -2.78924145e-02 -7.34028161e-01
-1.75577268e-01 6.68660551e-02 -2.23270997e-01 -2.60926820e-02
2.35080719e-01 2.82712221e-01 -3.24758083e-01 -7.02415228e-01
1.84295610e-01 3.39451790e-01 8.11051965e-01 -3.65403175e-01
8.18903089e-01 5.77817321e-01 -1.30542085e-01 -6.75822854e-01
-9.21222091e-01 -1.73944578e-01 -6.76426828e-01 -2.37420127e-02
1.11267400e+00 -1.31333745e+00 -2.16363117e-01 1.76980466e-01
-1.07339692e+00 -1.52199209e-01 -9.04287584e-03 6.27122760e-01
-6.82727754e-01 5.69866359e-01 -9.17160213e-01 -5.72512507e-01
-2.79728174e-01 -1.19309139e+00 1.15172482e+00 4.36065525e-01
-8.23742803e-03 -5.63929498e-01 -1.08200267e-01 5.29221058e-01
5.78958511e-01 2.26409465e-01 3.79789680e-01 1.63613975e-01
-1.09930015e+00 3.47133756e-01 -3.88777047e-01 1.21732749e-01
3.53579193e-01 -1.70399696e-01 -7.94286489e-01 -4.83122766e-01
-6.52925074e-02 8.06808323e-02 8.51373672e-01 5.07814884e-01
1.22258317e+00 -4.73405778e-01 -8.49233344e-02 8.93063784e-01
1.46539772e+00 1.45095691e-01 9.27138627e-01 4.19877410e-01
9.46509182e-01 3.36161882e-01 1.00052869e+00 3.95358264e-01
3.77955258e-01 1.03646445e+00 3.03328007e-01 -1.48591399e-01
-4.12687808e-01 -3.09580505e-01 4.87931669e-01 6.78479314e-01
-6.64669394e-01 -2.77062412e-02 -3.11128289e-01 4.31547135e-01
-1.99578750e+00 -1.61087263e+00 6.03575706e-02 2.29782438e+00
8.42590630e-01 -1.02018334e-01 1.10179603e-01 -7.07135126e-02
1.04499578e+00 5.03143370e-01 -4.62110966e-01 1.18118972e-01
-2.72582084e-01 4.22687195e-02 2.18257278e-01 6.95061445e-01
-1.05181885e+00 1.01085281e+00 6.47782612e+00 7.98312545e-01
-1.02233970e+00 2.68377751e-01 6.93177938e-01 -3.69488657e-01
-2.94624776e-01 -3.69038463e-01 -8.98128808e-01 5.91060221e-01
7.74670303e-01 6.62753955e-02 5.82764268e-01 5.75953662e-01
4.84901339e-01 2.84920149e-02 -8.95437062e-01 1.19895101e+00
2.32050687e-01 -1.51284826e+00 3.13824385e-01 -7.99195245e-02
9.86873448e-01 -3.45700890e-01 9.79412794e-02 -9.71956924e-03
-2.38070235e-01 -1.10243559e+00 6.07158005e-01 6.45162702e-01
1.15216446e+00 -9.38505948e-01 5.75806439e-01 1.25236912e-02
-1.58875632e+00 -2.19404489e-01 -4.61843699e-01 8.16653133e-04
5.30509770e-01 3.57490629e-01 -1.12838782e-01 7.67392755e-01
9.74813521e-01 8.20188999e-01 -3.51571977e-01 6.31475389e-01
5.10088168e-02 6.97145909e-02 1.89637259e-01 7.87360728e-01
-4.37430926e-02 -6.82526827e-02 6.49896502e-01 9.94437635e-01
5.46818852e-01 4.76474613e-01 1.90389246e-01 7.57548451e-01
6.15503825e-02 -3.22905034e-02 -4.76463228e-01 2.04182535e-01
4.69860107e-01 1.13763940e+00 -3.60655040e-01 -4.06458557e-01
-7.94478893e-01 1.27744603e+00 2.33669728e-01 5.28431773e-01
-1.00308609e+00 -2.99086068e-02 7.61720181e-01 4.06102270e-01
7.18472183e-01 -3.02626312e-01 9.74118859e-02 -1.39683676e+00
1.33466393e-01 -9.23676252e-01 3.00366223e-01 -9.97734964e-01
-1.00511813e+00 9.80027854e-01 -8.86555538e-02 -1.43561888e+00
-6.11164272e-01 -3.39500420e-02 -2.10274652e-01 8.06675076e-01
-1.81138015e+00 -1.05831182e+00 -5.49952567e-01 7.69426525e-01
1.03989625e+00 1.64212845e-02 4.26614851e-01 5.24759710e-01
-4.22150612e-01 4.42033470e-01 -1.26504511e-01 -4.38115234e-03
8.79794657e-01 -7.97552347e-01 2.17130646e-01 1.15660012e+00
-1.54031947e-01 5.53448498e-01 6.65988982e-01 -7.14151382e-01
-1.32014263e+00 -1.11484754e+00 6.30191386e-01 -2.54649639e-01
1.62594780e-01 2.62279343e-02 -1.16072786e+00 5.56653142e-01
1.55301303e-01 2.35496819e-01 3.79049748e-01 -4.13582325e-01
-2.15958223e-01 -1.12709284e-01 -1.07040834e+00 5.59672415e-01
1.26406360e+00 -4.74396706e-01 -5.38177073e-01 -3.60632688e-01
9.87991154e-01 -5.84659040e-01 -1.10845160e+00 6.54490530e-01
6.30281508e-01 -1.35713577e+00 1.35365283e+00 -1.97136268e-01
7.48441577e-01 -8.01036477e-01 -2.97505349e-01 -8.28375340e-01
-8.11213613e-01 -7.35100865e-01 -5.41341305e-01 1.25631034e+00
-1.15026295e-01 -1.39131889e-01 5.12323260e-01 3.96421224e-01
1.70053896e-02 -5.12401462e-01 -8.04741621e-01 -4.96405303e-01
-2.99005121e-01 1.03573509e-01 5.56297600e-01 9.73385394e-01
-2.08589107e-01 3.20555329e-01 -9.19522166e-01 4.98204201e-01
7.57691860e-01 3.09764713e-01 5.64321518e-01 -8.85518610e-01
-4.23865229e-01 -1.39027923e-01 -2.75361031e-01 -1.21450472e+00
1.15671799e-01 -2.10230380e-01 5.82963936e-02 -1.46968043e+00
4.66013134e-01 1.03094853e-01 -4.06339347e-01 3.97355072e-02
-5.15475392e-01 6.37285113e-01 2.56966621e-01 3.65652710e-01
-8.42586875e-01 4.41314459e-01 1.54899514e+00 2.57642239e-01
-3.13927621e-01 -3.60706657e-01 -7.10632324e-01 6.02542818e-01
7.16553867e-01 -4.35571112e-02 -3.75481546e-01 -5.63235700e-01
-2.26877257e-01 5.36063612e-01 3.52891207e-01 -1.03720224e+00
7.64094368e-02 -3.05980116e-01 7.88368523e-01 -7.34807730e-01
5.11543274e-01 -6.82591736e-01 4.13669497e-01 6.35443479e-02
-3.43277067e-01 2.74389535e-01 -2.56926306e-02 6.53600693e-01
-4.45775002e-01 1.62246466e-01 1.03287911e+00 -2.57391810e-01
-1.02421784e+00 5.34098804e-01 -2.83060640e-01 -1.41174451e-01
8.66069257e-01 -2.73769587e-01 -4.16601807e-01 -5.49139619e-01
-5.90978444e-01 -3.93560901e-02 9.31274235e-01 6.95956767e-01
9.37562704e-01 -1.45019269e+00 -7.99924016e-01 2.34485343e-01
-2.60033309e-01 2.17013106e-01 7.93138385e-01 6.42278433e-01
-3.28655034e-01 3.56225193e-01 -6.01317227e-01 -4.57821220e-01
-1.36969078e+00 9.43667531e-01 1.26360372e-01 -1.39992669e-01
-9.47588205e-01 4.55668092e-01 5.64539969e-01 4.81308550e-01
5.63093536e-02 -4.62732129e-02 -4.48320329e-01 -2.69984484e-01
1.26407146e+00 4.27928030e-01 -3.75893205e-01 -1.04234934e+00
-1.40288323e-01 9.97669041e-01 -2.13351116e-01 -3.55944745e-02
1.31087494e+00 -7.16616333e-01 1.61635846e-01 2.62203217e-01
9.84146118e-01 6.55716956e-02 -1.80911410e+00 -3.36836606e-01
-3.94431800e-01 -1.06161869e+00 2.29570851e-01 -3.34209234e-01
-1.25343919e+00 5.41091502e-01 6.93919122e-01 -1.30865425e-01
1.52924335e+00 -1.07827276e-01 1.00518215e+00 -1.42123640e-01
5.33740461e-01 -9.52711344e-01 1.51413351e-01 3.34017515e-01
8.47000182e-01 -1.24703133e+00 8.86908099e-02 -5.71556330e-01
-5.22083461e-01 1.14815032e+00 8.54296744e-01 -1.40911154e-03
8.35213140e-02 2.67073900e-01 -1.18167132e-01 2.53099144e-01
-8.66836131e-01 -2.76191115e-01 3.28520834e-01 7.48922229e-01
6.62100673e-01 -4.69295710e-01 -2.66310275e-01 4.17622209e-01
2.30138093e-01 2.47282133e-01 6.70183420e-01 5.80267429e-01
-5.15005469e-01 -9.31803823e-01 -4.97122347e-01 -1.16645187e-01
-7.68764377e-01 -2.15716466e-01 1.55528128e-01 5.45716941e-01
1.03824874e-02 9.51191962e-01 2.49347463e-01 -3.01035196e-01
2.93480679e-02 -2.95249552e-01 5.85948646e-01 -3.42233747e-01
-8.93730745e-02 5.04031360e-01 -9.78735536e-02 -1.05901647e+00
-8.85681629e-01 -5.09612322e-01 -1.16615009e+00 -3.34992170e-01
4.13955078e-02 -9.18062404e-03 1.02468856e-01 6.07674122e-01
6.40508890e-01 7.53108680e-01 6.23302042e-01 -1.48867989e+00
4.18336280e-02 -7.87892878e-01 -3.51203561e-01 4.59950149e-01
7.38238871e-01 -5.96873343e-01 -1.39947414e-01 3.89035046e-01] | [11.027881622314453, -1.887801170349121] |
2e4dc6e0-6273-4d5b-9c4a-5498db646a09 | optimization-of-residential-demand-response | 2305.08077 | null | https://arxiv.org/abs/2305.08077v1 | https://arxiv.org/pdf/2305.08077v1.pdf | Optimization of Residential Demand Response Program Cost with Consideration for Occupants Thermal Comfort and Privacy | Residential consumers can use the demand response program (DRP) if they can utilize the home energy management system (HEMS), which reduces consumer costs by automatically adjusting air conditioning (AC) setpoints and shifting some appliances to off-peak hours. If HEMS knows occupancy status, consumers can gain more economic benefits and thermal comfort. However, for the building occupancy status, direct sensing is costly, inaccurate, and intrusive for residents. So, forecasting algorithms could serve as an effective alternative. The goal of this study is to present a non-intrusive, accurate, and cost-effective approach, to develop a multi-objective simulation model for the application of DRPs in a smart residential house, where (a) electrical load demand reduction, (b) adjustment in thermal comfort (AC) temperature setpoints, and (c) , worst cases scenario approach is very conservative. Because that is unlikely all uncertain parameters take their worst values at all times. So, the flexible robust counterpart optimization along with uncertainty budgets is developed to consider uncertainty realistically. Simulated results indicate that considering uncertainty increases the costs by 36 percent and decreases the AC temperature setpoints. Besides, using DRPs reduces demand by shifting some appliance operations to off-peak hours and lowers costs by 13.2 percent. | ['Amir Khorsandi', 'M. M. Ardehali', 'Reza Nematirad'] | 2023-05-14 | null | null | null | null | ['energy-management'] | ['time-series'] | [-1.76166907e-01 -1.20925987e-02 1.09401926e-01 -2.75194377e-01
-5.20323694e-01 -6.00647151e-01 1.71482369e-01 6.57327399e-02
2.81960100e-01 1.17948186e+00 2.08637759e-01 -1.51232421e-01
-5.32759309e-01 -1.09901822e+00 -1.71353534e-01 -1.17332065e+00
1.91487297e-01 5.54728925e-01 -3.22565287e-01 1.50933648e-02
-2.40505368e-01 4.87269431e-01 -1.56196523e+00 -2.17214420e-01
1.11786556e+00 1.00485027e+00 1.46474361e-01 -4.80162539e-03
4.36299562e-01 2.19993651e-01 -4.17301178e-01 4.35321957e-01
3.00547093e-01 -2.45082870e-01 -3.13813657e-01 -1.44477606e-01
-1.03836572e+00 -8.86875391e-01 4.34295118e-01 8.91146421e-01
6.46482229e-01 7.18785703e-01 6.30215824e-01 -1.56441736e+00
-2.91227609e-01 7.15252459e-01 -4.51976478e-01 -1.30841613e-01
3.81676376e-01 1.79742008e-01 5.71315885e-01 -2.25152988e-02
-3.60223204e-01 9.01529491e-01 2.59319901e-01 4.43733521e-02
-1.46617174e+00 -5.62097192e-01 1.65739536e-01 1.62787184e-01
-1.70664859e+00 -8.28290060e-02 9.62229431e-01 -5.76207303e-02
1.42431545e+00 1.04598927e+00 1.25133157e+00 4.61694986e-01
8.64080116e-02 5.16094744e-01 1.47387457e+00 -3.53482693e-01
8.95910859e-01 2.98310786e-01 4.15155925e-02 -3.97220075e-01
4.36487883e-01 2.63395399e-01 5.66489458e-01 -1.96117997e-01
-3.03142667e-02 1.58720687e-01 -5.10577381e-01 6.76191598e-02
-2.67964512e-01 5.70021510e-01 2.81303167e-01 5.14576972e-01
-5.25819838e-01 -1.92623645e-01 4.90992628e-02 -1.97565556e-01
1.58565462e-01 2.59065062e-01 -4.93473887e-01 -6.91412687e-02
-9.44370210e-01 -7.33654061e-03 8.34863722e-01 8.73275042e-01
3.95556092e-01 2.58696616e-01 -1.31503120e-01 5.50210238e-01
5.46303272e-01 8.50777447e-01 4.44458574e-01 -9.31727767e-01
-1.28231511e-01 4.97137338e-01 6.51147008e-01 -4.10926759e-01
-6.47112072e-01 -1.63442165e-01 -9.17661846e-01 2.60801524e-01
-1.42049119e-01 -4.80786413e-01 -7.96683848e-01 1.45424390e+00
3.76774043e-01 -4.56853420e-01 -1.13034002e-01 9.11803782e-01
-1.71238586e-01 1.22666955e+00 3.05988044e-01 -8.78816664e-01
1.35044587e+00 -3.54950070e-01 -1.10959995e+00 1.30769927e-02
2.82446593e-01 -4.74036783e-01 6.10826492e-01 2.12462172e-01
-9.88479853e-01 -1.17095098e-01 -1.11847353e+00 4.11247432e-01
-7.37057090e-01 -1.55305758e-01 1.06385157e-01 9.80845869e-01
-9.07806575e-01 4.01838034e-01 -9.79778647e-01 -3.94657850e-01
-9.78314877e-02 2.88523585e-01 4.09372658e-01 2.47412249e-01
-1.37610734e+00 1.15513051e+00 5.10419846e-01 4.78659838e-01
-4.61243451e-01 -6.98845863e-01 -7.17146397e-01 4.13382500e-01
1.73542112e-01 -5.77824414e-01 1.08192921e+00 -4.75004524e-01
-1.72661293e+00 -1.91618577e-01 2.68674821e-01 -3.69237155e-01
4.74459946e-01 6.21826462e-02 -6.40475094e-01 -2.09354609e-01
-1.97188973e-01 -5.68319038e-02 1.85621321e-01 -1.24297309e+00
-6.39711201e-01 -4.67062414e-01 -2.81301022e-01 2.88624406e-01
1.11567587e-01 -3.39023888e-01 6.52156055e-01 -1.82439312e-01
-1.26758531e-01 -9.55477297e-01 -2.75900036e-01 -6.45010352e-01
-6.85601830e-01 -3.43976825e-01 9.43394601e-01 -1.01875997e+00
1.28402007e+00 -1.88670301e+00 -2.88658023e-01 8.50745082e-01
-7.54318595e-01 -1.65193066e-01 7.41676807e-01 6.99051797e-01
-2.54058093e-01 2.85667717e-01 -5.60223281e-01 -1.24786356e-02
7.34470189e-01 4.58815575e-01 -1.03881553e-01 4.06922847e-01
-3.43508869e-01 3.44434530e-01 -5.16687691e-01 1.16306819e-01
1.04598057e+00 5.17851830e-01 -1.08621068e-01 1.76889971e-01
-8.07159245e-02 2.88479519e-03 -6.18871689e-01 7.24262416e-01
9.27340806e-01 3.81673664e-01 4.83435690e-01 -4.80659693e-01
-4.77867901e-01 -4.30642925e-02 -1.52662027e+00 9.00554180e-01
-8.01637828e-01 1.04815230e-01 2.43761212e-01 -1.16818357e+00
6.25149429e-01 4.22166258e-01 8.58182728e-01 -8.33549917e-01
4.88671213e-02 1.35210350e-01 -3.61142695e-01 -4.09336269e-01
1.63642094e-01 -1.76687181e-01 4.81518172e-02 4.43966627e-01
-7.39562988e-01 -4.72999990e-01 -1.14149682e-01 -4.63035077e-01
6.29024684e-01 1.19335391e-01 4.80204821e-01 -7.44943857e-01
3.80671561e-01 -3.80611748e-01 7.82553375e-01 -2.38913730e-01
-1.21234171e-01 -3.91421840e-02 -6.51575327e-02 4.82898578e-02
-8.38243008e-01 -9.73419666e-01 -4.14760888e-01 6.11213088e-01
1.21802457e-01 3.17630559e-01 -8.42628121e-01 6.75074831e-02
1.36655629e-01 2.08025789e+00 -1.18232049e-01 -2.80143142e-01
-4.65561748e-01 -1.18168259e+00 -4.22035903e-01 5.69296122e-01
7.31927037e-01 -2.73396760e-01 -1.12608111e+00 3.05155426e-01
-2.69353122e-01 -5.57001472e-01 -2.36558586e-01 4.49245274e-01
-3.98892671e-01 -7.93099523e-01 -4.62393522e-01 -1.18385963e-01
8.17071497e-01 -2.24011485e-03 7.14226723e-01 -1.10491842e-01
-1.52874902e-01 3.23119938e-01 -8.13316926e-02 -3.33882779e-01
2.21431211e-01 -1.40355915e-01 1.37156680e-01 -3.78333151e-01
3.24198484e-01 -7.80373931e-01 -8.64391387e-01 6.36890054e-01
-6.92144394e-01 -9.16530043e-02 1.64989084e-01 4.41475779e-01
6.36638224e-01 9.72657025e-01 9.13695574e-01 -4.77203995e-01
5.27804315e-01 -6.18319094e-01 -9.49982643e-01 5.18683195e-01
-1.26481509e+00 -6.13806397e-02 7.59033740e-01 1.99158937e-02
-1.57800817e+00 1.49288669e-01 7.92078301e-03 1.73902228e-01
-2.85305262e-01 1.45289786e-02 -9.53051209e-01 4.58471894e-01
-1.43375158e-01 -3.63457464e-02 -5.36479056e-01 -4.26279455e-01
1.34984419e-01 5.48966825e-01 1.59829393e-01 -6.87299967e-01
1.07532549e+00 1.32848814e-01 3.44326440e-03 -5.06272197e-01
-1.36248425e-01 -1.25330791e-01 -3.00737888e-01 -4.09047216e-01
8.70639205e-01 -8.86070848e-01 -9.77773786e-01 2.41331458e-01
-3.73808503e-01 -4.21763122e-01 -3.93085480e-01 5.04542828e-01
-5.10886192e-01 1.56608105e-01 6.36568815e-02 -1.54612970e+00
-4.74315405e-01 -1.19188309e+00 3.69985610e-01 6.39115930e-01
-3.87333810e-01 -6.99586511e-01 -2.80937910e-01 4.37432051e-01
6.16409719e-01 8.84251654e-01 9.05738413e-01 -2.63592601e-01
-6.42334938e-01 -8.64267647e-02 3.18155080e-01 4.48656708e-01
5.03896832e-01 2.07687795e-01 -1.00049210e+00 -6.78877890e-01
3.20987076e-01 2.18842298e-01 -1.00393161e-01 5.75968981e-01
1.15633333e+00 -6.99867249e-01 -5.14041007e-01 2.35472724e-01
1.88424730e+00 1.18337262e+00 5.92297792e-01 3.78787607e-01
-2.28488237e-01 6.85557544e-01 6.68231666e-01 1.01429939e+00
5.55225790e-01 5.28738856e-01 6.02921128e-01 2.15912834e-02
7.15112448e-01 6.53562043e-03 2.63028353e-01 4.07413572e-01
-1.59516692e-01 -4.61568952e-01 -3.98998201e-01 5.79253197e-01
-1.78941011e+00 -9.88510132e-01 4.55100276e-02 2.36189055e+00
6.66848719e-01 -1.46140382e-01 1.14955790e-02 4.15970743e-01
8.19490671e-01 -1.90470833e-02 -6.37423217e-01 -7.38221407e-01
1.12589866e-01 6.44677952e-02 7.91963756e-01 3.82563323e-01
-2.77489185e-01 -3.13683450e-01 5.38358355e+00 5.26323199e-01
-4.64217037e-01 3.74346823e-02 1.02025282e+00 -4.57273394e-01
-6.05811477e-01 3.11734062e-02 -3.47395033e-01 1.02653956e+00
1.33244765e+00 -6.35756433e-01 8.53676081e-01 9.69571769e-01
9.66458797e-01 -7.69294858e-01 -1.24251401e+00 5.98939240e-01
-5.09663999e-01 -6.50187254e-01 -6.11037254e-01 1.61170498e-01
8.57481480e-01 -7.03307867e-01 -3.99165690e-01 2.95469314e-01
6.03278160e-01 -7.09682226e-01 5.17377317e-01 7.77941287e-01
5.44102676e-02 -1.50424874e+00 1.13707459e+00 4.32976156e-01
-1.32621622e+00 -3.99416566e-01 -1.06197082e-01 3.21959630e-02
8.85680616e-01 7.43113518e-01 -2.90435314e-01 6.68741822e-01
1.10675669e+00 -3.48115593e-01 1.40999511e-01 6.47908270e-01
-1.95695639e-01 3.57513428e-01 -1.17342257e+00 -2.82946765e-01
-8.54513496e-02 -8.46244097e-01 -1.00839278e-02 6.28100097e-01
6.94242358e-01 1.00899374e+00 9.86258313e-02 1.04265904e+00
5.10704637e-01 -1.46875918e-01 -3.08165491e-01 4.23754126e-01
9.71368551e-01 1.21266234e+00 -5.15659809e-01 -1.35443941e-01
1.88204255e-02 7.15816259e-01 -7.08311081e-01 5.79512179e-01
-1.01676583e+00 -4.52408999e-01 6.20379448e-01 3.02258253e-01
-1.46988124e-01 2.25407243e-01 -4.83784109e-01 -3.70121926e-01
1.32570386e-01 -6.45503998e-02 3.16783667e-01 -9.41324949e-01
-1.17663169e+00 -1.95894122e-01 6.05450213e-01 -7.24003255e-01
-6.18105471e-01 -1.31639108e-01 -9.00650501e-01 1.03992760e+00
-1.48304725e+00 -8.27274740e-01 -1.85012177e-01 5.18481195e-01
4.44248676e-01 4.84994382e-01 8.22062075e-01 4.81629461e-01
-1.07831419e+00 2.40552902e-01 8.56441319e-01 -5.88911414e-01
-8.47212132e-03 -1.28265238e+00 -5.32579422e-01 8.13367963e-01
-1.16593480e+00 1.38289779e-01 1.02303159e+00 -4.43475336e-01
-1.26430953e+00 -8.93630624e-01 3.55614185e-01 3.91172260e-01
3.95335406e-01 -3.68364938e-02 -6.86441064e-01 5.94943941e-01
7.22157896e-01 -6.82399452e-01 7.21677780e-01 -4.02229071e-01
5.28462708e-01 -5.27947187e-01 -2.06187081e+00 4.78766501e-01
3.98985356e-01 -2.92660952e-01 -4.31773454e-01 5.16005814e-01
3.94397795e-01 8.40332434e-02 -1.22018313e+00 3.26316357e-01
4.36555117e-01 -7.55100667e-01 6.39504433e-01 3.25371385e-01
-6.77008152e-01 -5.32590687e-01 -5.07109225e-01 -1.71179914e+00
-3.00368786e-01 -6.34089768e-01 2.93506961e-02 1.90657043e+00
2.65956283e-01 -1.22739267e+00 6.03040718e-02 1.88946033e+00
-3.74139547e-01 -4.80477154e-01 -1.48351490e+00 -5.84642231e-01
-2.21380904e-01 -1.54162973e-01 1.30934274e+00 8.52353096e-01
3.27400923e-01 -2.51339585e-01 -7.53928572e-02 6.91335380e-01
7.94769406e-01 2.26724483e-02 8.93076230e-03 -8.76500368e-01
1.25362188e-01 -2.93228984e-01 2.96967655e-01 -2.08459944e-01
-1.33793475e-02 -1.75835952e-01 1.47762090e-01 -1.95742786e+00
-1.64704710e-01 -2.36453682e-01 -3.73417526e-01 4.51674581e-01
3.71211857e-01 -6.46720588e-01 1.32621348e-01 -1.85437039e-01
4.74198252e-01 9.15126026e-01 5.02061069e-01 -2.65812457e-01
-3.52080256e-01 5.09885252e-01 -3.17908585e-01 6.67909920e-01
1.10543239e+00 -3.08922846e-02 -8.44380021e-01 2.19877008e-02
1.52353691e-02 1.13314182e-01 -1.66914426e-02 -9.07377839e-01
-4.49930411e-03 -6.20181262e-01 5.09768009e-01 -1.00230002e+00
2.33437881e-01 -1.86995399e+00 1.42658341e+00 6.84752703e-01
2.72354424e-01 8.63440111e-02 1.22673355e-01 3.79277587e-01
3.91633421e-01 -1.92140818e-01 9.16477740e-01 1.08793033e-02
-3.80236745e-01 -2.41630584e-01 -6.74430966e-01 -9.05149579e-01
1.78393543e+00 -4.09010738e-01 -4.37487274e-01 -3.04089546e-01
-7.02256441e-01 8.33130598e-01 5.51513433e-01 5.99891953e-02
-2.36836057e-02 -1.41625786e+00 -1.61130041e-01 -1.73314109e-01
-5.31693101e-01 1.22487083e-01 5.86418986e-01 4.38604921e-01
-3.15792143e-01 5.64477444e-01 -8.56210291e-02 -2.37633631e-01
-7.66295135e-01 6.93326533e-01 7.26412416e-01 -4.80494350e-02
-3.06882888e-01 4.41069566e-02 -2.05898672e-01 -9.59079489e-02
-1.97999865e-01 -7.52715945e-01 6.03689775e-02 4.97728914e-01
2.75904447e-01 1.08720016e+00 1.08266190e-01 -3.37375045e-01
-4.54362482e-01 4.90852416e-01 6.50679231e-01 1.43832862e-01
1.27189660e+00 -6.58530712e-01 -9.45454687e-02 4.06718194e-01
9.37315345e-01 -3.02244395e-01 -1.13733888e+00 4.80319381e-01
-1.50910333e-01 -2.26618022e-01 3.69308293e-01 -1.18408620e+00
-1.03870404e+00 -7.87066482e-03 1.02999616e+00 8.91529262e-01
1.85265172e+00 -4.58622068e-01 7.64171064e-01 1.04560778e-01
5.94749749e-01 -2.03033686e+00 -9.41318572e-01 -5.03508866e-01
7.78908372e-01 -8.37015688e-01 2.87727356e-01 -9.65602025e-02
-3.98195475e-01 8.15846503e-01 4.65160042e-01 1.94567516e-01
9.97201443e-01 3.88337910e-01 -6.91166878e-01 3.33843857e-01
-3.77917886e-01 1.99497655e-01 -3.52071732e-01 5.38561523e-01
7.89469387e-03 8.31645489e-01 -4.69173670e-01 5.94421566e-01
-1.43108338e-01 -1.08672522e-01 4.30745721e-01 8.78171742e-01
-5.22224069e-01 -5.31474411e-01 -7.09699333e-01 3.78262579e-01
-1.03766076e-01 4.22739893e-01 5.23239136e-01 1.01365697e+00
2.80949384e-01 1.38002157e+00 9.91479233e-02 9.61205214e-02
7.70700812e-01 1.79681778e-01 -2.70895511e-01 1.90729406e-02
-2.87097335e-01 1.54643238e-01 2.53981203e-01 -3.30139518e-01
-3.71362597e-01 -8.17308307e-01 -1.64337671e+00 -8.38428199e-01
-5.34176290e-01 4.07310992e-01 8.78114223e-01 9.83671427e-01
4.10784483e-02 6.67370975e-01 1.32372534e+00 -7.09109128e-01
-6.49095058e-01 -8.99418831e-01 -1.03212428e+00 -9.49051306e-02
4.91674116e-04 -6.02786064e-01 -9.70077753e-01 -4.71799314e-01] | [5.715348243713379, 2.483076572418213] |
daaa3a07-1154-45c4-8a52-6c8edd27620e | humandiffusion-a-coarse-to-fine-alignment | 2211.06235 | null | https://arxiv.org/abs/2211.06235v1 | https://arxiv.org/pdf/2211.06235v1.pdf | HumanDiffusion: a Coarse-to-Fine Alignment Diffusion Framework for Controllable Text-Driven Person Image Generation | Text-driven person image generation is an emerging and challenging task in cross-modality image generation. Controllable person image generation promotes a wide range of applications such as digital human interaction and virtual try-on. However, previous methods mostly employ single-modality information as the prior condition (e.g. pose-guided person image generation), or utilize the preset words for text-driven human synthesis. Introducing a sentence composed of free words with an editable semantic pose map to describe person appearance is a more user-friendly way. In this paper, we propose HumanDiffusion, a coarse-to-fine alignment diffusion framework, for text-driven person image generation. Specifically, two collaborative modules are proposed, the Stylized Memory Retrieval (SMR) module for fine-grained feature distillation in data processing and the Multi-scale Cross-modality Alignment (MCA) module for coarse-to-fine feature alignment in diffusion. These two modules guarantee the alignment quality of the text and image, from image-level to feature-level, from low-resolution to high-resolution. As a result, HumanDiffusion realizes open-vocabulary person image generation with desired semantic poses. Extensive experiments conducted on DeepFashion demonstrate the superiority of our method compared with previous approaches. Moreover, better results could be obtained for complicated person images with various details and uncommon poses. | ['Tieniu Tan', 'Zhenan Sun', 'Kunbo Zhang', 'Binghao Zhao', 'Jianxin Sun', 'Muyi Sun', 'Kaiduo Zhang'] | 2022-11-11 | null | null | null | null | ['virtual-try-on'] | ['computer-vision'] | [ 2.78562009e-01 -2.13783056e-01 3.55958462e-01 -3.93754691e-01
-7.16225028e-01 -3.23009312e-01 9.22076583e-01 -5.37140012e-01
-2.76271433e-01 6.43312633e-01 4.04932082e-01 4.42313880e-01
7.88740888e-02 -9.16969419e-01 -4.61228400e-01 -7.89979994e-01
6.69991851e-01 4.00069267e-01 -4.27360944e-02 -4.73794907e-01
-7.35767335e-02 1.96808666e-01 -1.56412756e+00 2.22621813e-01
9.75818574e-01 7.72740066e-01 4.35108930e-01 4.12524760e-01
-1.16640605e-01 1.76238343e-01 -5.60274661e-01 -7.15286076e-01
3.27352554e-01 -6.09933615e-01 -3.31919491e-01 6.02489233e-01
6.05800807e-01 -3.60105932e-01 -4.10021007e-01 1.14619935e+00
9.63514268e-01 1.51485115e-01 5.06785095e-01 -1.18753767e+00
-9.59011257e-01 2.54542798e-01 -8.05865169e-01 -3.20052475e-01
7.98006296e-01 6.14201307e-01 3.52708250e-01 -1.02659655e+00
7.35242844e-01 1.70600557e+00 3.19930613e-01 8.81515920e-01
-1.06093371e+00 -7.99347997e-01 2.53464609e-01 -9.13552791e-02
-1.64043570e+00 -3.96490067e-01 8.00621033e-01 -5.29811978e-01
3.14319342e-01 3.62150162e-01 1.03343117e+00 1.32505870e+00
2.11442024e-01 7.14634895e-01 1.30782664e+00 -2.90649861e-01
-2.74987906e-01 2.99131751e-01 -5.05279183e-01 9.16514933e-01
2.90995657e-01 2.21034452e-01 -8.08777273e-01 1.10761501e-01
1.33177781e+00 2.56677002e-01 -2.65512884e-01 -3.51138741e-01
-1.72035933e+00 5.90876937e-01 3.74997675e-01 1.10367060e-01
-5.08528531e-01 7.66923791e-03 2.08875656e-01 1.71213858e-02
3.34759712e-01 2.25713044e-01 4.85805832e-02 1.18497215e-01
-8.81386220e-01 7.38828897e-01 1.39041454e-01 1.12949407e+00
4.65749979e-01 1.13945924e-01 -7.58935452e-01 7.74707735e-01
3.79633576e-01 1.14058495e+00 6.40144050e-01 -6.87117755e-01
5.86999774e-01 5.80266297e-01 3.69475007e-01 -1.22608769e+00
-1.07297219e-01 -2.69280374e-01 -1.29757297e+00 3.34730707e-02
2.27836132e-01 -2.37006530e-01 -1.06177306e+00 1.79783154e+00
5.99419773e-01 -1.73416689e-01 -7.47579336e-02 1.44893730e+00
9.48690772e-01 6.60078645e-01 1.98657662e-01 -2.20352113e-01
1.75567865e+00 -9.90572155e-01 -9.49530423e-01 -2.84034908e-01
-1.47107586e-01 -9.81465638e-01 1.12287521e+00 2.25147113e-01
-1.32152522e+00 -9.56685483e-01 -7.40188837e-01 -1.82700261e-01
-2.59438515e-01 3.28964978e-01 4.96389300e-01 5.90563655e-01
-7.76984155e-01 8.32891986e-02 -1.47171929e-01 -3.85574847e-01
2.27073088e-01 1.46640718e-01 -6.04587555e-01 1.41336126e-02
-1.44441640e+00 4.94993776e-01 3.12695891e-01 1.83941722e-01
-6.87054634e-01 -3.57713073e-01 -9.09165382e-01 -3.05761576e-01
3.35644931e-01 -1.41883540e+00 9.59892809e-01 -9.25130725e-01
-1.58324707e+00 1.12729084e+00 -1.78590283e-01 -3.53000127e-02
9.89335239e-01 -2.20458597e-01 -4.49693710e-01 1.10985704e-01
2.64860123e-01 9.79247153e-01 1.33538127e+00 -1.36908948e+00
-6.35180295e-01 -5.50597012e-01 -5.47995418e-02 7.43604779e-01
-4.07581121e-01 -1.28591701e-01 -9.00216281e-01 -1.20617318e+00
2.19531972e-02 -8.91471446e-01 -3.04475307e-01 1.97873265e-01
-6.49459839e-01 -1.18106127e-01 5.72645187e-01 -8.41618538e-01
1.00999522e+00 -1.86438525e+00 5.58314562e-01 2.17969060e-01
2.61136591e-01 5.69119230e-02 -8.10315683e-02 1.31865740e-01
1.72740638e-01 -1.32056221e-01 2.12356690e-02 -5.94153583e-01
6.84780329e-02 -3.42834026e-01 -1.18825413e-01 1.83009803e-01
-3.85654569e-02 1.17257309e+00 -7.84280419e-01 -9.15048242e-01
5.53933382e-01 5.95835149e-01 -2.65747070e-01 3.25507283e-01
2.77271755e-02 8.47649157e-01 -6.90622211e-01 5.54760039e-01
6.64210141e-01 -5.76684922e-02 -3.29874188e-01 -6.01367831e-01
1.46405473e-02 -5.83139777e-01 -1.20709121e+00 2.14357710e+00
-4.85110283e-01 1.98960733e-02 2.95901466e-02 -1.87938854e-01
1.07806957e+00 2.17844754e-01 3.89955401e-01 -7.76618183e-01
2.08985418e-01 2.26506288e-03 -4.39231694e-01 -6.12551749e-01
8.10056269e-01 -2.13836297e-01 -3.34243327e-01 3.56607616e-01
-2.56506383e-01 -3.12402070e-01 1.25757113e-01 1.72141135e-01
1.01121530e-01 3.16817254e-01 -2.91733369e-02 -1.93905875e-01
7.68888712e-01 -1.67735055e-01 2.91118056e-01 3.60677898e-01
4.40266244e-02 1.05157697e+00 -2.61837840e-01 -2.60089219e-01
-1.21044481e+00 -1.16191864e+00 1.60238549e-01 6.71332300e-01
6.58591628e-01 -3.33142221e-01 -1.09377337e+00 -3.71167660e-01
-2.17529178e-01 3.46539378e-01 -5.48149765e-01 -2.08149374e-01
-5.02059579e-01 -4.57684696e-01 2.85320878e-01 3.32389325e-01
1.03261304e+00 -1.25876999e+00 -2.94148326e-01 2.85496507e-02
-5.14459908e-01 -1.09701455e+00 -1.14922321e+00 -9.77153718e-01
-4.17510152e-01 -5.86047292e-01 -1.53604317e+00 -9.59650040e-01
9.61412311e-01 3.35276216e-01 7.15564907e-01 -1.24210052e-01
-5.34133017e-01 4.02586162e-01 -1.41710639e-01 -1.57481283e-01
-7.23505318e-02 -1.29502922e-01 3.39892685e-01 4.57689553e-01
7.89031908e-02 -3.19604188e-01 -9.68916953e-01 2.89360255e-01
-8.59151721e-01 7.43303597e-01 7.70280838e-01 9.36992943e-01
6.74878240e-01 -1.16400057e-02 3.21516454e-01 -4.10062164e-01
9.92183506e-01 8.18612240e-03 -2.69329667e-01 4.13954645e-01
-2.90362895e-01 -1.31388873e-01 5.45494854e-01 -6.63512588e-01
-1.37282586e+00 8.25948417e-02 1.55764073e-02 -5.28898895e-01
-1.64151073e-01 -4.12942655e-02 -6.11785352e-01 1.07889608e-01
4.90200430e-01 7.26359129e-01 1.97956890e-01 -2.59357959e-01
6.83207333e-01 6.77323401e-01 7.51095831e-01 -7.36717105e-01
1.20603406e+00 5.39499402e-01 -2.87667990e-01 -5.22632778e-01
-6.12357020e-01 2.19736248e-02 -6.02204382e-01 -4.99992698e-01
1.33930933e+00 -1.21214402e+00 -7.35367954e-01 8.61976266e-01
-1.21019709e+00 2.38868743e-02 -3.18817794e-01 2.72876948e-01
-4.92950112e-01 1.92208573e-01 -5.07833064e-01 -6.44182265e-01
-7.60473847e-01 -1.18981349e+00 1.59877551e+00 6.81681991e-01
-8.00448358e-02 -7.23729312e-01 -3.32778215e-01 8.13187420e-01
2.82331526e-01 3.09520096e-01 2.53113002e-01 1.84124410e-01
-7.61781752e-01 -5.54648302e-02 -3.91694725e-01 6.49952367e-02
2.30354995e-01 -5.07758141e-01 -6.22615993e-01 -3.57967883e-01
-3.65077585e-01 -2.66173869e-01 5.06626487e-01 1.43147096e-01
1.10981023e+00 -2.07906306e-01 -4.09376502e-01 5.74572146e-01
9.62123692e-01 -5.75966993e-03 6.58248544e-01 1.29053712e-01
1.11130178e+00 7.73536801e-01 9.29387510e-01 6.11317515e-01
5.47932923e-01 9.16980028e-01 -2.16411382e-01 -4.25556391e-01
-4.50675845e-01 -7.91105330e-01 2.90128589e-01 5.52378774e-01
-2.66476065e-01 -2.68867642e-01 -4.75828588e-01 3.47731948e-01
-1.86044979e+00 -1.02851379e+00 -2.56686453e-02 2.16484499e+00
8.18638742e-01 -1.08870924e-01 1.85670286e-01 -2.83584207e-01
1.01780879e+00 1.12937875e-01 -5.87574482e-01 2.40528211e-01
-3.04918826e-01 -1.70961395e-01 8.71165916e-02 4.34559315e-01
-9.16670382e-01 1.16685653e+00 4.72338533e+00 1.36978412e+00
-9.38558698e-01 1.76589370e-01 6.87679172e-01 -2.24332929e-01
-4.97099012e-01 -4.61263806e-01 -8.72757733e-01 6.26551330e-01
3.29703577e-02 -2.70902067e-01 3.05204868e-01 5.12483120e-01
4.88662332e-01 4.77643572e-02 -6.48366809e-01 1.64162660e+00
4.22510773e-01 -1.18379033e+00 5.70347965e-01 9.42646340e-03
9.18172479e-01 -9.75359619e-01 4.47754472e-01 5.02647236e-02
-4.44933735e-02 -9.79625523e-01 1.08933938e+00 1.12405360e+00
1.26993179e+00 -9.00062144e-01 4.18071538e-01 1.18116543e-01
-1.42847061e+00 1.99193791e-01 -2.00428218e-01 2.82984793e-01
7.19977796e-01 5.49530447e-01 -1.50820792e-01 6.93860888e-01
7.01631308e-01 4.97911632e-01 -6.06544614e-01 4.27289426e-01
-3.09601352e-02 -3.01195174e-01 -1.62692033e-02 -7.27367699e-02
-9.48108956e-02 -4.39162254e-01 5.60034931e-01 1.02331686e+00
5.21532059e-01 2.78441817e-01 4.65542108e-01 9.89345849e-01
2.40366757e-01 3.04463923e-01 -4.56932902e-01 1.14079766e-01
3.59392792e-01 1.36228859e+00 -4.23801333e-01 -5.49315989e-01
2.09673364e-02 1.65280640e+00 -7.96797872e-02 2.99234241e-01
-9.40930545e-01 -3.51272315e-01 5.95213830e-01 4.41702247e-01
1.52195608e-02 -2.24057823e-01 -1.07781120e-01 -1.37745869e+00
1.76743627e-01 -1.07362473e+00 4.27237153e-02 -1.14627731e+00
-1.42979860e+00 6.76093042e-01 7.79723004e-02 -1.42286599e+00
-2.69590944e-01 -2.25852668e-01 -4.56081182e-01 1.27186608e+00
-9.71456766e-01 -1.69215429e+00 -8.05626869e-01 1.08144319e+00
7.33683646e-01 -3.09008420e-01 5.31140506e-01 4.23886567e-01
-4.79721129e-01 7.47377455e-01 -4.11216080e-01 1.00270383e-01
9.32378888e-01 -8.16393375e-01 5.61941147e-01 8.02704751e-01
-3.86855379e-02 8.56781006e-01 5.66705227e-01 -9.63420510e-01
-1.30905282e+00 -1.25109422e+00 6.03233576e-01 -3.84804428e-01
1.14187144e-01 -3.33976030e-01 -3.11147690e-01 2.04922006e-01
1.41820326e-01 -3.21498275e-01 2.17535183e-01 -3.75120670e-01
3.79349999e-02 -2.06556022e-01 -1.10961676e+00 1.13183737e+00
1.40581894e+00 -3.80742610e-01 -3.09594542e-01 3.72279584e-01
7.67052770e-01 -7.02089846e-01 -7.84593344e-01 2.30689913e-01
6.96249008e-01 -9.32810903e-01 1.23132408e+00 -2.01252654e-01
4.25679982e-01 -6.64824545e-01 1.22744337e-01 -1.16833591e+00
-3.82225245e-01 -7.88821340e-01 1.11184590e-01 1.46151853e+00
5.83693758e-02 -5.20696998e-01 6.23633146e-01 6.65913343e-01
1.79713175e-01 -5.49151182e-01 -5.13669968e-01 -5.29316306e-01
-2.72209823e-01 2.62508355e-03 8.84643495e-01 6.99156046e-01
-3.35177124e-01 4.34006065e-01 -9.92340088e-01 1.29235927e-02
8.57652128e-01 2.79664129e-01 1.16574621e+00 -9.50574875e-01
-2.26936638e-01 -3.66191745e-01 -3.51717353e-01 -1.33102822e+00
-1.81346014e-01 -6.26734555e-01 -3.34812626e-02 -1.62587130e+00
4.30063337e-01 -3.25893968e-01 2.24806845e-01 9.59218517e-02
-4.60386097e-01 4.80280489e-01 4.03771490e-01 2.26850882e-01
-4.52367693e-01 8.58201623e-01 2.04308319e+00 -1.48235455e-01
-9.63347927e-02 -2.46278018e-01 -8.06623936e-01 6.07982934e-01
6.17495358e-01 7.96734765e-02 -5.59684396e-01 -3.76925915e-01
-1.80444568e-01 2.00139090e-01 7.55593896e-01 -1.01855457e+00
2.18769267e-01 -2.98832595e-01 7.68336415e-01 -5.28908789e-01
7.87775815e-01 -5.49724102e-01 4.10287827e-01 5.02353609e-01
-2.47629777e-01 9.77069438e-02 -1.91897377e-01 5.80694318e-01
-1.32792622e-01 3.85602355e-01 6.45803630e-01 -3.72809410e-01
-7.21945167e-01 9.01578069e-01 2.01231912e-01 -3.63778733e-02
1.13328862e+00 -3.92311603e-01 -1.16335288e-01 -5.93470514e-01
-7.32752681e-01 2.86804765e-01 5.05189478e-01 9.17095482e-01
1.00191021e+00 -1.80773389e+00 -1.01058292e+00 4.31928784e-01
2.35464007e-01 6.59429953e-02 8.64694834e-01 7.00981617e-01
-3.39991063e-01 2.35108361e-01 -2.83041477e-01 -5.44448912e-01
-1.36497760e+00 4.29847896e-01 3.21195275e-01 -1.98300481e-01
-6.96943283e-01 8.95423234e-01 6.18843853e-01 -4.44654047e-01
-1.21511444e-01 3.47497940e-01 -2.81461235e-02 -5.35856374e-03
7.37470746e-01 8.99591893e-02 -5.28582156e-01 -1.12737942e+00
-3.81606631e-02 1.05248189e+00 -1.37170687e-01 -5.35403252e-01
7.92419136e-01 -5.21969914e-01 9.40432996e-02 1.54486969e-01
7.13073790e-01 -8.57506469e-02 -1.37152278e+00 -2.63932735e-01
-8.95631015e-01 -7.98112631e-01 -1.98947236e-01 -6.68659806e-01
-1.13072157e+00 8.20962906e-01 8.27463329e-01 -2.66412646e-01
1.10463655e+00 -1.81522280e-01 1.18579173e+00 -5.22298105e-02
6.82859600e-01 -1.06497169e+00 4.83116478e-01 -6.56404719e-02
1.29886019e+00 -1.13960195e+00 2.18475834e-02 -5.34519553e-01
-1.00903153e+00 7.69013524e-01 1.11496699e+00 1.56540826e-01
1.71726972e-01 -1.61849231e-01 1.72833681e-01 -1.00338414e-01
-2.71711975e-01 -2.26091385e-01 6.97553933e-01 7.98761249e-01
1.55811727e-01 2.20404059e-01 -3.18375140e-01 6.49472535e-01
-6.30452275e-01 -4.26131785e-02 3.93368602e-02 4.93611872e-01
-3.18816364e-01 -1.00351644e+00 -7.70226657e-01 2.47182712e-01
6.14494048e-02 -1.63879097e-01 -2.50808597e-01 6.15962923e-01
5.75840056e-01 8.88510108e-01 -1.42379984e-01 -4.84247625e-01
4.45673078e-01 -2.57970512e-01 6.73039317e-01 -3.65646809e-01
-3.99703115e-01 1.28364071e-01 -1.47006258e-01 -5.21212697e-01
-3.69330406e-01 -7.28296340e-01 -1.04478383e+00 -4.10166174e-01
-9.38851610e-02 -1.30406991e-01 3.52759421e-01 8.07920277e-01
3.65642041e-01 4.78170484e-01 4.55763876e-01 -1.06010664e+00
-1.51166424e-01 -9.01739120e-01 -5.17260551e-01 8.24513078e-01
-2.96103433e-02 -5.57783365e-01 1.61862954e-01 4.51194614e-01] | [11.980626106262207, -0.8390780091285706] |
a886ad20-6422-455c-8eaf-a2bf79ccae6b | large-scale-news-classification-using-bert | 2107.06785 | null | https://arxiv.org/abs/2107.06785v2 | https://arxiv.org/pdf/2107.06785v2.pdf | Large-Scale News Classification using BERT Language Model: Spark NLP Approach | The rise of big data analytics on top of NLP increases the computational burden for text processing at scale. The problems faced in NLP are very high dimensional text, so it takes a high computation resource. The MapReduce allows parallelization of large computations and can improve the efficiency of text processing. This research aims to study the effect of big data processing on NLP tasks based on a deep learning approach. We classify a big text of news topics with fine-tuning BERT used pre-trained models. Five pre-trained models with a different number of parameters were used in this study. To measure the efficiency of this method, we compared the performance of the BERT with the pipelines from Spark NLP. The result shows that BERT without Spark NLP gives higher accuracy compared to BERT with Spark NLP. The accuracy average and training time of all models using BERT is 0.9187 and 35 minutes while using BERT with Spark NLP pipeline is 0.8444 and 9 minutes. The bigger model will take more computation resources and need a longer time to complete the tasks. However, the accuracy of BERT with Spark NLP only decreased by an average of 5.7%, while the training time was reduced significantly by 62.9% compared to BERT without Spark NLP. | ['Anantha Yullian Sukmadewa', 'Kuncahyo Setyo Nugroho', 'Novanto Yudistira'] | 2021-07-14 | null | null | null | null | ['news-classification'] | ['natural-language-processing'] | [-1.11951303e+00 -1.05253704e-01 5.27954638e-01 -6.38755620e-01
-5.69134235e-01 -4.60218549e-01 5.92928588e-01 2.94668227e-01
-6.39800906e-01 5.27402580e-01 4.13737804e-01 -1.44682646e-01
3.43624912e-02 -1.35524893e+00 -7.24100292e-01 -6.21905327e-01
2.20133677e-01 1.05745196e+00 5.92950312e-03 -8.03245697e-03
-3.44296545e-02 2.99516261e-01 -1.39931381e+00 1.06796336e+00
8.57197762e-01 8.29962373e-01 5.32805681e-01 6.96561337e-01
-9.38585401e-01 9.40716863e-01 -1.05773258e+00 -1.74221277e-01
6.51561201e-01 3.89759123e-01 -5.75223207e-01 -6.65191770e-01
-3.70232202e-02 -2.19999701e-01 -9.77342576e-02 6.06292427e-01
9.06055510e-01 1.51863217e-01 2.10440874e-01 -1.20156908e+00
-3.03711593e-01 8.75440359e-01 -4.22105551e-01 3.38626623e-01
1.39247105e-01 6.49076747e-03 5.16630948e-01 -6.88492417e-01
4.53711540e-01 1.36004686e+00 8.47417772e-01 -1.73918065e-02
-7.97823548e-01 -9.05763686e-01 -4.25037652e-01 1.62010536e-01
-1.37037241e+00 -5.26756868e-02 -2.84052510e-02 -6.78739727e-01
1.49206948e+00 9.02867392e-02 5.31436324e-01 3.99331450e-01
5.83263934e-01 6.90837085e-01 1.15407848e+00 -1.97916821e-01
4.40040618e-01 3.48787516e-01 7.09881425e-01 3.44913788e-02
1.53726026e-01 -2.71435648e-01 -5.00572503e-01 -6.78095296e-02
-1.25351980e-01 4.17234778e-01 4.52757806e-01 8.68507862e-01
-1.03701854e+00 1.01316452e+00 2.81783879e-01 7.49862909e-01
-8.50380421e-01 -1.70673192e-01 7.32158601e-01 6.41931534e-01
9.22546625e-01 4.08516645e-01 -1.08881307e+00 -4.82077241e-01
-9.48414743e-01 4.69188929e-01 1.28099453e+00 7.54306793e-01
9.34939265e-01 -1.49263933e-01 -3.36293906e-01 8.20625365e-01
-1.65966123e-01 5.16708732e-01 8.17931831e-01 -3.69285882e-01
7.46342003e-01 1.07573020e+00 -3.31893750e-02 -1.10545063e+00
-1.09373200e+00 -1.75194591e-01 -8.79554808e-01 -2.05230221e-01
5.88967919e-01 -6.28789008e-01 -8.57457697e-01 8.30057085e-01
4.67198223e-01 -3.16378742e-01 3.21574062e-01 6.49975121e-01
1.17341495e+00 1.56575704e+00 2.39550114e-01 1.86430797e-01
1.51818895e+00 -1.22583175e+00 -8.77075434e-01 -2.00957581e-02
1.29955137e+00 -1.03916967e+00 1.50937748e+00 4.28428352e-01
-8.07907343e-01 -5.97333968e-01 -5.93722224e-01 -3.69078517e-01
-9.24685657e-01 2.97904700e-01 8.55932653e-01 5.59597969e-01
-1.03812659e+00 3.40982527e-01 -6.75630569e-01 -4.27478224e-01
1.67060018e-01 1.32209346e-01 -2.19155684e-01 -3.29561651e-01
-1.35543847e+00 6.53547406e-01 7.94074953e-01 -8.25798213e-02
-7.51889348e-01 -1.26065624e+00 -2.20440000e-01 3.88267756e-01
-8.65081847e-02 -3.32943410e-01 1.15668893e+00 -3.82280916e-01
-1.19905555e+00 5.93442798e-01 -2.00073868e-01 -6.26088381e-01
5.62135458e-01 -4.72857267e-01 -4.11543608e-01 -1.29018798e-01
3.20569485e-01 4.52159494e-01 7.67372325e-02 -2.85524875e-01
-9.21887279e-01 -6.31600797e-01 -6.91264570e-02 5.02879061e-02
-3.97575200e-01 5.79934180e-01 -1.48732260e-01 -3.61772776e-02
-3.74021024e-01 -7.57604539e-01 -2.09407270e-01 -7.98237562e-01
-1.47140056e-01 -8.11941683e-01 7.67786324e-01 -8.36645901e-01
1.05909157e+00 -2.08230853e+00 -9.09095347e-01 -1.77742809e-01
4.59930211e-01 3.44813317e-01 1.22082651e-01 8.78367305e-01
3.22852135e-01 3.82954925e-01 4.63662684e-01 -1.30209057e-02
-1.10848948e-01 2.58948982e-01 -2.91066855e-01 -2.03905292e-02
-3.57268661e-01 8.68004441e-01 -3.37263286e-01 -5.41901231e-01
6.28238767e-02 5.73714912e-01 -2.91171879e-01 1.72338337e-01
-3.89247388e-01 1.00416787e-01 -5.68621874e-01 4.63699885e-02
1.15537322e+00 -4.67139423e-01 -3.63385201e-01 2.27500468e-01
-4.37186509e-01 5.19909561e-01 -1.06863070e+00 1.40526605e+00
-7.05008090e-01 7.36752629e-01 1.33193061e-01 -9.92432237e-01
1.22140551e+00 5.70795178e-01 5.08534372e-01 -8.36319804e-01
-1.22351341e-01 2.88717672e-02 -5.00690788e-02 -9.04136121e-01
3.47949952e-01 4.12911475e-02 1.54252902e-01 9.37657297e-01
-2.83067971e-01 -1.21778287e-01 5.39252460e-01 4.72586334e-01
1.26690388e+00 -5.46694160e-01 -1.02113783e-01 -3.51404965e-01
1.75153390e-01 6.36512578e-01 6.41481280e-01 7.58392692e-01
6.88771084e-02 1.01049162e-01 6.61378682e-01 -1.27759135e+00
-1.12509227e+00 -3.46935064e-01 -2.65448332e-01 1.60669744e+00
-6.93066478e-01 -7.26163208e-01 -6.11231625e-01 -3.58460307e-01
1.16489448e-01 8.01205993e-01 -2.97195971e-01 5.92133045e-01
-4.09112215e-01 -1.36734271e+00 5.56710005e-01 2.72705883e-01
1.00143123e+00 -1.24246454e+00 -4.38334495e-01 1.65341556e-01
-3.10044765e-01 -1.26832020e+00 3.00353527e-01 -9.98344943e-02
-7.40284085e-01 -5.19311547e-01 -3.28925282e-01 -6.47362649e-01
2.44979918e-01 2.09780261e-01 1.24467552e+00 -1.62856549e-01
-4.00880337e-01 -3.89766574e-01 -6.29125535e-01 -1.15355778e+00
-4.28134650e-01 3.34758162e-01 -3.47281456e-01 -3.37422162e-01
1.13236284e+00 -1.71489626e-01 -3.32709640e-01 2.08167642e-01
-8.86314929e-01 5.37258029e-01 6.10531807e-01 2.76939124e-01
3.09843481e-01 7.18878925e-01 7.52488732e-01 -9.27674830e-01
1.03314114e+00 -5.61227560e-01 -5.70128024e-01 6.70215115e-02
-7.75442064e-01 -1.42730638e-01 9.29441452e-01 -1.58836097e-01
-1.29700208e+00 5.89774325e-02 -1.81044623e-01 1.84482768e-01
-1.88681573e-01 8.50606382e-01 6.89334497e-02 5.04612505e-01
9.47122872e-01 -1.37238935e-01 -2.30219781e-01 -8.13199997e-01
1.07888229e-01 1.16978979e+00 -4.04199272e-01 -2.07016528e-01
8.24945644e-02 5.15562892e-01 -5.00359118e-01 -9.89319503e-01
-9.34223235e-01 -7.74207056e-01 -4.72653657e-01 1.01344064e-01
8.59794438e-01 -1.23329270e+00 -8.95066738e-01 3.99644017e-01
-1.29563606e+00 -5.02168059e-01 -2.81451195e-01 7.26229668e-01
6.51699528e-02 -1.87643752e-01 -1.04292023e+00 -4.35775578e-01
-1.03205764e+00 -8.01606297e-01 8.39796901e-01 6.74657598e-02
6.41646609e-02 -7.59730101e-01 1.51312202e-01 6.48739219e-01
7.83684671e-01 3.83902015e-03 6.24406457e-01 -1.19410944e+00
-1.65900752e-01 -6.98479831e-01 -6.27935886e-01 1.32279202e-01
-5.23473442e-01 -6.29444271e-02 -1.26392424e+00 6.99722990e-02
4.84075427e-01 -2.09742695e-01 6.50347888e-01 5.45734584e-01
1.07077861e+00 -5.47139704e-01 -3.67244214e-01 3.01553130e-01
1.37096739e+00 1.35497048e-01 5.85319340e-01 6.28876209e-01
6.76821411e-01 8.67100298e-01 6.53593302e-01 5.99843323e-01
3.97185981e-01 1.14179276e-01 -2.10979596e-01 1.45565554e-01
2.67999202e-01 -5.63757904e-02 2.87439704e-01 9.77843881e-01
1.76738083e-01 -6.40014783e-02 -1.63783205e+00 3.92098606e-01
-1.89681494e+00 -7.46508300e-01 -8.02900791e-01 1.66249990e+00
8.96391511e-01 -1.64698780e-01 -1.41487941e-01 9.60978195e-02
6.43418908e-01 -1.17883295e-01 -1.71708688e-01 -6.45117581e-01
-4.72490899e-02 -1.21090263e-01 4.12083507e-01 3.44820052e-01
-7.92491734e-01 1.08472645e+00 5.86969614e+00 9.10503328e-01
-1.18294251e+00 5.65351725e-01 5.77713132e-01 -5.79353988e-01
8.59114155e-02 -3.95896696e-02 -1.31116581e+00 8.21644843e-01
1.59193850e+00 -5.01093566e-01 2.48464569e-01 1.09310782e+00
6.81074381e-01 -1.55404091e-01 -7.66669452e-01 1.01605523e+00
-3.88309985e-01 -1.44843054e+00 -2.40095899e-01 1.83465526e-01
8.35480094e-01 9.12703633e-01 -7.85908341e-01 8.29867244e-01
5.51895201e-01 -9.88594234e-01 1.45482525e-01 3.56938273e-01
-1.40761733e-02 -6.58887565e-01 1.31148267e+00 1.10040116e+00
-7.90552199e-01 -1.74716532e-01 -1.18278980e+00 -8.58111978e-01
4.78649363e-02 1.74415648e+00 -1.44417417e+00 1.71112195e-01
1.22319102e+00 3.59328121e-01 -4.08869505e-01 7.22444296e-01
-1.82378948e-01 9.13537323e-01 -9.12633359e-01 -5.38131595e-01
2.26818383e-01 -1.10080913e-01 2.43681315e-02 1.50037277e+00
6.09096549e-02 1.96390366e-03 3.77911419e-01 6.56905651e-01
-6.01333864e-02 7.92244017e-01 -4.46293592e-01 -1.46584034e-01
6.03984535e-01 1.64824820e+00 -6.35636806e-01 -9.42054510e-01
-3.27121496e-01 1.79635257e-01 2.06467032e-01 1.87616259e-01
-6.39503598e-01 -5.98609388e-01 2.99298540e-02 2.72706270e-01
-1.96541697e-01 -1.06548667e-01 -7.97760367e-01 -9.52111244e-01
8.61792341e-02 -5.08939862e-01 4.96897489e-01 -1.00099719e+00
-1.22208595e+00 6.21635437e-01 -9.23224762e-02 -4.15110439e-01
-1.44812822e-01 -3.87511522e-01 -7.41077006e-01 1.06834733e+00
-1.04685652e+00 -1.03986633e+00 -7.25959420e-01 5.55394888e-01
5.45124292e-01 -1.98672101e-01 8.12446058e-01 4.63043809e-01
-6.05270386e-01 1.13246329e-02 5.64332485e-01 2.20992804e-01
7.46030927e-01 -9.82385099e-01 5.93279183e-01 5.05055785e-01
-2.28070453e-01 3.06958228e-01 5.49005568e-01 -7.54720330e-01
-1.30102754e+00 -1.10859334e+00 1.70668209e+00 -3.70715201e-01
5.65744042e-01 -5.42316854e-01 -9.63199258e-01 6.25886083e-01
2.16598481e-01 -4.76183109e-02 7.05588996e-01 4.77378130e-01
-1.25261515e-01 -6.15713835e-01 -1.29336429e+00 1.39169842e-02
1.79264456e-01 -2.07117930e-01 -4.51860189e-01 1.19470656e+00
1.10671353e+00 8.13587848e-03 -9.98729408e-01 -2.31340807e-02
3.04784685e-01 -9.47308183e-01 5.75832844e-01 -3.50700676e-01
5.05560935e-01 9.54540297e-02 2.03671619e-01 -1.18431818e+00
-1.52292565e-01 -3.07652116e-01 4.26319867e-01 1.23121870e+00
5.94970286e-01 -1.02427900e+00 8.94277155e-01 1.14895451e+00
-2.23698709e-02 -5.77968001e-01 -5.13346314e-01 -7.56886423e-01
3.05539519e-01 -7.36164689e-01 9.10332918e-01 1.05930638e+00
1.19324960e-01 6.55132651e-01 9.60172787e-02 -2.05191121e-01
4.64411601e-02 1.56314731e-01 1.19798851e+00 -1.40573502e+00
1.68082695e-02 7.88990557e-02 -1.68689162e-01 -6.21449113e-01
-1.22629374e-01 -9.68019366e-01 -2.78782487e-01 -2.04224205e+00
1.94604069e-01 -6.74463332e-01 1.51370898e-01 7.57615924e-01
1.13535129e-01 -4.58635539e-01 2.40311265e-01 5.94089031e-01
-3.37670207e-01 1.52166247e-01 1.05830157e+00 -9.96470377e-02
-5.41862190e-01 6.61680996e-02 -4.71237987e-01 7.08712399e-01
1.32134652e+00 -7.98126936e-01 -1.95231169e-01 -8.52447629e-01
9.73223746e-01 -1.77486360e-01 -3.85296702e-01 -8.81473124e-01
6.04966044e-01 2.13620499e-01 4.94750112e-01 -1.00846565e+00
-9.37748998e-02 -6.13376915e-01 7.91548006e-03 4.47694987e-01
-2.86196589e-01 2.53218591e-01 4.26606238e-01 3.50631401e-02
-2.84031779e-01 -2.49952629e-01 3.70096773e-01 -2.87447751e-01
-4.35978323e-01 8.63325372e-02 -3.95301253e-01 1.09581351e-01
1.03841996e+00 2.44082004e-01 -4.72499877e-01 -2.12566301e-01
-5.98766267e-01 4.50708896e-01 -2.14654520e-01 2.07876652e-01
8.49015489e-02 -7.30356157e-01 -1.12952626e+00 2.46294796e-01
-3.28812838e-01 5.88484824e-01 4.26149577e-01 9.63483989e-01
-1.17478383e+00 8.71865988e-01 -8.99749100e-02 -5.51975787e-01
-1.00458241e+00 6.88359082e-01 -1.15600176e-01 -7.53809988e-01
-1.00719500e+00 7.32724428e-01 2.63791293e-01 -9.26394045e-01
1.12751730e-01 -4.90624905e-01 -2.55480170e-01 1.90722704e-01
1.07060266e+00 9.19685483e-01 5.67124009e-01 1.48290336e-01
-1.16578311e-01 1.74104050e-01 -2.99247354e-01 -1.11522712e-01
1.66351771e+00 6.36073574e-02 -5.66734254e-01 4.79889065e-01
1.21943021e+00 -5.40132783e-02 -6.51793599e-01 5.35659529e-02
4.68599573e-02 -3.92766297e-01 2.58181721e-01 -1.15527439e+00
-9.63295758e-01 1.21816385e+00 3.73702228e-01 6.06187701e-01
8.32879543e-01 -1.20758731e-03 9.15231884e-01 7.29034305e-01
2.12398872e-01 -1.37034166e+00 -4.64694887e-01 9.07862186e-01
1.04884207e+00 -1.26674402e+00 2.47845259e-02 -1.09504551e-01
-6.52646720e-01 9.73503530e-01 3.78952861e-01 -1.87400699e-01
1.15993059e+00 7.11331427e-01 2.56485254e-01 -5.14924765e-01
-9.89725471e-01 3.08586001e-01 -3.25934112e-01 1.96428373e-01
5.56827724e-01 4.35564697e-01 -6.26193166e-01 7.32608438e-01
-8.73016536e-01 5.05611002e-01 3.40147793e-01 6.65113211e-01
-6.67464614e-01 -5.31068385e-01 -5.75628698e-01 8.62444282e-01
-9.01820123e-01 -2.72550464e-01 -1.67785659e-01 5.64589739e-01
-2.99946796e-02 1.21498728e+00 3.89639765e-01 -3.79180372e-01
5.87883964e-02 3.56339157e-01 -5.20318985e-01 -6.17630482e-01
-9.98860657e-01 -5.22368960e-02 8.16353112e-02 -5.38505971e-01
1.76123511e-02 -2.81544119e-01 -1.68339932e+00 -1.03206134e+00
-1.13847673e-01 5.24828553e-01 1.47159457e+00 8.00382376e-01
1.01100707e+00 2.46700540e-01 3.08913767e-01 -3.46526980e-01
-3.97201836e-01 -1.67183781e+00 -6.29528642e-01 2.98868984e-01
-3.83257359e-01 7.72871170e-03 -5.85576594e-01 2.40264349e-02] | [10.465505599975586, 8.122106552124023] |
b621ef41-2c98-4ffb-88e2-366fe42cdacc | atrial-septal-defect-detection-in-children | 2306.03835 | null | https://arxiv.org/abs/2306.03835v1 | https://arxiv.org/pdf/2306.03835v1.pdf | Atrial Septal Defect Detection in Children Based on Ultrasound Video Using Multiple Instances Learning | Purpose: Congenital heart defect (CHD) is the most common birth defect. Thoracic echocardiography (TTE) can provide sufficient cardiac structure information, evaluate hemodynamics and cardiac function, and is an effective method for atrial septal defect (ASD) examination. This paper aims to study a deep learning method based on cardiac ultrasound video to assist in ASD diagnosis. Materials and methods: We select two standard views of the atrial septum (subAS) and low parasternal four-compartment view (LPS4C) as the two views to identify ASD. We enlist data from 300 children patients as part of a double-blind experiment for five-fold cross-validation to verify the performance of our model. In addition, data from 30 children patients (15 positives and 15 negatives) are collected for clinician testing and compared to our model test results (these 30 samples do not participate in model training). We propose an echocardiography video-based atrial septal defect diagnosis system. In our model, we present a block random selection, maximal agreement decision and frame sampling strategy for training and testing respectively, resNet18 and r3D networks are used to extract the frame features and aggregate them to build a rich video-level representation. Results: We validate our model using our private dataset by five-cross validation. For ASD detection, we achieve 89.33 AUC, 84.95 accuracy, 85.70 sensitivity, 81.51 specificity and 81.99 F1 score. Conclusion: The proposed model is multiple instances learning-based deep learning model for video atrial septal defect detection which effectively improves ASD detection accuracy when compared to the performances of previous networks and clinical doctors. | ['Yuqi Zhang', 'Qingli Li', 'Jiangang Chen', 'Jionglong Su', 'Angelos Stefanidis', 'Jinfeng Wang', 'Lijun Chen', 'Zhifang Zhang', 'Tongtong Liang', 'Xiaoxiang Han', 'Qiming Huang', 'Yiman Liu'] | 2023-06-06 | null | null | null | null | ['defect-detection', 'specificity'] | ['computer-vision', 'natural-language-processing'] | [-1.43603042e-01 2.18569171e-02 -1.58955932e-01 -7.57870823e-02
-2.61803687e-01 -4.43093359e-01 -3.76228839e-01 1.72389019e-02
-1.56596810e-01 5.77047765e-01 -4.79577258e-02 -5.80252290e-01
4.51568607e-03 -7.13918209e-01 -4.60332990e-01 -5.26403308e-01
-3.90888363e-01 6.61849082e-01 2.56734669e-01 5.34545064e-01
-1.15902543e-01 5.18507063e-01 -1.26469529e+00 5.22829711e-01
9.46300924e-01 9.05109525e-01 -7.69940317e-02 1.14449191e+00
4.63810675e-02 8.62952471e-01 -7.87340641e-01 6.64326921e-03
2.66600311e-01 -4.98535156e-01 -7.09299088e-01 1.54736629e-02
3.83394450e-01 -9.89009440e-01 -2.49107450e-01 4.41371143e-01
1.27388859e+00 -4.79496598e-01 6.46860838e-01 -7.75379658e-01
-3.34641665e-01 5.06703913e-01 -4.84912276e-01 7.23437846e-01
2.84983963e-01 6.15429021e-02 4.26829517e-01 -1.15414691e+00
5.07088006e-01 6.21106625e-01 9.60060596e-01 7.61047065e-01
-5.47734082e-01 -7.43473291e-01 -4.75721806e-01 4.44392907e-03
-1.12485516e+00 -1.81748018e-01 5.06498158e-01 -8.94080937e-01
6.67003930e-01 -1.07106362e-02 1.53313696e+00 6.91811740e-01
-5.24431951e-02 7.41760552e-01 7.09200084e-01 -4.47223157e-01
-1.05310798e-01 -1.95573226e-01 -3.62285644e-01 1.25283241e+00
4.24573928e-01 2.44790003e-01 -2.50503689e-01 -5.76973595e-02
1.28238618e+00 1.01570748e-01 -5.28459191e-01 -3.01910907e-01
-1.30792749e+00 6.45433962e-01 -1.26245171e-01 3.79139602e-01
-3.50743622e-01 -2.15339839e-01 5.91543019e-01 3.89492184e-01
2.21670076e-01 9.13802087e-02 -5.79285622e-01 -3.42659593e-01
-7.64205813e-01 4.44760136e-02 7.60778308e-01 7.10634768e-01
-5.89325055e-02 4.68865216e-01 -2.94365436e-01 1.22428346e+00
3.79586607e-01 6.43588781e-01 7.83634961e-01 -9.57233667e-01
3.87475073e-01 4.56731260e-01 -3.22792768e-01 -7.40873933e-01
-3.68513048e-01 -4.87619430e-01 -1.21704042e+00 2.28589531e-02
8.17412063e-02 -7.73596346e-01 -1.04184186e+00 1.31230927e+00
2.86792517e-01 7.17906535e-01 -4.03353274e-02 9.52990234e-01
1.56904292e+00 3.19749147e-01 -2.05200419e-01 -5.15857399e-01
9.78191853e-01 -8.11096668e-01 -3.44058186e-01 5.37032485e-01
1.04074800e+00 -5.81284285e-01 4.56821859e-01 6.20714664e-01
-1.32265854e+00 -4.10475582e-01 -9.47322249e-01 6.97944939e-01
1.68737710e-01 8.23155582e-01 4.87094462e-01 8.23361695e-01
-1.11165869e+00 6.05810106e-01 -7.81987011e-01 -9.07822773e-02
7.03658938e-01 4.56384867e-01 -5.36601365e-01 1.92488626e-01
-1.03404462e+00 4.98820394e-01 2.05025822e-01 -1.53114095e-01
-8.84068787e-01 -9.49533224e-01 -6.78167164e-01 -5.61850742e-02
1.11085279e-02 -9.82517302e-01 1.00011122e+00 -6.72580123e-01
-1.32352746e+00 1.39895499e+00 2.62456357e-01 -5.66212714e-01
4.41752702e-01 -6.12698533e-02 -3.45378458e-01 6.85910165e-01
6.37769559e-03 4.81343091e-01 8.60017598e-01 -5.79307616e-01
-7.23212540e-01 -2.82213002e-01 -1.67418778e-01 6.09220006e-02
-4.67633344e-02 1.91361055e-01 -2.62291223e-01 -1.03576624e+00
5.50306380e-01 -8.72469604e-01 3.05749774e-02 1.33109629e-01
2.23600324e-02 4.62143263e-03 6.38377666e-01 -1.12605166e+00
1.45839739e+00 -2.02184987e+00 -2.35307619e-01 1.24063514e-01
9.74885166e-01 1.24983871e+00 2.36410812e-01 5.99315464e-02
-4.82074887e-01 4.52951550e-01 -7.30028450e-02 1.04114763e-01
-9.38330591e-01 1.90808922e-02 4.62172270e-01 3.63349080e-01
-4.24310751e-03 6.60616696e-01 -5.53590715e-01 -7.75116265e-01
4.68587399e-01 3.20071012e-01 -7.47527897e-01 6.32876039e-01
6.15777791e-01 7.64890373e-01 -2.70771205e-01 7.42711663e-01
5.80129921e-01 -3.82731646e-01 8.89709741e-02 1.34639693e-02
3.05591524e-01 -7.69207999e-02 -1.17306244e+00 1.54583848e+00
-1.98040783e-01 5.52857280e-01 -3.70989367e-02 -1.32201469e+00
9.47342813e-01 1.08319819e+00 9.01801229e-01 -1.86978981e-01
3.53531897e-01 3.19634557e-01 4.55687016e-01 -1.16061878e+00
-8.13109994e-01 2.66392324e-02 7.27372825e-01 2.92425424e-01
1.42663077e-01 -6.35244250e-02 7.39805475e-02 5.05392589e-02
1.13436949e+00 -1.81648090e-01 4.36574489e-01 1.48357544e-02
8.16018999e-01 -5.91343760e-01 9.66249764e-01 8.68278384e-01
-4.44407582e-01 1.19120705e+00 5.59936285e-01 -1.18796122e+00
-9.69263494e-01 -9.24543202e-01 -1.98243067e-01 5.73598454e-03
-3.21300775e-01 -1.44655138e-01 -6.35471642e-01 -9.42655265e-01
2.55411156e-02 -2.18833461e-01 -4.31524634e-01 4.19976711e-02
-7.25053191e-01 -4.20585275e-01 7.66708493e-01 8.13563049e-01
5.20618200e-01 -9.72358823e-01 -9.80272591e-01 2.65929073e-01
8.68315548e-02 -8.87973309e-01 -1.87276125e-01 -5.93051612e-01
-9.35444176e-01 -1.56968760e+00 -1.14610314e+00 -9.92949128e-01
4.92964923e-01 -7.27029145e-02 1.04058945e+00 5.12712657e-01
-7.43733227e-01 4.19538528e-01 -5.39917469e-01 -6.64551914e-01
-6.77713513e-01 -2.78767377e-01 2.68007159e-01 -1.98616400e-01
-3.48417298e-03 -7.73853421e-01 -9.42363143e-01 9.20379460e-02
-3.76294225e-01 1.84122890e-01 4.89762187e-01 1.06363451e+00
3.52336198e-01 -5.55059791e-01 6.06483698e-01 -7.50307858e-01
5.12882710e-01 -4.08266187e-01 -5.60290396e-01 1.94412902e-01
-8.08306932e-01 -6.85402751e-01 4.11208093e-01 -2.38285616e-01
-4.87478644e-01 -2.28409097e-01 -2.82378942e-01 -1.06507576e+00
-3.14827949e-01 3.99040997e-01 5.39857447e-01 -4.01887260e-02
4.30361360e-01 2.45362930e-02 4.23347145e-01 -2.28847548e-01
-3.93491775e-01 7.55549312e-01 2.62148052e-01 -5.10109544e-01
2.71385521e-01 4.98816036e-02 1.10702023e-01 -7.54920065e-01
-4.81529683e-01 -4.55548286e-01 -6.58685446e-01 -4.72298950e-01
9.53554928e-01 -1.02431428e+00 -7.88953662e-01 5.62702477e-01
-1.26613283e+00 3.16253901e-01 -3.40776145e-01 1.13570559e+00
-3.04814309e-01 5.13346732e-01 -6.30785286e-01 -7.67439365e-01
-9.29002166e-01 -1.11254644e+00 7.44932771e-01 -1.15268687e-02
1.09306060e-01 -8.69340181e-01 2.13444993e-01 4.05334651e-01
2.70608872e-01 5.18241942e-01 7.12735713e-01 -9.63205934e-01
-2.08160371e-01 -2.21076235e-01 -2.35662073e-01 9.06201720e-01
9.48148668e-02 3.38411689e-01 -7.99511433e-01 -3.65759134e-01
-8.76754746e-02 -2.20492393e-01 6.45853043e-01 9.17161942e-01
1.47588968e+00 1.08683906e-01 -1.40355155e-01 9.14130270e-01
1.10575914e+00 8.02475691e-01 2.57147133e-01 -3.30656111e-01
8.13255489e-01 2.05062360e-01 4.29275364e-01 7.27643609e-01
2.18145207e-01 1.07502073e-01 1.20301798e-01 -4.10725325e-01
-1.73061758e-01 -6.01867139e-02 -3.15416723e-01 1.27021861e+00
-6.78357780e-01 -1.41082302e-01 -1.49888062e+00 6.44594014e-01
-1.50486183e+00 -8.10125530e-01 -1.58877000e-01 2.17425776e+00
3.98859888e-01 -1.53727055e-01 1.85948014e-01 4.56012458e-01
8.38960469e-01 -1.39649674e-01 -2.52771765e-01 -3.70249659e-01
-3.17292772e-02 5.70899427e-01 -2.78904922e-02 -5.07114120e-02
-1.19082546e+00 3.82142663e-01 6.35505533e+00 4.16138202e-01
-1.35854900e+00 1.11903854e-01 8.45990658e-01 1.01786807e-01
2.04676256e-01 -3.12217206e-01 -2.38718197e-01 4.86712068e-01
6.74743831e-01 1.67200446e-01 -5.09289019e-02 5.23326218e-01
1.16662107e-01 1.58506855e-01 -9.58148241e-01 1.51291919e+00
5.97837344e-02 -1.81215727e+00 -3.22404094e-02 -2.96211392e-01
5.84226191e-01 -3.60649563e-02 -3.43457013e-01 1.38142854e-01
-3.68020773e-01 -1.10060799e+00 1.94862634e-02 3.36359918e-01
1.55137610e+00 -2.24815279e-01 1.03389239e+00 3.29139262e-01
-1.11472285e+00 -1.67801842e-01 -3.70670930e-02 7.90558904e-02
-1.71358690e-01 5.35450637e-01 -1.35747170e+00 4.36061084e-01
8.84886980e-01 8.52084041e-01 1.06730506e-01 1.21963763e+00
9.61241424e-02 9.88336980e-01 -2.21007705e-01 2.15378359e-01
-3.16328704e-01 -4.66140881e-02 7.39954770e-01 8.19760621e-01
6.00491047e-01 4.77503479e-01 1.33238509e-01 6.47453666e-01
-8.54273420e-03 5.00490606e-01 -9.07963395e-01 7.02652857e-02
6.04947746e-01 8.43125522e-01 -5.39185286e-01 -5.53537250e-01
-7.77506471e-01 2.18471810e-01 -9.46282595e-02 1.13099433e-01
-7.81888962e-01 -1.28579631e-01 -7.36538991e-02 4.40812111e-01
2.17118964e-01 3.54428351e-01 -2.11235642e-01 -1.37330759e+00
-1.87974982e-02 -1.05126238e+00 5.75731933e-01 -6.36999667e-01
-7.14783728e-01 2.72832245e-01 -2.29630873e-01 -1.67527056e+00
-1.61707655e-01 -5.15184462e-01 -9.36637223e-01 8.01731169e-01
-9.59937513e-01 -9.68601108e-01 -4.11832362e-01 2.63467938e-01
6.57033861e-01 -9.38379049e-01 9.22388256e-01 5.48038006e-01
-6.02376521e-01 5.35199523e-01 -4.36219126e-01 5.73585987e-01
4.47404891e-01 -1.26049161e+00 1.68618653e-02 8.67838502e-01
-2.13893637e-01 3.86276573e-01 -4.19593137e-03 -6.90995574e-01
-7.01189220e-01 -8.87569368e-01 5.61732590e-01 -1.73152998e-01
-7.30906129e-02 4.70099717e-01 -6.58454835e-01 3.49198222e-01
-1.27996370e-01 5.67934453e-01 8.42319071e-01 -5.43648779e-01
2.46195868e-01 -1.25945691e-04 -1.35235202e+00 -5.37322164e-02
1.01368535e+00 -5.50085828e-02 -5.79047441e-01 1.21365488e-01
5.57246149e-01 -9.83566642e-01 -1.32976270e+00 1.23550308e+00
8.15697372e-01 -1.28697360e+00 8.61526787e-01 -7.27580011e-01
6.71613097e-01 9.57159384e-04 2.52202213e-01 -8.66662085e-01
-5.99700306e-03 -4.98817444e-01 -3.07291567e-01 8.21893334e-01
3.50159019e-01 -7.56346345e-01 9.45495903e-01 1.15413852e-01
-2.06763163e-01 -1.52660418e+00 -7.40857363e-01 -1.83671474e-01
1.55569734e-02 -2.59322554e-01 6.20528340e-01 1.06115925e+00
-3.73909444e-01 -1.61669143e-02 -2.27729380e-01 2.56721228e-01
1.88779280e-01 -1.15990378e-01 5.05831540e-01 -1.49621105e+00
-3.55252415e-01 -2.02605307e-01 -6.86683118e-01 -4.97526795e-01
-3.25500697e-01 -6.93787754e-01 -7.46466696e-01 -1.56063581e+00
1.87108368e-02 -6.81065321e-01 -3.49886835e-01 1.14237726e-01
-3.12095731e-02 -1.75499469e-02 -1.92684203e-01 5.73891625e-02
-1.24312609e-01 5.64350225e-02 1.59217167e+00 1.18605800e-01
-1.30494356e-01 5.93395710e-01 -7.73365749e-03 8.62196743e-01
8.15501750e-01 -2.46896580e-01 -7.04930782e-01 -4.29454029e-01
-1.35840774e-01 1.02282012e+00 1.85538724e-01 -1.08563745e+00
-7.69958720e-02 2.91350365e-01 7.03356981e-01 -6.08093798e-01
-4.85775955e-02 -5.36882043e-01 -1.17270760e-01 9.98461366e-01
-1.03492022e-01 1.56297296e-01 -8.97641927e-02 9.24452469e-02
-3.75144243e-01 -2.98980385e-01 8.61739397e-01 -3.28222722e-01
-2.72674412e-01 1.01488376e+00 -2.66710252e-01 5.51134825e-01
1.17923152e+00 -4.34992135e-01 3.34438905e-02 -2.32527837e-01
-1.23327255e+00 1.39945103e-02 -7.88781121e-02 3.53662223e-01
1.17320144e+00 -9.27751303e-01 -1.03161466e+00 8.21858943e-01
-1.10039949e-01 3.94914865e-01 3.67964238e-01 1.28950632e+00
-1.35497749e+00 2.71769553e-01 -1.27071008e-01 -1.13040972e+00
-1.64402199e+00 1.50406212e-01 7.55055368e-01 1.76179763e-02
-9.03872728e-01 1.26006234e+00 -6.67846203e-02 -4.82428491e-01
3.84994894e-01 -4.13281530e-01 -5.63933849e-01 -3.08324337e-01
2.74768740e-01 4.81254131e-01 2.93095596e-02 -2.36064732e-01
-2.84442306e-01 8.47407699e-01 1.68019056e-01 3.57234508e-01
1.13351548e+00 2.15141386e-01 -1.82150558e-01 5.29943779e-02
8.97290170e-01 -1.99417286e-02 -4.94389415e-01 9.37561840e-02
-6.69754326e-01 -5.61545193e-01 -1.46061242e-01 -6.33757353e-01
-1.53629863e+00 1.35503864e+00 1.26238406e+00 2.30503216e-01
1.06818974e+00 -1.88239425e-01 6.55687690e-01 -1.86386369e-02
1.68211415e-01 -7.42075920e-01 -8.42760578e-02 -1.51814967e-02
7.88110554e-01 -1.40410936e+00 -2.11255133e-01 -4.37091768e-01
-4.71909732e-01 1.32890141e+00 1.04345810e+00 -1.00368984e-01
9.86584961e-01 3.08317423e-01 3.72434676e-01 -1.46667823e-01
-8.41076314e-01 3.22926670e-01 1.50722921e-01 7.64808834e-01
6.67831779e-01 1.34532899e-01 -5.11537015e-01 6.35540247e-01
2.05278650e-01 3.33527058e-01 5.01663685e-01 7.79512584e-01
-4.61088240e-01 -8.51753354e-01 -4.17767838e-02 1.05476356e+00
-8.77768934e-01 -5.51642664e-02 1.43348187e-01 6.41402364e-01
6.71902776e-01 5.08054137e-01 -2.06555694e-01 -3.32877576e-01
1.73626497e-01 -5.30725233e-02 4.24005955e-01 -8.27213526e-01
-4.41056401e-01 2.09637284e-02 2.11912408e-01 -2.98277020e-01
-5.06867588e-01 -5.09627342e-01 -1.24649751e+00 4.53903005e-02
-3.15814108e-01 6.05957024e-02 3.90642911e-01 7.52361774e-01
3.97458643e-01 7.40709364e-01 8.85673046e-01 -7.64243379e-02
-3.17990690e-01 -8.60018253e-01 -5.31640470e-01 1.15539990e-01
6.02366626e-01 -5.91307282e-01 -3.61453027e-01 1.89913183e-01] | [14.165300369262695, -2.3084464073181152] |
3c57fa0a-e3b6-4835-92df-de41bd1ed1ea | photofeeler-d3-a-neural-network-with-voter | 1904.07435 | null | https://arxiv.org/abs/1904.07435v3 | https://arxiv.org/pdf/1904.07435v3.pdf | Photofeeler-D3: A Neural Network with Voter Modeling for Dating Photo Impression Prediction | In just a few years, online dating has become the dominant way that young people meet to date, making the deceptively error-prone task of picking good dating profile photos vital to a generation's ability to form romantic connections. Until now, artificial intelligence approaches to Dating Photo Impression Prediction (DPIP) have been very inaccurate, unadaptable to real-world application, and have only taken into account a subject's physical attractiveness. To that effect, we propose Photofeeler-D3 - the first convolutional neural network as accurate as 10 human votes for how smart, trustworthy, and attractive the subject appears in highly variable dating photos. Our "attractive" output is also applicable to Facial Beauty Prediction (FBP), making Photofeeler-D3 state-of-the-art for both DPIP and FBP. We achieve this by leveraging Photofeeler's Dating Dataset (PDD) with over 1 million images and tens of millions of votes, our novel technique of voter modeling, and cutting-edge computer vision techniques. | ['Agastya Kalra', 'Ben Peterson'] | 2019-04-16 | null | null | null | null | ['facial-beauty-prediction'] | ['computer-vision'] | [-1.86898842e-01 3.86059970e-01 -2.93776989e-01 -1.01896226e+00
-3.50654647e-02 -3.94998699e-01 9.75408673e-01 -5.51908389e-02
-2.29962856e-01 5.68348527e-01 -1.41432390e-01 -1.33488886e-02
3.02473009e-02 -9.47687984e-01 -6.98878229e-01 6.26073852e-02
7.08641186e-02 7.96078146e-01 -5.57318926e-01 -3.07691664e-01
2.50001818e-01 4.62516248e-01 -1.69253635e+00 1.91998824e-01
5.63179910e-01 1.27303338e+00 -7.09446728e-01 5.07077456e-01
1.00513451e-01 3.32410693e-01 -2.25839660e-01 -1.66870785e+00
4.30613190e-01 1.39672086e-01 -5.13360500e-01 -5.25848687e-01
1.30145788e+00 -8.42215776e-01 -4.70871210e-01 9.10708070e-01
4.50349897e-01 -4.36973751e-01 7.20076740e-01 -1.54954135e+00
-1.36377895e+00 6.94663286e-01 -9.93927538e-01 -2.21816421e-01
3.26063395e-01 -5.23239560e-02 8.94664884e-01 -1.05708027e+00
8.72319937e-01 1.76714873e+00 7.95730054e-01 1.10548615e+00
-1.06834853e+00 -1.19243670e+00 -8.87077227e-02 2.65836835e-01
-1.24454641e+00 -7.56137729e-01 9.65301931e-01 -4.36524898e-01
5.96482866e-02 3.31315130e-01 9.71916556e-01 1.65190613e+00
-5.38009256e-02 9.55171645e-01 1.14035332e+00 -1.63637310e-01
8.31480995e-02 1.31297722e-01 4.18293756e-03 5.18509448e-01
3.61418217e-01 2.78174281e-01 -7.75942862e-01 -1.85246632e-01
6.94918692e-01 -1.87162742e-01 3.15820366e-01 -3.03624216e-02
-2.99173564e-01 8.43823552e-01 7.20492959e-01 4.01843479e-03
1.05780736e-01 3.60385388e-01 1.26385912e-01 1.74164191e-01
6.29194379e-01 3.68232876e-01 2.14147210e-01 -2.47212172e-01
-1.10242510e+00 7.19876945e-01 5.29642165e-01 6.37650609e-01
6.32442236e-01 -1.35175735e-01 4.61141914e-02 1.00228894e+00
3.37254077e-01 5.23571134e-01 8.62423629e-02 -1.40478992e+00
8.77536237e-02 6.10912442e-01 3.10180306e-01 -1.51064968e+00
-4.62817997e-02 -6.10943586e-02 -8.61541271e-01 6.27079070e-01
6.86499476e-01 1.99253082e-01 -8.98192883e-01 1.80230737e+00
3.25769663e-01 -1.28266737e-01 -7.49465287e-01 1.06738806e+00
1.08939946e+00 2.49816298e-01 1.45662948e-01 3.07283103e-01
1.21416962e+00 -4.04023528e-01 -1.48037747e-01 -5.16065955e-01
-1.90461919e-01 -5.29389322e-01 8.08687508e-01 5.85976303e-01
-1.47206795e+00 -4.80697423e-01 -1.23164213e+00 -5.83312869e-01
-3.61764371e-01 -5.02597615e-02 1.21727824e+00 1.19432843e+00
-1.19988656e+00 1.10845590e+00 -2.10617542e-01 -1.55364722e-01
1.26720631e+00 5.09747028e-01 -7.00730026e-01 -2.48563722e-01
-1.04143786e+00 1.09952736e+00 -5.05332828e-01 2.27198601e-01
-6.19703293e-01 -9.13276553e-01 -6.66432440e-01 7.19966963e-02
-3.25947374e-01 -5.89919150e-01 1.24065745e+00 -1.38663960e+00
-1.18476605e+00 1.49301565e+00 -2.12696970e-01 -2.76626021e-01
1.06181955e+00 -6.54972643e-02 -3.21278155e-01 -1.24555163e-01
1.66196987e-01 1.23963630e+00 1.00237083e+00 -1.40820444e+00
-1.66938648e-01 -1.10837269e+00 -1.16470590e-01 -6.15964420e-02
-5.69861889e-01 9.94316638e-02 -3.17018121e-01 -6.35429770e-02
-4.47892323e-02 -8.04092646e-01 1.34024844e-01 7.89035618e-01
-2.02124715e-01 -3.06890279e-01 7.28371143e-01 -8.08163285e-01
7.70939648e-01 -1.90402317e+00 -3.07619274e-01 3.09751302e-01
6.49425685e-01 4.49787170e-01 -1.03112832e-01 -9.66157988e-02
9.29856747e-02 4.62125957e-01 4.32421029e-01 -9.46924210e-01
3.41925204e-01 -2.33744700e-02 -7.23865852e-02 3.20853084e-01
-1.95836532e-03 1.09144366e+00 -9.37496662e-01 -6.69978797e-01
-3.34563144e-02 6.81759894e-01 -1.98079780e-01 -2.91964591e-01
1.75714746e-01 1.64846912e-01 -2.43135914e-01 1.02814412e+00
1.20776200e+00 1.27735004e-01 1.10741936e-01 3.57578844e-01
1.98027194e-01 -2.27685735e-01 -6.29722118e-01 1.13395679e+00
-3.88929367e-01 8.73993576e-01 1.63362533e-01 1.08761549e-01
1.30626297e+00 -2.83375651e-01 3.22349876e-01 -1.17669618e+00
3.47864360e-01 3.49952936e-01 -1.93560660e-01 4.22470868e-02
9.11377907e-01 -9.31679532e-02 1.02342352e-01 4.14536268e-01
-2.21831769e-01 4.79075909e-02 8.93652346e-03 3.37972045e-01
7.71928072e-01 6.64489865e-02 -1.70741454e-01 3.11379824e-02
-3.88764329e-02 -4.50857848e-01 6.90280735e-01 6.69500828e-01
-4.82210398e-01 1.02508903e+00 5.96593618e-01 -9.88617659e-01
-1.59219265e+00 -1.10717142e+00 -2.58698553e-01 8.26938391e-01
1.15485609e-01 -1.13486335e-01 -8.73138487e-01 -3.63718331e-01
6.51133239e-01 4.73406821e-01 -1.05395257e+00 -4.51460555e-02
-4.00463641e-01 -3.13962579e-01 7.57162273e-01 3.01714242e-01
4.15926516e-01 -7.87893832e-01 1.05623879e-01 -1.65655799e-02
1.35529995e-01 -7.25628138e-01 -1.50723562e-01 -8.51534069e-01
-4.25120324e-01 -8.86890411e-01 -9.61423278e-01 -2.34215915e-01
6.31344676e-01 3.48181948e-02 1.54642320e+00 4.80201960e-01
-2.68780828e-01 -1.32243410e-01 1.76174253e-01 -6.50862873e-01
-2.31227219e-01 -2.50680268e-01 2.75408298e-01 3.11982393e-01
7.45497882e-01 -1.09071410e+00 -1.03706563e+00 2.32095733e-01
-2.23775223e-01 3.06596719e-02 2.85311282e-01 4.03130442e-01
-5.48181385e-02 -7.31636226e-01 5.25841057e-01 -9.70491469e-01
6.68259740e-01 -3.99585605e-01 -5.02902806e-01 1.30458921e-01
-8.67039502e-01 -5.29369712e-01 4.88051504e-01 -4.30536836e-01
-1.08730328e+00 -1.44124264e-02 -6.31430298e-02 -6.77085578e-01
1.78848058e-01 -2.17973769e-01 7.33023956e-02 -3.86713237e-01
8.65048349e-01 -2.72400707e-01 -4.65043373e-02 -3.93121034e-01
3.55057240e-01 8.44463766e-01 8.92639577e-01 -6.59654617e-01
8.04528415e-01 6.55697942e-01 2.38396645e-01 -5.36616206e-01
-7.26357222e-01 -7.81920776e-02 -6.15665793e-01 -8.72109354e-01
4.89924103e-01 -8.42939258e-01 -1.55768681e+00 7.41174579e-01
-1.15428960e+00 2.40538940e-01 3.06887686e-01 -2.16384590e-01
-1.21826410e-01 4.13201034e-01 -6.96184337e-01 -1.19772196e+00
-5.59511483e-01 -5.94244838e-01 1.06416416e+00 7.09649503e-01
-5.25109053e-01 -4.15829003e-01 5.84212840e-02 1.02744687e+00
3.59892070e-01 4.06858832e-01 4.98665065e-01 1.82525273e-02
-4.30165291e-01 -5.85945427e-01 -7.79454052e-01 3.95466350e-02
-5.31667650e-01 4.69573975e-01 -1.58973885e+00 7.66099542e-02
-6.97796047e-01 -6.41650319e-01 8.76958907e-01 2.99750626e-01
1.33637547e+00 -1.93629712e-01 -3.39383125e-01 5.72213411e-01
1.15390646e+00 -2.35478759e-01 9.86393154e-01 8.02991912e-02
5.55037260e-01 5.58995128e-01 4.57031399e-01 5.68069100e-01
7.64637649e-01 7.11359739e-01 7.86657512e-01 7.95519650e-02
2.81084236e-03 -8.35865676e-01 -8.78234878e-02 -7.80864358e-02
-6.30449057e-01 -4.24868660e-03 -7.60308385e-01 5.66893816e-01
-1.81343925e+00 -1.17519259e+00 -2.20950231e-01 2.25846148e+00
7.89808154e-01 9.58832502e-02 1.30099252e-01 -1.50814384e-01
8.81849885e-01 1.62085563e-01 -6.28032327e-01 -1.08119226e+00
-3.11012734e-02 2.55900979e-01 3.58664930e-01 2.63071179e-01
-8.98924410e-01 7.41648972e-01 5.95352983e+00 7.23808348e-01
-8.98205221e-01 -3.68857086e-02 1.58409381e+00 -3.40061754e-01
-6.05082691e-01 -1.31734341e-01 -8.89516950e-01 8.59595895e-01
7.45084167e-01 -2.22416576e-02 5.97003818e-01 1.24820757e+00
1.22681521e-01 -1.61252692e-01 -1.43070638e+00 1.28897226e+00
5.59870899e-02 -1.40370393e+00 -2.27874234e-01 2.70850539e-01
6.97224379e-01 -4.07553166e-01 5.93735933e-01 1.95452467e-01
5.65512598e-01 -1.69822383e+00 8.00618470e-01 5.83096802e-01
9.20854151e-01 -8.00338507e-01 3.18973929e-01 1.07353888e-01
-6.07313871e-01 -2.19397664e-01 -6.53676569e-01 -3.63353640e-01
-4.16335091e-02 6.65205419e-01 -9.12641883e-01 -4.67978343e-02
1.02859449e+00 5.17650008e-01 -6.81082129e-01 7.75084972e-01
-1.45069808e-01 9.55223888e-02 -4.56289560e-01 -1.28526986e-01
3.61533202e-02 -1.47010282e-01 7.57307187e-02 6.15605175e-01
3.63585949e-01 -3.96403484e-02 -6.35023117e-01 1.14279246e+00
-8.70044470e-01 -1.63852394e-01 -5.62776029e-01 -2.48138048e-02
7.85814822e-01 1.66552866e+00 -2.25734711e-01 4.82561029e-02
-3.24974731e-02 9.23755646e-01 6.21232212e-01 -1.81952685e-01
-5.21875799e-01 2.54181802e-01 9.50947523e-01 4.28342909e-01
-3.24838996e-01 1.13740094e-01 -9.58892047e-01 -9.25681055e-01
3.36615831e-01 -4.81219918e-01 6.99768886e-02 -1.17285883e+00
-1.63492954e+00 1.31237268e-01 -6.82607234e-01 -8.81146371e-01
-1.19372904e-01 -5.08017838e-01 -8.43515515e-01 8.19420397e-01
-1.49565649e+00 -1.87765932e+00 -4.71023232e-01 7.84790665e-02
5.42180762e-02 -2.56196856e-02 6.20081663e-01 5.37888050e-01
-3.17902386e-01 1.04008722e+00 9.33483895e-03 2.40524948e-01
9.61996675e-01 -1.17975771e+00 8.83027911e-01 3.38593155e-01
-4.70802709e-02 5.42739213e-01 4.68887776e-01 -6.19056880e-01
-1.21908498e+00 -5.95020354e-01 1.41429448e+00 -7.73377001e-01
2.94572741e-01 -7.76242137e-01 -2.70198822e-01 6.46560431e-01
-1.72426566e-01 -2.75713764e-02 3.32631737e-01 5.58044553e-01
-7.13465989e-01 -4.10933197e-01 -1.52217531e+00 8.96995723e-01
1.31303346e+00 -2.74376690e-01 3.31798918e-03 1.22031346e-01
-1.25136942e-01 -3.33341420e-01 -8.31841171e-01 3.30114923e-02
1.26927257e+00 -1.25159669e+00 1.28902209e+00 -5.37901700e-01
1.40227997e+00 4.26450610e-01 4.38478500e-01 -9.07869399e-01
-4.01426077e-01 -6.24340594e-01 6.45172447e-02 1.48766959e+00
3.88206452e-01 2.03682762e-02 1.52592480e+00 1.55732083e+00
2.15228006e-01 -7.73384035e-01 -1.13084948e+00 -3.69063437e-01
3.68926167e-01 -2.23685384e-01 8.75462949e-01 1.09750319e+00
-2.36135982e-02 -7.13862255e-02 -1.02769566e+00 -2.54305631e-01
9.84950662e-01 9.67327952e-02 1.00428867e+00 -1.78919041e+00
6.17236607e-02 -6.40528560e-01 -5.50398648e-01 -5.78829110e-01
1.53264642e-01 -5.66363990e-01 -4.10872608e-01 -1.04287732e+00
6.10798061e-01 -7.65680254e-01 7.60606378e-02 4.30761397e-01
1.08217806e-01 9.24837708e-01 3.65390599e-01 1.92137361e-02
-2.67933100e-01 2.06756920e-01 1.41838479e+00 -4.62191820e-01
3.59160304e-01 1.76858500e-01 -1.01671505e+00 7.50526130e-01
3.92540097e-01 -1.04061842e-01 -5.54342493e-02 -6.50348842e-01
8.61693621e-01 -3.48071344e-02 5.87000191e-01 -6.09684467e-01
1.60164684e-01 -9.10752863e-02 1.29368579e+00 -4.62163925e-01
1.17661536e+00 -5.80755711e-01 1.57096475e-01 -2.23541111e-02
-2.36820728e-01 -2.75710911e-01 -3.09839994e-01 2.92857170e-01
2.29148760e-01 -1.84046477e-01 9.85248208e-01 -3.52450103e-01
-6.82457149e-01 6.90226614e-01 3.14719141e-01 -2.95019418e-01
6.82256162e-01 -5.27264893e-01 -6.90036654e-01 -8.64457905e-01
-5.35714328e-01 4.05604094e-01 9.87765074e-01 6.10794961e-01
7.18821347e-01 -1.59917963e+00 -8.88506532e-01 -1.89474262e-02
6.54748678e-02 -3.86163801e-01 4.25826758e-01 5.15840873e-02
-4.97619361e-01 -1.03116617e-01 -7.73594797e-01 -2.99634039e-01
-1.56257999e+00 1.57466248e-01 1.46290228e-01 2.01380968e-01
-1.53593391e-01 1.21570456e+00 -1.84362024e-01 -4.79196846e-01
1.97120644e-02 6.05003715e-01 -1.70778915e-01 1.21950954e-01
5.83934009e-01 4.89595324e-01 -1.13197327e-01 -7.43626833e-01
-2.95245856e-01 3.41979623e-01 -3.06860030e-01 -2.98758987e-02
1.46308684e+00 1.03114080e-02 -7.29675740e-02 -5.25077507e-02
9.45141017e-01 -1.59656256e-01 -1.32527518e+00 1.91477880e-01
-3.95091861e-01 -1.29940486e+00 -3.05114686e-02 -1.04098213e+00
-1.16474617e+00 1.00334752e+00 4.73402739e-01 -7.74450600e-02
5.45218885e-01 -2.73389876e-01 1.06242049e+00 8.72290954e-02
5.90006232e-01 -1.44018984e+00 -9.19833258e-02 7.02027325e-03
7.45525777e-01 -1.76828969e+00 3.45051378e-01 -3.38175833e-01
-5.52303851e-01 1.02025437e+00 9.37306225e-01 5.51629364e-02
4.06179219e-01 -1.77133858e-01 -7.75699969e-03 -7.87702203e-02
-4.62281466e-01 5.68587899e-01 2.46047661e-01 7.20458865e-01
2.34794915e-01 3.55796754e-01 -2.83246756e-01 7.13036120e-01
-6.23012960e-01 2.69527912e-01 5.90702295e-01 3.79070252e-01
-1.49072722e-01 -1.34585893e+00 -2.42639825e-01 5.09653449e-01
-5.19320071e-01 5.71853258e-02 -1.22848856e+00 7.00417161e-01
1.69544488e-01 7.80511558e-01 2.36873746e-01 -5.33142149e-01
-1.63459122e-01 -8.52982849e-02 6.15240395e-01 -7.95133412e-02
-7.13586092e-01 -6.79512680e-01 4.17198300e-01 -6.87015057e-01
7.07234964e-02 -6.66060627e-01 -5.33758581e-01 -1.41494656e+00
8.49953815e-02 -3.69014889e-01 1.12770319e+00 5.89058697e-01
5.00264823e-01 -7.31779516e-01 8.47582579e-01 -1.21164799e+00
-3.44524920e-01 -7.50451207e-01 -6.81109846e-01 6.95785284e-01
-1.18392721e-01 -5.87846518e-01 6.01221062e-02 -4.16714311e-01] | [13.28288745880127, 1.044864296913147] |
edfbe0b1-40e2-4c3d-b654-cef50341c6b9 | multimodal-fusion-transformer-for-remote | 2203.16952 | null | https://arxiv.org/abs/2203.16952v2 | https://arxiv.org/pdf/2203.16952v2.pdf | Multimodal Fusion Transformer for Remote Sensing Image Classification | Vision transformers (ViTs) have been trending in image classification tasks due to their promising performance when compared to convolutional neural networks (CNNs). As a result, many researchers have tried to incorporate ViTs in hyperspectral image (HSI) classification tasks. To achieve satisfactory performance, close to that of CNNs, transformers need fewer parameters. ViTs and other similar transformers use an external classification (CLS) token which is randomly initialized and often fails to generalize well, whereas other sources of multimodal datasets, such as light detection and ranging (LiDAR) offer the potential to improve these models by means of a CLS. In this paper, we introduce a new multimodal fusion transformer (MFT) network which comprises a multihead cross patch attention (mCrossPA) for HSI land-cover classification. Our mCrossPA utilizes other sources of complementary information in addition to the HSI in the transformer encoder to achieve better generalization. The concept of tokenization is used to generate CLS and HSI patch tokens, helping to learn a {distinctive representation} in a reduced and hierarchical feature space. Extensive experiments are carried out on {widely used benchmark} datasets {i.e.,} the University of Houston, Trento, University of Southern Mississippi Gulfpark (MUUFL), and Augsburg. We compare the results of the proposed MFT model with other state-of-the-art transformers, classical CNNs, and conventional classifiers models. The superior performance achieved by the proposed model is due to the use of multihead cross patch attention. The source code will be made available publicly at \url{https://github.com/AnkurDeria/MFT}.} | ['Jocelyn Chanussot', 'Antonio Plaza', 'Behnood Rasti', 'Danfeng Hong', 'Ankur Deria', 'Swalpa Kumar Roy'] | 2022-03-31 | null | null | null | null | ['remote-sensing-image-classification'] | ['miscellaneous'] | [ 4.57859457e-01 -3.71437937e-01 4.83816862e-02 -4.36082602e-01
-8.75953436e-01 -3.98356318e-01 5.90796411e-01 -9.85675752e-02
-3.12838167e-01 6.91160858e-01 -4.77274545e-02 -5.25695741e-01
-1.50586039e-01 -9.31872368e-01 -8.29657197e-01 -9.72955525e-01
3.28853458e-01 -7.33478963e-02 -2.97898669e-02 -3.54968488e-01
5.44739477e-02 5.48811078e-01 -1.73004043e+00 3.07948053e-01
1.22764444e+00 1.47054398e+00 4.38772142e-01 2.42649525e-01
1.04977973e-02 6.88033998e-01 -2.74827540e-01 -3.97999138e-01
3.59613061e-01 -1.21894926e-01 -7.67449558e-01 -1.00132637e-01
6.54251337e-01 -2.05623269e-01 -4.56401110e-01 1.28014219e+00
5.22640884e-01 2.06546172e-01 6.80917859e-01 -1.52648306e+00
-8.34144533e-01 6.81995630e-01 -7.80835569e-01 1.10993132e-01
-7.12669566e-02 2.24241689e-01 1.13352394e+00 -8.83123994e-01
1.98482741e-02 1.12570095e+00 8.78471255e-01 2.13521734e-01
-9.71683562e-01 -9.93687987e-01 2.67513897e-02 5.95966518e-01
-1.51578188e+00 -2.10606858e-01 7.32805669e-01 -2.24120811e-01
9.50092077e-01 1.74391910e-01 5.36998570e-01 1.08536553e+00
1.51553884e-01 9.54324424e-01 1.25556087e+00 -4.11497712e-01
-2.04459149e-02 1.01850353e-01 1.13656849e-01 8.10647666e-01
1.98390305e-01 8.42355490e-02 -4.26857680e-01 1.89146191e-01
5.07976532e-01 2.44377866e-01 -4.54102516e-01 -5.05804718e-02
-8.54239166e-01 8.44475389e-01 1.19678366e+00 3.72912765e-01
-4.74804103e-01 2.01720089e-01 1.83108106e-01 9.91007984e-02
4.91185516e-01 7.65183344e-02 -2.57656008e-01 3.66380692e-01
-7.94648170e-01 -1.05270989e-01 2.22474068e-01 9.15053129e-01
9.40001726e-01 3.55258942e-01 -1.87268108e-01 9.96363223e-01
2.94669718e-01 7.16455638e-01 4.62835968e-01 -4.82098639e-01
5.80094278e-01 7.78882205e-01 -2.64433593e-01 -6.38817072e-01
-2.86286175e-01 -6.68655097e-01 -1.12971866e+00 1.47823393e-01
-1.78303160e-02 -8.10375845e-04 -1.26526296e+00 1.66046953e+00
-5.67298234e-02 3.22693139e-01 2.16331899e-01 7.78838038e-01
1.08618045e+00 7.28639364e-01 1.46907240e-01 3.42938900e-01
1.31853890e+00 -8.69994581e-01 -3.35594565e-01 -2.04323456e-01
2.74054170e-01 -5.24303436e-01 1.08429253e+00 3.58536869e-01
-6.32943213e-01 -7.98545718e-01 -1.11173475e+00 -7.00189695e-02
-7.81147659e-01 4.14391786e-01 4.62259531e-01 6.93334222e-01
-1.13139164e+00 4.38895643e-01 -7.23914087e-01 -4.93836850e-01
8.73026729e-01 3.86474311e-01 -4.12716657e-01 -3.74419987e-01
-1.19091797e+00 8.09539855e-01 4.29617316e-01 4.63408887e-01
-1.02244031e+00 -6.58641458e-01 -1.06030345e+00 2.61313230e-01
1.54758289e-01 -5.63551426e-01 1.01611769e+00 -9.96494174e-01
-1.05608046e+00 4.31761116e-01 -7.61349574e-02 -3.87030602e-01
3.52290794e-02 -1.01615578e-01 -1.08780213e-01 1.62009373e-01
2.99908191e-01 9.82369483e-01 7.60683239e-01 -1.24365783e+00
-7.76777923e-01 -4.67496276e-01 1.39294684e-01 2.25968987e-01
-7.12603748e-01 -3.19509149e-01 9.14127976e-02 -5.81832767e-01
1.38079841e-02 -7.92161465e-01 2.42064632e-02 -2.91693449e-01
-5.81830740e-01 -4.90281790e-01 1.04686034e+00 -5.59569180e-01
8.25169265e-01 -2.19983530e+00 9.20912847e-02 4.96734045e-02
-8.48484859e-02 5.46202004e-01 -2.80062944e-01 4.75278050e-01
-2.27962926e-01 2.15524197e-01 -5.77379405e-01 -3.21637064e-01
1.62844300e-01 2.24356964e-01 -1.77472100e-01 4.07534957e-01
4.09379184e-01 9.50384915e-01 -7.29785442e-01 -1.11672424e-01
4.91262645e-01 6.85518265e-01 -8.26202780e-02 -5.62335700e-02
-8.10277537e-02 3.90491754e-01 -3.33179623e-01 1.05980384e+00
8.54858220e-01 -6.09473176e-02 -2.75306642e-01 -5.98484516e-01
-3.26393068e-01 -4.54006158e-02 -8.37712884e-01 1.53107619e+00
-3.27349186e-01 5.59905887e-01 -2.19990462e-01 -1.30758786e+00
8.85158241e-01 3.98058236e-01 2.42541134e-01 -6.89018071e-01
3.73822778e-01 3.06519210e-01 -1.84780702e-01 -4.10281837e-01
3.58359963e-01 -2.91134983e-01 2.31763303e-01 -3.44807915e-02
2.93857574e-01 -9.13048759e-02 1.14085324e-01 -1.52658626e-01
7.52218187e-01 2.29554191e-01 -3.22032860e-03 -2.27698907e-01
6.18353605e-01 3.60999256e-02 4.65661913e-01 6.17131650e-01
-1.40489742e-01 4.43893254e-01 1.11209355e-01 -2.45930105e-01
-7.66723096e-01 -8.12025785e-01 -3.35291624e-01 8.41877162e-01
1.55066848e-01 -3.64912078e-02 -4.97248620e-01 -5.14444232e-01
-5.72959073e-02 6.96467876e-01 -5.74452102e-01 -3.38053674e-01
-5.81164621e-02 -9.68648970e-01 9.08570230e-01 8.22191060e-01
1.37332106e+00 -1.02487528e+00 -5.65174758e-01 -7.58297145e-02
-4.83350456e-01 -1.08735251e+00 3.70522439e-02 4.65113968e-01
-6.72297657e-01 -8.94987166e-01 -7.66263664e-01 -6.99381351e-01
2.65460461e-01 5.22679687e-01 7.10986853e-01 -2.59361058e-01
-3.72960687e-01 3.06022376e-01 -5.64309597e-01 -6.34247124e-01
2.32316777e-01 2.04731435e-01 -2.64745533e-01 2.73525953e-01
3.87745351e-01 -5.27842760e-01 -5.52530587e-01 1.38180807e-01
-1.31410027e+00 -3.53591181e-02 7.92149365e-01 1.14051032e+00
3.56144518e-01 3.80450696e-01 4.55648273e-01 -5.55321574e-01
2.46399209e-01 -5.96123457e-01 -4.43230271e-01 3.02421719e-01
-3.85616153e-01 -1.77630603e-01 5.68754315e-01 -1.06017143e-01
-1.26560307e+00 7.98987374e-02 -2.30193824e-01 -7.61227310e-01
-4.03319508e-01 8.39713931e-01 -2.35813543e-01 -3.36113423e-01
5.12624264e-01 3.76721084e-01 -1.37532532e-01 -4.12134618e-01
1.55896753e-01 8.35245967e-01 4.21363682e-01 -4.95875806e-01
8.93980443e-01 2.58762956e-01 1.55239567e-01 -9.49735165e-01
-8.68575811e-01 -4.67899472e-01 -5.98774672e-01 -1.88261926e-01
7.74495661e-01 -1.11416209e+00 -6.44011021e-01 8.66774499e-01
-7.80260444e-01 -2.81623453e-01 -7.47360736e-02 4.47731912e-01
-2.04090834e-01 1.98257491e-01 -4.55249697e-01 -9.91131425e-01
-3.85577500e-01 -1.32871878e+00 1.20751202e+00 6.13253474e-01
5.24916410e-01 -8.89085948e-01 -5.08551717e-01 6.58791780e-01
5.55984020e-01 4.35760677e-01 1.02857709e+00 -3.87991965e-01
-6.45045400e-01 -2.61611436e-02 -5.74576259e-01 8.31526816e-01
2.67392427e-01 -9.25619677e-02 -1.45536029e+00 -4.23108548e-01
-2.03369096e-01 -4.64800000e-01 1.38498390e+00 4.36234921e-01
1.16376889e+00 -6.38788426e-03 -2.60711193e-01 7.21979558e-01
1.83749294e+00 3.61543268e-01 8.14152002e-01 3.72747421e-01
7.91990161e-01 2.81849593e-01 4.97787952e-01 3.82334501e-01
5.52925229e-01 2.97199339e-01 9.06863689e-01 -3.99914891e-01
-7.17216507e-02 6.28651399e-03 4.17794526e-01 2.05491081e-01
-3.82333815e-01 -4.09264505e-01 -9.25075889e-01 5.60274243e-01
-1.72739971e+00 -9.33644056e-01 -1.17826752e-01 2.01196265e+00
4.07463938e-01 -2.95883626e-01 -1.60653010e-01 4.74623621e-01
5.78180611e-01 3.58490422e-02 -4.90880251e-01 2.14364734e-02
-3.39388371e-01 6.66667640e-01 6.46039367e-01 2.07981795e-01
-1.33871841e+00 9.77049887e-01 4.40360546e+00 9.23195779e-01
-1.42655981e+00 5.51688485e-02 5.00299871e-01 3.10234308e-01
-1.39926791e-01 -2.41668686e-01 -6.77031398e-01 1.63893625e-01
6.64260566e-01 2.46093184e-01 4.51703340e-01 4.47388142e-01
-3.14616808e-03 -1.74998283e-01 -7.69740462e-01 8.25470746e-01
1.13888897e-01 -9.63254035e-01 3.29386234e-01 -2.62544304e-02
4.89381462e-01 3.30227733e-01 4.77433085e-01 4.35325235e-01
1.49736762e-01 -1.12805700e+00 7.62164295e-01 4.07576978e-01
9.05292988e-01 -7.03070343e-01 1.16235387e+00 1.60954684e-01
-1.62524664e+00 -5.69017351e-01 -5.51530659e-01 2.28940666e-01
-3.41092765e-01 2.86554217e-01 -5.89579105e-01 1.24599564e+00
1.07999718e+00 9.58865285e-01 -8.23272765e-01 1.06368339e+00
-2.09094435e-01 7.48575807e-01 -2.85971105e-01 1.39036790e-01
6.88554823e-01 -1.71811610e-01 3.15352827e-01 1.11760247e+00
5.29130936e-01 2.64867187e-01 1.79324582e-01 7.94130802e-01
-1.40817001e-01 -6.03025258e-02 -7.21327841e-01 -4.31167670e-02
2.60007590e-01 1.44372404e+00 -4.70963240e-01 -2.30835930e-01
-4.69261140e-01 6.76224172e-01 1.12468950e-01 4.79906946e-01
-1.03562856e+00 -6.41125262e-01 4.65977132e-01 -1.03686847e-01
4.25541162e-01 1.12539329e-01 -6.69030622e-02 -1.10803735e+00
-3.22329737e-02 -7.24191904e-01 4.63538349e-01 -1.15564430e+00
-1.50232422e+00 9.01235700e-01 1.76619500e-01 -1.22004938e+00
3.83764952e-01 -8.31171155e-01 -5.58936834e-01 1.03529990e+00
-2.18054271e+00 -1.85729611e+00 -7.98583269e-01 9.83277977e-01
4.91612852e-01 -2.41546109e-01 6.25702679e-01 5.76512158e-01
-7.40770698e-01 6.82533205e-01 -9.07594860e-02 2.45411187e-01
4.49989617e-01 -1.21408165e+00 -1.62492126e-01 8.48937988e-01
-5.78311197e-02 3.69639605e-01 2.14993894e-01 -3.12255055e-01
-1.29786253e+00 -1.52035725e+00 5.60223222e-01 4.54991460e-02
4.67398167e-01 -1.12883210e-01 -8.58008683e-01 8.87388229e-01
6.01660311e-01 -5.02321869e-02 6.09353423e-01 -2.47062087e-01
-6.01019740e-01 -5.79235613e-01 -1.23569679e+00 2.32240528e-01
8.34976852e-01 -4.98039126e-01 -2.60347575e-01 2.57036597e-01
4.58177298e-01 -1.21018797e-01 -6.75833464e-01 7.28192270e-01
3.73817652e-01 -1.04532170e+00 9.34682548e-01 -1.85914978e-01
4.42770869e-01 -3.65234226e-01 -5.74426711e-01 -1.41352463e+00
-4.49769646e-01 -5.98942116e-02 5.60259819e-01 1.30465114e+00
3.50138813e-01 -1.01964033e+00 3.38936388e-01 1.20250866e-01
-6.57368362e-01 -5.94001710e-01 -9.70775843e-01 -7.47483850e-01
1.00599647e-01 -3.69580746e-01 6.54497683e-01 8.97502959e-01
-2.64660746e-01 3.93575191e-01 -2.77507186e-01 5.42201221e-01
7.91192055e-01 4.12998684e-02 3.14919323e-01 -1.34586179e+00
1.01570874e-01 -5.01236022e-01 -4.75016117e-01 -6.54253840e-01
1.90146431e-01 -1.16564357e+00 4.20479141e-02 -1.73017788e+00
2.23675117e-01 -5.26709497e-01 -5.84461987e-01 1.14186800e+00
-3.21420878e-02 6.45186961e-01 3.35782140e-01 -8.20551366e-02
-1.21033989e-01 9.02507484e-01 1.09456503e+00 -5.63182414e-01
1.96336687e-01 -1.45099670e-01 -7.88042605e-01 3.48706722e-01
9.83732224e-01 -2.60933071e-01 -4.26357746e-01 -6.30864859e-01
-2.35716492e-01 -3.54332030e-02 8.08950007e-01 -1.24860609e+00
3.00586760e-01 8.10338855e-02 6.37348294e-01 -6.92045271e-01
6.51404679e-01 -8.45392406e-01 1.89103201e-01 3.85182679e-01
7.34783262e-02 -3.48901376e-02 4.75993037e-01 3.54847729e-01
-4.70711380e-01 -2.10967660e-01 9.58635390e-01 -2.58044064e-01
-7.57730186e-01 4.99915034e-01 -3.65747660e-01 -5.65108180e-01
8.76015902e-01 -3.13220769e-01 -6.02667868e-01 -1.81180552e-01
-4.66436118e-01 2.90648937e-01 -1.25642410e-02 3.52255523e-01
7.27798700e-01 -1.20963681e+00 -7.80427098e-01 2.55258888e-01
2.74015546e-01 7.81502873e-02 3.03702176e-01 9.96833563e-01
-2.44476691e-01 4.18427408e-01 -4.12912339e-01 -7.90230095e-01
-1.05896103e+00 1.71571076e-01 5.65581560e-01 3.33019346e-02
-4.37844485e-01 9.60752785e-01 2.40983069e-01 -5.18167377e-01
2.10199162e-01 -4.48686838e-01 -3.63174230e-01 1.90204173e-01
2.18688071e-01 2.20322341e-01 2.72584200e-01 -6.50703251e-01
-3.25208008e-01 6.19768858e-01 1.75320446e-01 1.27485573e-01
1.54736221e+00 4.94077168e-02 -2.24536732e-01 2.42272452e-01
9.69406784e-01 -6.73582613e-01 -9.08896685e-01 -4.15632695e-01
-2.81530321e-01 -3.62770528e-01 2.74840206e-01 -9.74275947e-01
-1.39316845e+00 1.22234368e+00 8.84735644e-01 1.15768775e-01
1.58305371e+00 -2.04256743e-01 7.28509009e-01 3.47410798e-01
3.76360089e-01 -6.32686019e-01 -4.71252128e-02 5.82853973e-01
7.96038508e-01 -1.31416857e+00 -4.47883844e-01 -2.86595345e-01
-6.87132120e-01 1.15030861e+00 6.50840044e-01 7.33062997e-02
6.51860774e-01 1.34342015e-02 -4.03662361e-02 -1.66872993e-01
-5.00270605e-01 -5.72796345e-01 2.44867772e-01 5.75687945e-01
3.53266090e-01 7.33462945e-02 2.25913212e-01 4.66367602e-01
-4.20113327e-03 -1.03290076e-03 4.07040983e-01 1.02935004e+00
-4.75653052e-01 -9.52200353e-01 -4.81338412e-01 5.05501270e-01
-2.34179050e-01 -3.78662467e-01 -2.38633066e-01 6.82621300e-01
5.27364075e-01 1.16140330e+00 -1.74913853e-01 -5.11866689e-01
4.22604263e-01 1.69947952e-01 4.15116459e-01 -4.56872463e-01
-7.45670557e-01 6.55486342e-03 -1.92003086e-01 -2.26468667e-01
-6.54638052e-01 -6.71656728e-01 -9.76998746e-01 -2.84885347e-01
-5.07105649e-01 1.70010865e-01 5.80054581e-01 8.43204021e-01
9.24574658e-02 6.46856308e-01 6.88175142e-01 -1.02772701e+00
-4.13408965e-01 -1.22036052e+00 -7.07500696e-01 8.53484944e-02
4.18141931e-01 -8.84488046e-01 -9.46329236e-02 -8.29989091e-03] | [9.888689994812012, -1.4925333261489868] |
91ea774c-fa1a-45eb-89af-f6e9b8c90946 | time-series-clustering-with-random | 2305.10457 | null | https://arxiv.org/abs/2305.10457v2 | https://arxiv.org/pdf/2305.10457v2.pdf | Time Series Clustering With Random Convolutional Kernels | Time series data, spanning applications ranging from climatology to finance to healthcare, presents significant challenges in data mining due to its size and complexity. One open issue lies in time series clustering, which is crucial for processing large volumes of unlabeled time series data and unlocking valuable insights. Traditional and modern analysis methods, however, often struggle with these complexities. To address these limitations, we introduce R-Clustering, a novel method that utilizes convolutional architectures with randomly selected parameters. Through extensive evaluations, R-Clustering demonstrates superior performance over existing methods in terms of clustering accuracy, computational efficiency and scalability. Empirical results obtained using the UCR archive demonstrate the effectiveness of our approach across diverse time series datasets. The findings highlight the significance of R-Clustering in various domains and applications, contributing to the advancement of time series data mining. | ['Rubén Cuevas', 'Jorge Marco-Blanco'] | 2023-05-17 | null | null | null | null | ['time-series-clustering'] | ['time-series'] | [-3.93373251e-01 -7.97363460e-01 -5.38195707e-02 -2.94770092e-01
-4.39850777e-01 -8.05889606e-01 2.92039216e-01 3.31452012e-01
-3.39682400e-01 2.98065037e-01 1.09842628e-01 -5.08962214e-01
-5.84021389e-01 -6.25421286e-01 -1.77941188e-01 -7.26504624e-01
-1.07303584e+00 2.13887289e-01 -4.32839185e-01 -3.32051784e-01
6.16023131e-02 5.90626895e-01 -1.63630903e+00 -5.79860285e-02
8.16224277e-01 1.04694462e+00 -2.09915906e-01 3.19997936e-01
-1.15539536e-01 6.34805143e-01 -6.30285680e-01 2.70973951e-01
3.91821593e-01 -5.80421612e-02 -3.11945468e-01 -1.36806533e-01
-4.39535469e-01 -9.86903906e-02 -1.42358407e-01 6.32964909e-01
3.45850438e-01 5.34558296e-01 3.26208234e-01 -1.46246970e+00
-5.82346439e-01 6.61244452e-01 -9.25649107e-01 7.04740584e-01
-2.50870973e-01 -1.29518852e-01 9.60771322e-01 -4.63888824e-01
3.00911874e-01 7.05152035e-01 1.08808827e+00 2.23144084e-01
-1.16505325e+00 -8.75066519e-01 1.81107610e-01 3.70173663e-01
-1.51627207e+00 -2.47196943e-01 9.30548668e-01 -4.47938979e-01
1.11745095e+00 2.68548876e-01 6.43228889e-01 5.84291816e-01
-1.76392063e-01 5.14651000e-01 9.32895780e-01 -1.08884081e-01
6.29167676e-01 -4.69857752e-01 3.78832310e-01 -8.62243846e-02
4.94894050e-02 -8.66112113e-02 -2.17075720e-01 -1.77946717e-01
3.25146198e-01 4.56516832e-01 9.78965592e-03 2.08830133e-01
-1.21649003e+00 7.47832537e-01 1.91137776e-01 5.31979799e-01
-5.58817327e-01 5.31609263e-03 7.99001753e-01 5.38606524e-01
9.03593302e-01 4.76646900e-01 -8.16192806e-01 -4.43953365e-01
-1.08470047e+00 3.98149937e-02 5.39936721e-01 5.72964668e-01
4.10229951e-01 2.38495290e-01 4.47683752e-01 6.21189117e-01
-1.36808857e-01 5.45826852e-01 6.61454558e-01 -7.23750532e-01
3.05398256e-01 9.13410664e-01 -1.03232674e-01 -1.10722566e+00
-1.02083445e+00 -5.65982759e-01 -1.15408480e+00 -1.55095518e-01
1.32081494e-01 -3.56040061e-01 -7.63176143e-01 1.88282371e+00
5.09170175e-01 4.30424392e-01 8.51581618e-02 8.70840967e-01
4.99434620e-01 8.05935204e-01 6.02533184e-02 -5.64806879e-01
1.06375086e+00 -3.86480153e-01 -7.09208965e-01 2.89535761e-01
4.75652933e-01 -3.18358630e-01 7.62943983e-01 3.45071226e-01
-6.39310300e-01 -4.14194256e-01 -8.33886862e-01 2.34924525e-01
-6.48625612e-01 1.47480980e-01 8.87645602e-01 3.68940026e-01
-1.10757387e+00 6.94925249e-01 -1.30473948e+00 -5.41702390e-01
4.30005223e-01 4.00198311e-01 -2.40199581e-01 4.83478546e-01
-1.11051178e+00 2.59589076e-01 3.13979715e-01 1.10222094e-01
-3.75714779e-01 -1.06823087e+00 -5.52401543e-01 2.55611151e-01
-3.76392230e-02 -4.80964892e-02 1.05950165e+00 -6.45158052e-01
-7.12362766e-01 4.87741232e-01 -8.72781351e-02 -8.71820569e-01
3.18061411e-01 -1.23113900e-01 -8.88711989e-01 2.52816141e-01
2.39790827e-01 1.70888945e-01 6.41433060e-01 -5.94214082e-01
-6.12030864e-01 -6.10684335e-01 -5.55669069e-01 -1.00064531e-01
-8.71725500e-01 1.10326812e-01 -1.30136356e-01 -7.61623502e-01
1.26898170e-01 -7.39343107e-01 -6.21703506e-01 -3.59297276e-01
6.79213032e-02 -4.63325232e-01 1.20871401e+00 -6.11136794e-01
1.43008363e+00 -2.36929965e+00 -9.57903415e-02 4.38815504e-01
4.90980297e-01 -2.10759751e-02 1.14639536e-01 7.29437768e-01
-3.59585017e-01 2.28372246e-01 -2.44523466e-01 -1.47461012e-01
-8.40390623e-02 1.07424073e-01 -7.96462715e-01 6.12497747e-01
2.60910690e-01 8.85235488e-01 -8.20481062e-01 -1.02478519e-01
4.35681462e-01 4.31278467e-01 -2.05259502e-01 -2.65989713e-02
-2.40601286e-01 6.19489968e-01 -3.34909499e-01 5.71243167e-01
4.26709890e-01 -7.97839820e-01 2.93068737e-01 1.75675042e-02
-5.42528331e-01 1.25632519e-02 -8.66506338e-01 1.37812567e+00
1.26849150e-03 7.64686942e-01 -1.76845580e-01 -1.39381874e+00
8.40034306e-01 3.34296077e-01 1.36129677e+00 -9.47293937e-01
2.05363512e-01 1.12151451e-01 -1.87849011e-02 -5.46183825e-01
6.29038990e-01 6.00705482e-02 -1.51692465e-01 9.18697238e-01
-4.33833748e-01 4.09018487e-01 3.44143629e-01 -3.08172554e-02
1.23868752e+00 -1.70792565e-01 1.29025325e-01 -4.09428209e-01
-1.20695047e-04 5.24241686e-01 6.87027633e-01 3.41642231e-01
-3.72103900e-01 2.88434058e-01 3.31086189e-01 -9.01566267e-01
-1.22487998e+00 -6.91590071e-01 -1.64440781e-01 1.08656180e+00
-3.85675192e-01 -4.78280455e-01 -3.31226021e-01 -1.55917078e-01
1.86787918e-01 2.75952458e-01 -9.27659392e-01 1.50965035e-01
-6.57721758e-01 -1.20188475e+00 5.57905853e-01 5.84353328e-01
1.86680287e-01 -9.49564755e-01 -9.94331360e-01 5.07796764e-01
-3.35723460e-01 -1.02095830e+00 -9.97430086e-02 3.18660796e-01
-1.21933770e+00 -1.08280373e+00 -2.56261230e-01 -3.85129273e-01
4.66709495e-01 7.73604214e-01 1.11043108e+00 1.05826827e-02
-5.95257521e-01 2.67363697e-01 -6.43840492e-01 -6.66036248e-01
-3.46258134e-02 3.50699991e-01 2.72567809e-01 -6.85509890e-02
9.83067036e-01 -1.00020421e+00 -7.93487370e-01 2.24907383e-01
-1.04989707e+00 -2.94260830e-01 2.47485843e-02 4.28146005e-01
5.24653852e-01 6.87811196e-01 1.10321116e+00 -4.56890941e-01
6.57343507e-01 -1.18712246e+00 -6.94868207e-01 6.45425394e-02
-7.73017645e-01 -4.86543804e-01 9.70360518e-01 -3.89721751e-01
-4.08276260e-01 -1.94354102e-01 2.52634645e-01 -8.08319211e-01
-1.15901642e-01 1.14277649e+00 6.92042589e-01 5.33883750e-01
7.49641061e-01 3.26226056e-01 3.68565530e-01 -4.58591789e-01
2.51706541e-01 5.57984591e-01 5.52314460e-01 -4.35562134e-01
6.77038789e-01 1.08737350e+00 -1.11573339e-01 -1.06480646e+00
-6.44764483e-01 -8.43403339e-01 -6.33377194e-01 -2.87162602e-01
6.22877657e-01 -1.18480253e+00 -7.91032016e-01 3.31101447e-01
-4.19511855e-01 -1.71057463e-01 -3.56661528e-01 4.68892902e-01
-4.42964993e-02 2.52023220e-01 -5.40327430e-01 -9.66044247e-01
-5.46148539e-01 -3.94371599e-01 7.99381614e-01 1.55849501e-01
-4.43791449e-01 -1.17454267e+00 1.76112130e-01 2.30583251e-02
6.27548933e-01 8.56554985e-01 9.02353227e-01 -8.67358685e-01
-4.45162505e-02 -2.28249595e-01 -2.82483851e-03 -8.41332972e-02
4.38043922e-01 1.65064573e-01 -8.89200926e-01 -5.28165400e-01
2.10834686e-02 -2.33561605e-01 5.85756242e-01 4.83222783e-01
1.50058687e+00 1.05694449e-02 -2.29472786e-01 6.04468048e-01
1.23255324e+00 4.04234618e-01 2.62952179e-01 3.92330080e-01
5.74621558e-01 7.45746970e-01 5.38103759e-01 1.00020874e+00
6.13043964e-01 3.87591869e-02 2.65450686e-01 -3.19362819e-01
5.25262892e-01 1.89697385e-01 1.89961493e-02 1.29322684e+00
-1.06248423e-01 1.01361714e-01 -1.38149917e+00 9.36955333e-01
-2.05746055e+00 -1.28485608e+00 -2.40257546e-01 1.93245161e+00
6.37212873e-01 -1.50058210e-01 4.42407459e-01 6.31160498e-01
5.21178126e-01 1.18462443e-01 -1.05102718e+00 6.90408498e-02
-3.72089267e-01 3.21192443e-01 2.93767333e-01 -3.36652189e-01
-1.46304131e+00 5.51788509e-01 6.87213421e+00 3.04115802e-01
-1.40140963e+00 -7.74871372e-03 6.35746360e-01 -5.20305455e-01
-8.61440226e-03 -4.61666912e-01 -2.66725004e-01 4.92916197e-01
1.48579836e+00 -6.20069921e-01 6.48215592e-01 7.66907752e-01
8.61841977e-01 3.49255592e-01 -9.24387932e-01 9.62743163e-01
-1.70620278e-01 -1.33492315e+00 -4.39918011e-01 6.41923100e-02
7.93850243e-01 6.05484486e-01 1.76792383e-01 3.27067167e-01
4.14103776e-01 -1.04465127e+00 4.62833822e-01 7.94978812e-02
5.94567776e-01 -9.23732281e-01 4.04816270e-01 8.27895403e-02
-1.58639967e+00 -2.42889225e-01 -1.56428501e-01 -4.65549409e-01
1.26071423e-02 7.74823189e-01 -4.20924455e-01 6.45937562e-01
1.33740127e+00 1.33572185e+00 -2.72412926e-01 7.74744511e-01
4.56633955e-01 1.02159739e+00 -6.96125090e-01 2.41500273e-01
2.51020342e-01 -4.27049309e-01 2.24161029e-01 1.02445638e+00
4.60178882e-01 2.52835244e-01 2.59070635e-01 7.12072670e-01
3.23462486e-02 1.36928976e-01 -6.89821780e-01 -5.23634911e-01
8.88830245e-01 1.33541977e+00 -1.13589215e+00 -2.43085608e-01
-4.58910227e-01 3.14752012e-01 2.72643656e-01 3.78525615e-01
-8.10848355e-01 -3.11848134e-01 8.79592061e-01 -1.09871037e-01
3.64184409e-01 -6.95321321e-01 -4.11034852e-01 -1.10868871e+00
1.87020808e-01 -9.27575767e-01 8.92953575e-01 -3.40020239e-01
-1.57881057e+00 5.19356489e-01 -1.22264370e-01 -1.78667283e+00
-1.96380109e-01 -3.72598320e-01 -5.58478773e-01 5.12216687e-01
-1.52739477e+00 -1.01393533e+00 -4.30064797e-01 7.75910437e-01
3.44234616e-01 -2.11073041e-01 6.93832099e-01 6.00152671e-01
-7.29413450e-01 2.85573930e-01 7.16917336e-01 1.81124777e-01
3.84378582e-01 -1.14559197e+00 6.97038591e-01 8.82766724e-01
-1.23821191e-01 7.96795011e-01 5.15942812e-01 -5.53026497e-01
-1.75794029e+00 -1.51007354e+00 5.47985554e-01 -3.01972806e-01
1.05080342e+00 -2.09153995e-01 -1.14611673e+00 4.53296006e-01
-7.30144307e-02 -9.40535963e-02 1.41750109e+00 3.95644724e-01
-5.42162299e-01 -2.36375257e-01 -8.74805629e-01 4.91221726e-01
6.45422041e-01 -5.31868160e-01 -4.31134433e-01 2.14422688e-01
6.99968219e-01 -7.73005653e-03 -1.29306936e+00 3.29020053e-01
4.80506152e-01 -7.60173678e-01 7.55231977e-01 -7.54696190e-01
4.25133824e-01 -4.93889332e-01 -1.94791391e-01 -1.11687982e+00
-2.76207060e-01 -7.43786633e-01 -2.29122341e-01 1.03643525e+00
1.47944584e-01 -5.13233900e-01 7.07260251e-01 4.72060591e-01
5.49622402e-02 -4.05600190e-01 -9.39136624e-01 -8.43418777e-01
2.02107713e-01 -9.12170768e-01 1.10095835e+00 1.81552052e+00
1.41861334e-01 2.39643957e-02 -3.41339797e-01 2.50013918e-01
7.75720656e-01 5.87694883e-01 6.97683394e-01 -1.65767574e+00
1.49933919e-01 -5.69743276e-01 -2.25190997e-01 -2.48655021e-01
8.79657343e-02 -8.32831204e-01 -3.78327250e-01 -1.16854501e+00
-2.11604774e-01 -6.57210767e-01 -6.73555672e-01 6.50338590e-01
-1.55907989e-01 3.20079654e-01 -2.56394327e-01 7.43225455e-01
-6.23633444e-01 4.45410669e-01 3.80554795e-01 -3.59562300e-02
-4.59949672e-01 -2.76057899e-01 -5.28474867e-01 3.44291061e-01
1.14362705e+00 -2.14624584e-01 -5.57353497e-01 -5.15569806e-01
2.77224898e-01 9.89594609e-02 1.42312586e-01 -8.88183773e-01
2.42792413e-01 -2.40901828e-01 4.02334899e-01 -1.06992435e+00
-1.39535099e-01 -1.01538253e+00 4.14595336e-01 4.20211107e-01
-1.72059461e-01 8.01084399e-01 5.23473799e-01 5.25254905e-01
-3.06815594e-01 6.13177538e-01 5.28281510e-01 -3.61034786e-03
-9.43077803e-01 5.30559421e-01 -2.83228725e-01 5.45178503e-02
1.13135171e+00 -8.99223611e-02 -2.91114867e-01 -2.66560107e-01
-5.71032107e-01 7.51064301e-01 2.78835475e-01 7.72453189e-01
3.66881490e-01 -1.34380102e+00 -7.13010371e-01 3.35988998e-01
2.39534453e-01 -3.09288092e-02 3.64327431e-01 1.07739234e+00
-5.27467847e-01 2.67453700e-01 -2.43085191e-01 -7.80609012e-01
-8.98377955e-01 6.78884208e-01 9.52054113e-02 -3.67369317e-02
-1.08555460e+00 5.03552966e-02 -3.76071662e-01 -3.95859927e-01
2.92744011e-01 -2.73789465e-01 -3.08771461e-01 4.65998590e-01
8.32436919e-01 5.85433006e-01 2.50852525e-01 -3.27750415e-01
-3.60551000e-01 4.18413490e-01 -3.24901454e-02 8.68028477e-02
1.98834467e+00 -2.27479100e-01 -4.86032069e-01 8.09460223e-01
1.00090468e+00 -4.96881962e-01 -1.07738698e+00 -1.95063546e-01
4.59940016e-01 -1.36276409e-01 -1.30008996e-01 -3.71185780e-01
-1.10549676e+00 7.66338229e-01 5.85703015e-01 8.11281145e-01
1.49062359e+00 -2.61106730e-01 7.92713642e-01 4.72631961e-01
2.47201636e-01 -1.23570228e+00 -2.71845460e-01 4.11875129e-01
3.65333408e-01 -1.32607079e+00 -1.09943017e-01 3.51475656e-01
-3.63131821e-01 1.08017623e+00 3.68320733e-01 -4.34981622e-02
9.44544435e-01 3.79484504e-01 5.11478066e-01 -4.68149126e-01
-1.03920937e+00 -1.45651847e-01 3.42615768e-02 4.04611647e-01
5.18978775e-01 2.29153216e-01 -1.67734668e-01 4.14209038e-01
-4.09291327e-01 -3.31497304e-02 2.46366501e-01 1.04816115e+00
-1.46561801e-01 -6.81348264e-01 -3.23243290e-01 7.23867655e-01
-7.36279488e-01 1.62284076e-01 -2.23630950e-01 7.37248600e-01
-2.88978964e-01 1.30874503e+00 5.46901941e-01 -5.30245543e-01
1.89729780e-02 2.81898547e-02 -3.02049696e-01 -9.69906747e-02
-9.65833366e-01 2.60623038e-01 -4.16791856e-01 -4.73901391e-01
-7.24852562e-01 -8.41564298e-01 -1.40292645e+00 -6.76590145e-01
1.28472090e-01 5.63005328e-01 8.16082954e-01 8.37165177e-01
7.79918194e-01 5.65570414e-01 1.10454631e+00 -5.55380940e-01
-4.13624555e-01 -8.18496883e-01 -6.03794932e-01 6.17137611e-01
5.31882584e-01 -2.67980725e-01 -3.15274268e-01 3.60384993e-02] | [7.145292282104492, 2.989281415939331] |
0f786c51-4b75-44d0-b6a9-3d2657a53b43 | araspot-arabic-spoken-command-spotting | 2303.16621 | null | https://arxiv.org/abs/2303.16621v1 | https://arxiv.org/pdf/2303.16621v1.pdf | AraSpot: Arabic Spoken Command Spotting | Spoken keyword spotting (KWS) is the task of identifying a keyword in an audio stream and is widely used in smart devices at the edge in order to activate voice assistants and perform hands-free tasks. The task is daunting as there is a need, on the one hand, to achieve high accuracy while at the same time ensuring that such systems continue to run efficiently on low power and possibly limited computational capabilities devices. This work presents AraSpot for Arabic keyword spotting trained on 40 Arabic keywords, using different online data augmentation, and introducing ConformerGRU model architecture. Finally, we further improve the performance of the model by training a text-to-speech model for synthetic data generation. AraSpot achieved a State-of-the-Art SOTA 99.59% result outperforming previous approaches. | ['Haidar Harmanani', 'Mahmoud Salhab'] | 2023-03-29 | null | null | null | null | ['synthetic-data-generation', 'synthetic-data-generation', 'keyword-spotting'] | ['medical', 'miscellaneous', 'speech'] | [ 2.87108302e-01 3.05611581e-01 1.40090391e-01 -1.86379746e-01
-1.14265239e+00 -6.24921739e-01 5.42715609e-01 -2.16664344e-01
-4.94240165e-01 4.37014520e-01 2.30272755e-01 -4.08215672e-01
2.45008841e-01 -2.46076763e-01 -3.38462561e-01 -4.46988255e-01
1.09809190e-01 7.30946898e-01 2.29928847e-02 -5.02583921e-01
1.99038275e-02 3.61807764e-01 -1.40463316e+00 3.81346285e-01
7.25615442e-01 1.00093532e+00 2.51675963e-01 1.26296329e+00
3.33910733e-02 5.62775910e-01 -8.92077863e-01 -3.53948116e-01
-1.95885869e-03 -3.26567680e-01 -7.87973940e-01 -1.55207098e-01
1.12707637e-01 -5.45643628e-01 -2.52653480e-01 5.31150043e-01
9.09675956e-01 -9.97017100e-02 3.08635414e-01 -1.28489506e+00
-3.47095490e-01 1.09210193e+00 -1.64240226e-01 -3.57597731e-02
8.26910436e-01 1.78263094e-02 9.85271573e-01 -1.43375766e+00
1.09968334e-01 1.13459539e+00 5.07807314e-01 4.16873246e-01
-9.72165763e-01 -6.26112938e-01 -7.67228901e-02 1.20264098e-01
-1.56022966e+00 -1.01479959e+00 6.55268610e-01 1.06958538e-01
1.34329867e+00 6.60490513e-01 4.50323910e-01 1.42856705e+00
-4.45386320e-01 1.05182791e+00 7.51187325e-01 -8.39133501e-01
7.62825459e-02 1.42945871e-01 -1.86230481e-01 4.94192541e-01
-5.28540373e-01 -7.60893822e-01 -1.20345151e+00 -4.46139812e-01
9.19393823e-02 -5.36077619e-01 -4.83648390e-01 4.77998465e-01
-1.42858887e+00 6.90630198e-01 -2.71646351e-01 3.28028411e-01
-3.54693115e-01 -8.01907778e-02 4.75992680e-01 3.37571591e-01
3.28138411e-01 5.24311364e-01 -6.65785730e-01 -9.80350196e-01
-1.15235245e+00 2.71879464e-01 1.18815386e+00 9.41278219e-01
3.14171910e-02 4.24412996e-01 -2.34943897e-01 1.04824936e+00
1.91782743e-01 8.46747220e-01 9.30910826e-01 -5.80516517e-01
7.13437319e-01 2.58310497e-01 3.81039947e-01 -5.05982816e-01
-3.14531356e-01 8.30779672e-02 -5.82885087e-01 -2.70510405e-01
4.02315289e-01 -2.95831770e-01 -1.08989024e+00 1.16896677e+00
5.01590706e-02 -4.00791802e-02 3.55864018e-02 7.88029730e-01
6.56366050e-01 1.12259734e+00 -2.08417058e-01 -3.53203565e-01
1.40519118e+00 -1.07860339e+00 -1.15940487e+00 -2.74385214e-01
3.92763138e-01 -1.02245438e+00 1.78375912e+00 9.08794582e-01
-1.01452315e+00 -1.61195219e-01 -1.16062772e+00 -4.06244248e-02
-5.64291358e-01 7.38068819e-01 3.78247797e-01 9.66337085e-01
-1.08668315e+00 1.74413875e-01 -8.40457916e-01 -2.86489159e-01
-8.64130259e-02 5.49542427e-01 -1.50026008e-01 3.21940988e-01
-1.42808867e+00 6.52678490e-01 3.18416506e-01 2.39586860e-01
-5.24371564e-01 -4.81289417e-01 -7.20526338e-01 1.24377623e-01
5.29857337e-01 9.78947952e-02 1.64160490e+00 -5.67804396e-01
-2.24432087e+00 4.28964764e-01 -1.04197919e-01 -6.00117862e-01
3.67041707e-01 -5.08070529e-01 -9.06771421e-01 2.82649428e-01
-2.14700818e-01 6.54720783e-01 1.17953229e+00 -7.53522635e-01
-5.76846361e-01 -1.87879696e-01 -4.68111098e-01 3.59381318e-01
-8.87517333e-01 3.54562849e-01 -6.37615502e-01 -9.21729565e-01
-1.19034618e-01 -9.74365890e-01 1.97132349e-01 -3.18561405e-01
-6.66718245e-01 -3.02484185e-01 1.31296611e+00 -1.25470829e+00
1.57820761e+00 -2.04359460e+00 -5.30855581e-02 3.38365644e-01
-2.52386332e-01 7.14992702e-01 -9.89288464e-02 5.43988109e-01
1.68184772e-01 2.55775172e-04 -6.41221553e-02 -4.10001159e-01
1.12720288e-01 -4.57970090e-02 -6.09908521e-01 1.00076236e-01
3.99030924e-01 8.83433759e-01 -6.32094860e-01 -1.57495499e-01
2.58847903e-02 3.97575289e-01 -1.35695770e-01 4.62834507e-01
-3.23811144e-01 -2.84143370e-02 6.39878735e-02 8.41299176e-01
1.81712985e-01 5.94924949e-02 1.30283117e-01 7.46787414e-02
7.76977912e-02 5.16902804e-01 -1.51579690e+00 1.59797764e+00
-5.76840162e-01 5.97247422e-01 2.49590173e-01 -5.79901457e-01
6.75173879e-01 7.05533266e-01 1.76972434e-01 -6.05373919e-01
1.25432968e-01 7.22185910e-01 -9.16893855e-02 -3.60906124e-01
9.83240843e-01 5.65987229e-01 -1.79945916e-01 7.83909857e-01
2.20846105e-02 -2.90198863e-01 -4.72134165e-02 4.04633462e-01
8.00343692e-01 -1.46118045e-01 5.01574241e-02 -5.46700731e-02
4.13487226e-01 -2.40287125e-01 -2.73135245e-01 7.23096550e-01
2.97872603e-01 6.36853576e-01 1.35740787e-01 -6.27710624e-03
-9.15038586e-01 -6.91494763e-01 1.87085509e-01 1.38658988e+00
-4.58014876e-01 -5.95957339e-01 -9.64643419e-01 -4.92231309e-01
-3.03251207e-01 8.11413348e-01 -7.75747150e-02 -7.01977313e-02
-5.36477327e-01 -6.27964318e-01 1.36936128e+00 3.75668138e-01
2.81249940e-01 -1.21980822e+00 -6.70770407e-01 4.41874653e-01
-5.09860158e-01 -1.51561391e+00 -8.63170326e-01 3.09481621e-01
-2.21938848e-01 -4.50640142e-01 -1.03085434e+00 -8.25563312e-01
2.50657380e-01 2.11484358e-01 8.06801796e-01 -2.17033207e-01
-1.86050713e-01 4.29601163e-01 -6.73176944e-01 -7.37268746e-01
-5.63139677e-01 4.65344787e-01 4.21345860e-01 1.22410379e-01
2.98249453e-01 -2.44203329e-01 -2.67105103e-01 1.90972432e-01
-7.17734873e-01 -2.07254924e-02 4.97376233e-01 8.64249229e-01
-3.96593511e-02 -1.53192118e-01 9.18700337e-01 -4.86612171e-01
1.07805479e+00 -9.73165408e-02 -3.61133337e-01 2.99081624e-01
-6.78793728e-01 -9.98543799e-02 7.87802815e-01 -8.31745267e-01
-7.08495796e-01 2.88050383e-01 -3.54391247e-01 -1.77681874e-02
2.38777980e-01 5.05358815e-01 -3.91228199e-01 5.66261448e-02
6.10836744e-01 4.49420840e-01 1.61233246e-01 -4.69328523e-01
4.71738398e-01 1.70928895e+00 6.12031102e-01 -8.40544775e-02
5.83389163e-01 8.01789090e-02 -7.45775640e-01 -1.29775584e+00
-4.33474809e-01 -5.16098976e-01 -3.16731185e-01 -9.43393558e-02
3.57116908e-01 -1.00183725e+00 -7.65464962e-01 9.11558747e-01
-1.11070144e+00 -5.41502118e-01 -5.30173890e-02 1.19943246e-01
-2.81533509e-01 2.42307574e-01 -4.88734514e-01 -1.18401003e+00
-9.32409585e-01 -9.27853107e-01 1.36183643e+00 1.07161433e-01
-5.61602890e-01 -4.11585331e-01 -3.58946234e-01 3.71149600e-01
5.13377428e-01 -4.45405364e-01 4.59759086e-01 -1.08802760e+00
7.03464448e-02 -4.62149590e-01 -2.07075700e-02 3.05483758e-01
2.08169386e-01 -1.00665092e-01 -1.36781859e+00 -2.65743464e-01
-1.98219165e-01 -5.84026694e-01 2.42937505e-01 -1.60805449e-01
1.20201707e+00 -7.06556857e-01 -9.00275819e-03 1.14397639e-02
6.23684108e-01 2.57816881e-01 4.67524588e-01 -8.48380774e-02
5.97794414e-01 3.04883838e-01 7.36648977e-01 7.07245409e-01
3.93164665e-01 9.87488508e-01 -8.47476199e-02 8.77061784e-02
-4.73411828e-02 -2.00987235e-01 6.44541621e-01 1.11537755e+00
5.10776520e-01 -6.24332726e-01 -1.13049042e+00 6.75973952e-01
-1.73806226e+00 -5.01372337e-01 1.53980022e-02 2.11700559e+00
1.37287164e+00 1.61776051e-01 4.23101366e-01 7.59691954e-01
3.66374910e-01 -1.28419504e-01 -3.34192455e-01 -6.16201282e-01
-9.35237929e-02 4.02858555e-01 3.79538715e-01 6.89078808e-01
-9.26966131e-01 1.15891469e+00 6.35847282e+00 9.76366162e-01
-1.35298061e+00 2.37880293e-02 6.91763878e-01 -9.10585597e-02
-1.75282583e-02 -5.15414476e-01 -8.37392509e-01 6.46468401e-01
1.50540912e+00 3.35245058e-02 9.44297671e-01 7.24092364e-01
4.71028030e-01 -2.23949015e-01 -9.90101516e-01 1.38062096e+00
4.59384918e-01 -9.61746216e-01 3.03552561e-02 -1.61117300e-01
2.13491261e-01 -8.01518857e-02 3.18707302e-02 1.86269611e-01
6.98900819e-02 -1.23754287e+00 9.16279972e-01 1.10336035e-01
1.26877952e+00 -7.23803639e-01 6.71295464e-01 3.50101680e-01
-9.61360753e-01 -1.64126623e-02 4.02018338e-01 4.23766598e-02
1.21850058e-01 2.54479051e-01 -1.80698204e+00 1.00715101e-01
7.61083543e-01 -1.66082472e-01 -5.29360235e-01 5.37704825e-01
-9.57474262e-02 9.19116318e-01 -6.48661375e-01 -4.98908728e-01
2.29235247e-01 2.14980900e-01 4.25491333e-01 1.47862828e+00
4.40875024e-01 -2.10006461e-02 7.43701532e-02 2.04674855e-01
-1.28216252e-01 1.90565720e-01 -3.25683236e-01 -4.34862822e-01
8.81978929e-01 1.31881785e+00 -7.38061607e-01 -2.60527670e-01
-6.54893816e-02 1.48491037e+00 -1.00564517e-01 2.27537498e-01
-7.54416943e-01 -9.35629129e-01 2.50434816e-01 -3.03437188e-02
9.01508182e-02 -6.14126503e-01 -1.90735787e-01 -7.91361392e-01
3.37348431e-01 -1.53035307e+00 1.16469055e-01 -8.90851200e-01
-6.73061371e-01 6.15295947e-01 -3.88383359e-01 -9.90823150e-01
-8.15115929e-01 -5.65097332e-01 -2.16683313e-01 1.00186300e+00
-1.14000165e+00 -1.36371434e+00 -3.31538081e-01 5.99101603e-01
8.86671841e-01 -3.07839006e-01 1.28282464e+00 3.38736385e-01
-9.93309021e-02 1.09656560e+00 -9.88812558e-03 1.22261561e-01
8.61488879e-01 -1.34484434e+00 7.53068209e-01 9.06087101e-01
6.43172741e-01 4.52506542e-01 7.35286593e-01 -4.07680988e-01
-1.84527421e+00 -6.52580976e-01 1.06216991e+00 -4.06805217e-01
8.95122230e-01 -9.73482847e-01 -7.22825587e-01 2.48899430e-01
3.34352136e-01 -4.34602499e-01 5.55766225e-01 -2.47515708e-01
4.75706579e-03 6.22416884e-02 -1.00878191e+00 8.64221334e-01
5.63904762e-01 -8.73148859e-01 -1.37467667e-01 4.38802123e-01
9.65905845e-01 -5.74784696e-01 -5.09058118e-01 -5.95387667e-02
7.11537838e-01 -1.97797447e-01 7.05918014e-01 -3.26465458e-01
-2.20999509e-01 -1.94236338e-01 -1.32744998e-01 -1.35823500e+00
4.54803050e-01 -1.55426347e+00 -4.98817503e-01 1.45418620e+00
8.18408668e-01 -4.39417303e-01 7.03674197e-01 6.09473884e-01
3.44906636e-02 -5.55288851e-01 -1.14801836e+00 -6.04396701e-01
-5.44729352e-01 -7.34222054e-01 5.33106565e-01 7.11927474e-01
3.91162306e-01 7.28869259e-01 -7.42518187e-01 1.79376334e-01
-8.13746303e-02 -5.57556331e-01 6.62268579e-01 -8.13033402e-01
-1.24518767e-01 -1.09889172e-01 5.71917966e-02 -1.10355198e+00
-1.20978743e-01 -5.25082767e-01 3.34865570e-01 -1.18727362e+00
-4.84009892e-01 -2.84260988e-01 3.56852859e-01 8.78566086e-01
-1.14694446e-01 4.13626045e-01 1.92666307e-01 -1.10644777e-03
-4.64572638e-01 5.78814447e-01 5.59481740e-01 -2.85101146e-01
-6.38950706e-01 8.80141705e-02 -6.65573955e-01 3.48746091e-01
9.73398924e-01 -1.03471242e-01 -4.64082241e-01 -2.47208670e-01
3.83321524e-01 -1.06398724e-01 9.06192139e-02 -8.07680905e-01
4.07294869e-01 3.12217206e-01 4.54655848e-02 -4.77527261e-01
7.12156177e-01 -9.11342740e-01 -4.37247634e-01 2.39356786e-01
-1.93090931e-01 3.55297804e-01 2.43968263e-01 2.76665777e-01
-1.16795190e-01 -1.98760495e-01 4.14514214e-01 1.80452585e-01
-4.82381582e-01 -1.78704813e-01 -8.68414342e-01 -7.27375448e-02
6.27100408e-01 -1.09784706e-02 -1.33746564e-01 -8.55159402e-01
-7.02855825e-01 1.96600333e-01 -1.12141073e-01 8.16356540e-01
7.68850029e-01 -1.11304414e+00 -5.27257264e-01 4.47231770e-01
5.75012006e-02 -3.77469696e-02 -4.46686357e-01 6.33089483e-01
-5.33972621e-01 4.50754076e-01 3.81625414e-01 -4.56711560e-01
-1.41017187e+00 -2.99673956e-02 1.25504255e-01 8.78399145e-03
-3.37088108e-01 9.08042610e-01 -1.11230505e+00 -4.58046049e-01
7.56629467e-01 -3.03786427e-01 -1.74685881e-01 3.45984101e-01
9.49455619e-01 4.83961582e-01 8.31050396e-01 -3.77029449e-01
-3.54901314e-01 -1.58675194e-01 -1.52048975e-01 -7.10320592e-01
1.07713914e+00 -6.83358461e-02 -2.13399921e-02 5.87790251e-01
7.82479465e-01 2.87779093e-01 -6.53789520e-01 -1.18022479e-01
2.95922011e-01 -1.04134023e-01 9.18737333e-03 -1.28445327e+00
-3.24989468e-01 5.95751524e-01 7.72877872e-01 3.47699612e-01
9.44820940e-01 -1.86372682e-01 1.13764298e+00 8.05939913e-01
2.04634875e-01 -1.72641420e+00 2.92354852e-01 7.56236196e-01
1.01043141e+00 -1.24382436e+00 -3.40764523e-01 -3.09654832e-01
-8.78202319e-01 1.16646135e+00 2.92268038e-01 4.30043340e-01
4.72576559e-01 7.90885925e-01 2.57423162e-01 9.41723734e-02
-5.20779490e-01 1.68145880e-01 3.87849331e-01 5.82546353e-01
3.34676266e-01 9.27056596e-02 1.35628479e-02 8.38084877e-01
-5.54113150e-01 -2.39320546e-01 6.29985929e-01 8.54577124e-01
-1.42121866e-01 -1.03567314e+00 -5.17295122e-01 5.56881547e-01
-6.84271157e-01 -5.29632688e-01 -7.50025630e-01 3.52298051e-01
-5.84470451e-01 1.33052170e+00 -1.61627769e-01 -5.86761355e-01
2.68449157e-01 5.75409651e-01 5.16198426e-02 -5.68031311e-01
-6.37829065e-01 2.64255345e-01 4.74066198e-01 -3.21756452e-01
7.97057003e-02 -5.66146731e-01 -1.18791056e+00 8.52395818e-02
-5.65205634e-01 2.72966325e-01 8.93539071e-01 7.91680336e-01
5.01671135e-01 3.17216933e-01 7.40425944e-01 -1.01362586e+00
-8.43025088e-01 -1.45998657e+00 -3.57536823e-01 -1.55496255e-01
2.71990895e-01 -7.83769414e-02 -1.81058139e-01 1.20487072e-01] | [14.240655899047852, 6.568187713623047] |
d7d5e45d-0ec1-485d-94ab-fb2da879d881 | supervised-prediction-of-aging-related-genes | 1908.08135 | null | https://arxiv.org/abs/1908.08135v4 | https://arxiv.org/pdf/1908.08135v4.pdf | Supervised prediction of aging-related genes from a context-specific protein interaction subnetwork | Background. Human aging is linked to many prevalent diseases. The aging process is highly influenced by genetic factors. Hence, it is important to identify human aging-related genes. We focus on supervised prediction of such genes. Gene expression-based methods for this purpose study genes in isolation from each other. While protein-protein interaction (PPI) network-based methods for this purpose account for interactions between genes' protein products, current PPI network data are context-unspecific, spanning different biological conditions. Instead, here, we focus on an aging-specific subnetwork of the entire PPI network, obtained by integrating aging-specific gene expression data and PPI network data. The potential of such data integration has been recognized but mostly in the context of cancer. So, we are the first to propose a supervised learning framework for predicting aging-related genes from an aging-specific PPI subnetwork. Results. In a systematic and comprehensive evaluation, we find that in many of the evaluation tests: (i) using an aging-specific subnetwork indeed yields more accurate aging-related gene predictions than using the entire network, and (ii) predictive methods from our framework that have not previously been used for supervised prediction of aging-related genes outperform existing prominent methods for the same purpose. Conclusion. These results justify the need for our framework. | ['Tijana Milenković', 'Qi Li'] | 2019-08-21 | null | null | null | null | ['human-aging'] | ['miscellaneous'] | [ 2.74944514e-01 1.66827440e-01 -4.98639971e-01 -1.11042887e-01
-1.80705279e-01 -1.00991838e-01 4.60628904e-02 4.78352100e-01
-2.81643897e-01 1.28748858e+00 1.95245832e-01 -1.52134597e-01
-4.20895159e-01 -8.74428511e-01 -4.34780687e-01 -7.96471000e-01
-4.72111374e-01 3.70947003e-01 2.55956829e-01 -3.89851421e-01
-1.74680650e-01 1.28915936e-01 -1.66953206e+00 -1.96116328e-01
1.13836682e+00 5.24292827e-01 -2.80127442e-03 4.54826683e-01
3.23095471e-01 6.63531348e-02 -4.14811969e-01 -2.04200104e-01
-2.22070247e-01 -5.53939879e-01 -6.73344374e-01 -3.14512551e-01
1.13100326e-02 5.04293621e-01 -1.12064622e-01 9.13314641e-01
7.52213717e-01 -4.70945686e-01 5.61034799e-01 -1.56348860e+00
6.28093481e-02 3.52002472e-01 -2.06778198e-01 8.11487250e-03
3.70504409e-01 9.51948315e-02 1.30603957e+00 -3.94067377e-01
9.05809343e-01 1.10986936e+00 9.27535236e-01 5.09315133e-01
-1.49040627e+00 -3.23139697e-01 1.10162869e-02 2.68515706e-01
-1.28840446e+00 -1.77067399e-01 6.30894482e-01 -4.00861025e-01
1.02277899e+00 3.73741716e-01 1.23283315e+00 1.05418682e+00
6.35363340e-01 5.09654880e-01 1.12014079e+00 -4.11727846e-01
1.61649212e-01 -4.68669683e-01 4.07040477e-01 7.82602727e-01
7.29692042e-01 -2.02107713e-01 -5.48092663e-01 -3.86868030e-01
4.77700569e-02 -1.39000893e-01 -5.72327495e-01 -2.18847573e-01
-1.46569991e+00 4.22495633e-01 9.55625251e-02 6.47770464e-01
-3.05637360e-01 1.27728596e-01 6.13497317e-01 3.19177032e-01
6.51601911e-01 4.95689481e-01 -1.10319996e+00 -1.33587301e-01
-9.05334353e-01 3.03318918e-01 1.09539425e+00 5.45443416e-01
7.60514319e-01 -4.62946057e-01 1.34176850e-01 8.23849976e-01
3.72488976e-01 1.67331621e-02 6.77922606e-01 -6.81725502e-01
-2.51357466e-01 9.43442106e-01 -2.90436774e-01 -9.90529954e-01
-9.56832707e-01 -3.95932376e-01 -8.56336713e-01 -1.29250407e-01
6.43900812e-01 -1.70132056e-01 -4.59053248e-01 2.04095387e+00
3.29235375e-01 1.86857253e-01 1.19582182e-02 3.79184753e-01
6.16413891e-01 4.51561771e-02 5.27118683e-01 -5.34137368e-01
1.67712569e+00 -7.53713727e-01 -7.26606131e-01 -1.09265029e-01
1.00117731e+00 -3.73198420e-01 7.22758234e-01 3.62940073e-01
-5.96750975e-01 -2.78474361e-01 -1.16396940e+00 1.50065646e-01
-8.51235211e-01 1.00586683e-01 6.73514426e-01 5.22800863e-01
-1.16784406e+00 8.79347205e-01 -7.99875617e-01 -1.26486158e+00
2.86819279e-01 5.08678079e-01 -6.56434238e-01 -1.42552495e-01
-1.46621490e+00 1.12202585e+00 4.52413797e-01 -1.67979538e-01
-1.85303092e-01 -8.45447242e-01 -6.20555520e-01 -3.36733274e-02
-3.16727534e-02 -1.26311338e+00 4.77370948e-01 -6.86865449e-01
-9.83831167e-01 8.65395010e-01 -4.61159766e-01 -3.29394162e-01
-2.23677494e-02 -2.02093929e-01 -5.65348804e-01 5.71259521e-02
2.04627052e-01 6.21583819e-01 1.72175214e-01 -7.70958066e-01
-3.43385696e-01 -5.12121558e-01 -3.07721674e-01 -9.44910720e-02
-4.43349421e-01 -1.08182728e-01 -3.06809247e-01 -5.37653923e-01
9.14367586e-02 -1.06188011e+00 -4.11366016e-01 1.02651253e-01
-5.18437386e-01 -3.79941314e-01 7.71258831e-01 -7.60914624e-01
1.34820092e+00 -1.75999177e+00 3.43439907e-01 2.22369526e-02
3.93706679e-01 1.16699584e-01 -9.12821293e-02 6.36679769e-01
-6.52211905e-01 3.66139472e-01 -2.52338499e-01 -8.52963403e-02
-2.35413656e-01 2.43288130e-01 5.28157473e-01 5.99775970e-01
2.45937303e-01 9.29712296e-01 -1.07272792e+00 -6.16026044e-01
-2.49048755e-01 5.34322381e-01 -3.41183394e-01 -2.91755885e-01
-2.62778014e-01 3.31288934e-01 -4.33742017e-01 1.00101078e+00
2.51834631e-01 -2.53446043e-01 6.62258089e-01 -5.88131368e-01
9.62255448e-02 -8.32697749e-02 -4.74549562e-01 1.60146570e+00
1.59735948e-01 3.07864249e-01 -2.22541422e-01 -1.29616189e+00
9.92008626e-01 4.16876078e-01 9.87302959e-01 -1.11764103e-01
4.40636277e-02 3.13702822e-01 4.83198047e-01 -4.68892604e-01
-2.47646838e-01 -6.92619607e-02 7.71407485e-02 1.39245719e-01
1.95514336e-01 4.15303946e-01 6.67927623e-01 1.43488392e-01
1.84015512e+00 3.82014304e-01 8.64432275e-01 -4.69761342e-01
7.21555889e-01 -7.52541870e-02 9.86431718e-01 5.15987873e-02
-5.45708060e-01 3.16916823e-01 9.88876879e-01 -1.13598540e-01
-8.65417480e-01 -6.80159152e-01 -1.79610237e-01 6.85455084e-01
-2.09225282e-01 -9.25550699e-01 -5.80555558e-01 -7.32609093e-01
1.13050163e-01 1.94011822e-01 -7.03285992e-01 -3.13820392e-01
-3.15522522e-01 -1.44687891e+00 7.11989701e-01 8.35423246e-02
3.51646096e-01 -5.26084244e-01 1.60825178e-02 3.97658199e-01
-3.01211953e-01 -8.55471432e-01 -1.54870957e-01 3.48743439e-01
-1.10362363e+00 -1.73444998e+00 -5.94567299e-01 -6.51813149e-01
7.72120118e-01 -1.16206579e-01 1.06758022e+00 2.08905369e-01
-4.63221222e-01 4.37720150e-01 -2.28066102e-01 -4.94961619e-01
-3.43469441e-01 2.39063203e-01 3.16588074e-01 -3.51537257e-01
4.03085351e-01 -1.09292877e+00 -6.10933602e-01 4.90866244e-01
-5.35446942e-01 -1.66078508e-01 8.22031558e-01 8.23717356e-01
4.80771393e-01 2.69950658e-01 1.00821877e+00 -7.10281074e-01
5.62599897e-01 -6.22441113e-01 4.24007513e-03 2.20042780e-01
-9.32091296e-01 -5.79072610e-02 1.60315126e-01 -1.80550560e-01
-5.29378414e-01 2.05220267e-01 -1.25022188e-01 3.73130083e-01
-2.29871318e-01 7.63132274e-01 -7.04545617e-01 6.20328821e-02
4.81925577e-01 1.30228698e-01 5.91063499e-01 -3.07084352e-01
-8.03348869e-02 1.59609705e-01 4.68162969e-02 -5.40101945e-01
3.27672571e-01 1.38544753e-01 7.12637663e-01 -8.15704763e-01
-5.98052323e-01 -5.46385765e-01 -7.88711131e-01 -3.62832248e-01
7.71390617e-01 -8.29311967e-01 -6.49459660e-01 4.74577695e-01
-8.94086957e-01 4.51712795e-02 3.36604938e-02 4.28185344e-01
-5.31853020e-01 6.43190682e-01 -4.88725275e-01 -3.48588973e-01
-4.39257443e-01 -8.52486312e-01 7.92674899e-01 -7.65043870e-02
-7.07536161e-01 -1.19291234e+00 5.27879298e-01 2.59622067e-01
7.28951246e-02 6.79338455e-01 1.23495519e+00 -7.94563174e-01
4.39603682e-05 -3.44687402e-01 -1.09603919e-01 2.43957087e-01
5.66237688e-01 2.90587276e-01 -6.06373310e-01 -9.12794620e-02
-5.45973599e-01 1.65838778e-01 7.96890914e-01 2.67541409e-01
6.78434074e-01 -5.85552193e-02 -1.08750534e+00 3.93848633e-03
1.39392233e+00 -1.72015220e-01 9.08354759e-01 3.93850207e-01
4.26910758e-01 8.95646930e-01 8.58277082e-01 -4.51280773e-02
2.40412652e-01 6.41750157e-01 3.94562751e-01 3.15535851e-02
-8.44770446e-02 1.17152967e-02 4.29700077e-01 5.39552331e-01
-4.24252510e-01 -2.63222992e-01 -8.55780005e-01 5.19721448e-01
-2.06740093e+00 -6.54804111e-01 -6.37915909e-01 2.01450372e+00
1.01657510e+00 1.20888233e-01 2.58383453e-01 3.05926144e-01
7.99322009e-01 -1.71214953e-01 -5.42194486e-01 1.70226172e-01
-4.67339486e-01 2.39602655e-01 2.92594492e-01 2.67828442e-02
-8.52441728e-01 5.97580731e-01 7.00895119e+00 5.55750072e-01
-6.60056055e-01 -3.15581299e-02 6.03800833e-01 2.04030395e-01
1.02914810e-01 1.93859115e-01 -6.27469838e-01 4.89386439e-01
1.06310570e+00 -3.33936721e-01 -2.08761483e-01 6.86082900e-01
6.01023555e-01 -1.99511066e-01 -1.07959592e+00 5.66746950e-01
-2.69256115e-01 -8.66418362e-01 -3.83414656e-01 4.49550301e-01
3.28000814e-01 -1.00001410e-01 -6.10281646e-01 1.67856991e-01
-2.23819111e-02 -7.59840906e-01 -1.37780353e-01 1.01232839e+00
2.58937240e-01 -4.31212425e-01 9.49352562e-01 -5.06891869e-03
-1.06903195e+00 1.64545313e-01 -6.09396920e-02 -8.76367316e-02
1.14522502e-01 1.55899417e+00 -8.05631697e-01 7.49161065e-01
6.11812055e-01 1.28491127e+00 -7.51183033e-01 1.30273688e+00
-2.76046813e-01 8.95951867e-01 -2.31912479e-01 1.47018194e-01
-4.44632858e-01 -1.67484447e-01 5.45690894e-01 1.12537181e+00
5.95070779e-01 -2.24067867e-01 -1.51391178e-01 5.53876758e-01
1.90831020e-01 2.92019486e-01 -5.47650576e-01 -4.21985239e-01
1.79924011e-01 1.58788192e+00 -7.82122850e-01 -1.86579838e-01
-4.97132868e-01 8.29619408e-01 1.45286426e-01 1.42712474e-01
-6.26358569e-01 -1.57632738e-01 1.18542886e+00 3.11479181e-01
-6.56173751e-02 -2.31136039e-01 -6.71379045e-02 -8.51213813e-01
-3.11406881e-01 -6.86534643e-01 4.50810224e-01 -6.79431796e-01
-1.55991590e+00 2.62605902e-02 -2.21408799e-01 -1.13840950e+00
7.04557896e-02 -7.09642768e-01 -6.58444226e-01 5.93063056e-01
-1.35062730e+00 -1.24509966e+00 -1.01303719e-01 1.54478610e-01
-3.59012745e-02 1.22014001e-01 1.13233078e+00 4.16322976e-01
-9.59074140e-01 6.03455529e-02 -9.81639102e-02 -3.69249493e-01
1.07256031e+00 -1.44565690e+00 -8.42594355e-02 4.66865122e-01
-6.41430020e-01 7.81927705e-01 8.95744503e-01 -1.08693790e+00
-1.17141271e+00 -1.15364826e+00 1.36919785e+00 -2.17930123e-01
7.69794464e-01 -8.82244930e-02 -8.15648735e-01 4.44950670e-01
-9.53546434e-04 -5.38253486e-02 1.12177753e+00 4.26801115e-01
-1.73617691e-01 -2.02263981e-01 -1.05944240e+00 6.13934934e-01
1.37768722e+00 -7.36528188e-02 -2.72102207e-01 4.25735056e-01
6.73878610e-01 3.08800399e-01 -1.41849864e+00 5.92024386e-01
6.60695016e-01 -7.99843848e-01 1.09869516e+00 -5.71520150e-01
3.00950468e-01 -5.59111238e-01 1.34308740e-01 -1.29987037e+00
-3.70272636e-01 -2.38155916e-01 -2.25813940e-01 1.23843586e+00
5.13752103e-01 -8.98596227e-01 7.62587667e-01 3.65894049e-01
-1.44879013e-01 -9.80193555e-01 -9.14994478e-01 -6.54931903e-01
-2.01438323e-01 -9.20081139e-02 2.59965450e-01 9.99664426e-01
5.62894166e-01 5.40180922e-01 -2.40699857e-01 1.50711965e-02
5.64956844e-01 -3.22058946e-01 6.12938643e-01 -1.75067699e+00
-1.52604520e-01 -3.16514164e-01 -8.85366261e-01 -1.62254333e-01
3.70604604e-01 -7.69970715e-01 -2.13684574e-01 -1.76121116e+00
3.61279726e-01 -1.28682375e-01 -4.78592753e-01 8.62066984e-01
-4.37386572e-01 2.46721879e-01 -4.82002318e-01 4.68846001e-02
-4.04322058e-01 3.20757508e-01 1.12016976e+00 -2.15814769e-01
6.71171676e-03 -2.75295079e-02 -9.37426925e-01 5.92387795e-01
1.07669175e+00 -3.12425256e-01 -2.23651871e-01 8.21078002e-01
4.80869621e-01 -3.29908043e-01 2.80884326e-01 -1.19159579e+00
9.31844637e-02 -3.03282775e-03 3.35597962e-01 -3.73683631e-01
3.09351385e-01 -5.10434866e-01 5.36989272e-01 1.10632288e+00
2.13646635e-01 -3.03570867e-01 6.97619542e-02 8.10507655e-01
-1.14237450e-01 -1.86339945e-01 4.33459789e-01 -7.04194456e-02
-5.05922496e-01 1.94915712e-01 -7.59351850e-01 -5.13531685e-01
8.63191068e-01 -2.68398225e-01 -3.77732605e-01 -1.93791956e-01
-1.22265887e+00 4.53695282e-03 4.77738649e-01 2.40950197e-01
2.58862585e-01 -1.14362776e+00 -5.42280495e-01 -4.56993431e-01
2.47537434e-01 -4.46868002e-01 -5.55914454e-02 1.43714476e+00
-5.03996722e-02 4.59149927e-01 -2.47190282e-01 -3.71565372e-01
-1.74623168e+00 5.63765764e-01 3.40131074e-01 -6.20553315e-01
-1.40102610e-01 4.71922785e-01 -1.03145830e-01 -4.68898147e-01
-2.89688945e-01 -2.28173211e-01 -4.31382030e-01 3.85490894e-01
2.64560580e-02 5.91376126e-01 4.11031432e-02 -5.36925495e-01
-3.60108703e-01 3.87232184e-01 1.98615059e-01 5.98412991e-01
1.42884266e+00 -1.33323491e-01 -1.01833224e+00 4.56521183e-01
1.06182420e+00 -4.58894446e-02 -4.56971854e-01 1.97475687e-01
3.25417399e-01 9.60382447e-02 -3.91752928e-01 -7.91969657e-01
-9.21072423e-01 1.36920035e-01 6.24839604e-01 6.05943650e-02
1.23837137e+00 -2.92096082e-02 4.86849755e-01 1.71956941e-01
3.94956261e-01 -7.33251810e-01 -1.58960506e-01 1.51773363e-01
6.39233589e-01 -9.01657641e-01 3.79029602e-01 -1.13931394e+00
8.98576677e-02 1.19502115e+00 7.18933702e-01 2.58125037e-01
9.06498015e-01 -5.45429140e-02 -4.94992256e-01 -2.92230427e-01
-8.99767280e-01 -3.12636405e-01 1.35620877e-01 8.19977641e-01
9.28459227e-01 -2.74667591e-02 -1.25104165e+00 8.38342786e-01
1.85130406e-02 3.67553234e-01 2.31546327e-01 1.01973963e+00
-5.61938882e-01 -1.83837843e+00 -1.07959740e-01 8.54860067e-01
-2.88990796e-01 -1.20940827e-01 -4.94005233e-01 8.23485613e-01
4.91644740e-01 7.40182996e-01 -3.82180631e-01 -2.88636059e-01
2.01932207e-01 5.56627989e-01 4.87882495e-01 -4.03409243e-01
-2.31670633e-01 1.28817810e-02 6.91897869e-01 -5.98427296e-01
-8.32648516e-01 -1.00085509e+00 -1.26342630e+00 1.79349855e-02
-2.71813005e-01 1.21156238e-01 5.41085601e-01 8.78509402e-01
7.32240617e-01 7.51652181e-01 8.48414153e-02 -4.43712413e-01
2.80892253e-01 -1.04952145e+00 -9.23135340e-01 8.29727016e-03
-2.98042744e-01 -7.22000182e-01 -4.31321591e-01 2.06466496e-01] | [6.529954433441162, 5.498507022857666] |
50409c6f-4911-4d1a-9797-14e677c6fe80 | learning-3d-aware-egocentric-spatial-temporal | 1909.09272 | null | https://arxiv.org/abs/1909.09272v3 | https://arxiv.org/pdf/1909.09272v3.pdf | Learning 3D-aware Egocentric Spatial-Temporal Interaction via Graph Convolutional Networks | To enable intelligent automated driving systems, a promising strategy is to understand how human drives and interacts with road users in complicated driving situations. In this paper, we propose a 3D-aware egocentric spatial-temporal interaction framework for automated driving applications. Graph convolution networks (GCN) is devised for interaction modeling. We introduce three novel concepts into GCN. First, we decompose egocentric interactions into ego-thing and ego-stuff interaction, modeled by two GCNs. In both GCNs, ego nodes are introduced to encode the interaction between thing objects (e.g., car and pedestrian), and interaction between stuff objects (e.g., lane marking and traffic light). Second, objects' 3D locations are explicitly incorporated into GCN to better model egocentric interactions. Third, to implement ego-stuff interaction in GCN, we propose a MaskAlign operation to extract features for irregular objects. We validate the proposed framework on tactical driver behavior recognition. Extensive experiments are conducted using Honda Research Institute Driving Dataset, the largest dataset with diverse tactical driver behavior annotations. Our framework demonstrates substantial performance boost over baselines on the two experimental settings by 3.9% and 6.0%, respectively. Furthermore, we visualize the learned affinity matrices, which encode ego-thing and ego-stuff interactions, to showcase the proposed framework can capture interactions effectively. | ['Yi-Ting Chen', 'Chengxi Li', 'Yue Meng', 'Stanley H. Chan'] | 2019-09-20 | null | null | null | null | ['novel-concepts'] | ['reasoning'] | [-2.35820487e-01 -2.66926195e-02 4.74559031e-02 -6.89559102e-01
2.74712056e-01 -3.57193142e-01 7.54592180e-01 -3.79584104e-01
-3.93772095e-01 2.29976505e-01 4.25189674e-01 -4.13509399e-01
-3.18299532e-01 -7.47283638e-01 -6.59558654e-01 -6.62197590e-01
-5.48706017e-03 3.50277036e-01 3.36960644e-01 -5.96268296e-01
3.05693924e-01 7.47167587e-01 -1.70262897e+00 -4.13959473e-02
9.65492249e-01 6.79690063e-01 2.64617354e-01 7.90222228e-01
-3.82092334e-02 8.60412836e-01 -2.09231481e-01 -3.16682190e-01
1.43031821e-01 -2.84413621e-02 -4.10044998e-01 4.40398380e-02
2.82809645e-01 -2.93742478e-01 -9.95653570e-01 9.62953150e-01
3.39987993e-01 6.74733818e-01 6.71749651e-01 -1.55317342e+00
-5.72730899e-01 3.96638423e-01 -5.16889691e-01 4.90152091e-01
-1.07946269e-01 6.30776882e-01 6.75105989e-01 -7.90827036e-01
3.42143178e-01 1.50094652e+00 1.97884783e-01 3.62206936e-01
-8.00276279e-01 -6.49323761e-01 4.07242924e-01 7.97826290e-01
-1.15456665e+00 -4.47132677e-01 9.41947699e-01 -5.12522340e-01
1.03398335e+00 3.42914373e-01 6.40891135e-01 6.89009905e-01
6.82209611e-01 1.01678610e+00 6.39551520e-01 1.68019012e-01
-1.44987917e-02 1.53495431e-01 7.08075345e-01 4.37649608e-01
5.75122833e-02 3.00029039e-01 -3.55388314e-01 2.86882043e-01
5.58889449e-01 2.41206646e-01 8.23245868e-02 -5.91742337e-01
-1.13604355e+00 6.08895898e-01 7.37495899e-01 -1.15120932e-01
-4.65760559e-01 4.46488500e-01 4.83619839e-01 -5.71328439e-02
2.52029270e-01 -8.65770131e-02 6.16642088e-02 -3.69909823e-01
2.28602905e-02 3.05612713e-01 4.62664485e-01 1.31886256e+00
9.31922495e-01 7.01588169e-02 -3.69067878e-01 8.87064874e-01
2.13172480e-01 5.61268091e-01 1.22313783e-01 -7.23827481e-01
5.03987968e-01 7.27731109e-01 -3.32162045e-02 -1.29965508e+00
-5.74535429e-01 -6.64589584e-01 -6.91957057e-01 2.10772967e-03
-1.13435701e-01 3.49847488e-02 -9.86093581e-01 1.61302888e+00
3.92432898e-01 5.17785370e-01 1.83125719e-01 1.18792534e+00
1.08051705e+00 1.38422102e-01 1.54722720e-01 4.17337298e-01
1.62977302e+00 -1.34376192e+00 -9.58403111e-01 -4.64521468e-01
9.54554498e-01 -4.13017899e-01 9.31083560e-01 -1.59135073e-01
-1.02068245e+00 -8.52461398e-01 -7.88615704e-01 -2.03903124e-01
-6.06536806e-01 6.00853562e-02 8.69312227e-01 3.06767374e-01
-9.41815794e-01 -1.33873951e-02 -6.82288110e-01 -4.10380751e-01
4.37959611e-01 3.70634437e-01 -4.99362051e-01 -2.05307320e-01
-1.41893470e+00 1.04521537e+00 3.23718756e-01 3.70716274e-01
-1.07425320e+00 -5.40236950e-01 -1.08378899e+00 2.01370180e-01
4.59840924e-01 -8.04125369e-01 1.06252706e+00 -2.01866865e-01
-1.05343616e+00 6.96363568e-01 -4.27146584e-01 -6.07739449e-01
3.45993668e-01 -2.49802768e-01 -7.96630561e-01 -2.55592257e-01
1.31010577e-01 6.98049128e-01 5.01325548e-01 -1.26533794e+00
-8.70710552e-01 -5.64844847e-01 4.36132222e-01 7.04397678e-01
-3.53538580e-02 -2.69253612e-01 -7.33864725e-01 -2.35635057e-01
2.25072876e-01 -1.08674109e+00 -2.35836774e-01 -4.64804441e-01
-5.00325739e-01 -5.29640734e-01 1.34821224e+00 -2.89055496e-01
1.06586874e+00 -2.61815381e+00 5.72981276e-02 3.79690416e-02
7.53111303e-01 2.59015411e-01 -2.14098528e-01 2.42791861e-01
-8.27222243e-02 -4.09549206e-01 -3.00988983e-02 -3.71119887e-01
3.15003276e-01 5.62054634e-01 -7.12537467e-02 3.74609321e-01
5.32897897e-02 1.42175245e+00 -9.10310864e-01 -1.71002790e-01
6.78367019e-01 4.91302967e-01 -5.66145301e-01 1.23058982e-01
3.47483516e-01 3.14859718e-01 -5.24063826e-01 3.79625410e-01
1.08404171e+00 3.35055560e-01 -4.39982653e-01 -3.55034053e-01
-2.29160652e-01 3.02642614e-01 -8.77582908e-01 1.33936810e+00
-4.04748678e-01 8.86281610e-01 -9.20537114e-02 -9.21746910e-01
9.60800350e-01 -1.71089515e-01 1.45868018e-01 -8.56645644e-01
4.24509138e-01 -2.80855268e-01 4.65689629e-01 -6.75769508e-01
6.80685282e-01 2.27902472e-01 -1.78312078e-01 3.30431819e-01
1.62750087e-03 6.74976707e-02 2.04623878e-01 4.04139072e-01
1.10670602e+00 -1.80303514e-01 -8.08350667e-02 -4.78862524e-01
7.20498502e-01 3.94405760e-02 4.72814351e-01 7.67700911e-01
-7.65143991e-01 9.90282595e-02 4.11288649e-01 -7.21498728e-01
-5.00572681e-01 -1.13560355e+00 1.74551710e-01 1.15169609e+00
8.37896049e-01 -1.62279949e-01 -7.58243442e-01 -6.81788981e-01
2.07165852e-01 1.17460489e+00 -7.47522354e-01 -7.65601277e-01
-6.12343192e-01 -4.07598048e-01 3.34103435e-01 8.09693694e-01
1.00390601e+00 -9.85315859e-01 -5.60229897e-01 5.23812138e-02
-1.06041670e-01 -1.29547632e+00 -8.43134880e-01 -5.91120832e-02
-4.27734315e-01 -9.63234425e-01 -1.21544398e-01 -6.31329656e-01
5.72627842e-01 1.04676986e+00 9.51902986e-01 -7.78531656e-02
-2.88661867e-01 4.96715218e-01 -1.54788417e-04 -5.30189991e-01
2.02612206e-01 -1.07532449e-01 2.82849818e-01 3.52241367e-01
1.16615188e+00 -6.04031026e-01 -9.86198843e-01 7.99891829e-01
-4.49081331e-01 3.37917060e-01 6.69272602e-01 6.44413829e-01
1.76022872e-01 1.68725133e-01 2.07101122e-01 -7.58272111e-01
7.05899060e-01 -5.51390827e-01 -2.21790239e-01 -1.92046374e-01
-1.93289652e-01 -3.00258934e-01 3.95266920e-01 -2.31183305e-01
-1.21389794e+00 -4.58593592e-02 1.28617182e-01 -6.30260885e-01
-4.81439501e-01 3.60256940e-01 -4.44014311e-01 -1.65612306e-02
4.83349204e-01 2.78632432e-01 5.48816398e-02 4.34586033e-03
7.45073915e-01 6.55290961e-01 8.00794542e-01 -2.16133684e-01
7.87888944e-01 7.12087691e-01 -3.87179703e-02 -7.31867433e-01
-4.64185357e-01 -7.49825776e-01 -7.04168558e-01 -4.87770706e-01
1.04691410e+00 -9.76048768e-01 -1.23260462e+00 5.19151628e-01
-1.01085281e+00 -2.12944776e-01 -1.29692733e-01 7.71757960e-01
-6.52094841e-01 7.69463480e-02 -2.15636835e-01 -8.84117484e-01
1.28851980e-01 -1.22174525e+00 9.73311007e-01 4.76425976e-01
1.20861650e-01 -1.05366373e+00 -3.18317384e-01 5.32309473e-01
4.46755677e-01 -6.95643798e-02 5.33544183e-01 -3.77311379e-01
-7.58896708e-01 -1.87404856e-01 -6.40282154e-01 -2.00732369e-02
3.50696623e-01 -4.04596061e-01 -9.73693609e-01 -7.27763623e-02
-1.02015905e-01 4.55934554e-01 9.56809998e-01 4.61000323e-01
1.04941082e+00 1.17526107e-01 -7.95872509e-01 6.63241446e-01
6.17550015e-01 6.01085782e-01 8.38744819e-01 -2.21661441e-02
1.22757626e+00 9.88118827e-01 8.55994105e-01 1.54529616e-01
9.59304631e-01 6.86792493e-01 6.22704685e-01 -2.60571361e-01
-1.96197614e-01 -2.43137643e-01 1.31852165e-01 5.60027599e-01
-1.31747916e-01 -2.85017014e-01 -9.39957559e-01 8.42903614e-01
-2.30294299e+00 -8.75200748e-01 -4.74397153e-01 1.63050723e+00
-2.29766220e-01 3.55112731e-01 1.41072404e-02 -2.74356276e-01
7.05255687e-01 8.98690075e-02 -9.35863256e-01 -4.64283109e-01
-5.57277203e-02 -3.70453000e-01 6.68486893e-01 5.18721461e-01
-1.12551188e+00 1.32730377e+00 5.08480406e+00 6.61586702e-01
-8.39393198e-01 5.73308691e-02 4.33178812e-01 -1.29274026e-01
-2.04410151e-01 -3.15684415e-02 -8.95792842e-01 3.75703067e-01
6.65487647e-01 -3.48369896e-01 5.00473499e-01 8.92337024e-01
6.36828065e-01 1.64481010e-02 -8.61695409e-01 1.21570361e+00
6.78021312e-02 -1.02107656e+00 -4.23504692e-03 1.42309487e-01
3.54959875e-01 1.35132428e-02 2.34346747e-01 6.41995490e-01
5.15944004e-01 -1.00150537e+00 4.04615581e-01 6.74164057e-01
2.77880073e-01 -1.09815001e+00 8.30071688e-01 4.05513436e-01
-1.56399167e+00 -3.13725501e-01 -4.41632718e-01 -2.67757505e-01
4.55089867e-01 4.20684636e-01 -6.58245802e-01 5.30826509e-01
7.68249035e-01 1.02107716e+00 -4.70244884e-01 7.33424187e-01
-1.04349896e-01 2.76784420e-01 -1.35398591e-02 -9.38300639e-02
6.56102180e-01 -5.65299451e-01 7.65328765e-01 1.07641041e+00
7.27140112e-03 4.80013996e-01 -7.62085319e-02 1.19317687e+00
2.61490673e-01 -1.84247285e-01 -1.09278357e+00 9.91169587e-02
2.99695045e-01 1.51385152e+00 -3.98123413e-01 -3.96995246e-01
-2.51594692e-01 8.18675578e-01 2.98159868e-01 5.90191007e-01
-1.15857100e+00 -6.37202263e-01 1.45406890e+00 1.97261840e-01
1.45342834e-02 -4.03308064e-01 -2.29390174e-01 -7.36261308e-01
-9.73019823e-02 -3.02981913e-01 8.96346346e-02 -8.93891811e-01
-1.01589978e+00 6.20154142e-01 2.23701730e-01 -1.15580177e+00
-3.03228069e-02 -6.17567956e-01 -9.73696113e-01 1.03788471e+00
-1.36410820e+00 -1.24481952e+00 -9.25062418e-01 6.55879915e-01
6.08210921e-01 -2.86930650e-01 3.10991034e-02 4.48718131e-01
-6.89877868e-01 5.95240057e-01 -3.33594799e-01 1.75922439e-02
3.70919853e-01 -1.04885876e+00 9.56634223e-01 7.01828718e-01
-2.61123985e-01 9.07057166e-01 5.23483217e-01 -6.17719769e-01
-1.56407726e+00 -1.41911077e+00 6.70846701e-01 -6.88767314e-01
4.28362012e-01 -5.60427964e-01 -7.29808152e-01 8.42976511e-01
1.34417281e-01 1.06300302e-01 3.47689033e-01 1.87886015e-01
-4.31837142e-02 -8.23510140e-02 -9.14255321e-01 1.11396146e+00
1.69316280e+00 -6.14597380e-01 -4.93369699e-01 2.78598726e-01
8.02748919e-01 -5.85549116e-01 -2.94378161e-01 4.55729455e-01
5.60436666e-01 -1.10073364e+00 9.04593647e-01 -7.22694397e-01
2.29502358e-02 -6.86763823e-01 -1.11138567e-01 -1.50421882e+00
-7.09472001e-01 -3.71271461e-01 -1.64345428e-01 7.45739996e-01
-1.94363128e-02 -1.07698607e+00 6.84715986e-01 5.99041641e-01
-8.50504339e-01 -6.16500139e-01 -6.99753106e-01 -6.42924726e-01
-4.14138049e-01 -6.34950697e-01 9.52020109e-01 7.66341269e-01
-2.00557616e-02 5.38498461e-01 -2.81294048e-01 2.88626879e-01
5.23838520e-01 -2.56357372e-01 1.32699990e+00 -1.02280641e+00
7.87636489e-02 -5.82531631e-01 -9.27928150e-01 -1.55751884e+00
4.09938037e-01 -7.92218626e-01 4.41682152e-02 -1.61312664e+00
2.56951660e-01 -4.47710782e-01 -4.55631107e-01 1.70628265e-01
-2.93637395e-01 1.74595088e-01 -8.53066668e-02 -2.24039271e-01
-3.92870724e-01 6.80773914e-01 1.48020601e+00 -3.45449239e-01
-2.98306942e-01 -3.47570665e-02 -7.58334398e-01 5.88898242e-01
8.61628890e-01 6.87775984e-02 -9.46976006e-01 -5.17957628e-01
-3.92478377e-01 -1.92194223e-01 6.84364259e-01 -1.00610816e+00
5.71472585e-01 -1.94868788e-01 -3.89903970e-02 -1.10751271e+00
5.89416683e-01 -7.43768275e-01 2.07386851e-01 3.15528691e-01
-1.64978728e-02 -5.16514070e-02 4.44196552e-01 7.20468819e-01
-2.48860657e-01 4.37195688e-01 5.12420893e-01 1.54028594e-01
-1.09144187e+00 5.97838044e-01 -3.61432046e-01 -3.83725911e-01
1.44344783e+00 -5.41000962e-01 -4.85497832e-01 -6.18740857e-01
-8.24152410e-01 8.41564238e-01 1.46233421e-02 8.95653546e-01
7.56889403e-01 -1.54837883e+00 -5.23296535e-01 6.00153983e-01
4.27729160e-01 -1.09644808e-01 9.55176473e-01 1.12165010e+00
-3.04450423e-01 7.56844103e-01 -2.61672705e-01 -9.01219368e-01
-1.30152488e+00 5.23377597e-01 4.45172012e-01 2.37292826e-01
-7.06932604e-01 9.11306322e-01 1.21208417e+00 -6.91218913e-01
-7.35826120e-02 -2.61377543e-01 -5.28733730e-01 -3.31034094e-01
3.97936165e-01 4.83730942e-01 -1.71934068e-01 -1.12998998e+00
-3.95285726e-01 4.75118697e-01 -3.58005434e-01 1.62101805e-01
7.88105726e-01 -4.16512221e-01 7.84642771e-02 2.07411617e-01
9.93824422e-01 -5.39008200e-01 -1.33461821e+00 -1.71145052e-01
-3.00960183e-01 -6.80039704e-01 2.37923294e-01 -4.17737722e-01
-1.14408314e+00 1.16035008e+00 6.95972562e-01 3.37142423e-02
9.68835831e-01 6.63163746e-03 8.56963992e-01 2.80412555e-01
3.01800042e-01 -9.26104128e-01 -1.58405185e-01 7.62189746e-01
8.51911247e-01 -1.17498589e+00 -4.72658455e-01 -6.00522041e-01
-9.01342928e-01 6.33176863e-01 1.05797374e+00 -1.08171217e-01
7.52700150e-01 -4.15118001e-02 9.17097926e-02 -6.97824359e-01
-8.34440529e-01 -4.98751253e-01 5.59355795e-01 6.48230970e-01
-9.79747716e-03 5.13630927e-01 -8.53609666e-02 4.52624977e-01
-4.16240066e-01 -4.09518361e-01 2.28125229e-01 6.66450918e-01
-3.27821553e-01 -4.83646661e-01 4.00670208e-02 6.35562360e-01
3.69640976e-01 4.55858745e-02 -2.62872100e-01 8.15272093e-01
2.08576024e-01 1.11972928e+00 5.56711912e-01 -8.00420284e-01
8.21270585e-01 -1.99000642e-01 6.44539148e-02 -4.10899520e-01
-2.87821472e-01 -1.90832794e-01 6.99348748e-02 -7.61524081e-01
-7.18283355e-02 -5.32881260e-01 -1.33569252e+00 -6.11192584e-01
-2.34430432e-01 1.20180674e-01 5.36922932e-01 9.19455171e-01
7.17674315e-01 1.02949476e+00 6.14679813e-01 -8.41442049e-01
1.36541367e-01 -9.64805484e-01 -6.49968624e-01 4.75653768e-01
2.47524172e-01 -1.26424885e+00 -3.96104783e-01 -2.79276162e-01] | [6.41081428527832, 0.6455407738685608] |
f68151e9-edce-4904-be04-105aeab443da | solving-mathword-problems-automatically-with | 2208.05645 | null | https://arxiv.org/abs/2208.05645v2 | https://arxiv.org/pdf/2208.05645v2.pdf | Heterogeneous Line Graph Transformer for Math Word Problems | This paper describes the design and implementation of a new machine learning model for online learning systems. We aim at improving the intelligent level of the systems by enabling an automated math word problem solver which can support a wide range of functions such as homework correction, difficulty estimation, and priority recommendation. We originally planned to employ existing models but realized that they processed a math word problem as a sequence or a homogeneous graph of tokens. Relationships between the multiple types of tokens such as entity, unit, rate, and number were ignored. We decided to design and implement a novel model to use such relational data to bridge the information gap between human-readable language and machine-understandable logical form. We propose a heterogeneous line graph transformer (HLGT) model that constructs a heterogeneous line graph via semantic role labeling on math word problems and then perform node representation learning aware of edge types. We add numerical comparison as an auxiliary task to improve model training for real-world use. Experimental results show that the proposed model achieves a better performance than existing models and suggest that it is still far below human performance. Information utilization and knowledge discovery is continuously needed to improve the online learning systems. | ['Meng Jiang', 'Zijian Hu'] | 2022-08-11 | null | null | null | null | ['semantic-role-labeling'] | ['natural-language-processing'] | [-1.82625040e-01 2.80065715e-01 -3.29416305e-01 -4.38913792e-01
-7.53870979e-02 -4.83376831e-01 1.89386889e-01 9.25624669e-01
-1.65764675e-01 7.71143973e-01 -2.07605567e-02 -9.74087417e-01
-4.17760521e-01 -1.47830999e+00 -5.59361160e-01 2.78030843e-01
-1.95616670e-02 5.80310762e-01 5.01710057e-01 -5.12611508e-01
5.11748791e-01 3.61965060e-01 -1.56708372e+00 5.78182995e-01
1.23186910e+00 8.67257118e-01 3.23167741e-01 5.44142604e-01
-1.02761734e+00 1.51946187e+00 -6.01368845e-01 -5.21418154e-01
1.94915324e-01 -4.54245172e-02 -1.26459908e+00 -3.88009816e-01
5.37936866e-01 -2.04709068e-01 -3.42873424e-01 9.86376047e-01
1.59099340e-01 2.48839363e-01 5.78835309e-01 -1.52836108e+00
-8.35713983e-01 1.09870934e+00 -1.49357215e-01 1.67520165e-01
9.25869107e-01 -3.89461100e-01 1.17887986e+00 -3.81574243e-01
6.13642156e-01 1.13286197e+00 6.78081214e-01 1.44750401e-01
-4.90070492e-01 -5.38083673e-01 4.00375247e-01 8.96317303e-01
-1.42349362e+00 1.85789958e-01 6.62439704e-01 -4.86528754e-01
1.20761752e+00 4.91823524e-01 7.30576932e-01 2.53014147e-01
3.02249312e-01 6.91140771e-01 8.61564875e-01 -7.23003447e-01
-4.89521809e-02 4.39096868e-01 7.30417430e-01 1.45563102e+00
1.98281944e-01 -7.12183356e-01 -4.40724880e-01 2.81482823e-02
8.70420098e-01 -7.09610283e-02 -5.02476431e-02 -8.13949257e-02
-8.89111698e-01 5.68122029e-01 2.45819703e-01 2.79163539e-01
6.83013583e-03 -1.97513103e-01 2.30541453e-01 6.61016583e-01
2.09592953e-01 7.85042286e-01 -8.47046316e-01 -1.22534819e-01
-6.78184628e-01 2.58057356e-01 1.21594429e+00 1.38491988e+00
8.28668416e-01 -1.57885656e-01 -2.75321960e-01 7.70613790e-01
2.31241941e-01 6.03881031e-02 5.58265626e-01 -5.41551650e-01
7.61022151e-01 1.44003952e+00 -2.59269506e-01 -1.22158623e+00
-4.87344295e-01 -2.71790117e-01 -5.05976319e-01 -9.55029279e-02
5.64911962e-01 1.66412875e-01 -7.57695019e-01 1.10125625e+00
2.89044738e-01 2.65711457e-01 -2.07433224e-01 5.39964259e-01
1.29374409e+00 7.57740855e-01 1.89444333e-01 -5.40790297e-02
1.29254735e+00 -1.02575004e+00 -1.05746329e+00 2.98244566e-01
1.09238601e+00 -7.41566658e-01 1.12767744e+00 6.50112033e-01
-1.27778959e+00 -6.82011485e-01 -8.59937787e-01 -4.09270972e-01
-1.11984897e+00 8.49958956e-02 1.15115106e+00 8.56199145e-01
-8.01390707e-01 7.00119495e-01 -3.92122477e-01 -1.39833406e-01
4.08825427e-02 4.04638350e-01 -2.88194150e-01 -1.64993793e-01
-1.51461315e+00 9.46459055e-01 5.80448747e-01 -2.79774606e-01
-2.27647513e-01 -1.06119752e+00 -1.22828615e+00 3.98697883e-01
5.41976273e-01 -5.48764884e-01 8.76637578e-01 -5.13731122e-01
-1.27044332e+00 6.67145491e-01 1.71628803e-01 -2.19466999e-01
2.45289907e-01 2.97178663e-02 -5.83971024e-01 -1.12152778e-01
-1.34033144e-01 1.46512296e-02 2.56244242e-01 -7.21991718e-01
-6.97832823e-01 -1.23056896e-01 5.05362511e-01 2.36718833e-01
-5.65701067e-01 -7.19068944e-02 -3.23199451e-01 -5.55687010e-01
2.38087922e-01 -2.31079161e-01 1.25798941e-01 -1.97663873e-01
-6.88773915e-02 -8.74284863e-01 3.85078043e-01 -8.92108977e-01
1.91588986e+00 -1.49346602e+00 -1.23596795e-01 6.27212226e-01
4.42998230e-01 1.88940257e-01 -8.88930410e-02 5.56661427e-01
-3.63640100e-01 3.38284552e-01 5.14570773e-01 2.70829350e-01
8.63424912e-02 9.88862850e-03 -3.24423835e-02 1.61596432e-01
-1.12412497e-01 6.16835177e-01 -8.99833441e-01 -7.32517004e-01
2.54544228e-01 -2.59548128e-01 -7.86767781e-01 4.11771029e-01
-4.11776274e-01 -4.11671370e-01 -5.37942827e-01 8.13433588e-01
6.33389235e-01 -2.81333894e-01 2.36909822e-01 -1.89470515e-01
-1.14786625e-01 5.46972871e-01 -1.90632999e+00 1.48848903e+00
-5.80641627e-01 2.88620383e-01 -2.54230022e-01 -1.08946848e+00
1.17234647e+00 -4.65456257e-03 4.36888784e-01 -7.35240638e-01
-1.54034868e-02 -2.35081449e-01 2.20916029e-02 -9.50498223e-01
8.00782561e-01 4.71802801e-01 6.44150004e-02 2.36256644e-01
1.87650710e-01 -2.60476619e-01 3.88126194e-01 5.62690318e-01
1.33209634e+00 2.12762266e-01 2.76716173e-01 -2.64445394e-01
8.79027128e-01 5.21823764e-02 3.30144346e-01 7.24808812e-01
2.55757570e-01 -7.63595626e-02 5.32625794e-01 -5.26270390e-01
-7.72454441e-01 -8.98367703e-01 1.40006632e-01 1.36543655e+00
1.66513637e-01 -1.11258698e+00 -5.48490167e-01 -6.92133069e-01
7.62415975e-02 7.06750512e-01 4.83140387e-02 -5.96052185e-02
-4.21464056e-01 -1.24646328e-01 3.31928968e-01 2.83393621e-01
5.53047359e-01 -7.24497497e-01 1.47286937e-01 2.13365898e-01
-1.22803241e-01 -1.17302549e+00 -8.10956508e-02 -6.72224164e-02
-7.41967261e-01 -1.29159665e+00 5.62561229e-02 -1.12057483e+00
7.74374723e-01 7.64437690e-02 1.29764044e+00 8.27855945e-01
-4.96838093e-01 5.24078429e-01 -5.82931817e-01 -4.81962740e-01
-1.04569480e-01 2.62505472e-01 1.74467228e-02 -5.05828083e-01
3.40465039e-01 -2.87696600e-01 -1.95575152e-02 -8.71405676e-02
-6.39128923e-01 1.96276352e-01 1.26679853e-01 4.93765533e-01
1.99186453e-03 7.47787714e-01 2.97475785e-01 -1.22100163e+00
8.77438605e-01 -5.88891804e-01 -7.58276999e-01 8.59675467e-01
-8.80240917e-01 3.47609252e-01 8.94772828e-01 -2.13267952e-01
-1.20974815e+00 -2.21292898e-01 -7.13190660e-02 -2.53852922e-02
-6.71619847e-02 8.32005978e-01 -2.09359199e-01 -1.75662771e-01
4.79391783e-01 7.11500831e-03 -3.67239952e-01 -4.76181477e-01
2.84919083e-01 7.37597942e-01 1.57329738e-01 -1.18881321e+00
7.28850901e-01 -4.10969555e-01 1.29361823e-01 -7.93487132e-01
-7.27054894e-01 -7.00565875e-01 -5.39005399e-01 -3.67782205e-01
3.82285655e-01 -9.03430283e-01 -1.32032585e+00 2.44639948e-01
-1.17490256e+00 -4.12524909e-01 -1.45931430e-02 4.52925950e-01
-1.12368874e-01 5.51563561e-01 -8.69203448e-01 -8.28895450e-01
-8.16417709e-02 -6.63449407e-01 4.32300866e-01 2.93367594e-01
-3.44773494e-02 -1.34677637e+00 -2.74137318e-01 7.45647967e-01
3.13470066e-01 -4.29206938e-01 1.62801909e+00 -9.13206816e-01
-8.95141900e-01 -1.93492159e-01 -3.56919855e-01 1.58298817e-02
-4.04021069e-02 1.86342955e-01 -4.15746838e-01 2.25360423e-01
-3.45284253e-01 -3.33630085e-01 5.00939190e-01 -9.87738594e-02
1.88538718e+00 -3.64300340e-01 -1.10159315e-01 4.57999498e-01
1.52625251e+00 1.44100472e-01 9.86508369e-01 3.35136414e-01
7.66214550e-01 5.70800126e-01 6.54072881e-01 3.75474334e-01
7.68737674e-01 3.18888128e-01 9.67617407e-02 1.38766050e-01
-1.32030025e-01 -7.24097133e-01 7.74700195e-02 1.05358720e+00
-2.68175066e-01 -2.51810431e-01 -1.00734496e+00 3.52585614e-01
-1.85224164e+00 -9.12911177e-01 -5.15869021e-01 2.00583935e+00
1.11391497e+00 3.73101950e-01 -1.18238457e-01 2.58379191e-01
4.25580710e-01 -3.40785921e-01 1.22826882e-01 -6.01154089e-01
4.44655597e-01 5.62999964e-01 6.62590921e-01 8.86083007e-01
-7.24243939e-01 1.35005867e+00 6.26833105e+00 1.14717567e+00
-7.54631102e-01 -2.98129171e-01 3.18999588e-01 6.65185094e-01
-4.69319314e-01 -1.24819085e-01 -1.06638706e+00 1.08958177e-01
7.73820221e-01 -2.59178668e-01 7.30222642e-01 8.94799173e-01
5.69974631e-02 -1.21632181e-01 -1.07635689e+00 1.07837260e+00
1.05047889e-01 -1.47984350e+00 3.62172455e-01 -5.20759046e-01
5.06925404e-01 -7.65182793e-01 -4.15301681e-01 8.01932156e-01
3.19392204e-01 -1.18076408e+00 3.41950327e-01 7.68546641e-01
4.10952598e-01 -5.67799389e-01 6.39121711e-01 5.49146771e-01
-1.53433454e+00 -5.63633703e-02 -6.05581939e-01 -7.82626152e-01
-5.87164998e-01 3.51858824e-01 -1.02883720e+00 9.28739309e-01
3.25078487e-01 6.99878335e-01 -9.35152888e-01 1.13607180e+00
-1.98362321e-01 4.10049856e-01 -1.46023437e-01 -4.62524056e-01
-2.34450802e-01 -2.98589736e-01 -4.50429544e-02 1.07007730e+00
2.67296553e-01 3.73237759e-01 4.68875021e-01 8.54851604e-01
6.50783163e-03 5.81814766e-01 -6.34832263e-01 -1.25606030e-01
5.72856724e-01 1.32318485e+00 -6.36080503e-01 -5.15370071e-01
-6.26163542e-01 5.61904967e-01 7.32651591e-01 1.91713274e-01
-9.12543833e-01 -7.29965210e-01 2.35521957e-01 4.61753994e-01
-1.40331894e-01 -3.52530748e-01 -5.17532468e-01 -1.24234533e+00
-1.04857005e-01 -6.33997202e-01 6.72399282e-01 -8.66311908e-01
-1.28144133e+00 1.16836354e-01 2.93831825e-01 -8.65555406e-01
6.23278320e-02 -1.07635307e+00 -5.60995221e-01 8.32190096e-01
-1.52666926e+00 -1.02971053e+00 -3.99062246e-01 8.89311492e-01
4.35245156e-01 -5.60599387e-01 8.46528709e-01 5.02079248e-01
-3.99176478e-01 7.53211677e-01 -4.86535579e-01 3.89430612e-01
6.29512250e-01 -1.52776754e+00 -1.59827784e-01 5.55246115e-01
7.79937580e-02 7.05531299e-01 5.82723141e-01 -8.01756918e-01
-1.68508804e+00 -8.63858700e-01 1.29817379e+00 -2.76639819e-01
7.80722558e-01 -2.62934715e-01 -8.69765759e-01 6.37833834e-01
7.26804659e-02 -2.94144273e-01 7.39990890e-01 4.83545512e-01
-2.18795151e-01 -1.96150854e-01 -1.20666814e+00 5.16060472e-01
1.08971620e+00 -2.86695421e-01 -9.80394006e-01 7.20359564e-01
8.02680433e-01 -5.85690618e-01 -1.09040773e+00 1.37241066e-01
1.43315807e-01 -5.47065973e-01 1.03008521e+00 -9.93148088e-01
5.90771616e-01 -1.80397600e-01 1.45758897e-01 -1.17641783e+00
-4.89514649e-01 -4.35893834e-01 -2.79790014e-01 1.25750208e+00
5.31142294e-01 -4.35276210e-01 9.77519274e-01 9.36626971e-01
-1.74222544e-01 -6.78823590e-01 -2.71271884e-01 -6.50130987e-01
-2.10129008e-01 -5.56891561e-01 7.26897657e-01 1.41540122e+00
7.63074994e-01 4.51080859e-01 4.78738770e-02 1.61851272e-01
3.08601737e-01 1.03737533e-01 5.67684829e-01 -1.36425674e+00
-2.24838957e-01 -2.95305312e-01 -5.24504662e-01 -1.02892852e+00
3.38672876e-01 -1.32779741e+00 -6.86120272e-01 -1.95128393e+00
-1.91581368e-01 -7.64081061e-01 -1.80467367e-01 6.87145054e-01
-7.87802339e-02 -4.03146595e-01 -2.45587714e-02 -5.70293784e-01
-7.62651086e-01 1.14302419e-01 1.47824740e+00 -2.18787149e-01
-1.83928296e-01 -2.35475488e-02 -5.39790690e-01 7.85740793e-01
6.98577046e-01 -1.54235110e-01 -5.33445835e-01 -2.96419442e-01
8.07845652e-01 2.57328033e-01 -8.33627433e-02 -9.25137222e-01
7.94627786e-01 -6.33323967e-01 4.34511006e-01 -4.59374607e-01
-5.83209470e-02 -1.20084250e+00 -2.99239904e-01 2.36571610e-01
-4.92624730e-01 2.72637725e-01 9.98589024e-03 1.44512951e-01
-6.66707158e-02 -6.53922081e-01 1.33172154e-01 -3.67194146e-01
-1.02593446e+00 1.56858832e-01 -1.84643894e-01 1.94292003e-03
1.05449271e+00 -9.21635032e-02 -5.65129995e-01 -2.21978247e-01
-6.27728403e-01 4.22725379e-01 -8.11793506e-02 4.35104698e-01
7.43799686e-01 -1.17325401e+00 -3.51523072e-01 3.93950343e-01
-7.20618945e-03 -1.06470689e-01 1.00750670e-01 1.80784076e-01
-1.03203011e+00 3.91220391e-01 -3.03501040e-01 7.93450624e-02
-1.54719877e+00 5.82412779e-01 1.50048748e-01 -6.34855986e-01
-3.79526973e-01 6.90686524e-01 -7.50092089e-01 -7.19456732e-01
7.32264698e-01 -8.26604724e-01 -7.48737752e-01 -4.54168208e-02
5.09276748e-01 5.99276066e-01 2.60566443e-01 2.06045747e-01
2.16220692e-03 2.73406565e-01 -4.30483669e-02 6.28472984e-01
1.24339354e+00 1.45449042e-01 -5.58437109e-01 3.49489331e-01
7.27345049e-01 2.36963183e-01 -1.20306626e-01 -1.13022849e-01
2.71453429e-02 -6.01909697e-01 -2.48030182e-02 -9.81931746e-01
-8.25647175e-01 4.67979521e-01 2.81740576e-01 5.02341986e-01
7.76588678e-01 -2.43773639e-01 6.55626118e-01 8.80954862e-01
4.02409494e-01 -1.46170235e+00 5.62235638e-02 8.94858241e-01
4.69783872e-01 -1.10406184e+00 4.36423391e-01 -1.07494307e+00
-1.90950051e-01 1.57263696e+00 1.09937644e+00 -2.20581070e-02
8.20098102e-01 4.22694296e-01 -3.86390328e-01 -2.44670630e-01
-8.21243703e-01 -1.21938601e-01 5.57708740e-01 3.81224275e-01
8.67543697e-01 1.24502830e-01 -6.17538512e-01 9.32331502e-01
-4.69178498e-01 2.55752027e-01 7.15282261e-01 8.76110554e-01
-6.83198869e-01 -1.37485826e+00 -3.84699047e-01 6.58461809e-01
-3.44956875e-01 -3.17043781e-01 -1.27072588e-01 7.66880929e-01
3.73645961e-01 8.30125034e-01 1.07721224e-01 -3.30259651e-01
4.61970985e-01 4.05820847e-01 9.08390641e-01 -9.40688789e-01
-7.10417807e-01 -6.79777205e-01 3.17197353e-01 -3.68834972e-01
1.87219568e-02 8.48051235e-02 -1.48910487e+00 -6.24774218e-01
-2.31826544e-01 2.43674234e-01 5.84010839e-01 9.99326527e-01
-1.55277863e-01 9.14336443e-01 3.83772284e-01 4.54597436e-02
-4.50481445e-01 -7.59223640e-01 -7.09854662e-01 4.32522416e-01
-2.23420143e-01 -5.10295630e-01 -4.82591875e-02 1.03211053e-01] | [9.715221405029297, 7.45811653137207] |
e4bab3c2-9699-4a54-8169-5fc81b180e10 | sign-language-transformers-joint-end-to-end | 2003.13830 | null | https://arxiv.org/abs/2003.13830v1 | https://arxiv.org/pdf/2003.13830v1.pdf | Sign Language Transformers: Joint End-to-end Sign Language Recognition and Translation | Prior work on Sign Language Translation has shown that having a mid-level sign gloss representation (effectively recognizing the individual signs) improves the translation performance drastically. In fact, the current state-of-the-art in translation requires gloss level tokenization in order to work. We introduce a novel transformer based architecture that jointly learns Continuous Sign Language Recognition and Translation while being trainable in an end-to-end manner. This is achieved by using a Connectionist Temporal Classification (CTC) loss to bind the recognition and translation problems into a single unified architecture. This joint approach does not require any ground-truth timing information, simultaneously solving two co-dependant sequence-to-sequence learning problems and leads to significant performance gains. We evaluate the recognition and translation performances of our approaches on the challenging RWTH-PHOENIX-Weather-2014T (PHOENIX14T) dataset. We report state-of-the-art sign language recognition and translation results achieved by our Sign Language Transformers. Our translation networks outperform both sign video to spoken language and gloss to spoken language translation models, in some cases more than doubling the performance (9.58 vs. 21.80 BLEU-4 Score). We also share new baseline translation results using transformer networks for several other text-to-text sign language translation tasks. | ['Richard Bowden', 'Simon Hadfield', 'Oscar Koller', 'Necati Cihan Camgoz'] | 2020-03-30 | sign-language-transformers-joint-end-to-end-1 | http://openaccess.thecvf.com/content_CVPR_2020/html/Camgoz_Sign_Language_Transformers_Joint_End-to-End_Sign_Language_Recognition_and_Translation_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Camgoz_Sign_Language_Transformers_Joint_End-to-End_Sign_Language_Recognition_and_Translation_CVPR_2020_paper.pdf | cvpr-2020-6 | ['sign-language-translation'] | ['computer-vision'] | [ 4.18683976e-01 -3.32269907e-01 -2.87941962e-01 -4.31094527e-01
-1.38945997e+00 -7.56545305e-01 8.77005279e-01 -9.94104743e-01
-7.03535914e-01 4.92249101e-01 5.60638785e-01 -4.05010611e-01
4.25870895e-01 -2.36535162e-01 -8.88527572e-01 -6.40889108e-01
1.50087059e-01 8.56867373e-01 1.23302758e-01 -3.16680163e-01
-3.22865427e-01 5.55271581e-02 -1.26562846e+00 7.50239909e-01
8.10145259e-01 8.62249970e-01 -2.08735541e-01 1.01837111e+00
-3.34548503e-02 8.94661903e-01 -4.39011484e-01 -5.16422451e-01
4.23347414e-01 -8.00881028e-01 -7.81737506e-01 -1.63292632e-01
1.15877998e+00 -7.27536798e-01 -5.32397747e-01 3.76314551e-01
9.04478192e-01 -2.88508892e-01 6.28750324e-01 -1.03724825e+00
-7.48296082e-01 4.17716295e-01 -1.73915491e-01 -2.65635490e-01
2.31645375e-01 6.31766200e-01 1.17624009e+00 -9.81871188e-01
9.42779839e-01 1.17111313e+00 6.79033637e-01 7.76035368e-01
-8.83796215e-01 -7.50376523e-01 9.40947905e-02 2.15799227e-01
-9.79810894e-01 -4.93624866e-01 5.15009165e-02 -2.11228475e-01
1.56731975e+00 1.71234191e-03 9.13383424e-01 1.53251910e+00
5.87085783e-02 1.32708418e+00 1.30766976e+00 -4.69917685e-01
-2.57174283e-01 -6.56077206e-01 -2.11677179e-01 9.22004819e-01
-1.39817476e-01 5.27996659e-01 -9.41585958e-01 3.75751108e-01
5.71058154e-01 -2.84024924e-01 -1.64919138e-01 -1.99102268e-01
-1.66040194e+00 3.79553676e-01 3.49699497e-01 3.00420374e-01
-2.68531799e-01 8.43139052e-01 7.29936302e-01 8.94032776e-01
1.82884276e-01 3.77164618e-03 -5.30620754e-01 -6.05621994e-01
-1.05021334e+00 1.18064344e-01 7.67184258e-01 9.86639500e-01
9.73589893e-04 3.65137160e-01 -6.17228687e-01 5.88126600e-01
3.43576849e-01 1.30562830e+00 5.35796642e-01 -6.54992521e-01
8.42260361e-01 1.09578229e-01 -1.18004225e-01 5.24361059e-02
-1.24315776e-01 -3.06096405e-01 -4.55350757e-01 3.09525669e-01
7.23583162e-01 -1.89711034e-01 -1.84307468e+00 1.72472394e+00
-2.64290869e-01 1.46902293e-01 6.97626844e-02 1.03924870e+00
6.83936775e-01 4.64071572e-01 6.97504804e-02 2.79213756e-01
1.32010210e+00 -1.22024620e+00 -6.34890556e-01 -1.85075402e-01
7.95575142e-01 -9.43298936e-01 1.17495501e+00 2.53621370e-01
-9.76189435e-01 -8.19997713e-02 -8.37872982e-01 -1.89581186e-01
-2.30597109e-01 5.38856506e-01 4.43216950e-01 3.75430346e-01
-1.28722107e+00 2.98620462e-01 -1.28303862e+00 -8.38879704e-01
3.29607040e-01 5.06855845e-01 -2.94295073e-01 -2.59059429e-01
-9.60912526e-01 1.35745692e+00 -2.15967879e-01 2.28307456e-01
-1.03152645e+00 -5.68243086e-01 -5.63977838e-01 -3.39138955e-01
6.47166967e-02 -9.63747084e-01 1.73312914e+00 -8.97310913e-01
-2.13308501e+00 1.09850454e+00 -3.88512105e-01 -5.16226947e-01
1.20088375e+00 -4.95819271e-01 -3.20116460e-01 5.20194992e-02
-2.00444713e-01 7.61699677e-01 8.93463790e-01 -6.99320257e-01
-7.21034825e-01 2.98614036e-02 -3.97688717e-01 3.18949580e-01
7.20097870e-02 4.64040756e-01 -5.45591891e-01 -5.86058855e-01
-1.54031254e-02 -1.15730107e+00 2.83659428e-01 3.93439680e-01
-9.83563438e-03 -1.53169394e-01 7.09146082e-01 -1.04410040e+00
7.25184739e-01 -1.62382269e+00 3.89684081e-01 8.56579375e-03
-7.02059790e-02 5.43240011e-01 -5.87601185e-01 5.87227881e-01
3.26639444e-01 -2.41033569e-01 -4.33005422e-01 -7.00649261e-01
2.82974929e-01 5.85670054e-01 -6.08687639e-01 4.96067584e-01
1.72497317e-01 1.59799099e+00 -8.37683201e-01 -3.67995590e-01
2.35714272e-01 6.94209635e-01 -4.17382658e-01 -1.42766126e-02
-2.99488842e-01 5.65388978e-01 -1.21017270e-01 1.05657375e+00
-4.53183800e-03 -2.56929174e-02 2.31726319e-01 -5.49202263e-02
-9.90221724e-02 6.96731925e-01 -2.59563059e-01 2.07121277e+00
-7.17482507e-01 1.09068656e+00 -1.81047946e-01 -7.09831417e-01
6.07732415e-01 6.82389855e-01 4.54909891e-01 -1.08918452e+00
2.29585439e-01 8.94815683e-01 -1.83037519e-02 -5.11375010e-01
1.40854076e-01 -3.64050597e-01 -1.24088950e-01 6.71379030e-01
1.83913782e-01 -2.29009256e-01 1.93670124e-01 -2.44728893e-01
1.16154790e+00 6.81561172e-01 -2.38168910e-01 2.19554871e-01
8.03597458e-03 2.16424093e-01 1.73791304e-01 6.93001091e-01
-2.46959090e-01 7.18304753e-01 -5.40630221e-02 -3.35858732e-01
-1.21255958e+00 -1.28797054e+00 4.00882125e-01 1.15967941e+00
-4.14238036e-01 -8.44123438e-02 -4.47532356e-01 -8.31098557e-01
-1.01977199e-01 5.56132495e-01 -4.83963072e-01 -8.81125480e-02
-1.33492041e+00 -3.03783953e-01 1.37188649e+00 8.10996354e-01
6.57774746e-01 -1.06208742e+00 -6.57075942e-01 1.98559687e-01
-6.08846009e-01 -1.31704509e+00 -1.00043082e+00 6.94053099e-02
-8.15610290e-01 -6.96101606e-01 -1.29014826e+00 -1.12144756e+00
3.29099864e-01 -1.30423933e-01 1.02498519e+00 -1.63433135e-01
-1.29237473e-01 4.60516751e-01 -5.35103679e-01 -1.94927752e-01
-5.17851293e-01 1.60278320e-01 -3.29618864e-02 -2.12903038e-01
4.46372539e-01 -4.25110549e-01 -3.48354012e-01 4.33949679e-01
-6.19023561e-01 5.41104004e-02 9.91960347e-01 1.19469357e+00
3.83024991e-01 -1.20011306e+00 1.90380625e-02 1.41725957e-01
4.37141657e-01 4.44709837e-01 -4.48061198e-01 5.95757246e-01
-5.95013857e-01 4.79800642e-01 3.63514900e-01 -6.86359823e-01
-6.65048957e-01 1.00615270e-01 -2.70354990e-02 -4.95216876e-01
2.32209668e-01 4.38437790e-01 3.47796589e-01 -1.38320670e-01
5.27829289e-01 8.58524740e-01 2.88952410e-01 -4.09952432e-01
6.89071655e-01 6.62553608e-01 8.15351129e-01 -5.32993376e-01
9.05211687e-01 4.38640773e-01 -1.44752458e-01 -4.59152520e-01
-5.20366848e-01 -2.72542506e-01 -7.33608246e-01 -2.41078824e-01
8.47169101e-01 -1.00538993e+00 -7.56514013e-01 8.47712100e-01
-1.11866021e+00 -8.63588035e-01 -3.42855871e-01 6.93712652e-01
-9.22804534e-01 2.70051360e-01 -7.36791551e-01 -3.93566161e-01
-6.87123835e-01 -1.06686246e+00 1.42872274e+00 -5.38193345e-01
-3.42239767e-01 -5.52845001e-01 2.82064170e-01 3.87548506e-01
8.23859215e-01 1.36591077e-01 4.51224983e-01 -9.21198353e-02
-8.18210542e-01 -3.26415837e-01 -3.19900692e-01 5.67434669e-01
-7.83768855e-03 -1.92588761e-01 -7.53532827e-01 -3.76522213e-01
-6.64290488e-01 -6.32521808e-01 1.32880700e+00 2.31770799e-01
-2.63182987e-02 -3.82986486e-01 -9.48338509e-02 7.39266753e-01
1.02646267e+00 -6.32546395e-02 7.56524622e-01 1.31253272e-01
7.19892025e-01 1.69117004e-01 3.98814529e-01 -3.68569978e-02
5.55453479e-01 1.08583057e+00 -5.54999225e-02 -2.29751155e-01
-9.75519180e-01 -4.91504580e-01 1.12786710e+00 9.79750574e-01
-3.65989983e-01 -2.35346347e-01 -9.88636792e-01 7.57952690e-01
-1.96334493e+00 -1.06602025e+00 1.03113316e-01 1.92039955e+00
1.00165844e+00 -3.28756571e-01 2.19909549e-01 -2.81297248e-02
5.85906729e-02 8.47067088e-02 -4.71399128e-01 -4.12681282e-01
-3.36545378e-01 6.46450639e-01 7.64886618e-01 5.79554141e-01
-8.90786290e-01 1.52944744e+00 6.48109913e+00 5.16265571e-01
-1.73440492e+00 2.66399235e-01 -3.06394964e-01 -5.78461826e-01
8.59204098e-04 -8.26059580e-02 -5.09178936e-01 9.17017758e-02
9.11119282e-01 1.39831021e-01 7.43346334e-01 2.20908418e-01
3.10707718e-01 4.26892787e-01 -1.14742863e+00 1.07276762e+00
2.64246821e-01 -1.11758518e+00 3.36985618e-01 1.10035151e-01
7.30602384e-01 8.68088126e-01 1.99544087e-01 4.83132601e-01
7.09265828e-01 -1.09352779e+00 1.21424198e+00 4.58129495e-01
1.58015287e+00 -1.19558871e-02 6.33674324e-01 -1.52734831e-01
-1.42619538e+00 2.03500256e-01 4.15402561e-01 1.48045952e-02
6.80982769e-01 -1.15016133e-01 -9.78650749e-01 4.25110668e-01
4.00406092e-01 8.88435185e-01 -1.14788763e-01 9.39028263e-01
-7.84748971e-01 1.01809645e+00 -7.56908894e-01 -1.92261517e-01
6.93205714e-01 2.61142775e-02 5.32664120e-01 1.44209445e+00
6.01871490e-01 -2.57844448e-01 6.21551275e-03 5.44396818e-01
-1.92102164e-01 -1.00779004e-01 -4.29804474e-01 -2.53602415e-01
-2.68296804e-02 3.46696079e-01 -2.64114022e-01 -6.33244336e-01
-4.31496084e-01 1.68099868e+00 4.59175967e-02 6.42978311e-01
-8.71791661e-01 -1.76549211e-01 9.60676193e-01 -1.63893208e-01
6.80375874e-01 -5.44866145e-01 -2.33722270e-01 -1.52408803e+00
4.35207576e-01 -1.15516281e+00 2.24444687e-01 -6.20191395e-01
-1.09487224e+00 4.82937187e-01 -2.84123540e-01 -1.64423084e+00
-6.94113731e-01 -7.63062418e-01 -2.78531075e-01 8.04803252e-01
-1.76280928e+00 -1.96288073e+00 -2.04080150e-01 6.69344008e-01
3.43848258e-01 -8.79743621e-02 9.18648064e-01 5.68929136e-01
8.26723948e-02 1.16106260e+00 2.02152371e-01 5.60101271e-01
1.12201130e+00 -1.01764321e+00 9.01414812e-01 9.87938225e-01
4.67204452e-01 3.42095912e-01 4.37493443e-01 -5.45820177e-01
-1.50537014e+00 -9.70835984e-01 1.68432224e+00 -7.75621533e-01
6.77463412e-01 -4.01695967e-01 -1.27601236e-01 9.41736817e-01
2.93486118e-01 -1.17023170e-01 -2.88928077e-02 -1.45107180e-01
-9.63317156e-01 -5.52736074e-02 -7.89010644e-01 8.28343868e-01
1.64299548e+00 -1.01136410e+00 -6.15145683e-01 3.65814567e-01
5.00874341e-01 -5.87566555e-01 -5.37299991e-01 3.11412036e-01
1.30896699e+00 -3.48063290e-01 7.82626092e-01 -7.71028817e-01
4.08299148e-01 -3.85177702e-01 -2.97980607e-01 -1.19550765e+00
2.19416559e-01 -9.71191645e-01 -5.72163947e-02 4.86731559e-01
6.40183628e-01 -7.24690080e-01 7.02397823e-01 2.21436441e-01
-4.05247718e-01 -5.98756969e-01 -1.28831148e+00 -1.38291764e+00
3.51309419e-01 -5.53483546e-01 2.43973061e-01 6.51685417e-01
6.48481250e-02 2.45517701e-01 -8.72266591e-01 -2.75304437e-01
5.14177799e-01 2.13146612e-01 9.01230037e-01 -6.11162543e-01
-3.72287691e-01 -8.13451409e-01 -4.32517320e-01 -1.58928370e+00
-2.73666810e-02 -1.29002738e+00 3.87605011e-01 -1.89879227e+00
-1.99449867e-01 1.34859845e-01 -4.48415615e-02 1.16862297e+00
3.26458991e-01 3.92398030e-01 5.11342049e-01 4.02878225e-01
-3.94126922e-01 7.77203918e-01 1.38455045e+00 -4.68354672e-01
-3.77788618e-02 -2.64658868e-01 5.05171251e-03 1.00668401e-01
5.19048810e-01 -3.51960659e-01 7.89375901e-02 -1.25051987e+00
-3.49813253e-01 -1.67195335e-01 3.75007838e-01 -8.87790680e-01
2.43922234e-01 1.98865056e-01 -2.56172210e-01 -4.55150574e-01
4.92523879e-01 -5.53744078e-01 -2.05626190e-01 8.39920282e-01
-3.59825373e-01 4.56159301e-02 1.38798758e-01 1.77457079e-01
-2.76816934e-01 6.80000603e-01 6.99429154e-01 5.73839694e-02
-6.29910231e-01 2.66109198e-01 -5.26298821e-01 1.30364388e-01
4.81225193e-01 -1.91313669e-01 -3.70853812e-01 -6.07863069e-01
-5.01255035e-01 2.23926157e-01 2.39545822e-01 7.29834676e-01
5.80554843e-01 -1.34240353e+00 -1.19099998e+00 8.61663073e-02
3.94842744e-01 -5.54222643e-01 -3.89239132e-01 1.18227088e+00
-6.55173719e-01 8.38125885e-01 -2.68628985e-01 -7.51209080e-01
-1.48246062e+00 -2.70965815e-01 5.92469871e-01 -5.44640720e-01
-8.20932150e-01 9.22513485e-01 -4.69171852e-01 -7.69066155e-01
4.65256095e-01 -1.04347372e+00 6.75706267e-01 -2.19726831e-01
2.66745150e-01 9.66601670e-02 5.46312630e-02 -5.99683225e-01
-4.89257693e-01 1.03918529e+00 -4.09086421e-03 -8.14421356e-01
1.13564491e+00 2.00110555e-01 2.56146520e-01 1.24382049e-01
1.16378665e+00 -4.22601968e-01 -1.11523414e+00 -5.42264760e-01
-2.82009244e-02 -6.96105063e-02 -1.67644560e-01 -1.64468455e+00
-9.80854332e-01 9.45246518e-01 6.78654253e-01 -1.10923278e+00
9.43916142e-01 -5.82381263e-02 1.34647620e+00 8.56756091e-01
6.09730363e-01 -9.39549863e-01 -2.96969377e-02 1.14432108e+00
1.03778529e+00 -1.24602509e+00 -3.98577362e-01 1.94166407e-01
-6.68431818e-01 9.67650354e-01 1.03139400e-01 -6.73281774e-02
1.65134639e-01 4.08572018e-01 7.47205079e-01 -2.68423110e-02
-7.39089906e-01 -5.54053068e-01 5.06232798e-01 5.19609094e-01
4.39730883e-01 2.24927202e-01 -3.58646870e-01 -5.93475327e-02
-2.62745351e-01 5.26651025e-01 -5.52207232e-02 1.04892612e+00
-5.60491644e-02 -1.31879103e+00 8.49150866e-03 1.92277357e-01
-6.23837374e-02 -4.63061273e-01 -8.16029191e-01 7.75129735e-01
-2.45268822e-01 6.62071109e-01 -2.15226129e-01 -5.36491454e-01
6.11347437e-01 5.29970467e-01 9.56833303e-01 -1.23085111e-01
-7.75252998e-01 -5.10830199e-03 5.02269208e-01 -8.02861392e-01
-4.13321525e-01 -8.20106089e-01 -1.41232860e+00 -2.66685665e-01
4.04459164e-02 -4.99122441e-01 6.99533939e-01 1.22636890e+00
2.96613038e-01 2.94091523e-01 -1.60264000e-01 -6.77593112e-01
-9.60844100e-01 -1.12609863e+00 2.77796183e-02 3.81183028e-01
5.55539012e-01 -3.42157125e-01 -1.76065296e-01 2.30670616e-01] | [9.23388957977295, -6.559328556060791] |
1e962209-4713-4360-b1c4-f52568fdb20b | generalizing-gaze-estimation-with-outlier | 2107.13780 | null | https://arxiv.org/abs/2107.13780v2 | https://arxiv.org/pdf/2107.13780v2.pdf | Generalizing Gaze Estimation with Outlier-guided Collaborative Adaptation | Deep neural networks have significantly improved appearance-based gaze estimation accuracy. However, it still suffers from unsatisfactory performance when generalizing the trained model to new domains, e.g., unseen environments or persons. In this paper, we propose a plug-and-play gaze adaptation framework (PnP-GA), which is an ensemble of networks that learn collaboratively with the guidance of outliers. Since our proposed framework does not require ground-truth labels in the target domain, the existing gaze estimation networks can be directly plugged into PnP-GA and generalize the algorithms to new domains. We test PnP-GA on four gaze domain adaptation tasks, ETH-to-MPII, ETH-to-EyeDiap, Gaze360-to-MPII, and Gaze360-to-EyeDiap. The experimental results demonstrate that the PnP-GA framework achieves considerable performance improvements of 36.9%, 31.6%, 19.4%, and 11.8% over the baseline system. The proposed framework also outperforms the state-of-the-art domain adaptation approaches on gaze domain adaptation tasks. | ['Feng Lu', 'Haofei Wang', 'Ruicong Liu', 'Yunfei Liu'] | 2021-07-29 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Liu_Generalizing_Gaze_Estimation_With_Outlier-Guided_Collaborative_Adaptation_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Liu_Generalizing_Gaze_Estimation_With_Outlier-Guided_Collaborative_Adaptation_ICCV_2021_paper.pdf | iccv-2021-1 | ['gaze-estimation'] | ['computer-vision'] | [-1.77043974e-02 4.30876063e-03 1.28354564e-01 -5.64644814e-01
-2.51215398e-01 -1.66992873e-01 1.90694943e-01 -7.22621858e-01
-3.58159930e-01 9.39379692e-01 -2.76972890e-01 -7.86844734e-03
1.19161911e-01 -1.78964481e-01 -7.22121000e-01 -7.83951879e-01
3.56908768e-01 2.08558500e-01 1.91405535e-01 -1.48210451e-01
1.58436164e-01 1.05931861e-02 -1.89379311e+00 -5.40774092e-02
1.13930023e+00 9.79646981e-01 -1.22805797e-01 4.70632493e-01
2.92811126e-01 2.39352837e-01 -5.48790395e-01 -2.60731459e-01
1.28406808e-01 -4.69977051e-01 -4.41950917e-01 -1.47535771e-01
8.16235304e-01 -4.87662613e-01 6.18870221e-02 1.09639287e+00
6.47672832e-01 3.34217846e-01 7.67740905e-01 -1.79851770e+00
-1.04241323e+00 -3.63291860e-01 -1.21163142e+00 4.54920419e-02
3.95849675e-01 3.53518367e-01 3.86302441e-01 -9.84922588e-01
6.52728438e-01 1.22496474e+00 4.13488120e-01 1.21311986e+00
-9.46630657e-01 -1.40051293e+00 5.03370702e-01 2.89551079e-01
-1.29960287e+00 -6.93524122e-01 6.30348980e-01 -4.13552880e-01
7.15475738e-01 -6.72864020e-02 1.62371457e-01 1.30002654e+00
-1.72028020e-01 8.59698355e-01 1.35047829e+00 -4.22290981e-01
-1.60731878e-02 2.06581995e-01 2.22672105e-01 4.77133363e-01
2.68365979e-01 2.89577901e-01 -1.06221437e+00 2.44742632e-01
6.17067277e-01 -6.75675422e-02 -4.51225042e-01 -3.43786389e-01
-9.28295851e-01 3.59627843e-01 6.87767804e-01 -3.04921567e-01
-2.05645785e-01 -1.62834555e-01 1.49596140e-01 1.48067623e-01
8.57317150e-01 1.88393399e-01 -4.24220562e-01 -1.58884928e-01
-8.03293109e-01 3.67120504e-01 3.15718234e-01 1.21114600e+00
8.56109142e-01 -1.39271066e-01 -7.69408569e-02 8.60912323e-01
6.73783183e-01 8.21773529e-01 4.32413846e-01 -8.89604568e-01
4.43661332e-01 5.45263112e-01 4.07144636e-01 -6.55146241e-01
-5.41785717e-01 -9.98078734e-02 -5.33100665e-01 6.43184006e-01
5.55043876e-01 -6.19511962e-01 -1.18456423e+00 2.04108143e+00
5.65647066e-01 2.97586203e-01 -3.34813036e-02 1.09092104e+00
1.12513041e+00 5.51501334e-01 2.71455973e-01 -1.81985244e-01
1.04001200e+00 -1.24545324e+00 -7.09021032e-01 -4.11998570e-01
2.80411750e-01 -4.34022725e-01 1.31230843e+00 4.24023479e-01
-1.01333034e+00 -7.79771388e-01 -9.64245439e-01 -1.13306567e-01
-2.35722482e-01 2.90898860e-01 2.50622660e-01 6.42104864e-01
-1.34576547e+00 8.16576406e-02 -5.44512451e-01 -8.07386518e-01
6.00071788e-01 6.89141214e-01 -3.41280580e-01 -3.28830385e-04
-7.31707990e-01 6.58671260e-01 3.24743807e-01 1.25419319e-01
-7.01537907e-01 -6.37599051e-01 -7.42358744e-01 -5.00508584e-02
3.03248674e-01 -8.26060653e-01 1.44725621e+00 -1.31946063e+00
-1.82577026e+00 1.04771924e+00 -8.91717553e-01 1.48420297e-02
3.34984690e-01 -5.17312288e-01 -7.78564870e-01 -7.81714767e-02
-1.25012791e-03 1.03941643e+00 9.42453563e-01 -1.27440596e+00
-1.01768458e+00 -5.80134690e-01 1.97815210e-01 4.11340803e-01
-3.51651549e-01 2.36250266e-01 -4.55031931e-01 -6.04771823e-02
-2.44116038e-01 -1.15077817e+00 5.46508849e-01 2.13752151e-01
-3.53235215e-01 -6.50676787e-01 1.16242337e+00 -5.36796868e-01
1.04306257e+00 -2.25892687e+00 -9.98765137e-03 -1.09616533e-01
4.76461112e-01 5.28573990e-01 -1.11288749e-01 -3.70114855e-02
-2.15974957e-01 -1.99755073e-01 1.81962341e-01 -6.92214251e-01
1.87863968e-02 -8.52313861e-02 1.55388087e-01 2.66399652e-01
1.83956355e-01 7.79055357e-01 -8.73866916e-01 -3.00032347e-01
5.72137162e-03 3.69836777e-01 -4.60053146e-01 2.56111473e-01
-1.24076374e-01 7.15922356e-01 -1.15461983e-01 7.15739012e-01
9.09540415e-01 -4.80745316e-01 -2.79136091e-01 1.42245814e-01
4.68156347e-03 1.54414773e-02 -6.35739505e-01 1.70864725e+00
-2.02640742e-01 1.06765556e+00 -1.74965888e-01 -2.47635841e-01
9.64644670e-01 8.46694484e-02 7.07099065e-02 -7.44257689e-01
3.59546363e-01 -3.06956135e-02 1.51029274e-01 -5.63120723e-01
6.64490283e-01 5.99284470e-02 2.16207385e-01 4.37614083e-01
3.57124567e-01 7.15514660e-01 9.12980661e-02 -2.25938395e-01
5.35952508e-01 5.03005862e-01 9.29700285e-02 -1.26414075e-01
7.63126433e-01 -3.46749753e-01 6.07260227e-01 2.77423829e-01
-8.08455288e-01 7.52816975e-01 3.72662604e-01 -1.96396828e-01
-6.40161216e-01 -9.81922448e-01 5.37128747e-02 1.82625830e+00
3.54637116e-01 -2.59126420e-03 -1.10280442e+00 -9.60922360e-01
-2.03745276e-01 7.43747413e-01 -9.58568633e-01 -2.66728550e-01
-1.47649407e-01 -6.17924750e-01 2.90660232e-01 5.59362471e-01
8.97132397e-01 -1.15408850e+00 -2.49588028e-01 -3.09740841e-01
-2.19815299e-01 -8.26829612e-01 -5.03341675e-01 -4.78460222e-01
-5.66272497e-01 -1.17683136e+00 -1.04261756e+00 -8.39943469e-01
8.70810628e-01 4.36356306e-01 9.78678167e-01 -1.36597633e-01
4.79763091e-01 2.86262393e-01 -2.20916793e-01 -8.91442835e-01
5.98064549e-02 2.79075980e-01 3.17495137e-01 2.96868175e-01
1.29489851e+00 -3.48740548e-01 -8.90554130e-01 7.12885976e-01
-1.89019457e-01 -7.61036202e-02 3.64713222e-01 9.34541941e-01
3.77479166e-01 -6.02532506e-01 5.89653373e-01 -8.39513183e-01
7.20505953e-01 -3.57864022e-01 -5.76921821e-01 3.68619233e-01
-8.65264237e-01 -2.64698714e-01 3.43912654e-02 -5.41991353e-01
-1.63747680e+00 -1.80185899e-01 1.23690069e-01 -5.89498341e-01
-5.19460082e-01 1.65182687e-02 -4.00631636e-01 -3.38634938e-01
1.07937908e+00 -9.61395428e-02 -1.90702770e-02 -4.25265461e-01
2.52577305e-01 1.01805007e+00 6.94253743e-01 -3.11918080e-01
7.68692732e-01 3.12264144e-01 -2.71859914e-01 -5.12632489e-01
-8.42064202e-01 -3.43378723e-01 -6.81390345e-01 -3.39435369e-01
8.53209674e-01 -1.02835524e+00 -9.55473661e-01 1.00363815e+00
-9.12478864e-01 -4.29886580e-01 1.92742571e-01 2.74878204e-01
-3.19052309e-01 1.20383188e-01 4.02135327e-02 -7.22226620e-01
-5.01606345e-01 -9.69321430e-01 1.00919390e+00 1.10484838e+00
-1.87211365e-01 -7.69707441e-01 9.27381665e-02 2.41197124e-01
2.73632258e-01 1.26432046e-01 2.50878423e-01 -5.55008769e-01
-3.94459188e-01 5.28432429e-02 -6.00415230e-01 3.29798281e-01
1.29293591e-01 2.37888232e-01 -1.43678701e+00 -4.70684916e-01
-4.75896299e-01 -4.89465117e-01 5.65845907e-01 5.31055570e-01
9.90408063e-01 6.00030981e-02 -6.66723549e-01 9.90355611e-01
9.48520064e-01 3.54433149e-01 6.93720758e-01 6.63907528e-01
7.22545981e-01 5.88392556e-01 8.36043477e-01 1.81807965e-01
7.09729612e-01 5.16664326e-01 5.09780824e-01 -8.69292840e-02
-1.61995098e-01 -2.40299985e-01 2.92601347e-01 3.59578617e-02
-3.71681005e-01 -4.75213796e-01 -9.46707964e-01 5.79940796e-01
-1.81464672e+00 -6.87120557e-01 -7.60982260e-02 2.08315992e+00
6.84493065e-01 -1.62376165e-02 4.75323647e-01 -3.64439011e-01
8.21850836e-01 -1.36687115e-01 -1.07294548e+00 -2.51541376e-01
2.30631441e-01 4.87740226e-02 2.07245573e-01 1.18569043e-02
-1.07211733e+00 9.15733218e-01 5.60413027e+00 2.31002793e-01
-1.30400538e+00 2.72246808e-01 3.31149578e-01 -4.37697500e-01
3.98784012e-01 -4.76826280e-01 -1.11981022e+00 8.14823031e-01
9.58550870e-01 -7.29066432e-02 4.32479739e-01 9.45344329e-01
1.10852709e-02 -1.58814847e-01 -1.11755323e+00 1.33384430e+00
3.95596921e-01 -4.48494643e-01 -5.47052145e-01 -1.34203508e-02
8.58042240e-01 2.41385117e-01 6.62379920e-01 4.04076189e-01
1.26003504e-01 -9.29092109e-01 3.25655878e-01 6.08466566e-01
1.24724948e+00 -6.61308706e-01 6.60260260e-01 9.62527171e-02
-5.99512637e-01 4.63043600e-02 -3.49798918e-01 5.01177572e-02
4.42723036e-02 -2.71315008e-01 -8.93425882e-01 1.97244659e-01
1.19501173e+00 9.24059749e-01 -8.25507045e-01 1.23893797e+00
-6.15921497e-01 6.36701584e-01 -3.09957236e-01 -1.83829442e-01
-3.29042450e-02 2.92719603e-02 4.11237985e-01 6.88013971e-01
3.86331707e-01 -7.01999590e-02 -5.99167347e-01 9.31371450e-01
-2.94459254e-01 -3.19598049e-01 -3.37722450e-01 3.83767456e-01
5.51911473e-01 1.01202416e+00 -1.50767982e-03 -1.45449281e-01
-5.30984104e-01 9.70203698e-01 3.93407971e-01 1.00295258e+00
-7.95990229e-01 -6.08224511e-01 1.13983715e+00 6.49727434e-02
3.00511688e-01 1.52137786e-01 -2.18220264e-01 -1.11956501e+00
6.20567761e-02 -8.98251295e-01 1.69536293e-01 -1.43508160e+00
-1.45616865e+00 6.43225729e-01 1.29911244e-01 -1.29005349e+00
-5.06508827e-01 -8.25148404e-01 -5.32682955e-01 1.33909178e+00
-1.65064943e+00 -1.39992261e+00 -1.03361654e+00 1.02166498e+00
2.01109022e-01 -4.73581791e-01 6.17689192e-01 6.22025803e-02
-8.77762556e-01 1.24427378e+00 4.08921629e-01 1.19289339e-01
1.48653603e+00 -1.13941920e+00 5.47901511e-01 8.91256750e-01
-3.57560515e-01 7.07460046e-01 5.16175508e-01 -2.59293020e-01
-5.45787930e-01 -1.09021187e+00 8.15305412e-01 -6.77847922e-01
2.23068759e-01 -4.02639359e-01 -1.08142185e+00 9.49241161e-01
6.00203276e-01 -4.58646342e-02 7.55533695e-01 6.14498258e-01
-3.53939027e-01 -3.97018552e-01 -1.20759547e+00 6.63756549e-01
1.19854033e+00 -3.83651793e-01 -5.99552691e-01 1.72704589e-02
4.42027539e-01 -9.04188752e-01 -4.82900858e-01 1.73433334e-01
7.43715763e-01 -1.11561549e+00 7.33922362e-01 -5.96125007e-01
2.52262592e-01 -4.50433701e-01 4.15912598e-01 -1.45650506e+00
-3.59043598e-01 -6.48573101e-01 -3.09762180e-01 1.36207783e+00
3.40189815e-01 -1.05799961e+00 7.61805713e-01 9.29859102e-01
6.70218468e-02 -2.36962676e-01 -6.75460219e-01 -5.53884864e-01
7.81176090e-02 -1.81394126e-02 8.86725664e-01 7.57053077e-01
-2.24272177e-01 4.55615968e-01 -5.05792916e-01 2.43060961e-01
7.09207177e-01 -2.82248437e-01 1.37612700e+00 -1.42301285e+00
5.20873852e-02 -2.56529748e-01 -3.47878903e-01 -1.30409753e+00
2.71168649e-01 -1.01506099e-01 1.55401498e-01 -1.13853908e+00
5.99976145e-02 -8.88276845e-02 -6.38222158e-01 6.22923911e-01
-5.92381239e-01 3.11798751e-01 1.35819837e-01 3.61411750e-01
-7.77745664e-01 5.02578795e-01 1.32461870e+00 -6.16717599e-02
-4.75683182e-01 2.07747862e-01 -8.85586441e-01 6.40910506e-01
8.30959976e-01 -4.00712639e-01 -6.19174480e-01 -5.98551750e-01
7.55124092e-02 -6.05128706e-01 3.19852173e-01 -1.08684337e+00
3.90567303e-01 1.06251622e-02 5.72297931e-01 -7.19083846e-01
3.86224210e-01 -6.44523382e-01 -2.71892786e-01 -3.08244705e-01
4.23938408e-02 -1.97780788e-01 6.00732386e-01 6.16995215e-01
-9.88054946e-02 -2.07023621e-02 6.99521780e-01 3.30894709e-01
-1.02085268e+00 3.21361721e-01 1.56207517e-01 1.51191100e-01
1.05005169e+00 -5.75052738e-01 -7.64229417e-01 -3.73723626e-01
-7.43666947e-01 4.44979817e-01 8.45749378e-01 7.46331990e-01
4.99295592e-01 -1.15604866e+00 -4.22174990e-01 4.79775339e-01
6.65143728e-01 2.51893431e-01 3.39269340e-01 8.18324029e-01
-2.05793694e-01 3.70265841e-01 -6.38929129e-01 -9.29098010e-01
-1.53652573e+00 5.44404149e-01 4.61079478e-01 3.26228529e-01
-8.07330757e-03 1.23550713e+00 8.04845333e-01 -2.82809347e-01
2.05564544e-01 -2.88956493e-01 -3.64946544e-01 -3.16749841e-01
7.00665712e-01 4.25190896e-01 -2.80170351e-01 -8.58018458e-01
-4.04491663e-01 7.26505518e-01 -3.60386789e-01 1.74039111e-01
1.08354580e+00 -6.03473067e-01 1.31600201e-01 2.63320714e-01
9.17438388e-01 -4.55884099e-01 -1.76310158e+00 -4.28707600e-01
-3.26635480e-01 -5.45675516e-01 -1.32407933e-01 -1.33923209e+00
-9.80304420e-01 1.14315808e+00 1.11607134e+00 -4.09477323e-01
1.64260888e+00 -1.58781752e-01 5.76193333e-01 3.19043219e-01
3.06922287e-01 -9.69788551e-01 -7.00016767e-02 3.95551026e-01
7.45993674e-01 -1.65075552e+00 -1.83057427e-01 -6.08098395e-02
-8.24562967e-01 8.33702922e-01 1.37391078e+00 8.37534145e-02
6.29809141e-01 -5.71801305e-01 3.48257631e-01 -1.49808004e-01
-5.07164061e-01 -3.22654873e-01 7.76082873e-01 1.22742701e+00
2.46683806e-01 -1.37087002e-01 1.52116790e-01 5.18588603e-01
-1.42721027e-01 5.08079708e-01 2.22183615e-01 5.88030875e-01
-1.98572934e-01 -9.16394413e-01 -4.45881933e-01 3.07694525e-01
-2.26423144e-01 4.94256392e-02 -4.72875565e-01 1.18787444e+00
2.40570948e-01 1.07768452e+00 6.01182356e-02 -5.03076553e-01
4.34510350e-01 3.15494388e-01 3.89320850e-01 -5.49603283e-01
-2.52541631e-01 3.94138880e-03 -1.21342964e-01 -4.73915696e-01
-7.68219411e-01 -8.58790278e-01 -7.54211009e-01 -4.79415178e-01
-4.17354286e-01 -3.03106785e-01 3.24245811e-01 7.10381329e-01
7.55917549e-01 4.61664021e-01 2.01525033e-01 -8.41261029e-01
-2.98580155e-02 -1.48832583e+00 -5.91621995e-01 3.55936706e-01
6.90822005e-01 -1.21452045e+00 -1.76222399e-01 1.89078987e-01] | [14.167685508728027, 0.015059489756822586] |
248e16e1-b441-46d0-a8a8-6d6ea37b926a | topological-point-cloud-clustering | 2303.16716 | null | https://arxiv.org/abs/2303.16716v1 | https://arxiv.org/pdf/2303.16716v1.pdf | Topological Point Cloud Clustering | We present Topological Point Cloud Clustering (TPCC), a new method to cluster points in an arbitrary point cloud based on their contribution to global topological features. TPCC synthesizes desirable features from spectral clustering and topological data analysis and is based on considering the spectral properties of a simplicial complex associated to the considered point cloud. As it is based on considering sparse eigenvector computations, TPCC is similarly easy to interpret and implement as spectral clustering. However, by focusing not just on a single matrix associated to a graph created from the point cloud data, but on a whole set of Hodge-Laplacians associated to an appropriately constructed simplicial complex, we can leverage a far richer set of topological features to characterize the data points within the point cloud and benefit from the relative robustness of topological techniques against noise. We test the performance of TPCC on both synthetic and real-world data and compare it with classical spectral clustering. | ['Michael T. Schaub', 'Vincent P. Grande'] | 2023-03-29 | null | null | null | null | ['topological-data-analysis'] | ['graphs'] | [ 1.80535130e-02 -3.91302332e-02 1.99729159e-01 3.35136771e-01
-3.73786360e-01 -7.79559493e-01 8.83314550e-01 6.78787827e-01
-3.31101678e-02 1.75147176e-01 4.70124632e-02 -7.17968345e-02
-7.57944703e-01 -8.52653921e-01 -5.34717083e-01 -7.69031048e-01
-4.84032393e-01 8.27301502e-01 2.21030608e-01 -2.25616261e-01
3.94440025e-01 8.06837261e-01 -1.58193588e+00 1.77064538e-02
9.81719911e-01 8.00064862e-01 -2.46950984e-02 4.05030578e-01
-2.46254534e-01 9.65604261e-02 -1.53989732e-01 -1.04266040e-01
2.64793724e-01 7.90134072e-02 -7.05832422e-01 3.25070530e-01
4.01840776e-01 7.34920263e-01 1.04289219e-01 1.09726882e+00
2.95529626e-02 6.83716685e-02 8.82385433e-01 -1.34061265e+00
-7.69244209e-02 2.19173342e-01 -4.46166009e-01 -2.43454143e-01
3.72060806e-01 -4.38935608e-02 1.02022374e+00 -1.01360655e+00
9.56644714e-01 1.04403472e+00 7.38151193e-01 -3.20064187e-01
-1.83263731e+00 -3.75473425e-02 -2.67580599e-01 -3.24769318e-02
-1.56185949e+00 -2.07334667e-01 9.52307522e-01 -8.89997065e-01
2.45424896e-01 5.20555496e-01 9.48005021e-01 5.23634553e-01
-2.64982104e-01 1.67387918e-01 1.10519338e+00 -2.32360080e-01
5.17045259e-01 -2.68981367e-01 3.30599785e-01 6.26239359e-01
2.23112047e-01 -2.58009106e-01 -1.54072553e-01 -4.75005031e-01
5.60413718e-01 5.27535332e-03 -2.23875940e-01 -1.33441281e+00
-1.58077741e+00 6.79949224e-01 5.67298949e-01 4.63206500e-01
-4.18434143e-01 1.24498740e-01 4.02486891e-01 1.28722146e-01
3.23396683e-01 3.37946564e-01 5.97081892e-02 -1.18943483e-01
-8.14230859e-01 9.09024924e-02 7.22532451e-01 8.41621041e-01
1.24212146e+00 -3.46959025e-01 3.29793543e-01 4.57699805e-01
2.61510819e-01 6.44327223e-01 -1.92699805e-01 -9.44844961e-01
2.13251814e-01 1.15699995e+00 -1.51384085e-01 -1.68299770e+00
-5.54107189e-01 -3.07309389e-01 -9.65126276e-01 8.96655396e-02
4.98208255e-01 3.60030472e-01 -4.92669553e-01 1.51439488e+00
5.07297695e-01 2.77206331e-01 -2.72499830e-01 7.17663229e-01
2.94749349e-01 2.83161700e-01 -3.12747955e-01 -2.08534613e-01
1.02641368e+00 -7.96122253e-02 -1.73888400e-01 5.24553478e-01
5.97019136e-01 -5.19113243e-01 8.48016381e-01 3.35367173e-01
-8.55435014e-01 -2.61513829e-01 -1.00672567e+00 1.36582524e-01
-5.01266837e-01 1.24558240e-01 1.53301984e-01 4.37162310e-01
-1.23563540e+00 7.04088807e-01 -7.72103369e-01 -5.41563570e-01
1.21003054e-01 3.34797204e-01 -5.28553069e-01 9.33781713e-02
-5.35368860e-01 6.14844143e-01 2.91627049e-01 3.48173408e-03
-4.73887697e-02 -5.96991658e-01 -4.42616493e-01 9.29496214e-02
4.17893529e-01 -6.04661703e-01 2.87471950e-01 -6.85543835e-01
-1.10560668e+00 7.13211894e-01 -1.91320330e-01 -2.21238762e-01
5.90301573e-01 4.76919085e-01 -1.81966364e-01 5.68524480e-01
2.04454303e-01 1.13719642e-01 8.82391036e-01 -1.63103390e+00
-1.27079770e-01 -4.29948896e-01 -3.80375415e-01 -1.64426684e-01
1.20594706e-02 -5.04042327e-01 -3.01157027e-01 -4.23804075e-01
6.29375100e-01 -1.29772496e+00 -2.30280772e-01 -1.68376639e-01
-6.31150842e-01 -1.75742626e-01 1.11038661e+00 -2.04000518e-01
1.16597462e+00 -2.39701009e+00 7.05101907e-01 1.12141645e+00
6.95835114e-01 -5.06254053e-03 4.18998711e-02 8.92900825e-01
-2.94217497e-01 2.75170505e-01 -5.36303043e-01 -1.97840184e-02
9.56360325e-02 3.55255865e-02 -8.78391266e-02 8.99929404e-01
1.94104388e-01 6.37934744e-01 -1.15788007e+00 -4.78886396e-01
5.51070809e-01 3.08257222e-01 -4.76032585e-01 -6.04828477e-01
-1.16303384e-01 4.37832505e-01 -5.03862858e-01 3.94259244e-01
8.47386420e-01 -2.17725024e-01 1.96828380e-01 -3.72641593e-01
-3.02536398e-01 -1.36144385e-01 -1.44836581e+00 1.44008648e+00
-5.89964800e-02 4.62455094e-01 2.24682793e-01 -1.13674223e+00
8.63992691e-01 1.25599563e-01 1.06698489e+00 -1.54098392e-01
1.23362705e-01 3.00543845e-01 -1.45449907e-01 3.60057806e-03
4.08705264e-01 1.00755110e-01 -7.76721761e-02 3.85258883e-01
-8.52805525e-02 1.15943618e-01 2.12397784e-01 4.99619603e-01
1.28613496e+00 -3.62384886e-01 -3.47228930e-03 -9.69315648e-01
6.96584880e-01 8.17439333e-02 1.07474089e-01 1.74349025e-01
1.67914107e-01 6.25701845e-01 6.54088914e-01 -2.28441224e-01
-1.13461828e+00 -1.44166493e+00 -3.44487518e-01 4.14832443e-01
3.08477759e-01 -6.24221921e-01 -7.43345320e-01 -2.57518560e-01
2.49906003e-01 6.82484582e-02 -6.79215968e-01 -4.22744527e-02
-2.66531020e-01 -4.28016752e-01 1.86682597e-01 -1.85041159e-01
4.80695963e-02 -5.17064154e-01 -4.92909670e-01 7.85302147e-02
-2.09826276e-01 -9.59984660e-01 -4.09809828e-01 -7.00265095e-02
-9.71409023e-01 -1.44017494e+00 -2.17485055e-01 -4.04573530e-01
8.27025354e-01 4.70345289e-01 1.07364285e+00 3.73337179e-01
-1.22153141e-01 8.36330831e-01 -4.42357808e-01 2.17946172e-01
-3.70465040e-01 -7.01941084e-03 3.75599742e-01 7.75734305e-01
-6.77033812e-02 -8.87645245e-01 -2.97737777e-01 3.24811161e-01
-1.01715004e+00 -9.50936005e-02 2.96255171e-01 3.88972491e-01
5.52576363e-01 2.99610049e-01 1.17787309e-01 -5.32192707e-01
5.46051800e-01 -6.24481916e-01 -6.43694401e-01 1.40829831e-01
-3.81470650e-01 1.58474237e-01 5.20485699e-01 -2.22148135e-01
-2.63181299e-01 3.67244899e-01 5.20189703e-01 -6.63208723e-01
8.72097835e-02 7.01172113e-01 5.70970140e-02 -4.28744882e-01
4.63004529e-01 9.02497172e-02 3.18599701e-01 -4.73655909e-01
7.04991400e-01 1.48050636e-01 4.72357303e-01 -7.12653756e-01
1.17256892e+00 9.32491779e-01 7.14126408e-01 -1.32520556e+00
2.12188095e-01 -8.20671916e-01 -1.17743456e+00 -4.45908844e-01
7.44269907e-01 -3.00782591e-01 -1.17513573e+00 8.61172099e-03
-8.62689555e-01 2.65617251e-01 -3.38726342e-01 2.82307953e-01
-7.37308860e-01 7.52700865e-01 -1.26123473e-01 -6.99069917e-01
2.57239550e-01 -9.75276887e-01 1.07098949e+00 -6.08206809e-01
-1.76055104e-01 -1.18533206e+00 3.30190748e-01 1.81409866e-02
3.22759636e-02 8.66372049e-01 1.20093119e+00 -3.73808861e-01
-8.63664806e-01 -2.98373252e-01 -1.58692241e-01 -1.70839012e-01
1.40489906e-01 5.92448890e-01 -3.96118701e-01 -3.47489923e-01
2.44950186e-02 4.90839243e-01 7.17254460e-01 2.85924077e-01
6.54112935e-01 -2.58162498e-01 -4.61959928e-01 4.17831093e-01
1.72136152e+00 -4.09251511e-01 3.82600904e-01 1.15652226e-01
9.58428085e-01 8.53366852e-01 2.90069044e-01 1.59220740e-01
2.98779875e-01 8.71063709e-01 5.09790063e-01 -4.25980464e-02
3.07540745e-01 -2.38054067e-01 1.91448644e-01 1.07098138e+00
-4.08186793e-01 4.77871567e-01 -1.31096363e+00 6.33926094e-01
-2.06226850e+00 -1.00478077e+00 -8.19944441e-01 2.50224495e+00
3.46980602e-01 -1.54656067e-01 5.69902241e-01 4.89747167e-01
1.04103923e+00 -5.33188172e-02 -1.51011080e-01 4.53915484e-02
-2.49775305e-01 1.20305546e-01 6.15004182e-01 5.29775202e-01
-8.72670233e-01 4.80252504e-01 5.89841747e+00 6.78303778e-01
-1.05153644e+00 -1.25781521e-01 -8.96766875e-03 6.43858463e-02
-2.64761060e-01 2.90887058e-01 -6.32248148e-02 5.71686149e-01
7.68436611e-01 -4.03498143e-01 5.14544427e-01 6.44353449e-01
3.09588343e-01 -1.51977822e-01 -1.09854114e+00 9.33422327e-01
-3.26912165e-01 -1.57757819e+00 2.09431276e-01 7.55129337e-01
5.88293850e-01 1.64877579e-01 1.41441822e-01 -4.57596958e-01
7.49277845e-02 -9.34001744e-01 7.35167146e-01 5.76834142e-01
5.55830717e-01 -7.90625215e-01 3.28127414e-01 3.43979925e-01
-1.52270472e+00 6.31164759e-02 -1.35283485e-01 1.36879876e-01
-1.60399266e-02 7.48254836e-01 -7.70225525e-01 8.43828380e-01
4.62909371e-01 8.47805440e-01 -8.85293484e-01 1.12237179e+00
5.76515555e-01 2.17486113e-01 -4.97147858e-01 2.66049892e-01
2.69775301e-01 -1.02583981e+00 8.74105453e-01 1.12569094e+00
4.25271243e-01 3.59536125e-03 3.16262245e-01 9.33578670e-01
2.66978920e-01 4.31998372e-01 -9.17068541e-01 -2.60193869e-02
5.11965394e-01 1.48316026e+00 -1.21816766e+00 -2.15989739e-01
-1.81082606e-01 4.50343519e-01 1.94308773e-01 3.14482898e-01
-3.64718616e-01 -6.22633286e-02 6.99616313e-01 5.34171999e-01
3.57293785e-01 -7.95539498e-01 -5.29880881e-01 -1.10636175e+00
1.39782324e-01 -5.77488184e-01 -9.30386931e-02 -6.10604465e-01
-1.17790020e+00 2.97927886e-01 4.43582274e-02 -1.70010281e+00
1.80728167e-01 -5.64260066e-01 -6.40361309e-01 6.62271380e-01
-8.63173306e-01 -9.60343719e-01 -1.77639797e-01 8.17983150e-01
-3.87872189e-01 3.61919701e-01 5.77042282e-01 -2.55998950e-02
-1.57908872e-01 -2.36023962e-01 4.67355847e-01 7.74726737e-03
8.42990801e-02 -1.47582746e+00 3.16485971e-01 6.81754470e-01
1.58519760e-01 1.01492023e+00 5.57643533e-01 -6.53868616e-01
-1.75054240e+00 -9.69617605e-01 3.15948099e-01 -8.19362640e-01
1.03008807e+00 -5.18046558e-01 -1.04130244e+00 3.93173605e-01
-2.11979717e-01 -7.92374313e-02 3.54510814e-01 1.02624767e-01
-2.78163612e-01 1.04240045e-01 -1.09070027e+00 5.95325112e-01
1.08903766e+00 -7.73401439e-01 -4.05279666e-01 3.37478131e-01
3.40719014e-01 4.50488701e-02 -1.27148938e+00 3.92210990e-01
2.42506668e-01 -1.06304133e+00 1.14010000e+00 -1.46524444e-01
-1.46509018e-02 -8.42126369e-01 -9.28652212e-02 -1.30563915e+00
-5.29880941e-01 -6.53922558e-01 3.84287864e-01 1.06361187e+00
7.92563260e-02 -6.56160474e-01 7.50081003e-01 1.70757920e-02
4.25437279e-02 -3.52938980e-01 -1.17415714e+00 -8.76526058e-01
-3.26275453e-02 -3.67372751e-01 4.48783338e-01 1.22400737e+00
2.64017969e-01 3.37409854e-01 9.06307399e-02 3.63519102e-01
1.04834878e+00 1.76920265e-01 8.34288538e-01 -2.11728668e+00
2.22893775e-01 -5.98591506e-01 -8.38663697e-01 -3.86491686e-01
2.23658420e-02 -1.27298343e+00 -2.25508898e-01 -1.13645625e+00
-1.97063729e-01 -8.05663228e-01 1.26848742e-01 -1.07611120e-01
2.12235659e-01 3.59836757e-01 3.83019775e-01 5.52507997e-01
-5.18458903e-01 4.85450715e-01 9.71306860e-01 -7.48154372e-02
-2.63422549e-01 -2.17099607e-01 -1.54520884e-01 6.54341161e-01
4.04524028e-01 -6.73233494e-02 -2.02351257e-01 5.30297235e-02
4.80314761e-01 1.95440546e-01 6.11313999e-01 -1.13299298e+00
4.15561497e-01 -8.09346735e-02 -1.14873968e-01 -6.41792774e-01
3.34249586e-01 -1.13935697e+00 7.32294142e-01 3.99823725e-01
2.25871623e-01 2.10305959e-01 3.04100304e-05 9.11209524e-01
-1.31363660e-01 1.45017147e-01 6.71140611e-01 7.93252438e-02
-5.10734558e-01 1.72458991e-01 -3.35078150e-01 -1.84222370e-01
1.12925696e+00 -6.15710020e-01 -2.19226107e-01 -1.34988859e-01
-1.10823178e+00 1.56843036e-01 1.30734539e+00 1.23808049e-01
4.95709449e-01 -1.65107524e+00 -4.18949962e-01 1.84933290e-01
3.88472021e-01 -7.33431876e-02 -4.70894426e-02 1.35313547e+00
-6.13555789e-01 4.75067973e-01 -1.00387380e-01 -1.22289526e+00
-1.31239820e+00 8.56939733e-01 1.02546349e-01 9.39657390e-02
-7.62574553e-01 5.06458580e-02 6.24180473e-02 -4.24840808e-01
-1.78524300e-01 -5.95372319e-01 -3.57600413e-02 3.06982696e-01
-2.62378484e-01 6.74768507e-01 1.58926442e-01 -1.14065087e+00
-4.06751752e-01 1.17752159e+00 7.12566972e-01 -1.26667246e-01
1.14202309e+00 -1.95006326e-01 -5.84952950e-01 7.03141391e-01
1.34726799e+00 3.88168424e-01 -7.91084886e-01 -1.24468990e-01
5.79750240e-01 -4.23116595e-01 -1.99281022e-01 -9.81653035e-02
-7.71940470e-01 5.44197142e-01 1.93102583e-01 9.23822165e-01
9.20252264e-01 9.61966217e-02 1.98514715e-01 1.77016616e-01
6.76251113e-01 -9.11025643e-01 1.82308368e-02 3.71210307e-01
7.60281742e-01 -6.64387524e-01 5.51861897e-03 -9.21740830e-01
-2.70456553e-01 1.10128713e+00 -4.14593577e-01 -6.07665360e-01
7.88692474e-01 -2.18963891e-01 -2.51990974e-01 -5.59571803e-01
-4.26297098e-01 -4.11725432e-01 3.26416731e-01 5.52817822e-01
-1.20114423e-01 2.11437926e-01 -2.33358696e-01 -3.32118660e-01
-3.98275465e-01 -2.36581534e-01 6.14268482e-01 6.86551034e-01
-5.84719062e-01 -1.11446071e+00 -6.79511368e-01 4.15624052e-01
7.07799122e-02 1.34119228e-01 -7.61412323e-01 7.54623592e-01
9.62731093e-02 7.14989603e-01 1.78657368e-01 -5.30270338e-01
2.26152048e-01 -7.49878362e-02 2.79174387e-01 -5.50628304e-01
-1.87609389e-01 -7.35641569e-02 -2.11468697e-01 -6.00191176e-01
-6.51900113e-01 -9.54451680e-01 -1.22777414e+00 -5.23164034e-01
-1.07234664e-01 5.52177668e-01 9.31297302e-01 8.56054544e-01
5.50866604e-01 4.81395461e-02 7.97018230e-01 -1.14468348e+00
-3.75149213e-02 -4.84493077e-01 -9.48636770e-01 6.20738745e-01
5.45491695e-01 -1.02054214e+00 -6.00551963e-01 -9.35854912e-02] | [7.408784866333008, 4.453433036804199] |
483ec355-3e74-4b49-8c25-0bbba00a1e7a | ct-bert-learning-better-tabular | 2307.04308 | null | https://arxiv.org/abs/2307.04308v1 | https://arxiv.org/pdf/2307.04308v1.pdf | CT-BERT: Learning Better Tabular Representations Through Cross-Table Pre-training | Tabular data -- also known as structured data -- is one of the most common data forms in existence, thanks to the stable development and scaled deployment of database systems in the last few decades. At present however, despite the blast brought by large pre-trained models in other domains such as ChatGPT or SAM, how can we extract common knowledge across tables at a scale that may eventually lead to generalizable representation for tabular data remains a full blank. Indeed, there have been a few works around this topic. Most (if not all) of them are limited in the scope of a single table or fixed form of a schema. In this work, we first identify the crucial research challenges behind tabular data pre-training, particularly towards the cross-table scenario. We position the contribution of this work in two folds: (i)-we collect and curate nearly 2k high-quality tabular datasets, each of which is guaranteed to possess clear semantics, clean labels, and other necessary meta information. (ii)-we propose a novel framework that allows cross-table pre-training dubbed as CT-BERT. Noticeably, in light of pioneering the scaled cross-table training, CT-BERT is fully compatible with both supervised and self-supervised schemes, where the specific instantiation of CT-BERT is very much dependent on the downstream tasks. We further propose and implement a contrastive-learning-based and masked table modeling (MTM) objective into CT-BERT, that is inspired from computer vision and natural language processing communities but sophistically tailored to tables. The extensive empirical results on 15 datasets demonstrate CT-BERT's state-of-the-art performance, where both its supervised and self-supervised setups significantly outperform the prior approaches. | ['Junbo Zhao', 'Gang Chen', 'Sai Wu', 'Liyao Li', 'Haobo Wang', 'Guoshan Lu', 'Chao Ye'] | 2023-07-10 | null | null | null | null | ['contrastive-learning', 'contrastive-learning'] | ['computer-vision', 'methodology'] | [ 6.30960464e-02 2.28198245e-01 -6.04260564e-01 -4.59145606e-01
-1.10193717e+00 -7.58496761e-01 5.20732164e-01 4.33749139e-01
1.18003286e-01 9.05976772e-01 2.09001198e-01 -2.91182786e-01
-3.74322683e-01 -9.54678655e-01 -1.00572443e+00 -3.58496845e-01
-5.13849258e-02 9.72132206e-01 2.36172780e-01 -5.16200840e-01
-1.75979901e-02 7.04882964e-02 -1.80445385e+00 9.90266085e-01
8.27926993e-01 1.35744524e+00 -5.35991788e-02 -5.67333326e-02
-5.88430345e-01 1.13486552e+00 -2.19599575e-01 -1.01411188e+00
3.55107129e-01 -1.31119490e-01 -9.44279969e-01 9.71632823e-02
5.24958491e-01 2.19269052e-01 -2.10280359e-01 7.98120439e-01
7.71101937e-02 -3.79625440e-01 2.57403404e-01 -1.32804728e+00
-4.65486944e-01 1.16703427e+00 -4.29044396e-01 -3.16347957e-01
2.97522843e-01 -2.38215700e-01 1.37541449e+00 -8.67898226e-01
9.77793217e-01 1.19956613e+00 8.12176347e-01 2.47907579e-01
-1.21619344e+00 -5.47491729e-01 1.58681333e-01 1.37853056e-01
-1.29646933e+00 -4.83715951e-01 6.38854802e-01 -3.93633276e-01
6.47436678e-01 4.17907625e-01 2.67068595e-01 9.99270558e-01
1.03680462e-01 1.05390167e+00 1.12076116e+00 -3.79571497e-01
9.34585482e-02 5.98726988e-01 1.02784917e-01 5.38748443e-01
5.15712798e-01 -5.54489732e-01 -7.82421350e-01 5.00563309e-02
4.35894400e-01 -1.21808283e-01 5.21059297e-02 -1.05083978e+00
-1.40209389e+00 5.45300186e-01 1.86225802e-01 2.55805433e-01
-3.01324241e-02 -4.09198761e-01 9.05775428e-01 4.87741470e-01
3.45991433e-01 3.24542761e-01 -7.95957088e-01 -2.90589273e-01
-8.92770648e-01 2.19027832e-01 1.09325624e+00 1.47554147e+00
8.47005129e-01 -3.74763072e-01 -1.35317922e-01 8.57448816e-01
-4.48010005e-02 3.62802029e-01 1.68079197e-01 -5.37208200e-01
1.15819943e+00 1.17413545e+00 -1.06742203e-01 -8.95309985e-01
-3.74806464e-01 -2.86436826e-01 -1.08977616e+00 -2.86740780e-01
5.12390375e-01 3.74599129e-01 -5.59545279e-01 1.55097830e+00
4.80533332e-01 -5.92396617e-01 7.19188228e-02 4.32791680e-01
8.48373055e-01 3.85206699e-01 -2.64400363e-01 -3.21800113e-01
1.42121363e+00 -8.49187016e-01 -9.45063114e-01 -1.76755890e-01
7.36993432e-01 -5.44008017e-01 1.24735296e+00 6.87642992e-01
-9.85341370e-01 -5.33934891e-01 -1.11124170e+00 -3.11608166e-01
-9.54003036e-01 -6.36512414e-02 8.99650753e-01 9.19161141e-01
-7.16112018e-01 4.39789116e-01 -6.57866836e-01 -5.52078962e-01
5.26358604e-01 1.32000357e-01 -6.86265945e-01 -1.84796393e-01
-1.18926954e+00 8.22162449e-01 6.65828645e-01 1.25385150e-01
-5.82173407e-01 -7.50742912e-01 -8.99900198e-01 -7.58266672e-02
1.21421421e+00 -6.09262109e-01 9.04717684e-01 -3.47986192e-01
-1.03624392e+00 1.11549544e+00 -1.96516588e-01 -5.39207876e-01
6.53653443e-01 -2.52601326e-01 -3.71664345e-01 -3.33608657e-01
2.99273610e-01 3.52452338e-01 3.40633601e-01 -1.41188693e+00
-4.63829577e-01 -5.14181197e-01 1.33250639e-01 9.70346481e-02
-4.69732314e-01 8.65187682e-03 -9.10792887e-01 -6.70128107e-01
2.79299587e-01 -6.21872008e-01 1.20012239e-01 -1.07262455e-01
-6.64843321e-01 -1.61159366e-01 5.56378961e-01 -4.36263531e-01
1.65168691e+00 -1.92561972e+00 1.46060303e-01 -2.46806033e-02
2.83758342e-01 4.52365838e-02 2.91777849e-01 9.96368051e-01
-2.10733250e-01 1.49871692e-01 -3.93513352e-01 -5.33693790e-01
3.99509311e-01 3.89224112e-01 -7.05555141e-01 1.53199509e-01
1.80054009e-01 9.76234257e-01 -5.36917686e-01 -7.38897622e-01
1.45601109e-01 7.92207289e-03 -5.33401430e-01 6.71214014e-02
-5.51420093e-01 3.66568789e-02 -2.08117977e-01 1.01793444e+00
8.22295547e-01 -3.81988108e-01 6.29278839e-01 -3.45847905e-01
-1.18514985e-01 6.28631175e-01 -1.57269216e+00 1.80027413e+00
-1.41598843e-02 4.51596566e-02 -2.54324474e-03 -1.07620668e+00
9.19471204e-01 6.54853135e-02 7.47924626e-01 -8.63179445e-01
-1.10535853e-01 3.71245742e-01 -4.80424166e-01 -3.30408603e-01
6.66540205e-01 2.70350778e-04 -4.06264454e-01 3.56358081e-01
4.84203547e-02 5.74668124e-03 6.42766714e-01 3.85296553e-01
8.28012228e-01 3.70898575e-01 5.08279145e-01 -2.28292078e-01
7.93500543e-01 3.56731206e-01 4.91588414e-01 8.11941803e-01
7.14699328e-02 7.14085460e-01 8.15693438e-01 -4.67663735e-01
-9.86486495e-01 -9.19331670e-01 -4.94699448e-01 1.15614498e+00
1.16849154e-01 -1.01931965e+00 -6.76639616e-01 -7.87961483e-01
2.67203689e-01 3.89724553e-01 -6.78238451e-01 2.12055758e-01
-5.65887094e-01 -6.85340345e-01 4.34207439e-01 5.01570344e-01
5.21730959e-01 -1.08469093e+00 -3.23981233e-02 2.22012162e-01
-2.95789987e-01 -1.35203195e+00 -1.04682855e-01 6.65310979e-01
-9.33056831e-01 -1.13340449e+00 1.81807409e-04 -6.12665117e-01
2.14027449e-01 2.13316858e-01 1.59724760e+00 -1.80122539e-01
6.07879125e-02 1.25571281e-01 -4.56502795e-01 -3.97524476e-01
-3.95924717e-01 3.89002472e-01 -9.65752825e-02 6.68542460e-02
4.22598273e-01 -4.72735137e-01 -7.63217732e-03 5.16443312e-01
-1.01202333e+00 3.50515485e-01 6.61928415e-01 6.71241343e-01
8.90010238e-01 1.59405798e-01 2.79274046e-01 -1.53345704e+00
2.70362467e-01 -5.70606768e-01 -5.72365642e-01 6.40885413e-01
-8.42576444e-01 1.83891550e-01 7.48239279e-01 1.25640556e-01
-7.83663929e-01 -2.20673800e-01 1.03761695e-01 -1.54411227e-01
-3.14285010e-02 9.39550936e-01 -8.05186212e-01 5.76507032e-01
4.74193901e-01 1.94361836e-01 -1.43322334e-01 -7.54651189e-01
4.33049828e-01 7.25501180e-01 6.59592509e-01 -9.91940975e-01
1.11256063e+00 7.49726951e-01 -2.25072075e-02 -1.92723334e-01
-1.19346094e+00 -4.95183825e-01 -1.09147012e+00 2.70516854e-02
5.96865535e-01 -1.01054811e+00 -6.72466695e-01 2.99119174e-01
-6.73145533e-01 -1.89431146e-01 -2.26410076e-01 -1.58815220e-01
-6.89626157e-01 2.66782731e-01 -2.62976885e-01 -6.42020404e-01
-1.98381528e-01 -1.03639555e+00 1.19626379e+00 -2.74245203e-01
-1.46200787e-02 -9.84559476e-01 -6.24816008e-02 7.35986173e-01
2.25130215e-01 2.95058221e-01 1.25355887e+00 -8.17510307e-01
-9.34525192e-01 -8.71491134e-02 -3.73197019e-01 1.06081218e-01
2.26566255e-01 -1.73888773e-01 -9.68267679e-01 -2.05048248e-01
-2.93628909e-02 -6.39460683e-01 7.50063300e-01 -8.72341841e-02
1.19730902e+00 -2.86121637e-01 -3.10765505e-01 6.60707295e-01
1.52473700e+00 -2.58543286e-02 7.50562608e-01 7.90213883e-01
8.58744562e-01 7.91122973e-01 1.02193785e+00 4.62256372e-01
9.65220630e-01 8.07580829e-01 4.24123943e-01 3.51750404e-02
-9.20698885e-03 -6.65863693e-01 2.07076341e-01 1.13799798e+00
3.91573936e-01 -1.92857176e-01 -1.00828302e+00 4.72901583e-01
-2.14277864e+00 -9.42426383e-01 -2.53714383e-01 2.37491727e+00
1.23238564e+00 5.43924749e-01 1.76143348e-01 5.70141077e-01
4.08726215e-01 1.96670443e-01 -4.09664035e-01 -2.37199828e-01
-5.18855810e-01 8.80906731e-02 3.43127191e-01 -4.02135123e-03
-1.40212774e+00 8.32608104e-01 5.55043316e+00 9.65480208e-01
-7.65888572e-01 -1.95897117e-01 7.72827089e-01 1.53649896e-01
-3.18530113e-01 1.29383028e-01 -1.11409605e+00 2.68987387e-01
9.77066875e-01 -2.64162809e-01 4.65837479e-01 1.02413297e+00
-1.48204044e-01 1.38695585e-02 -1.53100657e+00 1.14989781e+00
7.38287866e-02 -1.60437930e+00 2.87014425e-01 -1.25430236e-02
4.97012258e-01 -1.93570703e-01 1.14551350e-01 7.32666075e-01
1.76709309e-01 -1.12637091e+00 1.07726967e+00 2.73117602e-01
9.67017174e-01 -4.84255731e-01 7.40050673e-01 4.39430237e-01
-1.42208350e+00 8.36400613e-02 -4.13976043e-01 1.20841131e-01
-3.48719180e-01 6.43940032e-01 -6.20369732e-01 1.25731575e+00
1.04032421e+00 1.03319180e+00 -9.01890218e-01 6.56415999e-01
1.87933207e-01 4.75833297e-01 -1.97836339e-01 3.01489353e-01
3.18698846e-02 -2.10948229e-01 7.09776860e-03 1.24000061e+00
-1.61106378e-01 -3.62536877e-01 1.87300295e-01 8.41434896e-01
-3.22926551e-01 2.30474010e-01 -5.50900459e-01 -2.32805349e-02
5.67258000e-01 1.26822674e+00 -6.05087519e-01 -3.62882316e-01
-6.40972972e-01 3.23812902e-01 4.82114911e-01 -7.22623318e-02
-7.40711153e-01 -1.85200036e-01 3.78704250e-01 2.69063413e-01
5.51761568e-01 4.75787222e-02 -9.70099568e-01 -1.39894152e+00
7.08855450e-01 -1.41616070e+00 8.61576259e-01 -6.03657126e-01
-1.48448420e+00 6.46545947e-01 2.57287145e-01 -1.36515725e+00
-1.46179765e-01 -7.08813846e-01 2.31685743e-01 6.26275718e-01
-1.53874135e+00 -1.36783922e+00 -3.96317124e-01 6.32533491e-01
4.11208838e-01 -1.32502168e-01 7.93978274e-01 4.91114646e-01
-6.03663743e-01 7.98622549e-01 3.51876229e-01 3.34302813e-01
9.97953653e-01 -1.52519143e+00 4.33301628e-01 6.41477466e-01
2.58259416e-01 9.29231286e-01 5.02298594e-01 -6.29796088e-01
-1.97347271e+00 -1.14583921e+00 1.13335431e+00 -1.12951314e+00
1.00533569e+00 -1.14922440e+00 -1.12956452e+00 9.28206205e-01
9.46224183e-02 -6.64745923e-03 4.50590014e-01 4.78888035e-01
-8.11542928e-01 -7.49841332e-01 -1.01031947e+00 3.80812973e-01
1.20729899e+00 -6.42634749e-01 -6.39491737e-01 3.17579299e-01
6.75412238e-01 -6.51480973e-01 -1.11332941e+00 5.18051386e-01
4.89777207e-01 -1.26514769e+00 8.85039926e-01 -5.44307768e-01
5.74462354e-01 -2.71153480e-01 -4.84562635e-01 -6.50422752e-01
4.22475189e-02 -8.63957405e-01 -2.11691156e-01 1.77420795e+00
3.86169612e-01 -3.86711329e-01 9.83534396e-01 4.48839962e-01
-1.11363262e-01 -7.78260827e-01 -8.35650623e-01 -8.46857011e-01
2.64320850e-01 -5.96077144e-01 8.14326823e-01 1.08304965e+00
2.48503476e-01 2.60649353e-01 -3.83787960e-01 -6.88068271e-02
6.23664558e-01 4.17970121e-01 1.23571014e+00 -1.44340205e+00
3.23460698e-02 -1.91147506e-01 -1.17440946e-01 -1.01745236e+00
-1.61533341e-01 -1.12556148e+00 -2.95628667e-01 -1.50856590e+00
5.05426466e-01 -7.67462313e-01 -3.25875759e-01 6.92010939e-01
-6.03736117e-02 -4.72847931e-02 1.75331458e-01 4.69280839e-01
-1.01272666e+00 3.50783914e-01 1.03214085e+00 -1.73608720e-01
-6.76483335e-03 -2.97243327e-01 -1.06093788e+00 2.02690318e-01
3.63696873e-01 -6.07611895e-01 -4.82307434e-01 -2.36106709e-01
5.46809256e-01 1.53099954e-01 -4.61311974e-02 -9.89550769e-01
3.34289551e-01 -2.10106596e-01 -5.30779697e-02 -1.11372542e+00
1.22129098e-02 -1.07704663e+00 2.75699258e-01 -8.03097561e-02
-2.89697677e-01 1.53154477e-01 4.58208263e-01 4.72593337e-01
-5.64828157e-01 1.61139131e-01 4.13977265e-01 -1.67801589e-01
-5.08457124e-01 2.06239447e-01 2.41679594e-01 4.32851106e-01
7.31984854e-01 -3.28656137e-01 -5.38911521e-01 -6.65841028e-02
-2.41840973e-01 3.16057622e-01 4.62190926e-01 6.64594352e-01
1.18111908e-01 -1.24646080e+00 -5.55252671e-01 2.38456383e-01
6.49633825e-01 2.57219315e-01 -3.72079946e-02 8.14557374e-01
-2.96013385e-01 9.35061157e-01 -1.11077242e-01 -6.26187384e-01
-9.35177267e-01 1.04189324e+00 -2.29240581e-02 -7.95738041e-01
-5.74771106e-01 4.41791117e-01 9.89225060e-02 -8.40817750e-01
4.62947071e-01 -5.84117055e-01 -3.08540110e-02 3.45816076e-01
2.63125718e-01 -3.52760218e-02 7.26608932e-01 -3.08364242e-01
-4.40336257e-01 2.53325880e-01 -3.83179247e-01 2.96362609e-01
1.58399153e+00 -1.75560400e-01 -5.00834227e-01 8.50088835e-01
7.65371978e-01 1.17547639e-01 -9.05740082e-01 -5.11598051e-01
5.75530589e-01 -3.52262527e-01 -4.81048614e-01 -1.10263813e+00
-8.20648372e-01 7.01593518e-01 -1.09352209e-02 3.75759900e-01
1.00802422e+00 3.99781018e-02 5.91043174e-01 4.92353618e-01
9.22111154e-01 -8.30837905e-01 -1.44482464e-01 4.89265263e-01
7.87319958e-01 -1.37614501e+00 2.17003807e-01 -6.99014068e-01
-7.14300036e-01 1.06112158e+00 5.58480740e-01 4.00270194e-01
4.44277287e-01 5.07206678e-01 9.23743472e-02 -2.62215167e-01
-1.15962148e+00 -1.55387521e-01 2.93987960e-01 5.31491995e-01
5.02195418e-01 -2.51297414e-01 -2.64775613e-03 7.74213135e-01
-4.56169844e-01 8.85808393e-02 3.73632520e-01 1.13020837e+00
-1.77897096e-01 -1.39807403e+00 -5.25761187e-01 4.13984090e-01
-1.96139321e-01 -1.31344035e-01 -5.65357745e-01 1.13576257e+00
1.46926060e-01 8.29569340e-01 -1.21302731e-01 -4.07252938e-01
5.59082091e-01 2.39427432e-01 2.76240557e-01 -6.09643757e-01
-8.90915275e-01 -1.27972156e-01 1.87837139e-01 -6.13769710e-01
-3.67195547e-01 -7.90269136e-01 -9.39742267e-01 -6.13690555e-01
-4.17358354e-02 2.68502116e-01 4.36034560e-01 7.12980628e-01
2.83276737e-01 3.18409950e-01 3.62551451e-01 -1.47378147e-01
-5.87469161e-01 -8.52748275e-01 -5.86814582e-01 6.13392115e-01
9.69100073e-02 -8.11575890e-01 -1.28941149e-01 2.02922970e-01] | [9.619219779968262, 7.893829822540283] |
7f1f636f-c57a-4a11-9a9d-5d4dbb2b3088 | slsdeep-skin-lesion-segmentation-based-on | 1805.10241 | null | http://arxiv.org/abs/1805.10241v2 | http://arxiv.org/pdf/1805.10241v2.pdf | SLSDeep: Skin Lesion Segmentation Based on Dilated Residual and Pyramid Pooling Networks | Skin lesion segmentation (SLS) in dermoscopic images is a crucial task for
automated diagnosis of melanoma. In this paper, we present a robust deep
learning SLS model, so-called SLSDeep, which is represented as an
encoder-decoder network. The encoder network is constructed by dilated residual
layers, in turn, a pyramid pooling network followed by three convolution layers
is used for the decoder. Unlike the traditional methods employing a
cross-entropy loss, we investigated a loss function by combining both Negative
Log Likelihood (NLL) and End Point Error (EPE) to accurately segment the
melanoma regions with sharp boundaries. The robustness of the proposed model
was evaluated on two public databases: ISBI 2016 and 2017 for skin lesion
analysis towards melanoma detection challenge. The proposed model outperforms
the state-of-the-art methods in terms of segmentation accuracy. Moreover, it is
capable to segment more than $100$ images of size 384x384 per second on a
recent GPU. | ['Saddam Abdulwahab', 'Petia Radeva', 'Md. Mostafa Kamal Sarker', 'Farhan Akram', 'Domenec Puig', 'Adel Saleh', 'Syeda Furruka Banu', 'Santiago Romani', 'Hatem A. Rashwan', 'Vivek Kumar Singh', 'Forhad U H Chowdhury'] | 2018-05-25 | null | null | null | null | ['skin-lesion-segmentation'] | ['medical'] | [ 6.87631845e-01 1.93960413e-01 -8.49117860e-02 -2.30634391e-01
-1.06899142e+00 -2.93624282e-01 3.19655836e-01 1.92732349e-01
-7.90195286e-01 6.14883125e-01 -3.68290395e-01 -3.88962239e-01
2.50649095e-01 -6.48312509e-01 -5.90627193e-01 -8.66747618e-01
3.32957566e-01 -2.68597871e-01 5.04766822e-01 3.16261649e-01
2.97718704e-01 3.48396659e-01 -1.03398907e+00 2.66480416e-01
1.28615117e+00 1.24369681e+00 6.22785464e-02 8.53707731e-01
1.22239113e-01 7.70722806e-01 -3.81995618e-01 -8.23653638e-01
2.69070119e-01 -3.51828992e-01 -9.16313112e-01 7.43065700e-02
4.22031403e-01 -5.47365248e-01 -1.52905375e-01 1.13403630e+00
7.76490271e-01 -4.32335287e-01 6.59132719e-01 -6.17126942e-01
-1.89903304e-01 -1.74037889e-02 -9.44743633e-01 1.06393605e-01
7.28284344e-02 2.63886839e-01 4.53039736e-01 -2.56451368e-01
7.13853002e-01 7.86935866e-01 1.13744843e+00 6.34917915e-01
-9.11754608e-01 -3.04692864e-01 -2.68933207e-01 3.02710589e-02
-1.38422120e+00 4.78950106e-02 1.48324758e-01 -4.46668029e-01
7.34478593e-01 5.47880530e-01 6.13734603e-01 9.08636630e-01
6.80367589e-01 8.82394731e-01 1.23689568e+00 -3.06752354e-01
2.46397071e-02 2.25798339e-01 7.20235407e-02 1.04472303e+00
2.40140587e-01 -9.44491401e-02 -4.64107879e-02 -3.08187511e-02
8.54723990e-01 -2.12200224e-01 -9.95759666e-02 2.90170968e-01
-7.84002244e-01 7.32518435e-01 6.16552770e-01 -1.97913155e-01
-4.70173866e-01 1.36522144e-01 6.80856943e-01 -1.94058716e-01
6.25425041e-01 -1.61186338e-03 -2.30771787e-02 9.96056944e-02
-1.05447936e+00 -5.91926090e-02 7.06615627e-01 2.99763978e-01
1.71576515e-01 -4.54087138e-01 -6.08190238e-01 6.72213495e-01
1.93819821e-01 1.70780122e-01 4.63710338e-01 -6.07258916e-01
3.16317052e-01 7.97757864e-01 -1.60260350e-01 -5.48614085e-01
-4.41977322e-01 -4.22665715e-01 -1.19493711e+00 2.28765979e-01
3.64956498e-01 -5.61794519e-01 -1.25174212e+00 1.25486970e+00
5.19205153e-01 2.59223759e-01 -7.22495690e-02 1.03146398e+00
8.48602831e-01 4.50209379e-01 3.19161594e-01 8.96190181e-02
1.48619688e+00 -1.13468218e+00 -6.38777137e-01 4.73322347e-02
5.43051958e-01 -8.52408886e-01 5.79132795e-01 5.02599120e-01
-1.29446912e+00 -3.55062574e-01 -9.67106044e-01 -2.93180704e-01
-2.37597346e-01 6.26413286e-01 5.77404797e-01 8.09723675e-01
-1.10346401e+00 5.41347504e-01 -1.16877294e+00 -4.83079940e-01
8.25595140e-01 5.19009113e-01 -2.14875206e-01 3.20513658e-02
-9.99791026e-01 7.24442899e-01 1.68979600e-01 2.71418184e-01
-7.36197948e-01 -8.74680042e-01 -8.15389574e-01 -3.06948364e-01
8.13523084e-02 -7.67069995e-01 1.14284468e+00 -1.12899446e+00
-1.70404339e+00 1.27426243e+00 -2.43389815e-01 -8.15531790e-01
9.50113058e-01 -8.27774927e-02 -1.86970860e-01 4.33409572e-01
-2.82330483e-01 8.18126082e-01 7.73068011e-01 -8.74913573e-01
-6.94778621e-01 -2.50310659e-01 -1.00009821e-01 2.22864151e-01
-1.50196150e-01 8.38345736e-02 -6.32626951e-01 -4.78018343e-01
-3.40591192e-01 -8.09720278e-01 -4.08887416e-01 3.74741286e-01
-9.15917277e-01 4.90213670e-02 4.88513827e-01 -1.46302426e+00
1.40412128e+00 -1.95062435e+00 -1.79138064e-01 2.77433753e-01
2.50060529e-01 8.80665421e-01 1.65468808e-02 1.49287000e-01
1.10397227e-01 1.98734432e-01 -6.19586170e-01 -4.24435109e-01
-2.01561272e-01 -3.59405220e-01 2.49576911e-01 6.95052743e-01
4.51702207e-01 9.08336699e-01 -5.34680724e-01 -7.11363375e-01
3.02966535e-01 9.57525790e-01 -1.88996986e-01 1.58614546e-01
2.76754722e-02 2.40226328e-01 -1.77169129e-01 7.70709515e-01
1.08267760e+00 -4.28537279e-01 -1.46532897e-02 -1.44487292e-01
-4.33719978e-02 -1.08506069e-01 -8.59130263e-01 1.67394412e+00
-4.05361205e-01 6.62538469e-01 2.23538592e-01 -3.67779762e-01
4.95055348e-01 2.85953760e-01 3.60021889e-01 -5.18269181e-01
3.36524576e-01 2.07282409e-01 -1.64581299e-01 -7.67416179e-01
1.68065846e-01 1.54055348e-02 3.10613543e-01 -2.40632053e-03
-1.30209967e-01 9.13865492e-02 2.06313327e-01 -9.30109769e-02
1.11806750e+00 -2.84802821e-02 2.50569850e-01 3.10292393e-02
8.23739529e-01 1.08377086e-02 4.74951595e-01 2.86113292e-01
-4.76275861e-01 6.64469898e-01 8.16614211e-01 -1.74008176e-01
-9.56088424e-01 -1.04424500e+00 -4.67029870e-01 3.68593872e-01
3.53160560e-01 -2.21504644e-02 -1.32902527e+00 -8.11269701e-01
-1.57480210e-01 2.91476399e-01 -6.52043164e-01 1.14356622e-01
-2.89339513e-01 -1.11769509e+00 8.68955076e-01 4.66170669e-01
1.00494075e+00 -8.45232368e-01 -7.03014314e-01 -3.45592126e-02
-1.60991922e-02 -9.59406137e-01 -5.99847257e-01 -1.69999853e-01
-6.63509667e-01 -1.37369573e+00 -1.20695531e+00 -9.97233570e-01
1.02979577e+00 -3.01040143e-01 7.72401810e-01 -1.33699507e-01
-1.13794613e+00 -2.62626052e-01 -3.13172974e-02 -4.29023594e-01
-3.94235104e-01 2.23657742e-01 -7.65604377e-01 1.80578083e-01
3.98028553e-01 2.10599214e-01 -1.23515034e+00 -8.10905360e-03
-1.11033487e+00 3.02088976e-01 1.10113788e+00 9.42743003e-01
9.61886227e-01 7.80695751e-02 3.60458076e-01 -1.21832478e+00
7.04670250e-01 -3.22750896e-01 -6.61602557e-01 4.65809911e-01
-3.06378007e-01 -3.22765797e-01 5.14593601e-01 -4.20371741e-02
-1.24899888e+00 4.17899162e-01 -6.35012984e-01 4.99997362e-02
-1.10903844e-01 2.74079889e-01 3.83958310e-01 -5.25832355e-01
5.01685381e-01 3.02538186e-01 3.77705306e-01 -2.93436021e-01
-7.04615265e-02 9.33472276e-01 4.84092951e-01 2.60374516e-01
1.44822001e-01 4.90036041e-01 1.94020823e-01 -6.65089548e-01
-6.85870647e-01 -6.14722848e-01 -5.28485537e-01 -1.38245746e-01
1.16280901e+00 -8.95915985e-01 -8.16415727e-01 1.21881199e+00
-1.13446975e+00 -3.34938198e-01 -3.83883864e-02 2.31783643e-01
-2.57964671e-01 6.85465932e-01 -1.10877204e+00 -8.86923611e-01
-9.40722227e-01 -1.25972378e+00 1.26318133e+00 7.15674222e-01
1.04939900e-01 -1.23806489e+00 -1.19534805e-01 4.96660113e-01
3.40096831e-01 8.97944033e-01 5.89802980e-01 -1.98330328e-01
-1.54272139e-01 -6.02446795e-01 -6.56917095e-01 8.02168727e-01
-8.58597681e-02 2.37601504e-01 -9.26038444e-01 -4.54704106e-01
-1.63018659e-01 -3.85136992e-01 1.21045828e+00 7.46744514e-01
1.70292807e+00 -1.67794794e-01 -3.78100425e-01 1.01036465e+00
1.97285032e+00 1.24573343e-01 1.01289594e+00 -5.73168471e-02
5.28786063e-01 4.92552996e-01 4.37966615e-01 2.49371678e-01
3.83177787e-01 1.85136944e-01 5.24510562e-01 -8.08845878e-01
-3.17178369e-01 -9.78183821e-02 4.85142432e-02 3.71397555e-01
-7.29878172e-02 -3.41161340e-01 -8.34412158e-01 5.64936697e-01
-1.44460905e+00 -4.17469054e-01 -2.84845680e-01 2.15897584e+00
9.75361764e-01 7.66962580e-03 7.95231760e-02 -2.22312529e-02
9.48349476e-01 6.47661239e-02 -9.40865457e-01 -7.06566691e-01
2.23716646e-02 5.27772367e-01 1.00277758e+00 5.28854132e-01
-1.41894984e+00 7.63438821e-01 5.86062145e+00 1.25438309e+00
-1.24079180e+00 5.26467077e-02 1.18552554e+00 1.74770221e-01
1.65818542e-01 -4.71193999e-01 -7.14542210e-01 7.33870268e-01
8.04212093e-01 4.68705684e-01 5.67459203e-02 4.17472541e-01
1.57508980e-02 -4.49509263e-01 -4.85309452e-01 8.48640203e-01
6.58211112e-02 -1.30275619e+00 -2.06922591e-01 6.47160038e-02
7.41367102e-01 -2.13151813e-01 4.02003169e-01 -2.38585964e-01
-1.08492553e-01 -1.38555181e+00 -5.62941320e-02 7.50798106e-01
1.44203949e+00 -8.14351499e-01 1.23471630e+00 1.71731651e-01
-8.13188255e-01 2.33169347e-01 -2.32334360e-01 5.02361596e-01
5.88510521e-02 6.46427631e-01 -9.66685474e-01 6.27875566e-01
2.88186729e-01 5.36477029e-01 -6.15373075e-01 1.42232275e+00
-1.19356744e-01 7.45727539e-01 -1.48408636e-01 -9.14246142e-02
5.17721415e-01 -1.43804953e-01 2.02461526e-01 1.50397038e+00
2.34958857e-01 -1.84788987e-01 -3.47282052e-01 6.09611988e-01
-1.54473633e-01 6.99000657e-02 4.93820049e-02 9.47811604e-02
1.87414661e-02 1.46649325e+00 -6.14176273e-01 -4.58409414e-02
-2.07185328e-01 1.39307833e+00 -8.37183222e-02 1.65960014e-01
-1.08313334e+00 -8.41509998e-01 3.98385227e-01 1.61170676e-01
1.30327851e-01 3.73960793e-01 -4.66062546e-01 -7.86460578e-01
6.02821261e-02 -5.73379993e-01 3.54129225e-01 -3.65739316e-01
-1.27934504e+00 6.05245531e-01 -6.45431697e-01 -1.10587740e+00
1.08465426e-01 -8.35283101e-01 -7.74934053e-01 1.24726140e+00
-2.12737155e+00 -1.38403058e+00 -6.12925768e-01 4.15583313e-01
3.58516991e-01 6.98734447e-02 7.34608889e-01 2.04865232e-01
-9.56898808e-01 9.65072870e-01 2.24260136e-01 2.59763032e-01
6.62866950e-01 -1.30583191e+00 4.54068214e-01 7.93798149e-01
-6.33250415e-01 8.88270885e-02 9.81036499e-02 -6.10154450e-01
-1.00547850e+00 -1.43523443e+00 7.64721930e-01 1.31303132e-01
4.00985032e-01 -3.92249674e-02 -5.82253873e-01 1.69321194e-01
4.20165569e-01 1.33160770e-01 8.59600067e-01 -5.66987634e-01
7.54017010e-02 -5.44831939e-02 -1.76671338e+00 3.70811313e-01
5.38707376e-01 -2.40828857e-01 1.20373666e-01 5.83854496e-01
4.77024317e-01 -9.71322477e-01 -9.88158882e-01 5.28492987e-01
5.23282290e-01 -9.71839309e-01 8.95716250e-01 -2.54769444e-01
7.34385848e-01 8.40403736e-02 4.12961125e-01 -9.32964921e-01
-2.74163857e-02 -5.76047003e-01 -4.09848019e-02 9.22949851e-01
3.22707683e-01 -6.18191838e-01 1.08120930e+00 2.72516042e-01
3.78433801e-02 -1.59685183e+00 -9.83734727e-01 -3.13698292e-01
1.38948426e-01 7.92236477e-02 2.20165402e-01 4.11215067e-01
-4.45080787e-01 -2.96961695e-01 -1.15573369e-01 9.52514485e-02
7.59958744e-01 -3.97115886e-01 1.60997197e-01 -6.30223513e-01
-1.22728355e-01 -4.72960234e-01 -5.64993262e-01 -9.29878592e-01
-1.30663306e-01 -7.55263507e-01 -2.01241955e-01 -1.73574793e+00
3.44667286e-01 -1.49889752e-01 -2.64256299e-01 2.99513102e-01
-4.45334792e-01 5.63398004e-01 -1.24870189e-01 -1.98799625e-01
-4.47227597e-01 -4.19780351e-02 1.50973248e+00 -1.73220858e-01
-7.06856474e-02 1.72285378e-01 -5.08277655e-01 7.65583336e-01
8.79553318e-01 -5.64400107e-02 -5.79337403e-02 -3.98992032e-01
-1.68633848e-01 1.29096866e-01 5.76730371e-01 -1.04235196e+00
5.18835604e-01 1.30925834e-01 5.14302611e-01 -6.53531194e-01
2.11960986e-01 -4.83490974e-01 -8.10564235e-02 9.07159328e-01
-4.05202597e-01 -5.52744627e-01 2.16859296e-01 4.57050890e-01
-4.01064992e-01 -2.40514487e-01 1.20988536e+00 7.61252120e-02
-5.81645131e-01 4.67776299e-01 -1.23895824e-01 -1.53076187e-01
1.54932404e+00 -3.39166045e-01 -4.92017299e-01 2.41022840e-01
-5.33609569e-01 2.04715177e-01 2.66863793e-01 -2.34840691e-01
6.80530071e-01 -9.44986582e-01 -9.69466150e-01 1.10374205e-01
-9.98927280e-02 2.90223360e-01 8.39674354e-01 1.24991941e+00
-1.43232727e+00 3.67107302e-01 -1.13530338e-01 -5.33730030e-01
-1.53568649e+00 2.82866266e-02 7.07391798e-01 -6.96493506e-01
-4.39096451e-01 1.37537479e+00 -5.47121987e-02 -1.65816858e-01
3.56754184e-01 -2.71855146e-01 -5.48778586e-02 -1.95201799e-01
5.25540709e-01 5.10860443e-01 1.15160443e-01 -3.26131433e-01
-2.27910906e-01 4.98889238e-01 -3.00415844e-01 2.41922379e-01
1.06630504e+00 8.06845278e-02 -2.72677600e-01 -1.79858863e-01
1.32619417e+00 -3.66502672e-01 -1.71217239e+00 4.96209860e-02
-3.47705305e-01 -4.06604081e-01 2.00265542e-01 -1.30045986e+00
-1.11953056e+00 1.00924468e+00 1.05138409e+00 5.61150275e-02
1.54781175e+00 -5.52570462e-01 1.46482539e+00 -2.51252800e-01
-1.24964692e-01 -1.24059176e+00 -2.98442692e-01 4.26653475e-02
4.64988798e-01 -1.42699802e+00 -6.82826266e-02 -7.79829502e-01
-9.50968325e-01 1.12669313e+00 5.60424626e-01 -4.30203408e-01
5.78853488e-01 3.83798897e-01 8.39824453e-02 1.16892241e-01
-3.70496213e-01 -1.66279182e-01 5.13707340e-01 5.20994425e-01
4.54979718e-01 7.85808340e-02 -6.81515038e-01 5.39804220e-01
2.05935597e-01 2.16808856e-01 2.66239405e-01 6.87325418e-01
-1.22781964e-02 -9.68020499e-01 -5.99794686e-02 8.56049657e-01
-9.10529017e-01 -7.69867897e-02 -4.39219028e-01 7.36178756e-01
4.62502331e-01 7.09406435e-01 4.92893495e-02 -2.90268600e-01
-3.17361653e-02 -3.22767168e-01 4.14831609e-01 -3.94467741e-01
-9.84285414e-01 -1.05809882e-01 -4.18017656e-02 -5.17696619e-01
-2.83225358e-01 -4.05145079e-01 -1.10773671e+00 -3.11082602e-01
-2.18516707e-01 -3.76873702e-01 9.34122026e-01 5.42764366e-01
2.01784462e-01 7.34995127e-01 6.58921659e-01 -2.34216765e-01
-6.23387277e-01 -7.60614634e-01 -7.12911308e-01 2.61256933e-01
3.44425559e-01 -1.52561339e-02 -1.45458519e-01 1.93272546e-01] | [15.640033721923828, -2.955991268157959] |
97eee63c-cbbe-44ec-a5b5-a3656b2f7433 | the-value-of-semantic-parse-labeling-for | null | null | https://aclanthology.org/P16-2033 | https://aclanthology.org/P16-2033.pdf | The Value of Semantic Parse Labeling for Knowledge Base Question Answering | null | ['Ming-Wei Chang', 'Wen-tau Yih', 'Jina Suh', 'Matthew Richardson', 'Chris Meek'] | 2016-08-01 | null | null | null | acl-2016-8 | ['knowledge-base-question-answering'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.173574924468994, 3.834340810775757] |
b22e7e8f-bb80-4e36-be40-16b5a13f1981 | a-study-of-variable-role-based-feature | 2303.04942 | null | https://arxiv.org/abs/2303.04942v2 | https://arxiv.org/pdf/2303.04942v2.pdf | A Study of Variable-Role-based Feature Enrichment in Neural Models of Code | Although deep neural models substantially reduce the overhead of feature engineering, the features readily available in the inputs might significantly impact training cost and the performance of the models. In this paper, we explore the impact of an unsuperivsed feature enrichment approach based on variable roles on the performance of neural models of code. The notion of variable roles (as introduced in the works of Sajaniemi et al. [Refs. 1,2]) has been found to help students' abilities in programming. In this paper, we investigate if this notion would improve the performance of neural models of code. To the best of our knowledge, this is the first work to investigate how Sajaniemi et al.'s concept of variable roles can affect neural models of code. In particular, we enrich a source code dataset by adding the role of individual variables in the dataset programs, and thereby conduct a study on the impact of variable role enrichment in training the Code2Seq model. In addition, we shed light on some challenges and opportunities in feature enrichment for neural code intelligence models. | ['Mohammad Amin Alipour', 'David Lo', 'Bowen Xu', 'Md Rafiqul Islam Rabin', 'Aftab Hussain'] | 2023-03-08 | null | null | null | null | ['feature-engineering'] | ['methodology'] | [ 1.54645443e-01 3.71271104e-01 -1.62923485e-01 -4.58427310e-01
1.32646814e-01 -5.91242373e-01 1.97078228e-01 4.25754875e-01
-4.18337256e-01 4.43361551e-01 4.95216511e-02 -6.81545973e-01
-9.51747000e-02 -9.99300838e-01 -1.05790329e+00 -1.24608859e-01
-8.01814254e-03 -4.22790945e-02 8.20018947e-02 -4.54156160e-01
4.46177989e-01 3.89648795e-01 -1.97519946e+00 4.22864228e-01
1.00355923e+00 4.78424668e-01 2.50795186e-01 5.74140370e-01
-2.66754895e-01 9.83089328e-01 -6.52921557e-01 -5.07266223e-01
9.96315703e-02 -2.66704679e-01 -8.16048741e-01 -4.68774289e-01
4.06640798e-01 -1.63474709e-01 -8.23827684e-02 1.03603041e+00
3.25364649e-01 1.85863927e-01 2.64075011e-01 -1.07200921e+00
-8.15015376e-01 1.15354705e+00 -3.55149806e-01 2.79495180e-01
2.70760834e-01 7.44471028e-02 1.11310482e+00 -7.01412499e-01
6.64761424e-01 7.87929416e-01 8.82679760e-01 6.08270884e-01
-1.20892894e+00 -6.81769609e-01 3.43080878e-01 1.52892411e-01
-1.13362932e+00 -3.53207737e-01 8.25733960e-01 -7.77799845e-01
1.08975899e+00 2.46744037e-01 7.84641266e-01 7.33737767e-01
-9.61763933e-02 8.13127279e-01 7.80354917e-01 -7.56012380e-01
7.27943853e-02 4.65589046e-01 6.93741262e-01 9.58718240e-01
4.32130307e-01 7.04950690e-02 -3.72221917e-01 -8.45787823e-02
4.33727950e-01 -6.53548539e-02 -1.23758040e-01 -4.53601331e-01
-7.71898687e-01 1.03148162e+00 4.74754542e-01 6.91934764e-01
-8.38487893e-02 3.15755606e-01 4.86801207e-01 3.69681984e-01
1.71229467e-01 1.08547008e+00 -7.46933997e-01 -4.63169366e-01
-8.05723250e-01 1.69074818e-01 6.70018673e-01 8.32861543e-01
7.68709660e-01 6.42802715e-02 -1.90302745e-01 6.21141136e-01
1.76807046e-01 -1.16436653e-01 6.62076116e-01 -7.74534881e-01
3.05123687e-01 9.57337081e-01 -4.30294782e-01 -8.52172554e-01
-5.06634712e-01 -5.16173542e-01 -1.73509732e-01 8.22111890e-02
5.07177055e-01 -4.52049911e-01 -4.80250984e-01 1.87093771e+00
-2.38102283e-02 1.42783016e-01 -1.18182540e-01 6.56965733e-01
1.07705057e+00 2.40421891e-01 5.84057756e-02 3.51326734e-01
1.05458570e+00 -1.00675666e+00 -4.35809225e-01 -8.91360268e-02
1.08513713e+00 -5.85691094e-01 1.09754372e+00 4.41900164e-01
-1.03303099e+00 -7.46366739e-01 -8.92250955e-01 -8.39270428e-02
-4.07848924e-01 2.35766560e-01 1.25364554e+00 1.12599909e+00
-1.05768108e+00 7.12663293e-01 -7.32787251e-01 -1.51108146e-01
3.66653442e-01 5.17864823e-01 -2.15774849e-01 -6.75556064e-02
-1.19536185e+00 8.55331063e-01 2.44603768e-01 -3.74536902e-01
-5.46375871e-01 -9.19024885e-01 -8.63024831e-01 4.28526312e-01
2.72906005e-01 -5.78783453e-01 1.37731290e+00 -1.46618724e+00
-1.27170479e+00 5.61098456e-01 -6.06902242e-02 -4.42558676e-01
1.41764209e-01 -8.66573676e-02 7.27278888e-02 -2.68195361e-01
-2.71931708e-01 6.55943811e-01 2.56181121e-01 -9.44436848e-01
-3.52664024e-01 -3.61349702e-01 6.59958303e-01 -2.27970734e-01
-6.22242033e-01 7.68006267e-03 -9.54180881e-02 -5.71316898e-01
-2.43362010e-01 -1.03581369e+00 -2.64161885e-01 -3.70345503e-01
-8.55867192e-03 -2.15641096e-01 2.69116044e-01 -4.80082929e-01
1.28872550e+00 -2.23143888e+00 1.36181623e-01 1.23936772e-01
2.95017749e-01 2.78004199e-01 -2.51069516e-01 1.79843143e-01
-2.78922707e-01 5.83916187e-01 -1.16338879e-01 -2.86525059e-02
-6.03964478e-02 1.42024234e-01 -5.79245463e-02 1.46226594e-02
4.74312693e-01 9.92220819e-01 -7.51931846e-01 -5.29298633e-02
-2.33216435e-01 6.09973669e-01 -1.21272218e+00 6.99945316e-02
-2.62179881e-01 1.35136500e-01 -3.34031880e-01 4.13372219e-01
3.43881428e-01 -6.32700399e-02 -8.76551587e-03 1.74494103e-01
-3.39370787e-01 5.56992948e-01 -8.15435708e-01 1.48036814e+00
-6.23712063e-01 8.21801066e-01 -2.89541781e-01 -1.11890817e+00
9.68407929e-01 3.18476409e-01 2.81664968e-01 -6.40338600e-01
2.69213133e-02 1.22167572e-01 7.01029956e-01 -5.47475576e-01
7.25559473e-01 -1.36403218e-01 -6.10124785e-03 4.06590462e-01
3.87069173e-02 3.71761657e-02 3.34201783e-01 2.11548917e-02
1.09724021e+00 3.03483486e-01 1.81976438e-01 -3.88338804e-01
4.51311767e-01 6.81080818e-02 3.33961815e-01 7.53250360e-01
-3.16476636e-02 4.06795174e-01 7.59103537e-01 -3.05924594e-01
-6.94228053e-01 -4.13876057e-01 -7.98954070e-02 1.53289592e+00
-4.62979436e-01 -5.90847313e-01 -8.40635896e-01 -7.38675535e-01
1.18003473e-01 8.54001105e-01 -9.23143566e-01 -5.60377896e-01
-5.50154090e-01 -7.87145913e-01 8.32553148e-01 6.63909435e-01
4.97379638e-02 -9.20903087e-01 -1.02097166e+00 8.53739828e-02
1.91609427e-01 -5.77981830e-01 -3.17127667e-02 6.18031025e-01
-1.08205318e+00 -9.76136446e-01 -2.16276690e-01 -6.93213105e-01
7.08727837e-01 -1.05641000e-02 1.39774024e+00 6.78613961e-01
-1.72326326e-01 2.51520544e-01 -5.41642547e-01 -6.71297252e-01
-3.47070932e-01 3.37059766e-01 -2.33957842e-01 -5.99594474e-01
5.81004918e-01 -4.74418670e-01 -6.33963346e-02 -1.10151485e-01
-9.20153379e-01 -1.34335440e-02 5.17107904e-01 8.58839452e-01
1.42042488e-01 1.22984961e-01 6.05588138e-01 -1.36874485e+00
7.16391027e-01 -6.12648249e-01 -5.07280707e-01 2.52869558e-02
-8.02602768e-01 4.56798851e-01 6.94312334e-01 -4.60488915e-01
-8.51415038e-01 9.91953537e-02 -2.91299313e-01 -9.51465815e-02
-1.70003280e-01 9.94544804e-01 6.31289324e-03 -3.93655449e-01
7.80396938e-01 3.73278148e-02 -2.07921818e-01 -5.62939465e-01
9.26913321e-02 3.48550528e-01 -1.05611123e-01 -9.20754313e-01
5.52697122e-01 -2.49011233e-01 -1.67979784e-02 -5.44947624e-01
-5.42681873e-01 -2.26847008e-01 -6.95525110e-01 3.36798690e-02
5.42866945e-01 -7.29467869e-01 -4.67798948e-01 1.45203680e-01
-1.15018702e+00 -4.45265770e-01 -2.35837668e-01 3.04708004e-01
-1.99768946e-01 1.09700374e-01 -5.82705975e-01 -5.04040480e-01
1.40411377e-01 -1.53374517e+00 3.32113951e-01 3.22079420e-01
-3.86985958e-01 -9.90563393e-01 6.08382337e-02 3.57611388e-01
7.04190016e-01 1.56241301e-02 1.12988198e+00 -8.00823569e-01
-3.97537827e-01 6.60479814e-02 1.68457441e-02 2.01699495e-01
-1.75452769e-01 7.52132982e-02 -1.18519080e+00 1.13005685e-02
-3.97066996e-02 -1.51354045e-01 1.02546966e+00 2.19537467e-01
1.21299779e+00 -1.14398077e-01 7.35463621e-03 5.96919775e-01
1.34751165e+00 2.31447563e-01 5.70689619e-01 6.52162254e-01
7.38923132e-01 8.42108369e-01 3.91807944e-01 3.81679386e-01
2.03100637e-01 5.28314054e-01 4.02105868e-01 1.66101620e-01
-1.71978399e-02 -5.50764985e-02 5.27396262e-01 6.52887881e-01
-1.71361253e-01 -1.79245584e-02 -1.18388045e+00 6.09986603e-01
-1.37927806e+00 -6.86538935e-01 -4.07571614e-01 1.80314672e+00
1.01026964e+00 3.32962163e-02 6.09439425e-02 1.45799845e-01
4.53930974e-01 -1.40794531e-01 -2.19616488e-01 -9.64776576e-01
2.87615776e-01 4.90843743e-01 4.00148481e-01 3.32907438e-01
-7.38622010e-01 6.61497533e-01 5.61629629e+00 2.33143285e-01
-1.18635941e+00 -4.54945713e-02 4.53144819e-01 -2.05060482e-01
-6.44325376e-01 -3.82072246e-03 -7.83688426e-01 3.77678752e-01
1.23404026e+00 -9.49445441e-02 5.00833511e-01 1.01188529e+00
-1.39819875e-01 1.02437519e-01 -1.33703995e+00 3.64572316e-01
1.34968758e-01 -1.04861617e+00 -2.20155463e-01 2.57022157e-02
6.31401539e-01 5.14174290e-02 7.71332309e-02 8.38229120e-01
3.11552048e-01 -1.20267773e+00 6.89024627e-01 3.74269187e-01
2.18986541e-01 -9.54242349e-01 9.51422632e-01 3.37777436e-01
-7.16309488e-01 -3.40736717e-01 -2.45948136e-01 -5.45257926e-01
-5.06441534e-01 5.78441262e-01 -1.04662323e+00 1.23897322e-01
5.16197801e-01 4.37358916e-01 -1.16312790e+00 9.07394648e-01
-2.96241224e-01 9.25876439e-01 5.09308241e-02 -2.77759701e-01
1.05094731e-01 2.42445603e-01 9.42519754e-02 1.31823814e+00
3.04809690e-01 -2.40832567e-02 -1.29922435e-01 1.09851718e+00
-4.81711626e-02 2.51054823e-01 -7.45339096e-01 -2.87875384e-01
3.04320246e-01 9.70699131e-01 -5.50648987e-01 -5.76406866e-02
-7.37523675e-01 5.45145035e-01 4.91497606e-01 5.88961095e-02
-6.39259279e-01 -5.26496768e-01 5.64340591e-01 2.34632358e-01
4.65028286e-01 1.06113320e-02 -6.66226268e-01 -1.02484334e+00
-3.06720641e-02 -9.72991526e-01 1.62357599e-01 -5.37051558e-01
-8.05210650e-01 3.69532228e-01 -1.98249564e-01 -6.85632050e-01
-1.30861834e-01 -6.81050718e-01 -6.12538159e-01 9.67731476e-01
-1.49969232e+00 -7.74110079e-01 -1.22001879e-01 1.78868517e-01
2.62839228e-01 -2.44500071e-01 7.76885986e-01 2.92608440e-01
-5.95060408e-01 9.39250588e-01 -5.64453006e-02 3.00264150e-01
4.36907321e-01 -1.28735793e+00 5.10282755e-01 8.41293275e-01
4.95088026e-02 1.20161629e+00 7.21960366e-01 -5.13591290e-01
-1.56750393e+00 -1.02631533e+00 8.45423520e-01 -4.98035669e-01
4.56359804e-01 -2.44418681e-01 -1.03039014e+00 8.56860936e-01
1.66169509e-01 -1.99026644e-01 9.85774159e-01 4.63828593e-01
-5.77328503e-01 3.87063801e-01 -1.05627644e+00 4.36125934e-01
8.44392836e-01 -4.43066508e-01 -6.58261180e-01 -2.22114950e-01
9.94947910e-01 -2.19508082e-01 -1.02569163e+00 4.02261257e-01
6.20034754e-01 -1.05750680e+00 8.96166384e-01 -7.89090633e-01
8.92891824e-01 5.90822175e-02 -1.52074262e-01 -1.45057082e+00
-4.87635553e-01 1.50723845e-01 -7.77039528e-02 1.27520359e+00
6.78910494e-01 -5.11697769e-01 7.93467224e-01 7.75284290e-01
-3.85192305e-01 -8.69659960e-01 -4.89338100e-01 -4.84364629e-01
4.55228299e-01 -4.35023814e-01 8.07367504e-01 1.32136798e+00
1.89497635e-01 2.07003549e-01 1.99826717e-01 -5.06248400e-02
-1.35414019e-01 8.60922933e-02 7.14733183e-01 -1.55346906e+00
-8.73357773e-01 -6.88051641e-01 -3.94297361e-01 -4.83063340e-01
4.43460196e-01 -1.22666872e+00 -1.59512833e-01 -1.16666603e+00
3.07622612e-01 -6.15142822e-01 -3.53679687e-01 6.47635579e-01
-3.61281842e-01 -1.36996672e-01 4.75980908e-01 -3.03862602e-01
-3.18053335e-01 1.84924066e-01 8.28034341e-01 -3.61480787e-02
-2.31623501e-01 -8.57573599e-02 -1.12041032e+00 5.20974874e-01
1.29582191e+00 -5.54984927e-01 -5.12906432e-01 -7.26200938e-01
5.58556378e-01 -1.79681122e-01 1.18942209e-01 -7.93315411e-01
-1.21098068e-02 -7.64543414e-02 2.54181564e-01 1.25807583e-01
-6.45823032e-02 -8.18371415e-01 -1.79039519e-02 6.79512203e-01
-5.65055668e-01 3.80527794e-01 7.91803718e-01 1.67894050e-01
-2.58104146e-01 -1.06286967e+00 4.57048535e-01 -4.23506200e-01
-8.00666511e-01 -3.30985844e-01 -5.82455456e-01 -5.29334769e-02
8.22857261e-01 -1.06910311e-01 -3.62671316e-01 1.45796239e-02
-3.24619710e-01 -1.52954176e-01 7.18312323e-01 5.34779072e-01
3.25849533e-01 -9.46276844e-01 -4.45400149e-01 4.45633560e-01
1.18802533e-01 -1.60858646e-01 -6.68583158e-03 8.14812541e-01
-3.61120492e-01 5.17347991e-01 -4.11157727e-01 -2.31388256e-01
-1.42842531e+00 7.44717956e-01 2.39603654e-01 -2.84587383e-01
-7.60539100e-02 1.05270183e+00 -8.12613741e-02 -1.01065743e+00
5.56142107e-02 -5.92537165e-01 -5.59029877e-01 1.11619517e-01
4.27123874e-01 2.06801564e-01 3.39249283e-01 -2.71968842e-01
-2.30971411e-01 2.27568984e-01 -2.24253044e-01 1.58597142e-01
1.65639472e+00 4.17375565e-01 -2.71927267e-01 5.41781962e-01
1.20300269e+00 4.03216273e-01 -6.45130277e-01 1.78939536e-01
3.21745425e-01 -2.83877522e-01 2.62183130e-01 -9.87421036e-01
-1.16350484e+00 1.04994500e+00 6.15168571e-01 4.49921191e-02
1.09784174e+00 -3.13455373e-01 3.01099509e-01 6.15909576e-01
3.97227615e-01 -6.84458673e-01 -2.01283053e-01 9.06379044e-01
4.59104270e-01 -1.13428414e+00 -1.60077423e-01 -3.02367240e-01
-3.91364545e-01 1.21150982e+00 9.98294890e-01 1.29240960e-01
4.78385061e-01 4.51324999e-01 -2.61289984e-01 -2.33113781e-01
-9.81402397e-01 -3.79217379e-02 2.21653387e-01 4.46630388e-01
1.33646071e+00 8.78624059e-03 -4.74348724e-01 1.09405863e+00
-3.56250197e-01 2.87064850e-01 8.62479925e-01 1.15644193e+00
-3.66975129e-01 -1.38878751e+00 -4.06203508e-01 6.16409540e-01
-5.63063264e-01 -5.66876292e-01 -7.19902635e-01 7.96857655e-01
5.10691762e-01 6.93356514e-01 -3.00776470e-03 -6.11180723e-01
5.03595591e-01 4.07282352e-01 4.96683210e-01 -9.99588072e-01
-1.28231585e+00 -7.56779909e-01 1.47952378e-01 -1.31302774e-01
-1.02664351e-01 -8.05135429e-01 -1.32719529e+00 -1.81923345e-01
-3.78126502e-01 3.56613189e-01 7.22063363e-01 6.39015555e-01
4.20661151e-01 9.35011744e-01 9.95350331e-02 -4.14460003e-01
-4.26499158e-01 -9.64881003e-01 -1.93198591e-01 2.63982207e-01
1.77463800e-01 -6.19533837e-01 -2.81547695e-01 -5.60506899e-03] | [7.664283275604248, 7.719470977783203] |
2401892b-394a-4b55-bf00-2c021566921c | marine-debris-detection-in-satellite | 2307.04128 | null | https://arxiv.org/abs/2307.04128v1 | https://arxiv.org/pdf/2307.04128v1.pdf | Marine Debris Detection in Satellite Surveillance using Attention Mechanisms | Marine debris is an important issue for environmental protection, but current methods for locating marine debris are yet limited. In order to achieve higher efficiency and wider applicability in the localization of Marine debris, this study tries to combine the instance segmentation of YOLOv7 with different attention mechanisms and explores the best model. By utilizing a labelled dataset consisting of satellite images containing ocean debris, we examined three attentional models including lightweight coordinate attention, CBAM (combining spatial and channel focus), and bottleneck transformer (based on self-attention). Box detection assessment revealed that CBAM achieved the best outcome (F1 score of 77%) compared to coordinate attention (F1 score of 71%) and YOLOv7/bottleneck transformer (both F1 scores around 66%). Mask evaluation showed CBAM again leading with an F1 score of 73%, whereas coordinate attention and YOLOv7 had comparable performances (around F1 score of 68%/69%) and bottleneck transformer lagged behind at F1 score of 56%. These findings suggest that CBAM offers optimal suitability for detecting marine debris. However, it should be noted that the bottleneck transformer detected some areas missed by manual annotation and displayed better mask precision for larger debris pieces, signifying potentially superior practical performance. | ['Richard Jiang', 'Yijie Zhu', 'Ao Shen'] | 2023-07-09 | null | null | null | null | ['semantic-segmentation', 'instance-segmentation'] | ['computer-vision', 'computer-vision'] | [-2.67547488e-01 1.54710919e-01 4.81272012e-01 4.96475637e-01
-6.60352349e-01 -6.52623832e-01 6.48409903e-01 5.76883733e-01
-8.67929995e-01 4.31093961e-01 1.90722391e-01 -2.57529676e-01
-3.47144634e-01 -6.79977834e-01 -4.93210167e-01 -1.07216942e+00
-4.06731248e-01 1.62243620e-01 6.77613795e-01 2.32400242e-02
5.57290137e-01 6.39798045e-01 -1.74722767e+00 4.71166009e-03
9.40979362e-01 6.86786354e-01 8.18458080e-01 7.45153069e-01
-3.16856086e-01 1.27636299e-01 -1.04204714e+00 -4.42628711e-01
4.63022441e-01 3.81329879e-02 -4.78770912e-01 -5.25299609e-01
4.33459312e-01 -5.44469245e-02 2.09195301e-01 9.74914789e-01
7.41329074e-01 2.23203465e-01 9.03632283e-01 -7.54233360e-01
-5.05343735e-01 5.19165039e-01 -5.24611175e-01 7.04431713e-01
1.53727725e-01 3.74802291e-01 9.00319159e-01 -9.03292596e-01
2.83799559e-01 1.05678546e+00 8.64213824e-01 1.71407402e-01
-1.14927316e+00 -5.85097730e-01 2.30546281e-01 2.17622723e-02
-1.42068219e+00 -1.75726637e-01 2.24071085e-01 -6.37643278e-01
1.04530823e+00 2.97615767e-01 8.33121479e-01 4.65155393e-01
2.06852928e-01 5.85631728e-01 1.27426195e+00 -2.79893070e-01
1.55777037e-01 1.32384270e-01 -1.05910413e-02 1.74788311e-01
8.68933439e-01 -1.83461905e-01 -3.45338941e-01 -6.03159517e-02
6.15375578e-01 -2.83712119e-01 -3.91726345e-01 3.09210330e-01
-8.30944002e-01 8.00641716e-01 4.36428785e-01 4.64327991e-01
-4.25554007e-01 -2.10970253e-01 3.41673464e-01 -9.47343707e-02
4.96110290e-01 8.01078677e-01 -3.49551380e-01 3.59077044e-02
-9.33512747e-01 1.50187939e-01 5.76975822e-01 5.45281649e-01
5.03458440e-01 3.09906602e-01 -2.80513197e-01 8.34163129e-01
3.38466376e-01 1.10838652e+00 3.34351569e-01 -4.94840503e-01
3.53098601e-01 5.71656644e-01 2.53132820e-01 -9.89560723e-01
-7.70135581e-01 -4.62773353e-01 -3.28871638e-01 4.60198373e-01
6.23190284e-01 5.31811453e-02 -8.84995699e-01 1.16886806e+00
8.01173225e-02 -1.07833490e-01 3.69258374e-02 1.01882112e+00
9.39449072e-01 6.77880645e-01 5.24939358e-01 4.84690145e-02
1.64782298e+00 -3.96069109e-01 -4.67666239e-01 -3.71005177e-01
4.83877599e-01 -7.02778697e-01 9.98914301e-01 1.47540048e-01
-8.49737287e-01 -4.42514896e-01 -9.43172634e-01 3.25046629e-01
-5.82851827e-01 2.75305212e-01 2.21216634e-01 8.01503003e-01
-1.23079526e+00 2.72212446e-01 -6.70367539e-01 -4.91933346e-01
3.96067142e-01 2.59744287e-01 -3.48865569e-01 2.52338707e-01
-6.33993566e-01 1.14615297e+00 1.94347557e-02 3.84216934e-01
-1.14697254e+00 -6.81367636e-01 -8.15349340e-01 1.85803398e-01
7.55052119e-02 -1.94178268e-01 7.46044099e-01 -4.33186501e-01
-1.07918251e+00 7.46923149e-01 5.00254780e-02 -6.53303325e-01
4.36299860e-01 -6.53721631e-01 -1.25771523e-01 4.70559567e-01
4.19728369e-01 7.53403664e-01 5.42855918e-01 -1.49996126e+00
-1.00255728e+00 -3.07676405e-01 1.17276497e-01 1.90161362e-01
-2.03550458e-01 7.43145272e-02 1.38297424e-01 -5.29160142e-01
-1.10979699e-01 -8.87500763e-01 -2.27466933e-02 -1.66961059e-01
-8.04522559e-02 -2.22202212e-01 7.65391767e-01 -9.62108910e-01
1.18538487e+00 -1.97890413e+00 -2.40866095e-01 6.03284761e-02
1.20674938e-01 7.96973646e-01 -3.52223255e-02 4.16350782e-01
1.60863206e-01 3.60058576e-01 -4.34213728e-01 -1.95053399e-01
-1.27833515e-01 7.73558915e-02 8.11996928e-04 7.11346805e-01
4.13660854e-01 4.64639693e-01 -9.66678619e-01 -5.00548065e-01
3.65829349e-01 5.29320836e-01 -2.98878908e-01 -3.95718822e-03
1.93986505e-01 2.97912002e-01 -1.12781428e-01 9.91738617e-01
8.50329399e-01 3.50464225e-01 -2.47804612e-01 -8.43020529e-02
-9.12899673e-01 -1.17863290e-01 -1.10434806e+00 9.62021649e-01
-3.48980814e-01 1.13105047e+00 3.50198001e-01 -4.78420317e-01
1.10759068e+00 1.28615782e-01 2.45642886e-01 -8.18361163e-01
1.54634297e-01 1.53133631e-01 1.01370923e-01 -8.69527042e-01
8.61021876e-01 -1.26753077e-01 1.64182365e-01 4.86547127e-02
-3.28321517e-01 1.70532212e-01 9.45045650e-02 4.91776466e-02
8.73292089e-01 -3.27452309e-02 3.35781239e-02 -7.78848112e-01
4.16639000e-01 1.39271885e-01 1.56360909e-01 8.59378278e-01
-4.09686416e-01 7.24909544e-01 2.74003178e-01 -2.32268617e-01
-6.14915729e-01 -8.29867423e-01 -2.83142358e-01 1.13231480e+00
4.05761957e-01 -4.70431410e-02 -8.54430377e-01 -5.43991208e-01
1.17159653e-02 6.24948204e-01 -8.61539185e-01 1.85193494e-01
-6.04695141e-01 -9.60922301e-01 8.20920587e-01 7.06760287e-01
6.26888096e-01 -1.09483576e+00 -1.13015294e+00 1.07000656e-02
-8.95602182e-02 -6.03570104e-01 -5.48902154e-02 2.66976804e-01
-7.75610924e-01 -1.21593928e+00 -1.09681404e+00 -5.68398833e-01
6.91306174e-01 7.30399251e-01 8.17144275e-01 3.11108708e-01
-2.19435558e-01 3.15036386e-01 -6.04453564e-01 -6.84050143e-01
1.43601664e-03 -8.63921456e-03 -2.79543132e-01 -2.49102801e-01
1.19121157e-01 3.82208787e-02 -1.03269684e+00 3.79334331e-01
-9.28437471e-01 -6.13897085e-01 9.44437325e-01 6.97754860e-01
3.29340100e-01 -2.02778384e-01 6.69454277e-01 -3.77950102e-01
2.66986996e-01 -5.61447620e-01 -5.68096101e-01 -2.32598111e-01
-4.09310758e-01 -2.78033525e-01 2.70010620e-01 -3.93014342e-01
-1.00454259e+00 -1.57603860e-01 -3.19517940e-01 -9.00476724e-02
-3.20469439e-01 2.18155429e-01 -1.34944469e-01 -1.53810486e-01
5.96369207e-01 1.02358304e-01 1.13962844e-01 -6.75621033e-01
-2.25059077e-01 6.83062375e-01 4.61763203e-01 -2.05720171e-01
3.94699126e-01 6.96903110e-01 -2.91891187e-01 -1.25497484e+00
-3.97389948e-01 -7.08547115e-01 -6.03896677e-01 -3.67349088e-01
1.16369009e+00 -8.19264352e-01 -5.96228004e-01 4.11321163e-01
-7.72050560e-01 -3.77565533e-01 -9.65815112e-02 5.27836144e-01
5.43884933e-02 4.38312292e-01 -1.67818353e-01 -1.24504745e+00
-5.34656227e-01 -1.18089485e+00 1.42456317e+00 2.85492301e-01
-7.53209963e-02 -9.66529131e-01 -1.10338405e-01 1.37627214e-01
6.11862838e-01 4.37348224e-02 5.51357448e-01 -8.40111613e-01
-2.18148321e-01 -6.29411116e-02 -5.65520644e-01 -1.19108884e-02
3.65645997e-02 -1.58992186e-02 -1.12866950e+00 -3.54980111e-01
-3.76745433e-01 5.65604568e-01 8.79766107e-01 6.33759737e-01
3.07821184e-01 -6.00807257e-02 -4.18555081e-01 9.55827609e-02
1.53742909e+00 5.21292448e-01 8.71760547e-01 1.07744098e+00
2.76894510e-01 7.92018771e-01 7.11768150e-01 3.87848616e-01
5.38720727e-01 5.72368681e-01 9.76416886e-01 -1.08832330e-01
-5.36961555e-01 2.61619747e-01 5.63573658e-01 1.73632186e-02
-4.08544540e-01 -2.83385396e-01 -1.08074474e+00 1.14319742e+00
-1.25386357e+00 -8.74084651e-01 -6.37181342e-01 2.34542418e+00
-1.13391124e-01 7.71491230e-02 3.65135193e-01 3.30027014e-01
9.44937170e-01 2.99904346e-02 1.64674241e-02 -3.50362033e-01
-4.25575793e-01 -1.26266584e-01 9.42683995e-01 2.97823012e-01
-1.21674776e+00 8.33881140e-01 6.76145983e+00 6.58787847e-01
-1.10996890e+00 8.42020065e-02 1.80158600e-01 8.47212598e-02
3.28072580e-03 -3.71412933e-02 -1.13442588e+00 8.98195565e-01
9.07724977e-01 4.08095270e-01 -2.24908978e-01 5.20844162e-01
4.06405509e-01 -7.00075686e-01 -1.65571168e-01 3.95830989e-01
1.58979908e-01 -8.27975571e-01 1.11411870e-01 1.72966003e-01
5.29376447e-01 1.27022564e-01 -2.20062450e-01 1.18597820e-01
1.10362619e-01 -9.21961308e-01 1.03593683e+00 5.31369865e-01
4.74551141e-01 -5.11269212e-01 1.44415188e+00 1.40271381e-01
-1.40969527e+00 -3.14974010e-01 -4.56038535e-01 -3.02071005e-01
1.10878356e-01 2.72997677e-01 -9.53708112e-01 5.23804724e-01
1.30828500e+00 1.53131396e-01 -8.21910679e-01 1.75784516e+00
-1.37836114e-01 6.32888496e-01 -4.85192090e-01 -1.47873074e-01
4.33322191e-01 -3.28569524e-02 8.60754967e-01 1.79031229e+00
5.40958107e-01 -2.14971989e-01 -1.22340478e-01 2.62710154e-01
6.43659294e-01 2.19955862e-01 -5.85637629e-01 2.91617870e-01
6.43712223e-01 1.10886037e+00 -1.25663280e+00 -1.17722660e-01
-4.26136732e-01 4.41495031e-01 -1.41258299e-01 -8.26455466e-03
-8.46692741e-01 -4.50796336e-01 6.51443362e-01 3.65762562e-01
6.47949874e-01 -1.09597102e-01 -5.24272919e-01 -3.89020413e-01
-3.18469018e-01 -3.38426918e-01 4.57781047e-01 -7.66613781e-01
-8.69389117e-01 7.29729652e-01 2.80106723e-01 -1.35748851e+00
5.22668123e-01 -8.28234553e-01 -6.70477808e-01 8.90923798e-01
-1.47076213e+00 -1.34030342e+00 -6.86273992e-01 -1.35948524e-01
7.01974750e-01 9.69626382e-02 4.35022444e-01 2.51113296e-01
-4.19862837e-01 2.81591952e-01 7.51437098e-02 2.88819633e-02
4.79425728e-01 -1.43709600e+00 -2.37821247e-02 7.94970334e-01
-3.34756702e-01 3.91694516e-01 7.53228068e-01 -9.95055676e-01
-1.01250768e+00 -1.13134134e+00 6.02887750e-01 -4.52072918e-01
4.64459002e-01 3.06478471e-01 -1.17959535e+00 2.62968749e-01
1.55627474e-01 -4.51474607e-01 6.17712498e-01 -4.41950947e-01
1.12755820e-01 1.53665513e-01 -1.31234467e+00 5.23918033e-01
7.69372165e-01 -7.38135278e-02 -5.63825548e-01 -8.99419039e-02
1.24920890e-01 -2.40443841e-01 -9.56522286e-01 4.00155991e-01
7.29588568e-01 -1.00340426e+00 9.34617639e-01 1.12123698e-01
4.39502895e-02 -7.93449759e-01 -6.91751018e-02 -1.20058513e+00
-4.25154656e-01 -1.62187107e-02 4.73984659e-01 1.40063405e+00
5.35203636e-01 -7.81567872e-01 5.35519183e-01 1.46811143e-01
-8.05048645e-01 -4.27834362e-01 -1.00037372e+00 -8.09422672e-01
1.31877124e-01 -2.50479937e-01 4.00098175e-01 4.49856699e-01
-4.20846909e-01 -3.68799686e-01 -1.31425276e-01 7.45843768e-01
2.14468390e-01 -6.15616888e-02 5.93024790e-01 -1.34094846e+00
4.66202885e-01 -8.74538839e-01 -5.13000667e-01 -5.31985402e-01
-2.44747013e-01 -5.76790154e-01 3.22369754e-01 -1.96892726e+00
7.07202777e-02 -3.85936111e-01 5.55443242e-02 4.39533263e-01
-2.67995447e-01 6.67364359e-01 3.02776337e-01 3.55323642e-01
-3.01292837e-01 1.96773708e-01 8.95075858e-01 -1.80456251e-01
-4.10256118e-01 -7.53782243e-02 -8.43757510e-01 7.77820230e-01
7.75808811e-01 -3.68759245e-01 1.48767367e-01 -6.68467760e-01
-6.41625822e-02 -4.07236159e-01 3.82740051e-01 -1.21744061e+00
3.81027609e-01 1.47776797e-01 4.49086517e-01 -6.61539316e-01
2.62976974e-01 -7.20141351e-01 2.52561837e-01 7.58410990e-01
2.35781103e-01 2.20102847e-01 5.98413110e-01 4.52891171e-01
-3.43163833e-02 -6.54635012e-01 6.68468237e-01 -1.33303344e-01
-9.47344899e-01 -4.35090959e-01 -9.72132504e-01 -3.39765042e-01
1.07490873e+00 -8.17213535e-01 -6.58640087e-01 -6.21269979e-02
-7.64271080e-01 2.48168916e-01 5.15442967e-01 1.75668135e-01
4.68633324e-01 -4.93504226e-01 -7.04221964e-01 4.80416752e-02
2.41195168e-02 2.77632680e-02 3.97241235e-01 1.19485581e+00
-9.54976618e-01 3.81039053e-01 -2.27372006e-01 -7.26511061e-01
-1.29527569e+00 5.87585270e-02 2.42022887e-01 5.08789942e-02
-7.36285090e-01 8.09555829e-01 3.36177438e-01 1.05918236e-01
1.06154248e-01 -2.88880736e-01 -7.61574030e-01 6.13265514e-01
6.74164891e-01 1.15148997e+00 2.25104585e-01 -1.04220998e+00
-5.07452607e-01 9.83376563e-01 3.84348005e-01 2.53691941e-01
1.20318580e+00 -2.26921156e-01 8.80256966e-02 3.66170295e-02
3.12717021e-01 5.10560513e-01 -1.38143730e+00 4.91939992e-01
2.70704478e-01 -5.35890400e-01 5.69481961e-03 -9.16917264e-01
-9.87094343e-01 7.15745211e-01 8.32070589e-01 4.96332765e-01
9.52945054e-01 1.57972828e-01 3.76578689e-01 1.44813175e-03
3.22508633e-01 -7.98678994e-01 -2.20695317e-01 4.77587193e-01
9.60504889e-01 -1.11715674e+00 3.45734805e-02 -1.82562932e-01
-7.00317025e-01 8.43007505e-01 6.43882573e-01 -1.56707421e-01
3.73648852e-01 4.11029577e-01 1.49534613e-01 -4.45360839e-01
-5.96943013e-02 -8.61664057e-01 2.28417739e-01 9.54473317e-01
1.68182522e-01 -7.99011812e-02 -5.53175092e-01 7.62497544e-01
-1.11210905e-01 -7.16308773e-01 4.28040773e-01 8.35009634e-01
-1.02056313e+00 -3.94379318e-01 -9.98495579e-01 5.73670030e-01
-7.17932940e-01 3.25964876e-02 -3.86206985e-01 1.19604468e+00
4.55856740e-01 9.78194952e-01 3.76923025e-01 -1.43337443e-01
4.72162247e-01 -1.20025426e-01 -7.85192624e-02 -4.27311689e-01
-1.02661979e+00 2.65198141e-01 3.03305537e-01 -1.20708659e-01
-5.86943626e-01 -8.48941028e-01 -1.20800173e+00 1.39072537e-01
-7.26262748e-01 5.53455710e-01 8.68282616e-01 9.04742181e-01
2.05890343e-01 3.84705961e-01 1.34876445e-01 -1.38145661e+00
1.01714805e-01 -1.37463450e+00 -4.77710664e-01 -9.30575728e-02
2.29632169e-01 -1.06136870e+00 -5.68435431e-01 -5.09028919e-02] | [8.921570777893066, -0.8580042123794556] |
230b6223-b3f6-4f46-a21a-3c9272d8ba1a | abanicco-a-new-color-space-for-multi-label | 2211.08460 | null | https://arxiv.org/abs/2211.08460v1 | https://arxiv.org/pdf/2211.08460v1.pdf | ABANICCO: A New Color Space for Multi-Label Pixel Classification and Color Segmentation | In any computer vision task involving color images, a necessary step is classifying pixels according to color and segmenting the respective areas. However, the development of methods able to successfully complete this task has proven challenging, mainly due to the gap between human color perception, linguistic color terms, and digital representation. In this paper, we propose a novel method combining geometric analysis of color theory, fuzzy color spaces, and multi-label systems for the automatic classification of pixels according to 12 standard color categories (Green, Yellow, Light Orange, Deep Orange, Red, Pink, Purple, Ultramarine, Blue, Teal, Brown, and Neutral). Moreover, we present a robust, unsupervised, unbiased strategy for color naming based on statistics and color theory. ABANICCO was tested against the state of the art in color classification and with the standarized ISCC-NBS color system, providing accurate classification and a standard, easily understandable alternative for hue naming recognizable by humans and machines. We expect this solution to become the base to successfully tackle a myriad of problems in all fields of computer vision, such as region characterization, histopathology analysis, fire detection, product quality prediction, object description, and hyperspectral imaging. | ['Arrate Muñoz-Barrutia', 'Javier Pascau', 'Agapito Ledezma', 'Laura Nicolás-Sáenz'] | 2022-11-15 | null | null | null | null | ['fire-detection'] | ['time-series'] | [ 3.69616240e-01 -6.50728643e-01 4.43317043e-03 5.57063054e-03
-2.85567343e-01 -9.83920097e-01 2.98438638e-01 6.06385589e-01
-2.89035648e-01 6.67920351e-01 -4.63633984e-01 -5.81555307e-01
-1.92545190e-01 -7.57185936e-01 1.02519900e-01 -8.37451339e-01
4.43547964e-01 2.55677611e-01 -3.06997690e-02 -8.60568359e-02
4.48700637e-01 1.00229287e+00 -1.67238498e+00 8.47636387e-02
1.11480975e+00 1.34709108e+00 -7.96868652e-02 5.49952090e-01
-2.95192927e-01 5.48429489e-01 -3.46649528e-01 -3.55500519e-01
1.03703767e-01 -7.55766690e-01 -6.24811709e-01 8.66345987e-02
2.70907253e-01 -5.23672774e-02 4.61242855e-01 1.43483555e+00
1.66302532e-01 4.95139100e-02 9.79581714e-01 -1.19772625e+00
-1.14816940e+00 2.13650793e-01 -7.03612506e-01 -5.44317663e-01
2.11735189e-01 -6.47801012e-02 9.75155592e-01 -7.00174451e-01
4.81913745e-01 7.77895868e-01 2.47835040e-01 3.66922498e-01
-1.01932824e+00 -3.94026250e-01 -6.08714856e-02 5.03920555e-01
-1.54059184e+00 1.21821463e-01 7.06859767e-01 -6.42993510e-01
1.05350867e-01 7.54056215e-01 9.29711819e-01 4.19490606e-01
-3.60661820e-02 3.05543810e-01 1.97764480e+00 -5.92840374e-01
4.39606756e-01 3.19349587e-01 8.61922428e-02 8.65320027e-01
3.50662291e-01 -8.73112399e-03 -8.66003260e-02 9.59858578e-03
3.92336398e-01 7.35661462e-02 -2.04065859e-01 -2.29156747e-01
-8.09026599e-01 5.72449207e-01 7.05287814e-01 5.81560850e-01
-4.81026739e-01 -3.63102376e-01 -5.94779737e-02 -1.11937761e-01
4.00778025e-01 6.80263877e-01 -1.64678633e-01 2.08136901e-01
-7.94626415e-01 -4.10969108e-01 4.91778940e-01 4.22572762e-01
6.77034736e-01 -8.02694634e-02 -9.72747430e-03 9.79190230e-01
4.50246871e-01 8.98958027e-01 2.69749254e-01 -9.21214938e-01
-5.20884156e-01 7.36870408e-01 1.41657591e-01 -1.15038347e+00
-5.15407860e-01 3.05864383e-02 -8.52485061e-01 6.74334824e-01
5.50560117e-01 2.70125777e-01 -9.04652476e-01 1.24107552e+00
2.02274337e-01 -5.00130594e-01 5.54864779e-02 8.78689826e-01
6.26552224e-01 5.90325117e-01 2.05290943e-01 -1.06466852e-01
1.49977756e+00 -6.79731488e-01 -4.94287848e-01 -3.28438319e-02
1.97433144e-01 -8.49991977e-01 9.33005810e-01 8.19653392e-01
-6.52618766e-01 -3.99427861e-01 -9.80134964e-01 1.36007771e-01
-1.03887367e+00 5.31951666e-01 7.53953755e-01 9.96364117e-01
-7.62950897e-01 2.47167945e-01 -3.19681346e-01 -5.15494823e-01
3.50696057e-01 -8.95921066e-02 -4.41689730e-01 -2.56684065e-01
-7.89951205e-01 1.16030622e+00 4.24062282e-01 4.95924443e-01
-1.50290757e-01 -1.21548966e-01 -5.10149419e-01 -1.20857626e-01
1.05565354e-01 -1.38208151e-01 5.60281038e-01 -1.34953833e+00
-1.60524726e+00 1.10715997e+00 5.41224144e-02 7.66755315e-03
2.52104998e-01 2.08970636e-01 -8.16868722e-01 3.76297534e-01
-1.29791811e-01 6.36430264e-01 4.38675910e-01 -1.50195837e+00
-7.29924500e-01 -5.04990757e-01 -5.75407296e-02 -7.19030574e-02
-3.53672057e-01 5.86940581e-03 -2.38040298e-01 -1.53506413e-01
2.57358015e-01 -9.78280902e-01 5.72758950e-02 4.72925812e-01
-6.79181278e-01 -2.61101454e-01 4.64277506e-01 -8.14226747e-01
8.86440217e-01 -2.37015891e+00 3.68727371e-03 8.40992332e-01
8.11977908e-02 2.37940818e-01 -1.37608781e-01 1.90624837e-02
-3.97365727e-02 2.88898617e-01 -5.37197411e-01 5.27957141e-01
-1.08102292e-01 -1.85550496e-01 1.97288796e-01 5.32345295e-01
9.15571079e-02 5.00030696e-01 -9.83104467e-01 -4.82354909e-01
4.93323565e-01 4.25315261e-01 3.67228203e-02 -3.56341079e-02
-6.42518178e-02 1.73906580e-01 -2.84168929e-01 1.22253883e+00
8.78533423e-01 -1.37718469e-01 2.25908682e-01 -4.62387830e-01
-4.53044742e-01 -7.58250833e-01 -1.11360884e+00 1.20502996e+00
-4.29694653e-01 7.02846050e-01 3.11763976e-02 -5.78390539e-01
1.03972793e+00 -1.59528032e-02 5.89442253e-01 -7.77781010e-01
1.74952671e-01 5.89795351e-01 -4.84944060e-02 -3.87071609e-01
4.94115680e-01 -1.28055111e-01 1.83321610e-01 1.71443865e-01
-2.34515145e-01 -3.92031282e-01 5.75652122e-01 -3.60096335e-01
3.60105544e-01 1.52116939e-01 3.87431562e-01 1.18086236e-02
7.48777926e-01 1.36447057e-01 1.29785821e-01 2.23580554e-01
-4.56032455e-01 5.84129274e-01 4.91755366e-01 -6.55661449e-02
-6.99913502e-01 -1.07111049e+00 -2.69375235e-01 1.01716673e+00
3.74089658e-01 1.52536124e-01 -6.10672176e-01 -1.70335934e-01
3.60298492e-02 7.82888949e-01 -5.97725391e-01 -1.75522551e-01
8.57291892e-02 -8.13152611e-01 2.90580541e-01 1.72124550e-01
5.94012558e-01 -9.79452133e-01 -4.55510587e-01 -1.19753450e-01
-2.86762770e-02 -7.60861576e-01 -7.28344098e-02 2.06221923e-01
-1.99319169e-01 -1.59585607e+00 -9.23166335e-01 -5.23985684e-01
8.37477326e-01 3.22739452e-01 6.92574203e-01 6.48631379e-02
-6.94871128e-01 3.93415570e-01 -7.34329760e-01 -3.08505595e-01
-4.41621006e-01 -3.48666102e-01 -5.38270652e-01 3.49207282e-01
2.91107982e-01 3.08299512e-01 -6.90582335e-01 1.24310970e-01
-1.21132469e+00 6.64590299e-02 7.73906410e-01 5.39846718e-01
7.47178614e-01 1.09419987e-01 8.14097673e-02 -9.26603973e-01
3.54489952e-01 -1.91664264e-01 -6.57263994e-01 8.12966049e-01
-8.27959180e-01 -8.16178098e-02 5.96205711e-01 -1.67959873e-02
-9.36591208e-01 1.59263864e-01 -6.37902766e-02 2.56277591e-01
-5.18263459e-01 5.10261118e-01 -2.29781181e-01 -5.00919342e-01
5.62258959e-01 1.84571177e-01 5.73954470e-02 -2.16364384e-01
8.68351638e-01 8.31718922e-01 5.92539310e-01 -2.10554615e-01
7.19152153e-01 5.79854131e-01 3.09197366e-01 -9.54423070e-01
-5.91477096e-01 -5.92181921e-01 -6.94167554e-01 -5.79563320e-01
1.33501387e+00 -5.06431758e-01 -8.39934528e-01 7.13018596e-01
-9.73914146e-01 -8.94365534e-02 3.13659348e-02 3.36426288e-01
-1.76256344e-01 5.98885000e-01 -2.80092865e-01 -8.18700194e-01
-3.77773583e-01 -9.18416619e-01 6.20226264e-01 5.71169019e-01
1.62930518e-01 -9.37514842e-01 -1.27313703e-01 3.10333729e-01
5.08042991e-01 6.53969467e-01 1.34049928e+00 -1.67968899e-01
-1.31876826e-01 -2.87936628e-01 -6.04004264e-01 7.00849175e-01
3.22533995e-01 7.59050727e-01 -8.18768740e-01 2.15097934e-01
-5.06190538e-01 -2.84298599e-01 9.70281661e-01 2.07399547e-01
1.26240671e+00 1.39256954e-01 -4.11930233e-02 5.62911689e-01
1.89327335e+00 5.97109914e-01 4.12557364e-01 2.85098433e-01
5.59653699e-01 7.33129859e-01 4.57828939e-01 2.60491490e-01
1.41142040e-01 3.35875154e-01 7.54679143e-01 -6.54227376e-01
-5.76954931e-02 2.93605298e-01 -1.79448575e-02 2.58370370e-01
-2.69970983e-01 -2.70770103e-01 -7.89073467e-01 1.04859315e-01
-1.32380366e+00 -9.57781374e-01 -4.50962007e-01 2.04509258e+00
5.06143928e-01 -2.34619230e-01 9.64452326e-02 1.85058787e-01
1.06557345e+00 -2.87943035e-01 -4.25448269e-01 -4.11426902e-01
-5.97359061e-01 3.17671925e-01 6.73951209e-01 1.49762332e-01
-1.10906959e+00 6.93456590e-01 5.86339378e+00 6.57452404e-01
-1.58489025e+00 -2.72649318e-01 8.06540012e-01 5.48833430e-01
-2.49128237e-01 -2.02593729e-01 -3.06266313e-03 2.95322776e-01
4.89092290e-01 1.05280302e-01 8.84118974e-01 5.10967016e-01
2.68445928e-02 -4.99361306e-01 -5.85694551e-01 1.12940419e+00
3.85311097e-01 -9.95845079e-01 9.71480608e-02 -2.92750686e-01
6.28590345e-01 -3.30131620e-01 2.77164668e-01 -2.44938836e-01
6.78849444e-02 -8.06613207e-01 9.16727722e-01 8.24205875e-01
8.87694299e-01 -4.48341012e-01 4.56810385e-01 -3.89239669e-01
-1.18548560e+00 -2.21313044e-01 -3.09861153e-01 4.02370065e-01
-2.85245717e-01 4.54122573e-01 -1.90697640e-01 8.56632173e-01
3.67734969e-01 5.55168808e-01 -1.03993309e+00 1.51599741e+00
-2.60206163e-01 1.47023484e-01 -5.79681955e-02 -2.92796880e-01
1.35839581e-01 -8.16142440e-01 -1.68784872e-01 8.90753865e-01
6.43290937e-01 -1.03103004e-01 9.22124907e-02 9.95867252e-01
2.56604612e-01 4.89745021e-01 -1.29067272e-01 -4.49866414e-01
2.97582656e-01 1.59344804e+00 -1.41738176e+00 -1.97287366e-01
-4.39120471e-01 1.02097106e+00 -2.61117160e-01 5.83004594e-01
-7.82679200e-01 -6.05742693e-01 2.97012091e-01 -3.69535446e-01
-1.45715609e-01 -1.49424627e-01 -4.61660355e-01 -6.40427411e-01
-2.88118988e-01 -6.18459880e-01 3.97940785e-01 -1.13961041e+00
-1.24232984e+00 6.75530970e-01 -2.44234443e-01 -1.48934376e+00
2.25519747e-01 -1.29201937e+00 -3.58646303e-01 7.77790248e-01
-1.59522128e+00 -1.26051915e+00 -6.21743858e-01 4.57806110e-01
-1.69249311e-01 1.79938134e-02 1.19814157e+00 2.11675584e-01
-6.77748322e-01 5.66809699e-02 6.28569424e-01 9.81091186e-02
9.12774622e-01 -1.47864759e+00 -6.13164842e-01 5.69833457e-01
8.40570256e-02 5.85130043e-02 3.52530599e-01 -1.12389527e-01
-1.26982617e+00 -9.31987166e-01 3.37746263e-01 2.77414739e-01
6.66285396e-01 2.30104119e-01 -4.95852649e-01 -2.35891178e-01
2.29720995e-01 -4.78303060e-02 1.09144795e+00 7.30619160e-03
-6.39378548e-01 -5.11720479e-01 -1.08516920e+00 3.78953338e-01
3.90980989e-02 -6.09320819e-01 2.96945721e-02 4.73728895e-01
-2.17375442e-01 -3.08210813e-02 -9.12810266e-01 -5.97825237e-02
7.89642215e-01 -8.93414319e-01 7.27577507e-01 -9.90463048e-02
2.12352306e-01 -7.12231874e-01 -6.20713504e-03 -1.28193033e+00
-3.74759465e-01 1.73312783e-01 7.96456218e-01 1.10432482e+00
6.42105520e-01 -5.48830271e-01 3.47920537e-01 5.32562673e-01
-2.62197882e-01 -2.84862787e-01 -7.49161899e-01 -5.77052951e-01
-1.59338340e-02 -3.70146856e-02 5.07012486e-01 8.65746379e-01
-5.99740408e-02 -2.28312910e-01 -1.12426758e-01 1.02221631e-02
5.40222645e-01 6.24372244e-01 2.19200566e-01 -1.63442111e+00
4.39053110e-04 -1.12159228e+00 -4.30254430e-01 -8.60869363e-02
9.53603238e-02 -9.61207330e-01 1.44012207e-02 -1.91497111e+00
3.44762146e-01 -3.50603521e-01 -6.92080796e-01 5.78158736e-01
-1.25639871e-01 7.39522338e-01 5.41478813e-01 5.42881303e-02
-5.40010273e-01 -5.68579398e-02 1.30763698e+00 -6.80003285e-01
-1.21459194e-01 -1.97018236e-01 -8.63529444e-01 5.21338224e-01
7.82163799e-01 -1.15048736e-01 9.78286862e-02 -1.58763051e-01
4.20290142e-01 -3.08043271e-01 5.10456860e-01 -1.13127899e+00
8.53809118e-02 -6.16227150e-01 5.72804809e-01 -2.17584312e-01
2.29958192e-01 -1.19054008e+00 3.33115727e-01 6.57764971e-01
-2.53852844e-01 -1.67565256e-01 -1.59983616e-02 1.89433664e-01
-1.96350589e-01 -3.21381480e-01 1.08959961e+00 -2.57582832e-02
-1.26722145e+00 9.35858116e-02 -4.47876871e-01 -4.00208533e-01
1.27209234e+00 -2.79486090e-01 -6.96821451e-01 -3.39671820e-02
-7.41573453e-01 -2.52297163e-01 6.45018518e-01 1.45999849e-01
5.12166798e-01 -1.34674692e+00 -5.62155545e-01 7.32759684e-02
6.12285435e-01 -7.76676774e-01 4.60055977e-01 6.46578908e-01
-1.16339684e+00 2.67317206e-01 -8.28545928e-01 -4.08762276e-01
-1.16295075e+00 4.90493834e-01 2.69821107e-01 3.29529852e-01
6.05696924e-02 3.74036521e-01 -1.93346277e-01 1.30421907e-01
-1.89625006e-02 -4.67244804e-01 -6.63728297e-01 3.45160961e-01
1.48929298e-01 5.89373887e-01 3.63060758e-02 -8.27546656e-01
-3.84276420e-01 9.23638940e-01 6.13564193e-01 2.49699444e-01
8.73351574e-01 9.28860083e-02 -7.27328420e-01 6.93742037e-01
8.69740605e-01 1.41794458e-02 -7.23762333e-01 3.99576962e-01
-5.45821711e-02 -3.49697024e-01 1.80612445e-01 -1.43685579e+00
-1.01423073e+00 7.53927410e-01 1.12204516e+00 8.35081160e-01
1.52054286e+00 -1.83776885e-01 2.40684882e-01 2.13086769e-01
2.31690764e-01 -1.39224315e+00 -2.17677563e-01 3.90306413e-02
6.39954209e-01 -1.27471435e+00 -2.95221508e-02 -4.62164581e-01
-6.78187549e-01 1.50895762e+00 2.84905612e-01 3.76828015e-01
6.60226822e-01 -4.20005411e-01 4.80279952e-01 2.87461188e-02
-5.69443032e-02 -7.29609430e-01 7.00728178e-01 6.85026944e-01
7.57487118e-01 4.28779900e-01 -4.89471108e-01 6.39674366e-02
4.29981887e-01 -2.74425626e-01 5.18375099e-01 4.38849777e-01
-6.15214705e-01 -9.78168547e-01 -6.60371602e-01 5.03621340e-01
-2.19046161e-01 1.80463064e-02 -7.28704751e-01 8.50072980e-01
4.02632445e-01 1.23503888e+00 -9.03877057e-03 -1.81620285e-01
1.41841322e-01 2.86236815e-02 5.25226891e-01 -1.94319591e-01
-1.84089974e-01 1.33820757e-01 -2.69535929e-01 -2.19529510e-01
-7.66729414e-01 -4.62559015e-01 -1.16157019e+00 -5.27290776e-02
-4.13099408e-01 3.25017154e-01 1.14546227e+00 7.34298050e-01
-2.60178864e-01 3.28793287e-01 9.09103990e-01 -2.67306805e-01
5.10275923e-02 -7.21023858e-01 -1.14678383e+00 5.65999210e-01
1.36637706e-02 -5.61854243e-01 -3.04695129e-01 7.25077614e-02] | [10.347616195678711, -2.458803415298462] |
7fdacf5c-a344-4d53-ac7c-53efa434cb94 | text-coherence-analysis-based-on-deep-neural | 1710.07770 | null | http://arxiv.org/abs/1710.07770v1 | http://arxiv.org/pdf/1710.07770v1.pdf | Text Coherence Analysis Based on Deep Neural Network | In this paper, we propose a novel deep coherence model (DCM) using a
convolutional neural network architecture to capture the text coherence. The
text coherence problem is investigated with a new perspective of learning
sentence distributional representation and text coherence modeling
simultaneously. In particular, the model captures the interactions between
sentences by computing the similarities of their distributional
representations. Further, it can be easily trained in an end-to-end fashion.
The proposed model is evaluated on a standard Sentence Ordering task. The
experimental results demonstrate its effectiveness and promise in coherence
assessment showing a significant improvement over the state-of-the-art by a
wide margin. | ['Zhongfei Zhang', 'Yingming Li', 'Baiyun Cui', 'Yaqing Zhang'] | 2017-10-21 | null | null | null | null | ['sentence-ordering'] | ['natural-language-processing'] | [ 5.78411184e-02 -1.30140170e-01 -1.83073003e-02 -7.95488894e-01
-8.34012568e-01 -1.89382315e-01 9.01280880e-01 3.66893083e-01
-4.30657536e-01 4.49016541e-01 9.71533716e-01 5.35787344e-02
-9.84542668e-02 -5.55010021e-01 -3.14486653e-01 -4.68963921e-01
3.64541784e-02 3.48894417e-01 -2.20407650e-01 -2.48181164e-01
4.93902147e-01 -9.49182808e-02 -1.18909180e+00 4.12085921e-01
1.13014841e+00 8.24966609e-01 2.09357753e-01 6.60422385e-01
1.22591734e-01 1.02462053e+00 -6.41176879e-01 1.00567834e-02
-4.08790171e-01 -3.12580794e-01 -9.82673466e-01 3.70426551e-02
7.98144162e-01 -4.29591238e-01 -6.44495547e-01 1.05388618e+00
4.94223565e-01 1.58574596e-01 5.28087914e-01 -6.33768141e-01
-7.45510161e-01 9.21954095e-01 -3.22974205e-01 3.93630385e-01
5.78865051e-01 -9.65967774e-03 1.61749315e+00 -1.02039242e+00
3.49743724e-01 1.31523681e+00 4.20731634e-01 -5.16294129e-02
-1.11964917e+00 -2.79953063e-01 1.86554566e-01 2.20707431e-01
-1.24863219e+00 -3.82193238e-01 9.34036136e-01 -5.24279892e-01
1.37623096e+00 2.07467139e-01 5.21847188e-01 9.22388673e-01
5.25009811e-01 9.22546625e-01 7.74866700e-01 -3.13905060e-01
8.44058692e-02 -5.53225219e-01 7.54503310e-01 4.66569513e-01
4.22009379e-02 -2.70757407e-01 -7.42803693e-01 3.27294841e-02
1.00219101e-01 -1.70481846e-01 -2.49809042e-01 1.22737192e-01
-1.39343452e+00 9.49921966e-01 6.93133891e-01 5.74484944e-01
-2.93183476e-01 2.80744284e-01 6.65885389e-01 2.68417865e-01
9.81360555e-01 4.46109742e-01 1.30893931e-01 -2.41327062e-01
-1.24437034e+00 4.22646105e-01 6.46601796e-01 6.98848605e-01
3.17087203e-01 -8.34912881e-02 -6.67628884e-01 7.33006239e-01
5.00150502e-01 3.51552188e-01 5.00281751e-01 -7.06362188e-01
6.60900235e-01 4.58638012e-01 -1.99826032e-01 -1.34331322e+00
-7.75749445e-01 -6.31813645e-01 -1.08057201e+00 -5.09046674e-01
-1.99719876e-01 7.88196996e-02 -2.02494264e-01 1.80360925e+00
-1.55195817e-01 2.07537755e-01 5.93961403e-02 7.38733768e-01
1.21647584e+00 7.39604592e-01 -1.93475068e-01 -4.29648191e-01
1.25140500e+00 -1.07233298e+00 -1.13524354e+00 -9.83182490e-02
5.83099306e-01 -7.87591338e-01 1.16981244e+00 2.06098691e-01
-1.22652018e+00 -6.86617792e-01 -1.25010419e+00 -3.12584907e-01
1.08449526e-01 2.25647077e-01 6.18021786e-01 -4.73392941e-02
-1.19768214e+00 5.73753238e-01 -9.32911158e-01 -3.76346201e-01
2.14071199e-01 1.80046767e-01 -4.14768904e-02 1.02781333e-01
-1.47986531e+00 7.48768151e-01 5.13700724e-01 1.09122284e-01
-4.28833306e-01 -4.82167959e-01 -9.59322870e-01 5.42772949e-01
-1.09083682e-01 -1.18848825e+00 1.33076310e+00 -5.10150433e-01
-1.77563298e+00 6.63991511e-01 -2.39424899e-01 -5.26936352e-01
2.76362777e-01 -7.03515410e-01 -3.32230896e-01 2.49509975e-01
2.24689722e-01 4.05593634e-01 4.31426138e-01 -5.63622057e-01
-1.87359348e-01 -9.16881040e-02 7.03193396e-02 4.10952210e-01
-7.21341789e-01 1.55445874e-01 -1.92866385e-01 -6.76180840e-01
-5.53070456e-02 -7.57285416e-01 1.05467059e-01 -5.57848036e-01
-6.95592582e-01 -8.46380591e-01 5.99317491e-01 -3.30556273e-01
1.74132180e+00 -1.97105098e+00 2.75540292e-01 -2.53553569e-01
5.37402809e-01 -1.01195360e-02 -2.67375141e-01 8.51556778e-01
1.64696768e-01 2.84474101e-02 -3.02163929e-01 -5.76989651e-01
2.17089832e-01 -2.97722965e-01 -4.42348331e-01 5.29106259e-01
2.34632969e-01 1.02291536e+00 -9.27024066e-01 -4.32738602e-01
1.96385175e-01 2.25752950e-01 -6.83053732e-01 4.80969965e-01
-3.11306238e-01 3.92492324e-01 -2.23402441e-01 8.07557181e-02
7.00377226e-01 -6.80062234e-01 5.34542322e-01 -1.65318981e-01
-8.40506926e-02 8.95202696e-01 -6.03105009e-01 2.02348948e+00
-3.78369451e-01 1.30690229e+00 -5.65424442e-01 -9.12845910e-01
8.16241443e-01 1.74283937e-01 3.90507996e-01 -7.25205183e-01
1.70651361e-01 2.64343563e-02 2.61980355e-01 -4.71674204e-01
9.38303590e-01 1.82886213e-01 -5.45662761e-01 7.65359342e-01
4.38552350e-02 -1.38827518e-01 3.67365509e-01 4.20879841e-01
9.29563403e-01 -4.27935421e-01 5.54011703e-01 -6.77984357e-01
6.70319557e-01 -3.83481592e-01 1.49012104e-01 5.05383670e-01
-2.37730443e-02 5.48313379e-01 5.85861564e-01 -2.58553237e-01
-8.70743155e-01 -1.04931676e+00 -2.79001951e-01 9.42870736e-01
1.67760894e-01 -8.35753560e-01 -7.56355286e-01 -3.92702222e-01
-2.26573288e-01 8.11302722e-01 -5.53167880e-01 -1.65773675e-01
-6.72572315e-01 -4.69937623e-01 4.29637432e-01 6.14923596e-01
7.51521111e-01 -8.62686038e-01 -6.32104516e-01 1.12023376e-01
-6.90575719e-01 -1.36562085e+00 -7.11038113e-01 -2.40530774e-01
-7.85840929e-01 -7.86278427e-01 -3.00050139e-01 -8.65423977e-01
2.19634339e-01 4.41888928e-01 1.38893533e+00 -5.68614574e-03
9.62718502e-02 8.21978226e-03 -2.10220978e-01 6.15567155e-02
-7.82133266e-02 4.32179362e-01 1.23225421e-01 -1.91463768e-01
4.98553395e-01 -5.64173281e-01 -7.89492786e-01 -3.21203679e-01
-9.81601059e-01 1.16400532e-01 2.89675057e-01 1.12672853e+00
1.95365161e-01 -1.21663526e-01 7.71752417e-01 -8.29470992e-01
1.50575304e+00 -5.71562111e-01 -2.19814256e-01 1.65574342e-01
-5.58037341e-01 4.56026904e-02 5.02986073e-01 1.46268317e-02
-9.01868045e-01 -5.64570308e-01 -4.45527695e-02 -2.98497593e-03
7.02126920e-02 1.17664194e+00 6.49653450e-02 7.96745777e-01
3.30758601e-01 1.69284910e-01 -3.49456847e-01 -2.12750807e-01
6.15702510e-01 8.95514309e-01 6.96944714e-01 -3.23109239e-01
2.80057400e-01 2.95054257e-01 -3.25000703e-01 -8.27382803e-01
-1.22254610e+00 -5.46772718e-01 -9.72850144e-01 -9.95113850e-02
8.54581177e-01 -1.20956278e+00 -4.80427563e-01 4.20836657e-01
-1.60185790e+00 -5.00719482e-03 1.60454988e-01 6.19166791e-01
-4.87471491e-01 6.43487930e-01 -6.02810681e-01 -2.18690187e-01
-8.81925404e-01 -9.71590221e-01 1.25459909e+00 1.78939745e-01
-7.09875107e-01 -1.36355400e+00 5.62336206e-01 8.93607587e-02
4.45233703e-01 2.95469552e-01 1.01159108e+00 -8.25094998e-01
-3.10794055e-01 -8.14181045e-02 -2.14277089e-01 1.49742007e-01
9.81584191e-02 1.65878087e-01 -8.62093806e-01 -3.70502472e-01
-5.30314930e-02 -5.28278828e-01 1.18400419e+00 5.60363770e-01
1.01139295e+00 -3.09267581e-01 -5.37722893e-02 3.44977349e-01
1.35993361e+00 -4.76751745e-01 4.52790737e-01 2.14623660e-01
5.81988752e-01 3.50304872e-01 4.29449648e-01 7.69624591e-01
7.36886442e-01 7.52730668e-01 3.11501086e-01 2.12909896e-02
-2.03374609e-01 1.02094203e-01 1.94699705e-01 1.75470722e+00
5.81238866e-01 -4.57714051e-01 -1.06364810e+00 4.05223578e-01
-2.28919387e+00 -1.19927788e+00 -2.65178710e-01 1.65958130e+00
9.76484239e-01 3.52954894e-01 -1.57675087e-01 -1.39629290e-01
4.86230910e-01 7.50203550e-01 -2.97687858e-01 -6.79623187e-01
-1.75606787e-01 1.28375992e-01 -5.12591183e-01 7.48713732e-01
-1.32798707e+00 9.71796334e-01 7.00168419e+00 5.17204762e-01
-1.05525565e+00 -4.30619158e-02 4.23029512e-01 -1.67954624e-01
-5.83627820e-01 -2.69038141e-01 -4.76478875e-01 4.94490653e-01
9.59321678e-01 -4.60986137e-01 1.09259915e-02 4.98442858e-01
4.42927778e-01 -2.26717904e-01 -1.35438764e+00 9.67774153e-01
3.43315303e-01 -1.55202115e+00 9.51662362e-02 -2.64934838e-01
9.86426175e-01 2.44786218e-01 1.40455231e-01 2.73979515e-01
1.54221743e-01 -1.12005723e+00 6.11329198e-01 9.35875177e-01
6.72494829e-01 -8.55352521e-01 9.94868040e-01 4.18884158e-01
-1.19121051e+00 1.37318537e-01 -4.42894548e-01 -5.41933656e-01
1.33999929e-01 1.06314945e+00 -4.88929182e-01 6.11380100e-01
3.35588932e-01 1.20762718e+00 -6.67091310e-01 6.95728719e-01
-2.09196866e-01 5.71215332e-01 5.64092547e-02 -4.27747548e-01
4.04735953e-01 -4.72659729e-02 4.60531741e-01 1.61808956e+00
1.07899174e-01 -1.10588409e-01 3.40032786e-01 1.06738353e+00
-4.16255176e-01 1.46910092e-02 -6.65460825e-01 -1.39688686e-01
7.18522191e-01 1.20869780e+00 -2.52604336e-01 -3.87896955e-01
-3.40881765e-01 8.58815312e-01 8.15452158e-01 1.54278487e-01
-7.71161973e-01 -4.78139430e-01 4.63233203e-01 -5.18202543e-01
1.64375767e-01 -4.78688657e-01 -6.99191570e-01 -1.33164620e+00
8.16583112e-02 -5.40277421e-01 -2.06695776e-02 -5.60793400e-01
-1.53335738e+00 8.87575567e-01 1.79142401e-01 -1.26274729e+00
-6.58838332e-01 -6.48552030e-02 -1.03520870e+00 9.29660916e-01
-1.53129590e+00 -1.00300395e+00 -4.27951217e-01 4.06860471e-01
6.93215847e-01 -3.06821644e-01 9.46123838e-01 -8.95203501e-02
-5.47061861e-01 6.12703741e-01 4.54691380e-01 -6.54969364e-03
5.72786868e-01 -1.32101548e+00 6.52693868e-01 8.50277722e-01
1.43518165e-01 8.69103909e-01 5.89937449e-01 -3.80411774e-01
-9.71333027e-01 -1.05968356e+00 1.59952211e+00 -2.16435209e-01
7.82025754e-01 -4.19567436e-01 -6.57103419e-01 2.58809924e-01
1.12212670e+00 -3.63519311e-01 9.92611706e-01 4.98490661e-01
-4.15653974e-01 9.10472944e-02 -4.97207314e-01 6.00876153e-01
6.63768828e-01 -8.91802907e-01 -9.22117054e-01 5.10799825e-01
9.71378922e-01 -2.42766023e-01 -8.79395247e-01 4.32895005e-01
4.70202535e-01 -9.06333625e-01 5.57072282e-01 -3.74225497e-01
1.07570875e+00 -9.07827020e-02 -4.43696827e-01 -1.35293615e+00
-6.01767480e-01 -4.71798271e-01 -4.59405571e-01 1.13826060e+00
9.21121314e-02 -1.19485736e-01 1.59257874e-01 7.17093274e-02
-3.43482524e-01 -1.03029251e+00 -1.02053976e+00 -4.17786241e-01
4.31481987e-01 -2.50174105e-01 3.74641955e-01 9.63625908e-01
6.55131400e-01 1.11725140e+00 -2.77746022e-01 -2.14562804e-01
2.94768423e-01 4.58442926e-01 4.81177449e-01 -1.23711789e+00
-1.32259980e-01 -8.48020136e-01 -1.40274495e-01 -1.43153954e+00
6.19438350e-01 -1.18113446e+00 1.15223721e-01 -1.78927171e+00
7.21108258e-01 3.10299248e-01 -3.02866071e-01 -1.08070590e-01
-6.00135028e-01 -1.59065619e-01 3.43489736e-01 5.10747552e-01
-1.09261036e+00 1.07469022e+00 1.04697490e+00 -5.22217214e-01
7.60531947e-02 -2.48064294e-01 -6.77455723e-01 5.08055508e-01
9.70265746e-01 -2.58276314e-01 -4.91511703e-01 -8.40250671e-01
2.81557530e-01 -5.61236255e-02 1.80097669e-01 -9.96555209e-01
4.86602843e-01 2.99454033e-01 1.01241946e-01 -1.21936405e+00
1.06445309e-02 -1.70406014e-01 -4.70899969e-01 3.70682210e-01
-1.14052880e+00 2.31000334e-01 4.97496426e-02 6.03611887e-01
-6.30073607e-01 -2.97374725e-02 6.59496367e-01 1.79851308e-01
-2.37107396e-01 4.15402912e-02 -3.93936157e-01 2.58984357e-01
5.61650634e-01 4.37075615e-01 -5.74435651e-01 -5.13835549e-01
-2.82485932e-01 2.16738418e-01 1.03826784e-02 4.63651866e-01
8.54631484e-01 -1.61676538e+00 -1.08136845e+00 -3.08952145e-02
2.72111148e-01 -2.56211340e-01 1.99191123e-01 9.31060314e-01
-3.43849063e-01 1.00657666e+00 1.89356655e-01 -1.07802594e+00
-1.21014512e+00 2.13696986e-01 2.43200183e-01 -7.89544582e-01
-6.28930807e-01 7.29319572e-01 2.70797402e-01 -3.99341732e-01
2.99466878e-01 -5.93791425e-01 -4.84564334e-01 -6.93785474e-02
5.81600726e-01 1.02453366e-01 -3.45966034e-02 -5.61584115e-01
-4.15302217e-01 4.30361539e-01 -3.15721810e-01 6.26561120e-02
1.24633467e+00 -3.59257877e-01 -4.29555416e-01 6.91155910e-01
1.58621144e+00 -1.62402600e-01 -9.85386968e-01 -6.22619689e-01
2.73482382e-01 -2.34542340e-01 1.13266021e-01 -5.50341666e-01
-6.03661001e-01 1.16726804e+00 3.29584301e-01 2.94124454e-01
9.80180800e-01 8.94110575e-02 8.50095749e-01 6.61906123e-01
-2.16817752e-01 -1.16670430e+00 6.46723628e-01 1.18950045e+00
1.18071485e+00 -1.34785688e+00 2.31354818e-01 3.74421030e-02
-5.58911026e-01 1.54458761e+00 3.52329969e-01 -5.28695524e-01
7.59026051e-01 6.97074831e-02 -2.55558491e-01 -4.86685246e-01
-1.18089032e+00 -3.17197591e-02 7.47476697e-01 3.06131728e-02
1.14606631e+00 2.79038221e-01 -7.21566558e-01 5.58437407e-01
-3.83950710e-01 -2.15301856e-01 5.12558043e-01 5.37000477e-01
-7.27690697e-01 -7.59826779e-01 2.57915169e-01 3.52665126e-01
-2.17431173e-01 -6.23806894e-01 -5.27731836e-01 5.18305898e-01
-4.10570800e-01 1.33960080e+00 4.47379768e-01 -6.61745429e-01
1.42136171e-01 -2.16058999e-01 2.28406966e-01 -7.02652812e-01
-6.48234606e-01 3.60133082e-01 2.59329200e-01 -3.94185603e-01
-7.41983891e-01 -6.58362925e-01 -1.25944185e+00 -5.34898341e-01
-3.03226262e-01 -7.31894895e-02 2.12727815e-01 1.17292178e+00
4.24025387e-01 1.01214767e+00 1.07202315e+00 -5.34125626e-01
-8.08645427e-01 -1.63007224e+00 -2.17254758e-01 3.17531377e-01
3.70364666e-01 -2.92615116e-01 -2.25735139e-02 -1.54184774e-01] | [11.90483570098877, 9.294374465942383] |
772ed768-ba0b-40e5-81c7-ca6a992e3851 | ci-avsr-a-cantonese-audio-visual-speech | 2201.03804 | null | https://arxiv.org/abs/2201.03804v2 | https://arxiv.org/pdf/2201.03804v2.pdf | CI-AVSR: A Cantonese Audio-Visual Speech Dataset for In-car Command Recognition | With the rise of deep learning and intelligent vehicle, the smart assistant has become an essential in-car component to facilitate driving and provide extra functionalities. In-car smart assistants should be able to process general as well as car-related commands and perform corresponding actions, which eases driving and improves safety. However, there is a data scarcity issue for low resource languages, hindering the development of research and applications. In this paper, we introduce a new dataset, Cantonese In-car Audio-Visual Speech Recognition (CI-AVSR), for in-car command recognition in the Cantonese language with both video and audio data. It consists of 4,984 samples (8.3 hours) of 200 in-car commands recorded by 30 native Cantonese speakers. Furthermore, we augment our dataset using common in-car background noises to simulate real environments, producing a dataset 10 times larger than the collected one. We provide detailed statistics of both the clean and the augmented versions of our dataset. Moreover, we implement two multimodal baselines to demonstrate the validity of CI-AVSR. Experiment results show that leveraging the visual signal improves the overall performance of the model. Although our best model can achieve a considerable quality on the clean test set, the speech recognition quality on the noisy data is still inferior and remains as an extremely challenging task for real in-car speech recognition systems. The dataset and code will be released at https://github.com/HLTCHKUST/CI-AVSR. | ['Pascale Fung', 'Bertram E. Shi', 'Xiaojuan Ma', 'Qifeng Chen', 'Genta Indra Winata', 'Holy Lovenia', 'Rita Frieske', 'Cheuk Tung Shadow Yiu', 'Peng Xu', 'Elham J. Barezi', 'Tiezheng Yu', 'Samuel Cahyawijaya', 'Wenliang Dai'] | 2022-01-11 | null | null | null | null | ['audio-visual-speech-recognition'] | ['speech'] | [-2.90120631e-01 -3.31026345e-01 -3.50263789e-02 -4.63031739e-01
-1.21224308e+00 -4.34524924e-01 6.67563021e-01 -6.37017012e-01
-4.93769795e-01 3.71639818e-01 2.35841781e-01 -6.34787500e-01
5.06148517e-01 -1.98876664e-01 -7.14630544e-01 -8.12531233e-01
2.21298665e-01 1.97883353e-01 1.55536160e-01 -4.55269247e-01
-3.00633162e-01 5.38255215e-01 -1.95963132e+00 3.70289326e-01
7.16209829e-01 9.70029593e-01 6.10502779e-01 8.74062598e-01
1.08783282e-01 9.24647987e-01 -7.26086617e-01 -4.07760978e-01
-3.34958397e-02 -1.39135525e-01 -3.96621913e-01 1.82838384e-02
4.41726089e-01 -3.31160009e-01 -8.73409629e-01 1.03060853e+00
5.91097414e-01 2.89405584e-01 2.92751640e-01 -1.63672471e+00
-5.68437457e-01 2.62980014e-01 4.34056073e-02 -4.41035144e-02
2.31871083e-01 4.98036087e-01 7.11517274e-01 -1.01802087e+00
2.42724806e-01 1.34551466e+00 2.03079253e-01 7.43495166e-01
-8.21393430e-01 -1.05798292e+00 1.02442622e-01 8.47466826e-01
-1.49835527e+00 -1.03164876e+00 7.19856977e-01 -2.08781660e-01
8.40886950e-01 4.73788887e-01 5.01601160e-01 1.69593036e+00
-2.56925970e-01 1.27602613e+00 8.55929852e-01 -1.21125966e-01
1.67232811e-01 4.32680279e-01 1.61113888e-01 2.55331814e-01
-2.40998223e-01 3.44060004e-01 -5.67310572e-01 1.46016821e-01
-3.96283381e-02 -4.07500565e-02 -3.85584027e-01 4.73785847e-02
-1.12666726e+00 7.98707843e-01 2.52329499e-01 1.00683004e-01
-2.55412191e-01 2.85282806e-02 5.12193620e-01 4.18795586e-01
1.59840249e-02 -3.42651427e-01 -1.69825658e-01 -5.78133285e-01
-5.79673648e-01 4.11639363e-03 5.54250121e-01 1.28194296e+00
5.13907433e-01 5.96879125e-01 1.48981124e-01 1.09010649e+00
2.43717372e-01 1.19498229e+00 4.16941196e-01 -9.08237994e-01
5.30508935e-01 -3.24636064e-02 -3.63195129e-02 -5.44962168e-01
-3.01753972e-02 -1.45897433e-01 -8.17532837e-01 2.21100494e-01
2.44515225e-01 -5.81504442e-02 -9.11274493e-01 1.30756891e+00
6.08985536e-02 1.64678507e-02 2.34794617e-01 9.97156560e-01
1.09377122e+00 9.14992988e-01 -7.83427283e-02 1.48692578e-02
1.39355791e+00 -1.10406351e+00 -1.15141547e+00 -3.32424551e-01
5.93336344e-01 -8.34471405e-01 1.47473061e+00 7.22009301e-01
-5.97896039e-01 -8.91094506e-01 -9.50459719e-01 -5.66384457e-02
-3.10469240e-01 4.11226898e-01 2.34236792e-01 7.29459524e-01
-9.29508507e-01 -3.63790214e-01 -6.84817314e-01 -1.20758891e-01
3.59672159e-01 -2.02110205e-02 -6.94433212e-01 -3.91958654e-01
-1.22460580e+00 9.10430491e-01 8.45063999e-02 1.99123278e-01
-1.35038948e+00 -3.13063085e-01 -1.12624300e+00 -2.23578498e-01
5.67313552e-01 2.28983015e-01 1.60534716e+00 -7.13079870e-01
-1.55105734e+00 5.99936962e-01 -5.82257390e-01 -4.60013002e-01
5.10553181e-01 -3.10814351e-01 -1.04158485e+00 -7.09325820e-02
-9.14108306e-02 7.50208914e-01 9.47240591e-01 -1.37120402e+00
-9.57169473e-01 -2.19760850e-01 -2.27070957e-01 -6.72356188e-02
-3.32351886e-02 2.30691701e-01 -8.16188574e-01 -3.82350653e-01
-4.63596165e-01 -1.06787050e+00 3.19630772e-01 -3.30610573e-01
-3.69040608e-01 -2.19106615e-01 1.29676509e+00 -9.37816799e-01
8.92465413e-01 -2.88856745e+00 -3.30340713e-01 -1.74936712e-01
-1.31251201e-01 5.19332647e-01 -3.55883598e-01 2.59528965e-01
4.22760695e-02 -3.42888713e-01 -7.28396252e-02 -6.25528574e-01
1.12228222e-01 3.54371518e-01 -5.84833622e-01 5.80507755e-01
2.42897063e-01 7.83335567e-01 -6.26830876e-01 -2.49366611e-01
4.79273647e-01 8.23565185e-01 -2.73841947e-01 3.11130255e-01
2.59557217e-01 4.25031304e-01 -7.87799284e-02 7.71617591e-01
7.05686986e-01 5.86070776e-01 -2.66883165e-01 -8.06567818e-02
-1.13769941e-01 2.79738873e-01 -9.98557091e-01 1.21248233e+00
-6.74883068e-01 1.36135924e+00 5.79082727e-01 -9.56846595e-01
1.00395584e+00 4.30170566e-01 -5.29185310e-02 -1.03827155e+00
1.89655259e-01 3.31404567e-01 8.93528089e-02 -7.32900679e-01
6.26582801e-01 6.96894974e-02 -1.14021987e-01 -1.98999166e-01
-1.10651068e-01 -4.25966829e-01 -3.93845188e-03 1.94002137e-01
7.69134700e-01 -3.57462794e-01 -7.33197629e-02 1.47521138e-01
7.94941962e-01 -3.31567042e-02 2.66714185e-01 4.31589782e-01
-6.31109655e-01 4.85074699e-01 2.18506247e-01 2.21927203e-02
-7.94810951e-01 -9.20923233e-01 -1.56429440e-01 1.10730636e+00
2.73604458e-03 -2.39054427e-01 -6.94477320e-01 -4.63868469e-01
-1.16464160e-01 1.31031621e+00 -1.22675903e-01 -2.34661624e-01
-5.31632245e-01 -1.47155344e-01 8.70047212e-01 6.23891532e-01
5.10957956e-01 -1.07554710e+00 -2.48702660e-01 -1.42377540e-01
-4.58741426e-01 -1.61973166e+00 -6.86979413e-01 1.41251698e-01
-1.99679155e-02 -7.83971846e-01 -5.48214614e-01 -9.58884656e-01
1.14100158e-01 7.35859036e-01 7.09579766e-01 -1.97397470e-01
4.68452601e-03 4.00907248e-01 -4.74356383e-01 -6.63744748e-01
-8.37119997e-01 -3.75424087e-01 2.35373959e-01 2.62581915e-01
6.55496120e-01 -1.98295712e-02 -3.94945331e-02 6.12221599e-01
-5.89843452e-01 -9.80277658e-02 5.05939305e-01 9.23282743e-01
2.89583415e-01 1.45295769e-01 5.95454276e-01 -1.68577775e-01
5.34716785e-01 -2.55336344e-01 -7.13520050e-01 -1.84971020e-01
-7.84839913e-02 -3.30769986e-01 7.27649748e-01 -4.58823025e-01
-1.02964473e+00 2.27922156e-01 -5.89655638e-01 -6.22592270e-01
-7.37628937e-01 2.56832033e-01 -7.26097167e-01 2.76711404e-01
2.66693741e-01 5.33334136e-01 2.01617807e-01 -4.51504618e-01
4.89383906e-01 1.59463656e+00 9.75996256e-01 -2.51737814e-02
7.09879458e-01 4.34000999e-01 -5.36336541e-01 -1.53495085e+00
-1.21878237e-01 -6.28830969e-01 -3.05560559e-01 -3.60730171e-01
7.90961504e-01 -1.32873118e+00 -9.81945992e-01 7.66945899e-01
-1.18728948e+00 -5.83927989e-01 -3.09445038e-02 7.87090003e-01
-4.46796924e-01 2.87640095e-01 -4.21669990e-01 -1.16519678e+00
-3.74624273e-03 -1.79515421e+00 1.29185474e+00 -1.49080738e-01
5.33051640e-02 -4.00602490e-01 -2.57951200e-01 8.75549138e-01
3.60537738e-01 -4.95224774e-01 3.25966716e-01 -6.78391695e-01
-4.90200996e-01 -4.68585163e-01 -6.29039630e-02 8.82102311e-01
5.34779206e-02 1.55315157e-02 -1.46297622e+00 -2.59674370e-01
3.33018154e-02 -4.46957141e-01 8.70574355e-01 1.37382641e-01
1.02597058e+00 -6.29664361e-02 1.83415494e-03 4.16725218e-01
7.26173759e-01 6.88229680e-01 9.09642756e-01 1.14684701e-01
7.27235198e-01 6.84948266e-01 9.06643391e-01 1.06862359e-01
5.36946177e-01 9.22505617e-01 4.59735394e-01 -3.00559010e-02
-3.38571817e-01 -1.54354542e-01 1.05493367e+00 1.05000627e+00
1.13955937e-01 -2.65040994e-01 -9.36234593e-01 7.01005042e-01
-1.49116659e+00 -1.00312865e+00 -3.51628542e-01 2.08024812e+00
6.16127551e-01 2.02588718e-02 2.89446920e-01 3.73872727e-01
5.57872772e-01 3.15878391e-02 -3.90426278e-01 -2.26909563e-01
-3.34421068e-01 -7.96952918e-02 3.21323037e-01 7.25704074e-01
-1.08248091e+00 1.03377306e+00 5.35080957e+00 1.12213612e+00
-1.37657821e+00 2.71474808e-01 4.10463899e-01 -1.71102852e-01
-1.16867255e-02 -4.80580986e-01 -8.87764394e-01 5.80833673e-01
1.39256346e+00 9.84980091e-02 6.73249066e-01 1.13563097e+00
6.48307025e-01 -4.09197137e-02 -1.00096440e+00 1.49157560e+00
2.58729666e-01 -8.86326969e-01 -2.95287251e-01 1.11417942e-01
2.18603224e-01 6.01702511e-01 1.28798246e-01 7.23568022e-01
1.70965388e-01 -1.19618809e+00 1.18287599e+00 1.15284286e-01
1.02484155e+00 -9.11273777e-01 8.44940782e-01 4.97422397e-01
-1.19041979e+00 -1.64513186e-01 -7.42752701e-02 2.84855664e-01
2.74312407e-01 4.09933478e-02 -7.94204295e-01 2.62996674e-01
9.32498097e-01 6.13953352e-01 -4.96902376e-01 7.15433419e-01
-1.98932871e-01 9.48347449e-01 -9.07247737e-02 -8.98842737e-02
2.01774418e-01 -1.27201930e-01 5.73498249e-01 1.35400486e+00
2.46263906e-01 -2.71091551e-01 -9.17286146e-03 4.98930305e-01
1.23092137e-01 -4.60163504e-02 -9.28528845e-01 -1.02868021e-01
5.17734051e-01 1.14609408e+00 2.14212611e-02 -4.22572434e-01
-7.58815169e-01 8.05366278e-01 -1.35892019e-01 6.36557698e-01
-1.17539120e+00 -6.04410708e-01 1.03190982e+00 -7.76649788e-02
3.59892309e-01 -4.93664801e-01 1.26564816e-01 -1.00242245e+00
2.02666640e-01 -1.50595570e+00 -8.93545672e-02 -9.40689981e-01
-8.57476592e-01 8.79436493e-01 -1.15416594e-01 -1.29395318e+00
-3.13803434e-01 -8.19515169e-01 -4.31144655e-01 7.21474290e-01
-1.58849466e+00 -1.11344397e+00 -5.20896375e-01 8.16769600e-01
1.04721355e+00 -5.32824039e-01 5.19150853e-01 7.55298436e-01
-6.82075739e-01 8.02389741e-01 2.35629216e-01 2.26256639e-01
7.65835166e-01 -7.47963428e-01 4.63290095e-01 9.17072833e-01
1.67508617e-01 2.26998240e-01 6.80885196e-01 -2.71110624e-01
-1.78104210e+00 -1.25874126e+00 6.61080837e-01 -3.97709787e-01
7.23103642e-01 -6.66315079e-01 -1.05145812e+00 7.15994775e-01
3.75783920e-01 -4.56888080e-02 3.55951160e-01 -4.05934125e-01
-2.74482101e-01 -3.39651197e-01 -8.00539851e-01 6.23353124e-01
8.01810145e-01 -9.96659219e-01 -4.04118448e-01 1.17794015e-01
8.82835925e-01 -1.33994639e-01 -3.98459762e-01 8.75708684e-02
3.44482481e-01 -7.86328793e-01 8.00435364e-01 -4.30162489e-01
-1.80209592e-01 -4.55826491e-01 -7.08369732e-01 -1.41734779e+00
7.51281157e-02 -6.93521678e-01 1.05599485e-01 1.24917257e+00
3.03285420e-01 -7.84506917e-01 3.71487916e-01 3.46893340e-01
-5.97345412e-01 -2.45966017e-01 -1.21085060e+00 -1.01847291e+00
-4.45673801e-02 -1.37865841e+00 6.32235527e-01 6.06562972e-01
-1.83825389e-01 4.22553509e-01 -5.59877396e-01 2.68991202e-01
4.47030157e-01 -4.76812452e-01 1.32789803e+00 -8.08096051e-01
9.54895020e-02 -2.47961372e-01 -5.47596812e-01 -1.28966928e+00
5.85285187e-01 -7.43867576e-01 4.43126321e-01 -1.07637405e+00
-2.10002825e-01 -7.51155466e-02 1.53996959e-01 3.68655950e-01
-7.05549493e-03 9.33711901e-02 2.88956940e-01 6.15895800e-02
-3.31479847e-01 1.03776324e+00 9.33016300e-01 -5.88898480e-01
3.26522216e-02 1.05938867e-01 -3.29984516e-01 5.18558562e-01
9.06549156e-01 -1.48975313e-01 -3.26103538e-01 -2.26862356e-01
-5.82214773e-01 -7.41448402e-02 5.78332007e-01 -9.97651815e-01
2.64946222e-01 -5.72467931e-02 -2.74158418e-01 -8.99701893e-01
8.82534802e-01 -9.58085775e-01 4.09133360e-02 3.37377161e-01
-1.12324253e-01 -1.86666444e-01 4.36173260e-01 4.99202698e-01
-6.52122557e-01 9.25621837e-02 7.60599732e-01 3.75890076e-01
-1.07424295e+00 -4.11879756e-02 -8.25930119e-01 -8.56934786e-02
8.25918078e-01 -4.24151458e-02 -5.03943801e-01 -9.37458575e-01
-3.93512636e-01 2.98554987e-01 2.61265069e-01 9.08315659e-01
8.16780150e-01 -1.41574764e+00 -8.08671415e-01 5.99399149e-01
4.34383750e-01 -2.02423036e-01 3.84191304e-01 1.03699505e+00
-3.70982409e-01 8.50271404e-01 7.38934427e-02 -8.04244101e-01
-1.58082843e+00 5.63869655e-01 2.07567394e-01 5.50178468e-01
-5.86658418e-01 5.11858046e-01 2.76166528e-01 -5.03159344e-01
7.17649400e-01 -4.11805362e-01 -2.43980303e-01 -1.36008695e-01
8.27705741e-01 4.30788547e-01 2.76707500e-01 -1.40429580e+00
-4.37259018e-01 1.33959383e-01 1.68404371e-01 -2.42340580e-01
1.03731787e+00 -2.21371874e-01 3.49762470e-01 5.08676171e-01
1.34255958e+00 1.97675988e-01 -1.06625485e+00 -6.62533641e-02
-3.76964211e-01 -5.20301402e-01 4.09827441e-01 -4.59182888e-01
-1.15213859e+00 1.13526046e+00 8.09974253e-01 -2.22983062e-02
9.32684898e-01 1.76480159e-01 9.35665905e-01 6.96804523e-01
3.61728966e-01 -9.89930749e-01 -2.27542505e-01 6.44023716e-01
1.19297445e+00 -1.67816246e+00 -7.30632961e-01 -3.59679192e-01
-1.29744101e+00 7.62010038e-01 4.44927633e-01 5.74948430e-01
5.55133462e-01 3.72068137e-01 6.46490932e-01 7.97279775e-02
-7.35616267e-01 -4.77985889e-01 1.96702942e-01 9.60590780e-01
1.29466951e-01 3.74821216e-01 5.19926846e-01 6.36452734e-01
-5.19218683e-01 -2.65829653e-01 4.89025891e-01 6.51311576e-01
-2.36354649e-01 -6.74365938e-01 -7.35510230e-01 -6.75906315e-02
-7.57933557e-02 -2.04345360e-02 -2.70640284e-01 8.18750322e-01
-1.34594560e-01 1.65024889e+00 4.01354842e-02 -8.02045882e-01
7.72309840e-01 1.56854644e-01 -1.31607875e-01 -8.06753486e-02
-1.44567471e-02 1.97477803e-01 3.48228395e-01 -5.54637432e-01
-9.12140831e-02 -6.51791751e-01 -1.27180171e+00 -4.70543057e-01
-2.99260825e-01 7.57794306e-02 9.69478905e-01 6.99568748e-01
3.43848020e-01 5.34851432e-01 7.11478233e-01 -8.13380957e-01
-6.71211123e-01 -1.07000315e+00 -6.02193594e-01 3.90223354e-01
7.61877537e-01 -6.17009997e-01 -6.14155352e-01 1.51773944e-01] | [14.335892677307129, 5.112422943115234] |
94696cc7-0b3d-411a-aa82-22e9965771fd | a-spreader-ranking-algorithm-for-extremely | 2211.09657 | null | https://arxiv.org/abs/2211.09657v1 | https://arxiv.org/pdf/2211.09657v1.pdf | A Spreader Ranking Algorithm for Extremely Low-budget Influence Maximization in Social Networks using Community Bridge Nodes | In recent years, social networking platforms have gained significant popularity among the masses like connecting with people and propagating ones thoughts and opinions. This has opened the door to user-specific advertisements and recommendations on these platforms, bringing along a significant focus on Influence Maximisation (IM) on social networks due to its wide applicability in target advertising, viral marketing, and personalized recommendations. The aim of IM is to identify certain nodes in the network which can help maximize the spread of certain information through a diffusion cascade. While several works have been proposed for IM, most were inefficient in exploiting community structures to their full extent. In this work, we propose a community structures-based approach, which employs a K-Shell algorithm in order to generate a score for the connections between seed nodes and communities for low-budget scenarios. Further, our approach employs entropy within communities to ensure the proper spread of information within the communities. We choose the Independent Cascade (IC) model to simulate information spread and evaluate it on four evaluation metrics. We validate our proposed approach on eight publicly available networks and find that it significantly outperforms the baseline approaches on these metrics, while still being relatively efficient. | ['Mukesh Prasad', 'Dinesh Kumar Vishwakarma', 'Pranav Chandhok', 'Arjun Choudhry', 'Inder Khatri', 'Aaryan Gupta'] | 2022-11-17 | null | null | null | null | ['marketing'] | ['miscellaneous'] | [ 5.42772235e-03 4.01811212e-01 -4.13597047e-01 -7.42524490e-02
2.76801102e-02 -6.41015351e-01 1.04255331e+00 3.95776093e-01
-3.40119392e-01 5.45131862e-01 4.81783718e-01 -2.92579949e-01
-2.45613560e-01 -1.10865736e+00 -2.61259317e-01 -3.76549721e-01
-3.83611590e-01 4.07830536e-01 7.02257633e-01 -4.46187705e-01
4.27789003e-01 4.93323542e-02 -1.03854406e+00 5.52917942e-02
9.31468129e-01 5.28312743e-01 1.53860636e-03 3.36339533e-01
-2.40561306e-01 5.81660450e-01 -4.31322664e-01 -8.01586270e-01
1.50880277e-01 -4.42031205e-01 -5.49148202e-01 1.56076252e-01
-3.01379651e-01 2.19919737e-02 -8.90125409e-02 1.04352045e+00
3.80922079e-01 -4.47648391e-02 6.03530645e-01 -1.11257553e+00
-4.31759238e-01 1.22471547e+00 -9.63130534e-01 3.05860907e-01
1.72101229e-01 -3.54957640e-01 1.35466921e+00 -5.22994578e-01
8.32162321e-01 1.33967495e+00 4.54479426e-01 6.94300160e-02
-1.24899268e+00 -5.81709623e-01 2.91793019e-01 -9.82181132e-02
-1.18742275e+00 -9.83409211e-02 8.00608456e-01 -3.17078292e-01
5.18823266e-01 1.98982343e-01 1.05106425e+00 7.97911584e-01
3.05633903e-01 7.60246098e-01 1.13864887e+00 -2.84228176e-01
4.34502631e-01 5.00609577e-01 -1.27503812e-01 1.33412421e-01
2.94694185e-01 -2.17622042e-01 -5.23433864e-01 -6.11329556e-01
6.97345078e-01 -5.68780638e-02 7.44039342e-02 -2.00165033e-01
-1.02872920e+00 1.33172345e+00 8.05281520e-01 4.55705553e-01
-6.12496257e-01 -1.56864747e-01 1.14479577e-02 2.74431080e-01
9.30500686e-01 3.47483039e-01 7.69913271e-02 2.07340539e-01
-1.03258955e+00 3.00507247e-01 1.15103281e+00 4.04950142e-01
6.68146610e-01 -6.53506577e-01 -4.33013737e-02 7.24388123e-01
8.36329579e-01 2.27877781e-01 2.54057371e-03 -6.05005682e-01
3.12267274e-01 9.88472581e-01 -9.34061129e-03 -1.57321751e+00
-1.78467229e-01 -8.26055229e-01 -9.30966735e-01 -1.36957929e-01
1.54573873e-01 -4.09742236e-01 -3.46874118e-01 1.48299825e+00
6.83097243e-01 2.62081563e-01 -1.48901626e-01 7.84964800e-01
3.67708951e-01 7.57070780e-01 9.85294431e-02 -3.71750921e-01
9.20250416e-01 -8.70187581e-01 -4.92831260e-01 -1.83821201e-01
3.34127665e-01 -1.00627971e+00 3.05610031e-01 2.47354031e-01
-1.01640010e+00 5.35294190e-02 -8.12143922e-01 7.40404189e-01
-1.05085887e-01 -5.44238806e-01 8.82202804e-01 7.31163025e-01
-1.20870733e+00 5.93682587e-01 -4.29917842e-01 -6.17979884e-01
4.65483695e-01 3.74438882e-01 2.62356788e-01 8.70426372e-02
-1.36878932e+00 4.95722026e-01 -1.37596568e-02 -1.57548934e-02
-8.16309035e-01 -4.34307337e-01 -1.45758495e-01 -1.49616092e-01
5.56152284e-01 -7.03394115e-01 8.24124277e-01 -1.02586985e+00
-1.28973389e+00 2.91023791e-01 1.68486923e-01 -6.61924124e-01
5.61995506e-01 1.43660739e-01 -3.88887525e-01 6.35542274e-02
7.64484704e-03 6.85223639e-01 7.96182990e-01 -1.10947728e+00
-5.52440584e-01 -1.17640518e-01 3.18371713e-01 2.89502621e-01
-7.71271706e-01 4.13459003e-01 -3.93518031e-01 -4.62393045e-01
6.70497045e-02 -1.09491181e+00 -7.71735787e-01 -4.29254413e-01
-5.51455140e-01 -3.35434020e-01 6.10041082e-01 -3.16654891e-01
1.47144222e+00 -1.72568357e+00 2.87861407e-01 7.83133805e-01
5.54127455e-01 1.02607362e-01 7.18926371e-04 8.57679665e-01
5.67260623e-01 4.63271886e-01 1.69895645e-02 -9.77617353e-02
-2.98446119e-01 2.34369561e-02 1.85729060e-02 4.02209461e-01
2.25342028e-02 7.98392534e-01 -8.56457293e-01 -6.05359495e-01
-2.44979069e-01 7.97862589e-01 -8.89851034e-01 -1.70970485e-01
-4.25914943e-01 4.93195355e-01 -8.69320393e-01 2.48317182e-01
6.56180978e-01 -7.21977472e-01 5.45317352e-01 2.85350800e-01
-1.40205637e-01 7.36663342e-02 -1.14298892e+00 1.00147903e+00
-1.41412839e-01 2.73966491e-01 4.01998907e-01 -6.31440997e-01
9.50477660e-01 8.61965641e-02 6.38048112e-01 -3.12461138e-01
3.37907284e-01 5.15455566e-02 3.74599338e-01 4.31352966e-02
3.89405549e-01 4.90072258e-02 1.97165027e-01 9.47122931e-01
-4.23151195e-01 1.63188800e-01 3.11455578e-01 1.05181408e+00
1.08090091e+00 -4.80424255e-01 1.97465256e-01 -3.46496105e-01
4.69045430e-01 -1.16083391e-01 8.68422240e-02 6.24989629e-01
-1.18360735e-01 1.24998979e-01 5.82347333e-01 7.75007010e-02
-1.00787973e+00 -6.69387698e-01 1.88037664e-01 9.81708407e-01
3.53821367e-01 -5.10857105e-01 -7.03769565e-01 -4.33975130e-01
-1.23839259e-01 3.90401304e-01 -6.22837782e-01 2.56981760e-01
-2.66454071e-01 -1.14407897e+00 -2.09624544e-02 -1.27969503e-01
5.82429171e-01 -9.85391259e-01 4.48702611e-02 4.53729510e-01
-3.44524294e-01 -8.02244127e-01 -4.17644501e-01 -3.72488052e-01
-1.01743102e+00 -9.80213702e-01 -8.67793798e-01 -4.49594885e-01
7.81567514e-01 3.50199223e-01 9.46988702e-01 2.39059538e-01
5.45275547e-02 2.71841198e-01 -6.32570207e-01 -2.77858853e-01
-5.58702409e-01 4.59776729e-01 -1.11645505e-01 4.93762612e-01
2.04273894e-01 -8.09558749e-01 -1.05833018e+00 7.58826256e-01
-9.77941334e-01 -3.88372540e-02 7.11385787e-01 1.14977032e-01
-2.70852121e-03 2.18199357e-01 7.38224745e-01 -8.35042059e-01
1.25234616e+00 -1.05070937e+00 -3.22745144e-01 -5.33630624e-02
-8.87198806e-01 -2.26148441e-01 6.56858757e-02 -5.23599982e-01
-9.48461831e-01 -5.01771808e-01 9.65326726e-02 4.49031204e-01
5.13769090e-01 8.04943860e-01 3.04651678e-01 -1.34368688e-01
7.21506715e-01 -1.59025937e-01 2.11584702e-01 -4.00073797e-01
5.17178714e-01 7.77348638e-01 -5.87246954e-01 -2.19223514e-01
8.08003485e-01 7.09232867e-01 -2.01036975e-01 -8.77229631e-01
-7.73671448e-01 -6.09406948e-01 -2.61522919e-01 -5.01326978e-01
4.89981353e-01 -7.26523876e-01 -5.85808635e-01 2.70331174e-01
-9.32164609e-01 1.18038595e-01 1.57539770e-01 3.50802690e-01
1.79933250e-01 4.86562788e-01 -7.11144030e-01 -1.04726768e+00
-4.81335074e-01 -8.55681598e-01 3.68601650e-01 2.94911653e-01
-3.04987520e-01 -1.23075747e+00 3.66923332e-01 2.90490299e-01
1.06250584e+00 1.44307271e-01 3.92370850e-01 -8.47280622e-01
-7.98076332e-01 -3.38796109e-01 -3.93294990e-01 1.08248264e-01
-1.45613998e-02 1.29452646e-01 -3.57071459e-01 -2.66841590e-01
-3.46684396e-01 2.90553737e-02 7.82231629e-01 5.20678997e-01
1.68411583e-01 -4.26422834e-01 -7.52377212e-01 -2.64750302e-01
1.21973813e+00 -1.99819535e-01 6.48821771e-01 3.51784468e-01
2.17661023e-01 9.23041701e-01 3.86581331e-01 5.99224210e-01
4.36612129e-01 6.95886910e-01 6.82443917e-01 -2.00140253e-01
1.82934836e-01 -3.15499812e-01 4.57429558e-01 1.11037230e+00
-3.80296528e-01 -3.94700527e-01 -5.52526653e-01 6.16102993e-01
-1.87119114e+00 -8.77056062e-01 -2.49786258e-01 2.03541255e+00
7.95069575e-01 4.06481653e-01 5.61899006e-01 -1.02591887e-01
1.04935420e+00 2.95432061e-01 -2.45052621e-01 -1.05708130e-01
-1.10214181e-01 -8.84743407e-02 4.30622935e-01 7.49690950e-01
-8.28550398e-01 8.11768532e-01 6.33478165e+00 8.48377347e-01
-7.18233705e-01 2.30806768e-01 8.29174280e-01 -1.48792073e-01
-5.86453795e-01 3.01752955e-01 -1.05317330e+00 3.07639211e-01
9.29791987e-01 -4.22437161e-01 3.95255566e-01 6.38448060e-01
5.17386258e-01 -1.03703260e-01 -3.03701967e-01 7.23872110e-02
-1.58760428e-01 -1.36466813e+00 -7.21785650e-02 5.23567080e-01
1.14281535e+00 2.17138812e-01 -1.21127635e-01 -1.18142515e-01
6.42586172e-01 -5.83970368e-01 3.27229828e-01 2.77711928e-01
-2.71766670e-02 -7.80122757e-01 5.88287890e-01 4.84916687e-01
-1.02698672e+00 -5.97804971e-02 -4.90714252e-01 -1.73723191e-01
6.20164335e-01 1.15217423e+00 -8.65774930e-01 3.08457613e-01
4.23706710e-01 7.79790342e-01 -4.78480637e-01 1.23978353e+00
-2.28597850e-01 1.01789188e+00 -5.13541639e-01 -6.37895584e-01
4.64975923e-01 -4.72421229e-01 8.23647976e-01 9.22802091e-01
1.69960454e-01 -2.33821705e-01 2.19049498e-01 8.21887553e-01
-1.00616351e-01 6.76940918e-01 -2.42639691e-01 -3.54167521e-01
3.11809182e-01 1.55472660e+00 -1.22621167e+00 -5.98036051e-02
-2.94499874e-01 5.56158543e-01 7.77449906e-02 2.14003190e-01
-7.06298053e-01 3.96995619e-02 2.91592747e-01 7.91863084e-01
3.18514198e-01 -1.91335395e-01 1.49468914e-01 -7.96977043e-01
-3.13911974e-01 -6.73220813e-01 2.58693874e-01 -2.26448163e-01
-1.38731956e+00 7.27199793e-01 9.27105360e-03 -8.02003920e-01
2.06830092e-02 8.72656479e-02 -6.86320364e-01 5.63614130e-01
-1.46412349e+00 -9.90441442e-01 -5.12747839e-02 5.08482873e-01
2.36098543e-01 -6.02132566e-02 2.21575558e-01 4.30323780e-01
-2.99458504e-01 6.23223558e-02 4.82293069e-02 -3.16614628e-01
4.29712832e-01 -8.15662444e-01 4.73517507e-01 6.12151682e-01
8.45723078e-02 8.37071419e-01 8.31167579e-01 -1.12349880e+00
-9.58828628e-01 -8.31573367e-01 1.00277698e+00 -1.85521528e-01
1.17990983e+00 -3.91530901e-01 -4.29526746e-01 3.75237852e-01
5.47855556e-01 -5.74128628e-01 5.92803419e-01 2.58629948e-01
-2.77356766e-02 1.24071933e-01 -9.09933031e-01 7.82012105e-01
9.08272326e-01 1.88836694e-01 4.26603593e-02 6.07519090e-01
7.02575266e-01 4.03116703e-01 -8.16334546e-01 1.30012244e-01
4.34206665e-01 -1.04766512e+00 7.80734539e-01 -1.84606239e-01
5.92633009e-01 -2.83685327e-01 2.61792123e-01 -1.37933826e+00
-3.91391098e-01 -9.32425320e-01 6.22880757e-02 1.31561267e+00
9.16601777e-01 -8.72844517e-01 1.01174641e+00 2.84269959e-01
6.62490904e-01 -7.47095406e-01 -5.18939435e-01 -2.01168090e-01
-8.01780373e-02 -1.07715756e-01 3.57961833e-01 7.78030276e-01
5.41891977e-02 7.14808166e-01 -5.35165608e-01 -1.55878082e-01
7.35501230e-01 -5.93220852e-02 6.39558494e-01 -1.58441150e+00
-4.84536409e-01 -4.12039340e-01 -1.48117989e-01 -1.17668140e+00
-3.61742228e-01 -8.94241810e-01 -4.80966806e-01 -1.64244831e+00
6.79841161e-01 -8.84695411e-01 -4.75458382e-03 1.83577500e-02
6.86997101e-02 3.99159700e-01 1.41140521e-02 6.07261717e-01
-8.50989878e-01 2.75744468e-01 1.44346154e+00 3.92861385e-03
-4.24226671e-01 4.50674444e-01 -1.29382825e+00 5.23042917e-01
8.50423813e-01 -6.47664964e-01 -4.92148221e-01 1.94934681e-01
1.14881611e+00 -1.42249033e-01 1.02303706e-01 -6.35522366e-01
3.65404278e-01 2.97229681e-02 -1.24433435e-01 -4.89057869e-01
2.59473138e-02 -6.53326273e-01 3.90678674e-01 3.74001205e-01
-4.61924821e-01 -2.10592091e-01 -3.52620035e-01 1.04088461e+00
-9.47247371e-02 -1.24140821e-01 4.68842089e-01 -2.64556527e-01
-9.85035226e-02 2.35798717e-01 -4.50173408e-01 -1.17733739e-01
1.06405044e+00 -1.00538973e-02 -2.35712141e-01 -9.30911660e-01
-6.84162617e-01 2.50443161e-01 3.50300074e-01 5.03684819e-01
3.49936575e-01 -9.62864459e-01 -1.16857219e+00 -3.05697322e-01
-7.87456706e-02 -4.57761556e-01 -4.68415171e-02 1.16207325e+00
-2.85912961e-01 1.81023628e-01 1.56888664e-01 -5.87681770e-01
-1.28503168e+00 3.15427303e-01 4.95592020e-02 -6.69607520e-01
-3.22111368e-01 8.47163618e-01 -6.28853887e-02 -2.55318910e-01
5.16364649e-02 2.11949125e-01 -6.81025743e-01 2.75201112e-01
6.15781724e-01 3.62970144e-01 -2.54282355e-01 -5.23229778e-01
-2.30272636e-01 1.68595925e-01 -2.93617904e-01 -2.52514362e-01
1.42429984e+00 -3.79405797e-01 -5.25267601e-01 6.83905482e-02
8.42139781e-01 3.74554247e-01 -7.38486528e-01 -3.14779133e-01
2.29447242e-02 -3.44381064e-01 2.39144087e-01 -6.46347821e-01
-1.23715532e+00 2.94988811e-01 3.36833224e-02 9.73202467e-01
6.67310119e-01 3.36389303e-01 6.54506326e-01 -1.33986652e-01
3.90895545e-01 -7.51258969e-01 2.63917983e-01 1.57615259e-01
5.39817095e-01 -9.57663715e-01 4.29460734e-01 -8.50739181e-01
-7.45992780e-01 6.06214523e-01 9.37049985e-02 -1.73967853e-01
1.19712234e+00 7.36690387e-02 -1.78282127e-01 -4.85546142e-01
-7.53183365e-01 -7.31564090e-02 2.06251085e-01 1.72381327e-01
5.70646763e-01 2.53812540e-02 -8.77978563e-01 1.61117062e-01
6.44499063e-02 -1.30731747e-01 5.25442481e-01 7.09906101e-01
-8.23769927e-01 -1.41699505e+00 -1.97496399e-01 5.78579307e-01
-6.10712647e-01 -1.90911993e-01 -8.00228477e-01 5.59808135e-01
-3.06925684e-01 1.37789083e+00 -1.71506926e-01 -3.85632247e-01
-8.50566775e-02 -5.27029037e-01 -9.28535871e-03 -5.05112648e-01
-9.16752338e-01 4.70212221e-01 3.13791364e-01 -1.62804991e-01
-8.16620409e-01 -6.82550073e-01 -6.21928036e-01 -5.98653734e-01
-8.96251142e-01 5.70838213e-01 7.81429231e-01 7.38239706e-01
4.73850280e-01 2.37790123e-01 8.76095474e-01 -6.73505187e-01
-3.67730111e-01 -1.15195394e+00 -6.07323468e-01 1.74432367e-01
-4.70316023e-01 -5.87873399e-01 -7.43056655e-01 -4.73057896e-01] | [6.930668354034424, 5.37814998626709] |
ee3edb4c-0cbf-47c9-965b-3205ce261f72 | point-voxel-absorbing-graph-representation | 2306.05239 | null | https://arxiv.org/abs/2306.05239v1 | https://arxiv.org/pdf/2306.05239v1.pdf | Point-Voxel Absorbing Graph Representation Learning for Event Stream based Recognition | Considering the balance of performance and efficiency, sampled point and voxel methods are usually employed to down-sample dense events into sparse ones. After that, one popular way is to leverage a graph model which treats the sparse points/voxels as nodes and adopts graph neural networks (GNNs) to learn the representation for event data. Although good performance can be obtained, however, their results are still limited mainly due to two issues. (1) Existing event GNNs generally adopt the additional max (or mean) pooling layer to summarize all node embeddings into a single graph-level representation for the whole event data representation. However, this approach fails to capture the importance of graph nodes and also fails to be fully aware of the node representations. (2) Existing methods generally employ either a sparse point or voxel graph representation model which thus lacks consideration of the complementary between these two types of representation models. To address these issues, in this paper, we propose a novel dual point-voxel absorbing graph representation learning for event stream data representation. To be specific, given the input event stream, we first transform it into the sparse event cloud and voxel grids and build dual absorbing graph models for them respectively. Then, we design a novel absorbing graph convolutional network (AGCN) for our dual absorbing graph representation and learning. The key aspect of the proposed AGCN is its ability to effectively capture the importance of nodes and thus be fully aware of node representations in summarizing all node representations through the introduced absorbing nodes. Finally, the event representations of dual learning branches are concatenated together to extract the complementary information of two cues. The output is then fed into a linear layer for event data classification. | ['Bin Luo', 'Lin Zhu', 'Zhimin Bao', 'Xiao Wang', 'Chengguo Yuan', 'Bo Jiang'] | 2023-06-08 | null | null | null | null | ['event-data-classification', 'graph-representation-learning'] | ['computer-vision', 'methodology'] | [ 1.17630050e-01 2.69800752e-01 -1.21875614e-01 -1.54751047e-01
-4.61991191e-01 -2.05416709e-01 6.90966785e-01 8.45452011e-01
3.63430157e-02 3.86354208e-01 3.00531268e-01 3.81309167e-02
-5.87096848e-02 -1.40720308e+00 -8.90901566e-01 -7.21542478e-01
-2.43711978e-01 2.22718120e-01 4.42321956e-01 3.92051153e-02
-1.12229988e-01 7.53488719e-01 -1.47923899e+00 1.85526326e-01
7.29162037e-01 1.16918564e+00 2.05200408e-02 3.70524406e-01
-5.53200603e-01 1.14611638e+00 -5.04801035e-01 7.55293891e-02
8.89495686e-02 -3.41282904e-01 -3.83838892e-01 1.25502869e-02
1.08483076e-01 -3.13973844e-01 -8.39901268e-01 9.90720749e-01
4.61037934e-01 4.29063588e-01 7.55798638e-01 -1.38499355e+00
-5.73624909e-01 8.32702458e-01 -7.96115756e-01 3.21149349e-01
2.07969546e-01 3.27762626e-02 1.05654693e+00 -7.72917211e-01
4.49670821e-01 1.11598384e+00 4.87505168e-01 -4.75711329e-03
-8.15709233e-01 -6.16554558e-01 6.78194046e-01 1.32324830e-01
-1.49984264e+00 -5.39760925e-02 1.05703521e+00 -3.48409086e-01
1.07033193e+00 4.38419804e-02 9.96212482e-01 9.55308735e-01
1.62779361e-01 8.92282128e-01 4.02721345e-01 8.58184323e-02
4.04126912e-01 -2.95177728e-01 3.82244855e-01 6.62366867e-01
3.56966347e-01 -1.51920989e-01 -5.13435841e-01 -2.13419899e-01
8.32927585e-01 8.20305049e-01 -3.41771066e-01 -3.76248956e-01
-9.16798234e-01 8.80779207e-01 1.12225378e+00 2.39854515e-01
-7.95478702e-01 2.60086596e-01 6.09318852e-01 -3.50234956e-02
4.28628355e-01 -3.46125998e-02 -1.49041628e-02 1.89744309e-01
-8.52174699e-01 2.81985968e-01 5.70657134e-01 8.76395345e-01
9.74843681e-01 3.32204878e-01 -3.80235434e-01 5.32516360e-01
3.68081570e-01 1.66100830e-01 3.76045287e-01 -5.84088974e-02
6.81737304e-01 1.17723906e+00 -4.60545003e-01 -1.42107773e+00
-5.85630357e-01 -7.46517301e-01 -1.08098114e+00 2.34844461e-02
-4.56915284e-03 1.31602839e-01 -1.09614217e+00 1.56389463e+00
2.68719077e-01 8.14596355e-01 -1.62174627e-02 8.40083480e-01
1.28477037e+00 9.13664877e-01 2.28249475e-01 -1.38436288e-01
1.31979680e+00 -8.09756756e-01 -5.55465400e-01 -1.42746300e-01
3.83101702e-01 -1.93018809e-01 7.86892176e-01 1.03989922e-01
-1.02637184e+00 -5.81954360e-01 -1.16955054e+00 -2.06826866e-01
-5.86374879e-01 -2.63846159e-01 7.88698494e-01 1.38141111e-01
-7.84653008e-01 4.48065788e-01 -1.09278858e+00 -3.68569642e-01
6.46431029e-01 9.37854871e-02 -1.45681232e-01 -4.02742594e-01
-1.12220883e+00 2.72869378e-01 3.74473751e-01 5.52585088e-02
-1.02224469e+00 -8.35744262e-01 -1.26693356e+00 5.04355252e-01
2.34212950e-01 -7.69480944e-01 7.03481495e-01 -5.84483981e-01
-8.69751811e-01 3.27258021e-01 -7.76651353e-02 -4.37133849e-01
2.04632685e-01 1.57387197e-01 -3.23569208e-01 3.69829059e-01
1.00666493e-01 4.23145980e-01 7.70665944e-01 -1.26200199e+00
-6.48441672e-01 -4.10963118e-01 2.60400981e-01 2.01582119e-01
-4.01278585e-01 -4.38962549e-01 -5.30730665e-01 -8.94224763e-01
3.86230797e-01 -3.70716661e-01 -3.45366597e-01 -1.25353709e-01
-4.73545402e-01 -4.22362834e-01 9.14988697e-01 -3.46500635e-01
1.36814773e+00 -2.28338337e+00 1.65325165e-01 2.18020156e-01
7.07102478e-01 -1.54524401e-01 -1.33817241e-01 5.22348523e-01
-2.31855169e-01 -7.80546218e-02 -3.51386011e-01 -5.51881790e-01
-1.58558264e-01 2.72204161e-01 -4.72767651e-01 4.12859619e-01
5.74107051e-01 9.55825448e-01 -1.12401128e+00 -3.55638117e-01
5.47592282e-01 8.76497567e-01 -6.55375540e-01 3.02857339e-01
-1.61984205e-01 2.08632499e-01 -7.72188127e-01 5.17790139e-01
7.89366543e-01 -2.80924499e-01 -7.91920051e-02 -5.02611637e-01
1.16794929e-01 3.46990794e-01 -1.26744998e+00 1.60468113e+00
-1.86859578e-01 1.27323538e-01 -1.48770615e-01 -1.23521256e+00
9.52658772e-01 3.18117708e-01 7.96141207e-01 -4.94778782e-01
1.46402180e-01 -1.21250758e-02 -3.24112803e-01 -1.80179849e-01
4.72015798e-01 -2.41125271e-01 -1.34503648e-01 3.21061730e-01
1.12462908e-01 1.38587117e-01 -1.12778574e-01 5.42957187e-01
1.39741945e+00 -1.42763615e-01 4.19549048e-01 1.34759411e-01
3.22277099e-01 -8.45426098e-02 5.37293196e-01 6.60081983e-01
-2.12658942e-02 1.14493740e+00 6.75904155e-01 -4.05048519e-01
-5.30538619e-01 -1.38024485e+00 1.72451138e-01 7.86651790e-01
2.32012734e-01 -5.77768087e-01 -3.37376088e-01 -7.30255842e-01
1.46470591e-01 6.00323200e-01 -7.17871904e-01 -4.30718213e-01
-6.74243569e-01 -7.60270417e-01 4.15927827e-01 8.45082104e-01
3.84615183e-01 -1.10854793e+00 -4.69397485e-01 4.43874717e-01
4.41522710e-02 -1.10544980e+00 -1.66014954e-01 4.29496318e-01
-8.40322495e-01 -1.13370144e+00 -4.34661955e-01 -6.24364555e-01
7.25367129e-01 4.33792472e-01 9.16973352e-01 1.62663847e-01
-2.35305056e-01 2.54263371e-01 -6.45089626e-01 -4.06970114e-01
8.03634673e-02 2.29319349e-01 -2.26962671e-01 2.95768946e-01
3.31331491e-01 -1.05248666e+00 -7.38625228e-01 -2.14118406e-01
-1.21810448e+00 -4.76686768e-02 4.03446615e-01 4.66799587e-01
8.68818879e-01 3.23523909e-01 7.29103565e-01 -8.79327416e-01
3.98732573e-01 -1.09289932e+00 -1.45161882e-01 -1.22212559e-01
-1.97858602e-01 -2.14459613e-01 1.04155385e+00 -2.79798478e-01
-5.59262514e-01 -5.77636212e-02 -2.11150438e-01 -8.71729076e-01
-1.17949545e-01 9.32305217e-01 -2.53470331e-01 2.59237766e-01
3.92274916e-01 2.98405647e-01 -3.15496087e-01 -4.55475211e-01
3.10404032e-01 2.89312989e-01 3.50591630e-01 -5.05086303e-01
7.49409497e-01 6.53584778e-01 1.84978917e-02 -7.09478796e-01
-7.34890759e-01 -4.63937640e-01 -2.84374446e-01 -1.37229845e-01
9.36187744e-01 -9.62750435e-01 -3.69366616e-01 3.19262058e-01
-8.67046475e-01 -1.60190716e-01 -6.68906629e-01 5.10757565e-01
-2.53615260e-01 4.27481353e-01 -6.40038431e-01 -6.72952950e-01
-2.24886835e-01 -1.03328443e+00 1.22531450e+00 2.98766762e-01
1.30590022e-01 -9.29632366e-01 -1.40624836e-01 -2.80851901e-01
3.05346847e-01 6.59576416e-01 1.08253884e+00 -6.66885495e-01
-6.51524723e-01 -5.16921759e-01 -6.16281152e-01 -4.40760814e-02
2.89264083e-01 -2.43009686e-01 -8.63806963e-01 -3.64044636e-01
-6.34630248e-02 -1.37444228e-01 1.03946257e+00 5.06398499e-01
1.39190972e+00 -5.13633937e-02 -4.75362301e-01 8.34602833e-01
1.50076628e+00 -1.02702305e-01 6.03532791e-01 9.50857699e-02
1.23597658e+00 3.67684364e-01 1.21181488e-01 7.02600777e-01
6.74771428e-01 2.77739793e-01 8.07048023e-01 -2.12505177e-01
-4.34492737e-01 -5.76483667e-01 3.04208696e-01 9.00465071e-01
1.10726044e-01 -4.40044314e-01 -8.08743954e-01 6.77667916e-01
-1.91966820e+00 -8.09799194e-01 -1.37064382e-01 1.93455768e+00
2.44057626e-01 2.14969158e-01 3.15409079e-02 3.31201941e-01
7.12672055e-01 6.74684167e-01 -5.90170562e-01 -9.55155268e-02
-7.67945871e-03 3.36151659e-01 3.32136899e-01 3.71900760e-02
-9.93009806e-01 6.27343893e-01 4.87282610e+00 7.29878783e-01
-1.12094700e+00 4.21135873e-02 4.33223158e-01 -1.93733394e-01
-5.67586362e-01 -9.07710344e-02 -5.99744141e-01 3.47914040e-01
7.38455176e-01 -1.21660978e-01 2.07540989e-01 6.83520794e-01
9.42509696e-02 2.55076498e-01 -1.15202272e+00 1.02838051e+00
1.08564079e-01 -1.30591953e+00 5.66868603e-01 -6.93992674e-02
5.88733554e-01 3.82844210e-02 -2.48251259e-01 7.05750227e-01
3.12838942e-01 -1.19379032e+00 7.45515883e-01 3.84542882e-01
5.25935650e-01 -7.63613939e-01 5.74231863e-01 3.75122637e-01
-1.93489444e+00 -4.07499820e-02 -5.46150565e-01 -2.01415926e-01
3.80518228e-01 7.17128873e-01 -4.51804280e-01 1.14956582e+00
6.04422033e-01 1.08213329e+00 -4.74097699e-01 1.08238924e+00
-7.88614526e-02 7.06197858e-01 -5.04903138e-01 3.09544235e-01
5.38017452e-01 -3.98704335e-02 5.85431755e-01 1.16198635e+00
4.31704134e-01 6.79532513e-02 6.67885661e-01 1.01222765e+00
-1.96486413e-01 7.70247653e-02 -7.52172053e-01 -7.83455838e-03
3.60302359e-01 1.26184118e+00 -9.59782898e-01 -4.14756924e-01
-6.85770988e-01 7.25625753e-01 7.45132387e-01 5.19440413e-01
-8.84378076e-01 -3.06570411e-01 5.95407128e-01 2.98414618e-01
5.72809219e-01 -1.24888211e-01 -3.31392288e-01 -1.22083282e+00
3.03828288e-02 -4.99763221e-01 8.41701448e-01 -5.74048817e-01
-1.47208047e+00 6.18959785e-01 1.09255329e-01 -1.22934926e+00
1.10156551e-01 -2.86115974e-01 -8.69879663e-01 8.12494874e-01
-1.80037332e+00 -1.47133422e+00 -7.59489596e-01 7.72678435e-01
4.83334750e-01 2.78817434e-02 4.50466812e-01 4.69616532e-01
-7.89057732e-01 4.70670670e-01 -4.52876538e-01 3.04024071e-01
2.12902352e-01 -1.23032856e+00 3.22428703e-01 8.36893916e-01
3.05807218e-02 5.49885273e-01 1.62046462e-01 -8.51632893e-01
-1.50653815e+00 -1.68729258e+00 3.86670589e-01 -6.04961105e-02
6.24035835e-01 -2.94426024e-01 -1.20397329e+00 8.29253972e-01
-1.46436945e-01 5.91827869e-01 3.77984136e-01 -3.96526940e-02
-4.48090792e-01 -1.15750358e-01 -9.50228214e-01 4.34576154e-01
1.07082283e+00 -4.72749829e-01 -6.22129858e-01 1.53611064e-01
9.68582511e-01 -2.60391146e-01 -9.05213118e-01 5.45046628e-01
1.78674161e-02 -6.97498083e-01 1.04710400e+00 -6.21046662e-01
5.33610821e-01 -4.85899955e-01 -2.20088035e-01 -1.19376600e+00
-4.22192484e-01 1.07034836e-02 -4.43265170e-01 1.36508322e+00
3.91122997e-02 -8.10225308e-01 7.70610332e-01 9.15720612e-02
-4.56660628e-01 -9.94891047e-01 -8.32120717e-01 -3.73001337e-01
-7.71524012e-02 -7.30040610e-01 9.04272258e-01 1.10163581e+00
-1.85475931e-01 3.39777499e-01 -4.03214470e-02 5.87913811e-01
6.05777025e-01 3.38106453e-01 4.98970807e-01 -1.18440628e+00
-1.78750813e-01 -4.89631057e-01 -7.01793194e-01 -1.10991037e+00
1.16056148e-02 -1.34421492e+00 -1.90340325e-01 -1.98373985e+00
1.56026945e-01 -4.96087164e-01 -7.23514557e-01 3.81782472e-01
-4.24510002e-01 1.70287281e-01 1.28886327e-01 1.88951716e-01
-4.65841532e-01 7.73392618e-01 1.01830912e+00 -1.94255874e-01
-1.78776920e-01 -1.75188035e-01 -7.76137173e-01 6.11964643e-01
5.36944389e-01 -3.16629589e-01 -5.97415626e-01 -4.66739237e-01
1.89611256e-01 1.61747500e-01 6.52432919e-01 -1.06224203e+00
3.94776642e-01 9.35205296e-02 4.40700471e-01 -9.04299617e-01
1.19428799e-01 -8.06858122e-01 1.27903670e-01 8.83919150e-02
-1.25844935e-02 2.20854506e-02 1.47794768e-01 9.61331308e-01
-5.19588351e-01 3.25954445e-02 5.15065968e-01 -2.27052435e-01
-6.44995928e-01 8.37190747e-01 -1.66275099e-01 -8.00668299e-02
1.01514518e+00 -2.87407190e-01 -2.02862293e-01 -3.05574656e-01
-6.53516114e-01 4.08295035e-01 3.10598612e-01 3.29834849e-01
7.09201932e-01 -1.49963105e+00 -6.23694897e-01 3.78466934e-01
2.50022024e-01 6.39877081e-01 5.37577331e-01 8.45391870e-01
-5.20224035e-01 -9.49692130e-02 8.72159153e-02 -6.95822358e-01
-5.01920879e-01 7.16302395e-01 2.02741787e-01 -5.10398865e-01
-1.18188143e+00 6.47449613e-01 6.17838502e-01 -3.08733791e-01
3.36761743e-01 -7.24164724e-01 -3.86853307e-01 2.81825870e-01
4.74440962e-01 1.70494005e-01 1.75194472e-01 -4.70565349e-01
-3.73983234e-01 4.27043408e-01 6.26008436e-02 4.31486577e-01
1.51208389e+00 1.26082711e-02 1.08513720e-02 5.87277174e-01
1.35595870e+00 -1.18664227e-01 -1.14378703e+00 -2.39064023e-01
-4.00302708e-01 -1.90396145e-01 1.49631485e-01 -1.18456699e-01
-1.51313806e+00 1.08163464e+00 2.70616412e-01 5.44938684e-01
1.15246856e+00 6.10159636e-02 1.03357041e+00 -3.36698405e-02
1.93968624e-01 -5.43676615e-01 2.65823696e-02 2.85483629e-01
7.69648790e-01 -8.15536857e-01 3.25537659e-02 -6.89702749e-01
-2.81715631e-01 1.11831605e+00 5.53632259e-01 -6.22424960e-01
8.43951464e-01 2.86969483e-01 -3.10243070e-01 -6.71817541e-01
-4.50774699e-01 -4.08036947e-01 2.51201063e-01 5.65712273e-01
1.97013766e-01 7.86431953e-02 6.90842569e-02 8.92845333e-01
-1.06401339e-01 -1.48688897e-01 3.57114941e-01 8.20293844e-01
-3.62611353e-01 -6.02791309e-01 -1.83637813e-01 8.39082837e-01
-2.90378153e-01 -2.46427096e-02 4.78399619e-02 6.82871461e-01
4.52173129e-02 6.79939985e-01 3.44566524e-01 -3.52941781e-01
6.95761383e-01 -8.59305263e-02 9.59605277e-02 -9.64244604e-01
-7.78553426e-01 -9.51587129e-03 -3.43181103e-01 -6.02125704e-01
-1.77669317e-01 -5.73858082e-01 -1.57998180e+00 -1.16473690e-01
-1.36652235e-02 4.32249494e-02 1.59323767e-01 7.89024234e-01
3.33695263e-01 1.14076185e+00 5.45316577e-01 -9.93237436e-01
-1.60734043e-01 -8.08878422e-01 -8.01289558e-01 5.86641610e-01
3.06410372e-01 -7.94533789e-01 -4.78246838e-01 -3.61103714e-01] | [7.276482582092285, 6.221925735473633] |
bc339cac-d0c7-4194-bc24-2b38fbdc5583 | anomaly-detection-with-score-distribution | 2306.14403 | null | https://arxiv.org/abs/2306.14403v1 | https://arxiv.org/pdf/2306.14403v1.pdf | Anomaly Detection with Score Distribution Discrimination | Recent studies give more attention to the anomaly detection (AD) methods that can leverage a handful of labeled anomalies along with abundant unlabeled data. These existing anomaly-informed AD methods rely on manually predefined score target(s), e.g., prior constant or margin hyperparameter(s), to realize discrimination in anomaly scores between normal and abnormal data. However, such methods would be vulnerable to the existence of anomaly contamination in the unlabeled data, and also lack adaptation to different data scenarios. In this paper, we propose to optimize the anomaly scoring function from the view of score distribution, thus better retaining the diversity and more fine-grained information of input data, especially when the unlabeled data contains anomaly noises in more practical AD scenarios. We design a novel loss function called Overlap loss that minimizes the overlap area between the score distributions of normal and abnormal samples, which no longer depends on prior anomaly score targets and thus acquires adaptability to various datasets. Overlap loss consists of Score Distribution Estimator and Overlap Area Calculation, which are introduced to overcome challenges when estimating arbitrary score distributions, and to ensure the boundness of training loss. As a general loss component, Overlap loss can be effectively integrated into multiple network architectures for constructing AD models. Extensive experimental results indicate that Overlap loss based AD models significantly outperform their state-of-the-art counterparts, and achieve better performance on different types of anomalies. | ['Hailiang Huang', 'Songqiao Han', 'Minqi Jiang'] | 2023-06-26 | null | null | null | null | ['anomaly-detection'] | ['methodology'] | [-3.05935126e-02 -3.41035157e-01 -5.34453355e-02 -6.47579849e-01
-6.96019888e-01 -2.82444060e-01 2.38283738e-01 2.43145242e-01
-1.65677086e-01 4.50024277e-01 -2.24629194e-01 -9.97461304e-02
-2.20308945e-01 -7.83338845e-01 -2.39555031e-01 -8.56955886e-01
1.57993473e-02 2.88755506e-01 3.66698295e-01 -1.51100472e-01
8.96076337e-02 4.39435095e-01 -1.40809464e+00 -9.60261449e-02
1.34852874e+00 1.47633088e+00 -3.13223511e-01 9.27815735e-02
-5.23967087e-01 4.04168546e-01 -6.34525776e-01 -3.58248621e-01
3.49461168e-01 -3.91598403e-01 -2.14315027e-01 4.50707152e-02
3.92579556e-01 -4.17826742e-01 -8.74160454e-02 1.25520337e+00
4.60041225e-01 8.75937790e-02 7.38373578e-01 -1.58519411e+00
-5.68890512e-01 3.22149098e-01 -7.00307846e-01 6.45255685e-01
-7.15869069e-02 2.73332447e-01 1.10744929e+00 -9.35610294e-01
-1.69922620e-01 1.03164399e+00 6.63277328e-01 4.68041092e-01
-9.36918736e-01 -8.07653904e-01 6.80291176e-01 4.02675837e-01
-1.34464979e+00 -1.43895060e-01 1.01320624e+00 -3.67331654e-01
3.51349473e-01 2.56267220e-01 5.05063713e-01 1.09165514e+00
-3.87502834e-02 9.20541644e-01 4.95170236e-01 -1.14441290e-01
3.12475383e-01 4.58129458e-02 3.14012855e-01 5.03191888e-01
4.34272587e-01 -3.40521485e-02 -1.42409205e-01 -4.07366514e-01
4.91959006e-01 4.68028396e-01 -2.75327444e-01 -5.27349353e-01
-7.76344359e-01 8.53100240e-01 2.73143917e-01 1.53633878e-01
-2.71588475e-01 -4.44303840e-01 6.20085776e-01 4.10752863e-01
4.88368988e-01 8.99411812e-02 -5.75222075e-01 1.09003991e-01
-5.10620892e-01 2.40006283e-01 3.33761960e-01 6.82911336e-01
7.91370928e-01 4.27373022e-01 -4.45314765e-01 1.03488493e+00
6.06265724e-01 5.19166648e-01 7.11487830e-01 -2.18957216e-01
7.66087890e-01 1.16908145e+00 -1.45600066e-01 -1.16321814e+00
-1.94658279e-01 -6.41068101e-01 -8.67642164e-01 9.69077367e-03
4.01744872e-01 -1.76018283e-01 -1.02351630e+00 1.80011499e+00
4.28444505e-01 3.68763864e-01 -1.05566978e-01 8.60236287e-01
4.50947016e-01 4.04087692e-01 1.41465887e-01 -2.39471823e-01
8.69007468e-01 -5.59300125e-01 -6.88733697e-01 -3.39800000e-01
6.92395926e-01 -4.19211179e-01 1.39341581e+00 5.20834208e-01
-7.04722404e-01 -3.26128572e-01 -1.23937595e+00 5.52756011e-01
-3.30177337e-01 3.58978808e-02 4.90617335e-01 6.44245744e-01
-4.36659276e-01 3.67025226e-01 -8.57995152e-01 -1.44575313e-01
5.18437564e-01 2.13593096e-01 -1.26994699e-01 7.44557679e-02
-1.31605327e+00 5.58340371e-01 4.86214370e-01 2.27438867e-01
-7.75818229e-01 -6.88595593e-01 -7.80718148e-01 2.09888220e-01
3.51462573e-01 -1.21479489e-01 1.12351835e+00 -1.08551633e+00
-1.11484170e+00 5.23450196e-01 3.19918096e-01 -3.95407945e-01
5.32004237e-01 -3.28739315e-01 -1.00109339e+00 -1.58624202e-01
-9.29885805e-02 -9.01149213e-02 8.45537663e-01 -9.38840389e-01
-7.30296016e-01 -6.63354695e-01 -3.62661779e-01 8.08086023e-02
-8.32831144e-01 -1.15273371e-01 -1.19277269e-01 -8.04665625e-01
3.58745396e-01 -4.44459140e-01 -2.00951993e-01 1.75677389e-01
-3.60130876e-01 -2.25538373e-01 1.25376427e+00 -4.14228529e-01
1.63904881e+00 -2.31011248e+00 -4.88508970e-01 6.15020514e-01
1.00685060e-01 6.95260286e-01 -4.65761051e-02 1.42131329e-01
-1.07432142e-01 -2.51488805e-01 -6.95833325e-01 6.65958077e-02
4.17062975e-02 3.08481485e-01 -5.01635671e-01 3.85433018e-01
5.28852463e-01 3.77727777e-01 -7.67894447e-01 -4.36339498e-01
1.12416066e-01 6.58606514e-02 -5.92623711e-01 5.33521652e-01
-2.18035340e-01 4.47890908e-01 -9.39601421e-01 8.15597475e-01
8.26776326e-01 -4.17690128e-02 -2.96295702e-01 3.98915000e-02
2.86611915e-01 -1.29352331e-01 -1.47159135e+00 1.26274216e+00
-2.03779414e-01 8.89034569e-02 -4.48948704e-02 -1.35388100e+00
1.49602902e+00 2.01546445e-01 5.77731073e-01 -6.23100698e-01
6.92061037e-02 5.39866686e-01 1.39058515e-01 -4.48940128e-01
-1.13699391e-01 3.11997905e-02 1.86161026e-02 4.18833822e-01
-1.10338166e-01 4.93401915e-01 -2.37342399e-02 -9.91299972e-02
1.02573848e+00 -1.95955530e-01 2.21774474e-01 -1.19675227e-01
1.03919506e+00 -3.68336380e-01 1.01831782e+00 4.75820392e-01
-5.08105218e-01 7.03097641e-01 4.96681124e-01 -5.97309470e-01
-8.72525573e-01 -1.36656189e+00 -4.93313819e-01 9.50513422e-01
3.03482920e-01 4.36469950e-02 -4.92880762e-01 -1.20085275e+00
7.74177462e-02 6.64643407e-01 -2.55313218e-01 -7.66856194e-01
-6.76629722e-01 -1.30105913e+00 6.04491353e-01 7.05112815e-01
7.04951644e-01 -9.50485706e-01 -9.32345390e-02 1.80684954e-01
-6.59640273e-03 -6.21168375e-01 -4.82774854e-01 1.24500439e-01
-8.87112260e-01 -1.29539657e+00 -3.75644118e-01 -5.53225696e-01
8.01215351e-01 -6.28409907e-02 9.93791342e-01 1.09961264e-01
-1.22026041e-01 1.99076593e-01 -4.35616106e-01 -5.41906655e-01
-7.79624209e-02 -4.42306623e-02 1.63402379e-01 5.72497308e-01
8.26722980e-01 -8.76307786e-01 -7.26981163e-01 5.68151534e-01
-1.23397505e+00 -8.69622231e-01 6.96478188e-01 9.54168916e-01
6.51226699e-01 -1.54026315e-01 1.05387354e+00 -7.14738071e-01
6.65110469e-01 -8.69877219e-01 -5.04840136e-01 2.59522945e-01
-9.76421952e-01 -3.55959460e-02 9.33773041e-01 -4.60173070e-01
-1.08379829e+00 -3.69373769e-01 -1.93611085e-01 -5.72877526e-01
-2.10732475e-01 2.75461227e-01 -6.44278824e-01 1.47160903e-01
6.29157841e-01 3.24228048e-01 1.24902517e-01 -5.94157457e-01
-1.11855678e-02 6.84796274e-01 4.17942584e-01 -7.22987115e-01
8.13439310e-01 2.80489028e-01 -1.91912968e-02 -3.74391556e-01
-1.00592244e+00 -4.88571674e-01 -4.21330482e-01 1.29222676e-01
3.87165397e-01 -6.95145428e-01 -1.98435187e-01 7.16025174e-01
-7.97666788e-01 2.02806145e-01 -3.80228043e-01 4.90157545e-01
-1.73063710e-01 5.96864343e-01 -3.43786657e-01 -8.40383768e-01
-4.79806215e-01 -8.70461822e-01 7.89790750e-01 2.77936041e-01
1.66483283e-01 -9.92208958e-01 1.76619500e-01 -1.04869932e-01
4.62730587e-01 3.95100623e-01 1.03100419e+00 -1.51301503e+00
-3.59526545e-01 -6.28847301e-01 -3.11792940e-01 8.50350797e-01
2.68845856e-01 2.56565642e-02 -8.93324614e-01 -2.86527038e-01
1.50748268e-01 -2.65589595e-01 7.52826393e-01 9.64486878e-03
1.57546830e+00 -5.18878877e-01 -1.18479811e-01 6.64012969e-01
1.24448204e+00 4.04900998e-01 6.57890677e-01 3.92519802e-01
8.30138862e-01 3.60963732e-01 8.30300689e-01 6.07065797e-01
8.40895809e-03 6.13557696e-01 7.51207590e-01 1.30827606e-01
4.36326891e-01 -1.29018322e-01 3.20694178e-01 7.76238441e-01
2.85851449e-01 -2.54689485e-01 -8.49735796e-01 5.43731153e-01
-1.90610981e+00 -8.91210973e-01 5.49100600e-02 2.41275382e+00
7.27385283e-01 3.73913914e-01 2.43616909e-01 3.39417815e-01
9.66643929e-01 2.63183355e-01 -1.05867803e+00 -1.82718933e-01
-4.98246923e-02 -1.63239241e-01 9.19539481e-02 -1.09544881e-01
-1.19780815e+00 3.90095890e-01 5.57001781e+00 1.10359681e+00
-1.11015546e+00 -1.67984828e-01 6.47100210e-01 -6.16150461e-02
-3.01591814e-01 -3.25165331e-01 -7.92870879e-01 9.23897028e-01
6.74188852e-01 -6.29531592e-02 -6.83856010e-02 1.10477114e+00
1.36062112e-02 5.16270876e-01 -9.28875327e-01 8.33263159e-01
-1.20578751e-01 -4.76374537e-01 3.13307792e-01 -1.20801426e-01
5.36008894e-01 -1.55934334e-01 2.14809254e-01 5.31313062e-01
9.37200189e-02 -7.69401968e-01 3.25456858e-01 4.67350662e-01
4.71504122e-01 -9.14056599e-01 1.05530739e+00 2.94265091e-01
-1.11986387e+00 -4.30533439e-01 -4.37227637e-01 1.64844200e-01
1.50893852e-02 9.44823980e-01 -5.05433857e-01 4.66018498e-01
5.85503936e-01 7.70246744e-01 -5.84974587e-01 1.18904972e+00
9.70951747e-03 7.19600737e-01 -3.71317029e-01 1.97029546e-01
3.22762340e-01 -3.94478023e-01 6.69921517e-01 8.34570467e-01
4.91546631e-01 -9.99225229e-02 4.73800272e-01 7.21867740e-01
1.87701300e-01 5.12797654e-01 -5.29503286e-01 1.42413288e-01
7.49208987e-01 1.08368874e+00 -2.16111869e-01 3.26937474e-02
-5.83042681e-01 4.64734644e-01 3.07870626e-01 2.56867707e-01
-8.78865063e-01 -5.65448582e-01 8.09686065e-01 1.21794626e-01
2.44387284e-01 2.34998316e-01 -2.33602896e-01 -1.24903607e+00
5.57606041e-01 -8.35363090e-01 8.95200789e-01 -1.71719417e-01
-1.99819255e+00 4.99387890e-01 -4.42501232e-02 -1.86028874e+00
-9.16594267e-02 -6.10782206e-01 -1.26570392e+00 6.35497332e-01
-1.39553809e+00 -9.03922141e-01 -4.58994150e-01 7.30812550e-01
4.05000597e-01 -5.92457294e-01 5.23076057e-01 6.05031490e-01
-1.00407946e+00 1.03512466e+00 2.28593677e-01 3.22290570e-01
7.59009778e-01 -1.21315587e+00 1.02842189e-01 1.03548002e+00
-8.63198265e-02 4.20848846e-01 2.92887270e-01 -4.89275575e-01
-6.93096936e-01 -1.40965068e+00 1.60801873e-01 -2.37213135e-01
6.56749249e-01 4.54901569e-02 -1.48434687e+00 5.94254553e-01
-4.56203341e-01 4.37211126e-01 8.08861136e-01 1.12078503e-01
-4.98793632e-01 -5.66749811e-01 -1.49960124e+00 4.51359481e-01
9.95785177e-01 -1.50224552e-01 -5.31847537e-01 1.00517958e-01
6.01350009e-01 -2.26953432e-01 -6.55141413e-01 9.06952083e-01
2.24655375e-01 -1.07847559e+00 9.04170513e-01 -8.65036607e-01
8.45408663e-02 -5.07048130e-01 -1.11368418e-01 -1.21020353e+00
-1.48715302e-01 -8.90297741e-02 -4.57011014e-01 1.43920434e+00
3.58350366e-01 -1.03245127e+00 7.12978840e-01 5.27307272e-01
-4.24744815e-01 -1.17252719e+00 -9.66954172e-01 -8.67668986e-01
-1.03806652e-01 -4.76005584e-01 1.02269912e+00 9.68867064e-01
-2.26922452e-01 2.03444958e-02 -2.51419038e-01 3.80387813e-01
5.45557797e-01 -1.28597185e-01 5.16641200e-01 -1.64891779e+00
-8.17194954e-02 -5.75664103e-01 -7.77645409e-01 -6.38574243e-01
-1.95965674e-02 -6.78619921e-01 -1.16608419e-01 -9.16592717e-01
-1.65867522e-01 -6.82753563e-01 -9.55271959e-01 5.21663427e-01
-5.34242690e-01 6.53523281e-02 -3.44682932e-01 3.88059586e-01
-6.62472785e-01 1.00999069e+00 1.06777740e+00 -1.56637043e-01
-2.30432928e-01 3.62453073e-01 -7.15647817e-01 9.97007728e-01
8.77954185e-01 -3.50916862e-01 -6.87687457e-01 -2.30338633e-01
-1.24079742e-01 -3.52705806e-01 2.96368152e-01 -1.14290917e+00
-1.01969540e-01 -3.71767759e-01 4.48132157e-01 -5.57295680e-01
-7.11011291e-02 -9.82261837e-01 -3.36632192e-01 2.04463437e-01
-2.16171592e-01 2.00413004e-01 -5.15901521e-02 9.29452538e-01
-5.28924644e-01 -1.86010495e-01 9.20953810e-01 2.83654798e-02
-7.12728977e-01 8.96107316e-01 1.80100828e-01 4.26924348e-01
1.13322997e+00 -2.67083138e-01 -2.81864464e-01 -7.95473307e-02
-5.55146933e-01 4.86663043e-01 1.95944324e-01 4.14418250e-01
5.76133132e-01 -1.75779128e+00 -6.18659914e-01 6.63879216e-01
5.08855343e-01 3.22764128e-01 3.29588056e-01 7.27745533e-01
-3.34899694e-01 -2.18010880e-03 -2.32203931e-01 -7.53948927e-01
-7.47157693e-01 4.59605545e-01 4.39237803e-01 -3.76245797e-01
-6.39354289e-01 5.35604239e-01 3.28224480e-01 -5.48581779e-01
4.26807553e-01 -3.22250649e-02 -3.41575652e-01 -1.09234735e-01
7.56175637e-01 4.99026567e-01 2.67745286e-01 -3.41605335e-01
-3.22025865e-01 4.66591567e-01 -3.48228037e-01 4.22533035e-01
1.05869663e+00 -1.98770106e-01 1.09862082e-01 4.09221321e-01
1.01960671e+00 -1.03848264e-01 -1.35052156e+00 -5.54874420e-01
1.38600826e-01 -5.46024323e-01 -2.71624923e-01 -6.78705871e-01
-1.47279429e+00 7.71303296e-01 8.30371678e-01 3.73088181e-01
1.35167086e+00 -2.42460430e-01 1.10496950e+00 3.42563927e-01
7.31039867e-02 -1.05119944e+00 3.70755553e-01 3.69516134e-01
4.37629998e-01 -1.37113094e+00 -3.00616771e-01 -1.12122163e-01
-7.34906316e-01 9.97869134e-01 1.33712804e+00 -1.94983020e-01
6.70954883e-01 -1.51327044e-01 2.98869193e-01 -1.67489156e-01
-3.90702397e-01 6.61947578e-02 4.82993722e-01 7.44150102e-01
1.34166509e-01 -1.52400687e-01 -3.78596008e-01 1.04668677e+00
2.09641293e-01 -5.31626165e-01 6.62411600e-02 7.99699605e-01
-7.07165539e-01 -1.08031678e+00 -2.52454370e-01 8.14438045e-01
-5.85895300e-01 1.16361491e-01 -6.14925921e-02 5.53210080e-01
3.13737422e-01 6.83346152e-01 2.64056474e-01 -2.67350942e-01
6.00006759e-01 2.61274546e-01 -1.75601318e-01 -5.01235068e-01
-1.71346217e-01 -2.87604362e-01 -3.17361385e-01 -4.62475628e-01
-2.22057123e-02 -5.97017348e-01 -1.14474916e+00 1.45086953e-02
-4.65049326e-01 2.78553456e-01 9.88089293e-02 1.10830498e+00
3.54408413e-01 3.51440102e-01 1.03363705e+00 -2.13667959e-01
-1.08546019e+00 -1.01477838e+00 -8.87796104e-01 6.77616477e-01
4.17060882e-01 -8.30323339e-01 -7.12885857e-01 -3.33649606e-01] | [7.603653430938721, 2.3910908699035645] |
eb5cb522-dff3-43a2-811b-7de4e8f8923c | what-you-can-reconstruct-from-a-shadow | null | null | http://openaccess.thecvf.com//content/CVPR2023/html/Liu_What_You_Can_Reconstruct_From_a_Shadow_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Liu_What_You_Can_Reconstruct_From_a_Shadow_CVPR_2023_paper.pdf | What You Can Reconstruct From a Shadow | 3D reconstruction is a fundamental problem in computer vision, and the task is especially challenging when the object to reconstruct is partially or fully occluded. We introduce a method that uses the shadows cast by an unobserved object in order to infer the possible 3D volumes under occlusion. We create a differentiable image formation model that allows us to jointly infer the 3D shape of an object, its pose, and the position of a light source. Since the approach is end-to-end differentiable, we are able to integrate learned priors of object geometry in order to generate realistic 3D shapes of different object categories. Experiments and visualizations show that the method is able to generate multiple possible solutions that are consistent with the observation of the shadow. Our approach works even when the position of the light source and object pose are both unknown. Our approach is also robust to real-world images where ground-truth shadow mask is unknown. | ['Carl Vondrick', 'Simon Stent', 'Dennis Park', 'Chengzhi Mao', 'Sachit Menon', 'Ruoshi Liu'] | 2023-01-01 | null | null | null | cvpr-2023-1 | ['3d-reconstruction'] | ['computer-vision'] | [ 3.08049440e-01 3.60772222e-01 4.32652503e-01 -1.64038971e-01
-2.86972046e-01 -7.30504096e-01 7.25164056e-01 -1.32067248e-01
-6.25777096e-02 4.24871027e-01 -2.64622480e-01 -1.33581519e-01
6.30257055e-02 -6.77790046e-01 -8.68403375e-01 -5.93512595e-01
3.86058033e-01 1.22446227e+00 5.36136925e-01 1.72209531e-01
1.88427806e-01 1.02263343e+00 -1.66171718e+00 2.35480536e-02
6.05094969e-01 7.66775787e-01 6.82181597e-01 7.08377600e-01
-2.28716850e-01 1.92491814e-01 -4.11572486e-01 1.53375091e-02
5.00285029e-01 -1.53322071e-01 -4.93888080e-01 8.88105571e-01
5.24864316e-01 -4.39622998e-01 9.22654793e-02 8.10059488e-01
-5.54940663e-02 -1.62662119e-02 8.89339447e-01 -9.68734562e-01
4.85621020e-02 -2.56254584e-01 -5.05956352e-01 -4.15678680e-01
3.78994375e-01 1.62135605e-02 5.71246743e-01 -9.18975890e-01
8.53438675e-01 1.24033618e+00 3.86702925e-01 1.65713385e-01
-1.40514040e+00 -7.78084844e-02 3.54381800e-01 -3.54572609e-02
-1.22425687e+00 -2.84752667e-01 1.04525995e+00 -6.91090405e-01
1.75636336e-01 2.21658319e-01 6.95070326e-01 7.25127995e-01
1.36470258e-01 4.98244166e-01 1.14346778e+00 -6.77796900e-01
4.43499446e-01 4.32521850e-01 -9.96343698e-03 7.56047785e-01
2.73162872e-01 1.02970012e-01 -2.15657189e-01 -3.19816232e-01
1.03993213e+00 1.05387293e-01 -4.85496014e-01 -9.36678350e-01
-1.15642130e+00 5.13909876e-01 3.34471524e-01 4.82837856e-02
-3.97647351e-01 1.62790362e-02 -5.24597645e-01 -2.49036744e-01
5.40555537e-01 3.43957484e-01 -3.40053886e-01 5.26127338e-01
-6.79283082e-01 4.49806631e-01 9.42944586e-01 8.86585236e-01
9.76874173e-01 -5.71950004e-02 2.50392370e-02 2.14979336e-01
4.28058386e-01 7.46866822e-01 -2.80662715e-01 -1.38559985e+00
-4.23518270e-02 5.76137543e-01 6.52465641e-01 -7.49938130e-01
-2.19154850e-01 -4.13953364e-01 -3.13813180e-01 8.36154103e-01
6.18040860e-01 1.43841550e-01 -1.21244884e+00 1.23289347e+00
7.81120479e-01 4.46006268e-01 -1.89380318e-01 8.98839533e-01
5.91607273e-01 5.34186661e-01 -6.73142791e-01 -3.63019019e-01
1.19341671e+00 -4.43063557e-01 -5.02884150e-01 -5.05515039e-01
-2.36966148e-01 -7.72843063e-01 7.21445322e-01 3.63253564e-01
-1.09108281e+00 -4.20472354e-01 -7.87356973e-01 -5.17708343e-03
6.95145428e-02 6.55572861e-02 3.14538896e-01 3.25797647e-01
-6.87126875e-01 3.43375117e-01 -8.26366603e-01 -2.05558483e-02
3.01321059e-01 1.84263602e-01 -1.81340337e-01 -1.78689867e-01
-1.79866210e-01 1.02086174e+00 2.74782419e-01 9.01245326e-02
-1.08521736e+00 -5.73914409e-01 -6.67789042e-01 -3.80446911e-02
6.68151498e-01 -8.59613001e-01 1.11737061e+00 -7.58487880e-01
-1.25915551e+00 9.47042882e-01 -3.34803164e-01 -4.95131649e-02
7.24655628e-01 1.20448537e-01 4.76081729e-01 2.45419696e-01
2.20875889e-02 3.16633552e-01 9.29257035e-01 -2.00892091e+00
-2.59767771e-01 -6.82529211e-01 -8.07783846e-03 3.60313624e-01
4.23187345e-01 -4.92251307e-01 -5.77196598e-01 -1.64950058e-01
6.54482186e-01 -9.61640716e-01 -4.02610689e-01 7.45007038e-01
-5.42009175e-01 1.51780859e-01 1.05597007e+00 -4.23192471e-01
4.15827855e-02 -1.91354215e+00 2.14299098e-01 2.05840275e-01
1.06105534e-03 -1.39146047e-02 1.26750991e-01 2.16150552e-01
3.44700634e-01 -2.59492695e-01 -5.61848104e-01 -7.67960191e-01
-2.07640857e-01 5.82025588e-01 -4.58095163e-01 6.27396762e-01
-7.34519511e-02 5.96836329e-01 -6.92134798e-01 -3.13026041e-01
5.63649714e-01 7.78670430e-01 -2.38361821e-01 5.40560663e-01
-8.67407322e-01 8.91593277e-01 -5.20389616e-01 4.31092054e-01
8.46603751e-01 -1.48974612e-01 2.13981092e-01 -1.62884712e-01
-2.90823013e-01 -1.11036681e-01 -1.54139483e+00 1.48494422e+00
-5.63999772e-01 4.70400423e-01 4.00468290e-01 -4.32102203e-01
8.79054129e-01 1.83139861e-01 3.93363297e-01 -8.70628804e-02
2.63122648e-01 1.09239161e-01 -3.33473563e-01 -4.40378755e-01
1.42673239e-01 -3.06718051e-01 4.30397421e-01 4.32184458e-01
-3.31632435e-01 -8.06432962e-01 -1.81930631e-01 9.99350995e-02
7.29911447e-01 3.15946132e-01 1.50268286e-01 -1.35310963e-01
2.47786567e-01 -1.08236022e-01 5.35345256e-01 4.73943681e-01
6.34545803e-01 1.15394604e+00 4.07278568e-01 -4.30174619e-01
-8.91011059e-01 -1.12566686e+00 -2.88395613e-01 1.40416175e-01
5.82387507e-01 2.98109919e-01 -5.48072815e-01 -4.95358706e-01
1.80474818e-01 7.70865738e-01 -6.12279773e-01 4.69349146e-01
-4.87412751e-01 -2.69498795e-01 -5.27138352e-01 4.75325435e-02
5.92357069e-02 -8.92580628e-01 -1.14960384e+00 1.93855867e-01
-2.92668164e-01 -1.43912637e+00 -1.98458239e-01 -2.20339634e-02
-8.72462571e-01 -1.45543325e+00 -4.31527853e-01 -4.95559514e-01
1.05327809e+00 3.60183477e-01 1.09278357e+00 1.74262851e-01
-5.47143817e-01 5.44113338e-01 -2.69466564e-02 -6.21922672e-01
-5.74094772e-01 -7.54660010e-01 -3.13143283e-01 4.79750097e-01
-4.62678820e-01 -5.45317292e-01 -4.92913425e-01 4.57850903e-01
-8.71928811e-01 2.75774568e-01 1.70433775e-01 3.64668071e-01
9.00127411e-01 2.01377869e-01 -2.31188014e-01 -8.87138128e-01
-6.82795644e-02 -6.73167482e-02 -1.15415609e+00 2.92292774e-01
-1.66516274e-01 1.37811244e-01 2.44890228e-01 -4.57702518e-01
-1.33095014e+00 6.26762331e-01 3.38424027e-01 -7.17922211e-01
-4.88658607e-01 -1.71257719e-01 -4.15214777e-01 -4.25563566e-02
4.51905757e-01 2.46576555e-02 5.83926328e-02 -8.53672862e-01
2.39628091e-01 1.66511193e-01 4.66766864e-01 -5.34050345e-01
8.52685988e-01 1.15887952e+00 4.92300272e-01 -1.03058541e+00
-7.67423034e-01 -3.52245569e-01 -8.57684016e-01 -3.43585402e-01
6.83819592e-01 -5.87506473e-01 -6.79516613e-01 3.39513540e-01
-1.59783733e+00 -3.88015032e-01 -3.06033015e-01 4.15327251e-01
-6.15267754e-01 2.39488855e-01 4.20006737e-02 -1.08028412e+00
8.58346671e-02 -1.33527803e+00 1.34739709e+00 -1.98679753e-02
2.75458843e-01 -9.32274044e-01 -3.02436054e-01 2.97045738e-01
-5.77458553e-02 6.02367878e-01 9.29133892e-01 1.36155605e-01
-1.32460475e+00 5.28054647e-02 -1.01516880e-01 4.41272967e-02
6.51835799e-02 1.81192666e-01 -1.04275906e+00 -5.92989754e-03
2.68114179e-01 2.37635031e-01 6.30912840e-01 7.09656417e-01
9.28183734e-01 -2.10962400e-01 -5.85366309e-01 4.53060806e-01
1.55619943e+00 9.53500718e-03 3.98638934e-01 -2.08076388e-01
5.93563318e-01 9.41831946e-01 6.12119853e-01 3.94280046e-01
3.45706701e-01 1.07059085e+00 1.01736629e+00 -4.86425124e-02
-3.84216934e-01 -6.94196522e-02 -1.00672968e-01 -7.13928193e-02
-8.06283876e-02 -3.43431026e-01 -8.10053885e-01 3.11508238e-01
-1.57913768e+00 -6.19125128e-01 -6.20984256e-01 2.44259739e+00
3.06244880e-01 -9.09985509e-03 -1.49792627e-01 -3.40264067e-02
4.51183528e-01 -2.27021381e-01 -6.35348737e-01 5.08965664e-02
1.73113152e-01 -6.88369721e-02 4.08345878e-01 1.10508895e+00
-4.94801432e-01 6.29897475e-01 6.12139940e+00 9.03823078e-02
-8.78640115e-01 -7.03868866e-02 2.40811184e-01 1.95505559e-01
-7.15380371e-01 4.62002963e-01 -7.25909710e-01 2.87140459e-01
-2.08050758e-02 1.69871390e-01 3.55173290e-01 4.08222705e-01
1.82309091e-01 -5.70216775e-01 -1.16419089e+00 8.06554377e-01
3.03581446e-01 -1.18547678e+00 -1.78856730e-01 3.33994269e-01
6.92571580e-01 -2.09807992e-01 -1.50236413e-01 -5.03525913e-01
2.46102229e-01 -5.30731261e-01 9.38272715e-01 7.04086304e-01
5.65406501e-01 -3.97243410e-01 2.89373368e-01 8.82453442e-01
-8.44918549e-01 1.67070046e-01 -1.74632385e-01 -6.61910400e-02
5.22190273e-01 8.58925998e-01 -1.39607489e+00 2.83606678e-01
3.30043137e-01 2.73414314e-01 -2.99808174e-01 1.09048855e+00
-5.67352057e-01 6.13292083e-02 -6.24667764e-01 2.94910967e-01
-2.66083896e-01 -4.56430525e-01 9.44530666e-01 5.45375228e-01
2.66541153e-01 2.61908621e-01 5.01845896e-01 1.28574955e+00
2.06849352e-01 -3.38873893e-01 -6.38236523e-01 5.53403080e-01
2.07914472e-01 1.05288482e+00 -1.08730745e+00 4.12311032e-02
-4.95996932e-03 1.00056326e+00 1.95327118e-01 6.89984441e-01
-5.13069332e-01 3.18649977e-01 3.88601303e-01 5.86022735e-01
4.02739912e-01 -4.27927524e-01 -3.27561051e-01 -1.08048642e+00
2.81303108e-01 -2.81177342e-01 -1.35962129e-01 -1.19222331e+00
-9.23470199e-01 4.90810394e-01 3.12071830e-01 -1.11253345e+00
-2.09904090e-01 -6.63707793e-01 -5.20419657e-01 7.99373686e-01
-1.52708733e+00 -1.13967919e+00 -4.30296600e-01 2.74441004e-01
6.01629615e-01 3.85652781e-01 6.78515315e-01 -3.51592839e-01
9.94884446e-02 -4.47008461e-01 -1.67582080e-01 -3.11060667e-01
1.70892239e-01 -1.15133870e+00 -1.10918283e-03 7.31535792e-01
3.90665144e-01 1.93094999e-01 1.05711746e+00 -6.75560594e-01
-1.38063693e+00 -7.05953181e-01 4.81148422e-01 -7.00375378e-01
1.62736662e-02 -6.42327130e-01 -9.96835113e-01 6.56974971e-01
-1.23882085e-01 3.00056428e-01 -1.17634004e-02 -2.48373896e-01
-9.94109586e-02 -1.46781504e-01 -1.21966136e+00 2.96149343e-01
7.29844451e-01 -1.07118547e-01 -3.98255736e-01 4.75970060e-01
3.61536354e-01 -7.73711264e-01 -1.46879479e-01 4.36301321e-01
4.76686895e-01 -1.14060831e+00 1.05604684e+00 -3.85198295e-02
-1.56004271e-02 -5.95165730e-01 -1.54409166e-02 -1.18880820e+00
2.19908506e-01 -3.58318299e-01 -1.60132498e-01 8.10001016e-01
1.65533394e-01 -6.27505362e-01 8.03121626e-01 7.12756455e-01
-8.80685672e-02 -4.32542592e-01 -9.42299366e-01 -6.12260997e-01
-4.03657913e-01 -3.15490276e-01 3.81186455e-01 5.93980193e-01
-8.86146605e-01 2.25865409e-01 -2.20018476e-01 6.83277726e-01
1.12537372e+00 8.73787105e-01 1.03395164e+00 -1.67484868e+00
-5.19013047e-01 3.00760884e-02 -1.91984564e-01 -1.14678907e+00
2.39935756e-01 -5.73004901e-01 3.22219461e-01 -1.68707144e+00
1.03237845e-01 -7.14037836e-01 5.50704360e-01 2.81364411e-01
1.12258904e-01 8.03788453e-02 2.73474872e-01 5.61021604e-02
4.85105030e-02 5.05338609e-01 1.57480693e+00 9.02763233e-02
-3.95610720e-01 4.30452168e-01 -1.28378302e-01 1.05091631e+00
4.43008989e-01 -5.94981670e-01 -4.44457233e-01 -6.09367013e-01
-1.19880736e-01 2.56291777e-01 8.39826584e-01 -6.41888499e-01
2.22469028e-02 -3.20780963e-01 4.86898005e-01 -8.73321056e-01
1.00309217e+00 -1.37795842e+00 6.24203265e-01 2.72769034e-01
2.57168170e-02 -5.09830832e-01 -7.29621798e-02 6.67857885e-01
3.14010233e-01 -5.72411895e-01 8.26279342e-01 -4.93360192e-01
-1.96252912e-01 3.71374875e-01 -1.12780206e-01 -2.45048210e-01
1.02562237e+00 -2.87490577e-01 1.88937604e-01 -3.70971680e-01
-7.42237449e-01 1.02556422e-01 8.77149165e-01 2.08649009e-01
7.27782309e-01 -8.40230346e-01 -7.59943426e-01 4.96318430e-01
-6.71865195e-02 6.30044222e-01 -1.85565595e-02 3.66435915e-01
-6.50395215e-01 -1.78584643e-02 1.95557013e-01 -9.79854763e-01
-1.52897274e+00 4.37697232e-01 5.74774861e-01 2.57811904e-01
-8.60228837e-01 4.56762403e-01 5.66355646e-01 -2.13817552e-01
2.90935099e-01 -3.70194823e-01 1.77865550e-01 -2.48741686e-01
4.71543640e-01 2.21884415e-01 -2.31867773e-03 -6.48414850e-01
-8.98379683e-02 8.73363972e-01 3.01581919e-01 -2.56776035e-01
1.32708859e+00 -8.19168538e-02 -7.50897154e-02 5.63222885e-01
7.94900477e-01 1.71512291e-01 -1.66417432e+00 -2.83759415e-01
-4.80503410e-01 -9.23497558e-01 1.06963083e-01 -5.70229352e-01
-9.96589661e-01 9.29115295e-01 2.68688768e-01 -1.14875901e-02
7.29767978e-01 3.61244947e-01 2.75806040e-01 8.12366605e-02
7.12249577e-01 -5.91060817e-01 1.72233388e-01 3.25717300e-01
1.12011731e+00 -1.06160545e+00 3.53845358e-01 -8.98477554e-01
-2.59068668e-01 1.21307361e+00 5.26963830e-01 -2.94809155e-02
6.73737884e-01 2.96072483e-01 7.29549974e-02 -2.55710423e-01
-3.07265699e-01 -2.90928870e-01 5.35347700e-01 5.60143828e-01
-2.81670123e-01 -2.46875826e-03 6.22046925e-02 -1.88713744e-01
1.59500018e-01 -3.54810148e-01 7.74210632e-01 8.76596451e-01
-4.62744266e-01 -1.14211059e+00 -7.96176076e-01 3.90234200e-04
1.75834298e-02 5.26266038e-01 -4.30501223e-01 5.73297679e-01
2.87834406e-01 6.47280216e-01 1.92705855e-01 4.07123059e-01
3.53229016e-01 -1.51931405e-01 8.75967920e-01 -9.70042050e-01
2.21785337e-01 4.36070621e-01 -2.30589762e-01 -3.60565007e-01
-5.38813770e-01 -6.25849426e-01 -1.38091862e+00 3.98294449e-01
-5.04880905e-01 -8.74079838e-02 9.80273366e-01 9.04881597e-01
1.82069317e-01 6.54894710e-02 5.98269820e-01 -1.31994462e+00
-3.85817289e-02 -4.12913203e-01 -6.63885057e-01 3.11493129e-01
6.86558604e-01 -9.13468540e-01 -5.17611265e-01 2.22779751e-01] | [9.075675010681152, -3.031416177749634] |
47476eb7-75be-4f3d-97cb-88468d4d87ee | hierarchical-bilinear-pooling-for-fine | 1807.09915 | null | http://arxiv.org/abs/1807.09915v1 | http://arxiv.org/pdf/1807.09915v1.pdf | Hierarchical Bilinear Pooling for Fine-Grained Visual Recognition | Fine-grained visual recognition is challenging because it highly relies on
the modeling of various semantic parts and fine-grained feature learning.
Bilinear pooling based models have been shown to be effective at fine-grained
recognition, while most previous approaches neglect the fact that inter-layer
part feature interaction and fine-grained feature learning are mutually
correlated and can reinforce each other. In this paper, we present a novel
model to address these issues. First, a cross-layer bilinear pooling approach
is proposed to capture the inter-layer part feature relations, which results in
superior performance compared with other bilinear pooling based approaches.
Second, we propose a novel hierarchical bilinear pooling framework to integrate
multiple cross-layer bilinear features to enhance their representation
capability. Our formulation is intuitive, efficient and achieves
state-of-the-art results on the widely used fine-grained recognition datasets. | ['Xinyi Zhao', 'Xinge You', 'Chaojian Yu', 'Qi Zheng', 'Peng Zhang'] | 2018-07-26 | hierarchical-bilinear-pooling-for-fine-1 | http://openaccess.thecvf.com/content_ECCV_2018/html/Chaojian_Yu_Hierarchical_Bilinear_Pooling_ECCV_2018_paper.html | http://openaccess.thecvf.com/content_ECCV_2018/papers/Chaojian_Yu_Hierarchical_Bilinear_Pooling_ECCV_2018_paper.pdf | eccv-2018-9 | ['fine-grained-visual-recognition'] | ['computer-vision'] | [-1.29020602e-01 -5.50938725e-01 -1.26454055e-01 -4.93201047e-01
-7.95741677e-01 -5.99640310e-01 8.01116586e-01 1.02658644e-01
-2.01870590e-01 7.03725576e-01 5.14152765e-01 3.71039867e-01
-4.17720824e-01 -9.16064620e-01 -6.06545150e-01 -7.39998937e-01
1.18282951e-01 -2.85789054e-02 4.60386276e-01 -2.04377901e-02
4.99101162e-01 1.01405108e+00 -1.88354003e+00 9.17372346e-01
7.01267958e-01 1.59765446e+00 -2.45787911e-02 4.39371347e-01
-3.40276301e-01 6.61759496e-01 -4.49419916e-01 -3.76206040e-01
2.12800294e-01 3.37308794e-01 -7.38963246e-01 1.44578829e-01
6.85978711e-01 -1.48453459e-01 -1.94440112e-01 8.55108142e-01
3.92402261e-01 1.62979290e-02 7.35268056e-01 -1.34871054e+00
-9.12218213e-01 2.12825403e-01 -5.31026363e-01 -4.44949456e-02
9.88429934e-02 -1.71171620e-01 1.18128037e+00 -1.39666867e+00
1.65763587e-01 1.51934731e+00 8.06769252e-01 2.10515812e-01
-1.17125845e+00 -6.51689470e-01 3.88993710e-01 2.70079255e-01
-1.64300716e+00 -9.14120749e-02 6.22578859e-01 -5.59148848e-01
1.16051376e+00 2.83579290e-01 5.59433639e-01 6.75172091e-01
3.28818738e-01 5.58961511e-01 1.68362129e+00 -3.06332171e-01
-1.43908873e-01 -8.61346573e-02 3.88614327e-01 8.91281724e-01
2.55320013e-01 3.02376956e-01 -8.71353328e-01 -2.06476271e-01
8.96785796e-01 4.15123522e-01 6.55594617e-02 -3.58566403e-01
-1.18774772e+00 7.54459202e-01 7.86496520e-01 5.12499154e-01
-5.03629208e-01 1.11908615e-01 3.68150398e-02 -1.70111582e-02
4.59035248e-01 3.18647236e-01 -4.16564018e-01 1.28883585e-01
-9.55084741e-01 3.68744999e-01 6.24412000e-01 6.25341296e-01
1.08888388e+00 -3.92330706e-01 -9.02687967e-01 9.15363967e-01
4.63083059e-01 1.05453447e-01 3.88640910e-01 -4.93662924e-01
3.08118373e-01 1.03634596e+00 1.42825782e-01 -1.05150771e+00
-2.69651234e-01 -3.34517002e-01 -1.01757932e+00 3.27030987e-01
4.41200733e-01 5.22440732e-01 -1.21671748e+00 1.39247680e+00
-3.78489345e-02 -2.36990256e-03 -2.91979313e-01 7.49174356e-01
9.08007622e-01 4.12184507e-01 4.57301199e-01 4.23064411e-01
1.54789269e+00 -1.34510624e+00 -3.43485266e-01 1.14035763e-01
-1.35702074e-01 -1.02759159e+00 9.93816793e-01 1.36785612e-01
-9.14398551e-01 -9.22906339e-01 -1.08037817e+00 -3.45123589e-01
-1.01148474e+00 2.09405690e-01 8.18955362e-01 7.01641262e-01
-1.01470959e+00 5.65620959e-01 -5.71110725e-01 -1.64287344e-01
6.72600448e-01 4.82962072e-01 -7.44828761e-01 -2.93981969e-01
-9.83341038e-01 8.94767463e-01 4.18809086e-01 3.26280683e-01
-5.88195801e-01 -7.36326814e-01 -7.26278961e-01 3.70954603e-01
-1.84941426e-01 -7.36484528e-01 7.37689018e-01 -2.81095892e-01
-1.28358805e+00 8.13578606e-01 -2.92325795e-01 4.38854992e-02
3.27313483e-01 -1.69041619e-01 -1.63144380e-01 -9.94231105e-02
-1.99255198e-01 7.80960023e-01 1.01839030e+00 -1.13100982e+00
-5.95719397e-01 -4.78660852e-01 3.00654262e-01 1.42069273e-02
-4.03351009e-01 1.70566898e-03 -4.64328438e-01 -9.76246834e-01
1.13049634e-01 -4.88791019e-01 -2.27145776e-01 5.13207242e-02
-1.98071510e-01 -4.27783996e-01 7.50602961e-01 -3.09728205e-01
1.14773023e+00 -2.03922510e+00 -4.13732491e-02 2.21645504e-01
3.64785314e-01 2.27872789e-01 -2.10950747e-01 4.15604025e-01
2.08480567e-01 2.40072280e-01 1.02737777e-01 -3.56481791e-01
5.17594934e-01 1.61784336e-01 -3.63041222e-01 2.02449784e-01
6.45368278e-01 1.38920546e+00 -3.78827393e-01 -4.18594807e-01
4.45123672e-01 8.23545456e-01 -5.46058238e-01 4.25836533e-01
2.35605821e-01 1.87892854e-01 -5.99204421e-01 1.04569173e+00
1.04789841e+00 -4.32516873e-01 -1.84665307e-01 -8.69774878e-01
-2.95440465e-01 -2.26550102e-01 -1.18651855e+00 1.30881953e+00
-3.09357405e-01 -2.51678694e-02 -2.77332932e-01 -6.08355880e-01
9.82362568e-01 7.48480856e-02 5.46682589e-02 -7.91077793e-01
1.12770334e-01 1.39244810e-01 -3.85618150e-01 8.98607913e-03
7.21867323e-01 -3.88993710e-01 -3.86642337e-01 3.15600276e-01
2.76504219e-01 -2.20214784e-01 1.07023108e-03 -3.28055441e-01
7.59530306e-01 3.43512520e-02 6.07270360e-01 -2.33547166e-01
6.52147412e-01 -3.59715015e-01 4.45614457e-01 9.94497240e-01
-4.05667037e-01 7.57632911e-01 1.77356780e-01 -4.95296806e-01
-8.38997602e-01 -1.21981931e+00 -2.10664064e-01 1.36015701e+00
3.83342505e-01 -3.66750807e-01 -6.25577748e-01 -6.66885912e-01
3.26151431e-01 -1.82282522e-01 -1.08447969e+00 4.78190407e-02
-3.26233387e-01 -6.69190049e-01 7.09306180e-01 9.30113077e-01
8.99345517e-01 -1.14263296e+00 -2.12125525e-01 1.95960909e-01
-6.44700602e-02 -1.13415313e+00 -5.80426812e-01 3.10768396e-01
-6.47914708e-01 -1.01959980e+00 -1.03589284e+00 -6.40481472e-01
6.47957861e-01 2.70087242e-01 1.12750971e+00 5.27801178e-02
-5.81542075e-01 2.13233680e-01 -3.34525764e-01 -4.49302793e-02
3.81849796e-01 4.27961014e-02 -1.18862890e-01 3.65235120e-01
4.30077851e-01 -2.05039889e-01 -6.48274183e-01 4.32554990e-01
-9.27170157e-01 -9.59869176e-02 9.83618140e-01 1.33761358e+00
8.97315502e-01 -1.97735112e-02 3.15808564e-01 -5.42542338e-01
7.71774888e-01 2.58492175e-02 -4.12743539e-01 7.67757893e-01
-3.30794334e-01 2.39223287e-01 4.57691818e-01 -3.06209952e-01
-1.14753842e+00 -4.17798199e-02 -2.41386354e-01 -3.39654267e-01
-4.06042606e-01 2.48811901e-01 -2.91947067e-01 -8.84370625e-01
1.53842553e-01 2.28825331e-01 -6.08564556e-01 -7.35643387e-01
5.69444418e-01 5.93715072e-01 2.82660246e-01 -9.01623607e-01
7.19285905e-01 3.67172033e-01 -3.75408158e-02 -5.43135643e-01
-9.08194423e-01 -5.08406997e-01 -9.89824593e-01 1.57587439e-01
9.08193111e-01 -8.56302142e-01 -9.66845334e-01 9.20484245e-01
-1.08217597e+00 1.24665154e-02 -2.60195315e-01 2.33499750e-01
-4.44848806e-01 2.96394706e-01 -6.74178183e-01 -4.41348165e-01
-3.17851007e-01 -1.25131047e+00 1.42084122e+00 4.99046892e-01
1.08674556e-01 -7.80990064e-01 -2.01349154e-01 3.37741971e-01
6.80236816e-01 -4.37877215e-02 7.07846999e-01 -1.12922534e-01
-8.28241050e-01 -1.29319638e-01 -1.09239221e+00 4.01781052e-01
2.82402515e-01 -2.13717744e-01 -1.16842926e+00 -2.86771715e-01
-5.35024822e-01 -5.48282444e-01 1.08958900e+00 7.80327171e-02
1.33323884e+00 -1.17903031e-01 -1.92488104e-01 4.72734988e-01
1.36725879e+00 -3.36603016e-01 6.12646580e-01 1.91554025e-01
8.32206666e-01 4.51832265e-01 5.56276143e-01 5.92656553e-01
6.06557250e-01 7.11690426e-01 1.31706864e-01 -1.90967411e-01
-4.25726920e-01 -2.83437669e-01 -9.79107916e-02 4.81632262e-01
-1.82917967e-01 5.53260818e-02 -4.80194449e-01 3.83521736e-01
-1.79284453e+00 -8.03782940e-01 2.77320474e-01 1.71909928e+00
6.51307821e-01 -7.41097778e-02 -7.64736608e-02 7.74366334e-02
5.43948114e-01 3.09913337e-01 -1.25506565e-01 -4.49711055e-01
-5.26171505e-01 6.64594233e-01 3.87796819e-01 4.96869147e-01
-1.36760807e+00 1.26175463e+00 6.79953384e+00 1.02523005e+00
-1.13287115e+00 1.54359236e-01 4.60985273e-01 3.69806528e-01
-2.28577569e-01 -3.58329147e-01 -1.01708710e+00 1.55889586e-01
2.46933490e-01 2.03098148e-01 2.66583651e-01 5.03610551e-01
-3.02972227e-01 -3.13816145e-02 -9.10311460e-01 1.02949309e+00
8.59270841e-02 -1.46075177e+00 4.83006239e-01 6.33352529e-03
8.87332499e-01 -2.78435290e-01 1.55757904e-01 1.54884160e-01
2.57528663e-01 -1.48442781e+00 7.96650469e-01 9.93756652e-01
7.42891610e-01 -5.77895045e-01 7.04685032e-01 -5.15494086e-02
-1.71912611e+00 -7.27837607e-02 -4.08915669e-01 -2.30267331e-01
-6.23067841e-02 5.93961060e-01 -3.03199589e-01 6.98081076e-01
9.73120749e-01 4.09432888e-01 -8.36969495e-01 1.12044859e+00
-1.71581686e-01 2.33217180e-02 -1.44442111e-01 -1.06497174e-02
4.53017443e-01 3.27733666e-01 -1.60626411e-01 1.47263956e+00
3.23761851e-01 2.02695392e-02 3.25465947e-01 7.87029326e-01
-1.89471673e-02 -1.17636928e-02 -1.94265589e-01 -2.46392146e-01
1.30921915e-01 1.37056303e+00 -6.66323960e-01 -3.06383312e-01
-6.36074364e-01 9.92273927e-01 7.59809375e-01 2.39869207e-01
-4.98114586e-01 -4.74971741e-01 9.63451385e-01 -2.35621706e-01
8.42772305e-01 -4.36807632e-01 -4.66540277e-01 -1.34946585e+00
3.00945658e-02 -6.49848223e-01 4.76958871e-01 -5.19514859e-01
-1.88712478e+00 8.45960855e-01 -3.74243766e-01 -9.48042750e-01
1.16108790e-01 -1.02476180e+00 -2.55752683e-01 1.33105552e+00
-1.95646727e+00 -1.90294707e+00 -6.21195614e-01 1.00720334e+00
1.55222520e-01 1.72812149e-01 1.15106916e+00 4.59214658e-01
-1.09668761e-01 9.10859048e-01 -3.50964725e-01 -5.63676246e-02
6.49242640e-01 -1.09197843e+00 2.47623622e-01 6.90303087e-01
1.06192902e-01 8.26904118e-01 1.44278482e-01 -4.41085547e-01
-1.24863195e+00 -1.11876619e+00 1.15095603e+00 -3.71987045e-01
5.39272547e-01 -5.15948594e-01 -8.77105355e-01 3.28207344e-01
-1.60687983e-01 7.38247216e-01 8.25210154e-01 1.60823837e-01
-1.09556973e+00 -2.34657615e-01 -1.31234777e+00 3.03774029e-01
8.96426976e-01 -9.10139859e-01 -8.79873037e-01 -4.13032830e-01
3.08108121e-01 -1.33681819e-01 -1.13168335e+00 7.93168426e-01
1.09738302e+00 -1.08749485e+00 1.22733796e+00 -6.30240381e-01
1.94550622e-02 -5.97592294e-01 -7.34682858e-01 -1.13697577e+00
-9.26885962e-01 2.12303940e-02 -1.71195924e-01 1.07788432e+00
2.83230208e-02 -7.61241019e-01 7.39632130e-01 4.05506462e-01
7.49844238e-02 -8.35576177e-01 -8.46855402e-01 -6.90846205e-01
8.27681050e-02 -2.91301221e-01 8.49017501e-01 5.60331047e-01
-1.70173690e-01 -1.59666434e-01 -4.02883649e-01 3.20866615e-01
7.22488344e-01 6.99469090e-01 4.94109273e-01 -1.21778488e+00
-1.70879394e-01 -6.73326671e-01 -7.29348004e-01 -1.05263197e+00
2.15419699e-02 -6.47516191e-01 1.86922908e-01 -1.56444967e+00
5.46508074e-01 -6.50567889e-01 -1.02054167e+00 7.56601751e-01
-3.52200150e-01 1.24516928e+00 4.14270192e-01 9.47740301e-02
-6.29126549e-01 5.90359569e-01 1.48317361e+00 -3.42778802e-01
2.60289311e-01 -3.17309082e-01 -9.37991500e-01 4.43300247e-01
4.49531674e-01 4.00102548e-02 1.99400023e-01 -2.34210506e-01
-3.41200501e-01 -4.00181890e-01 6.44445479e-01 -8.74005020e-01
3.91943842e-01 -2.15683937e-01 9.39996362e-01 -7.12487102e-01
5.93843579e-01 -7.06702292e-01 -3.52895074e-02 2.91403621e-01
-1.77962482e-02 -2.36158967e-01 4.53349948e-01 1.93834931e-01
-8.05566609e-01 2.61416644e-01 8.09332430e-01 -2.20792428e-01
-1.14629662e+00 6.84568226e-01 -7.49062654e-03 -2.58570701e-01
8.81948650e-01 -3.01606983e-01 -4.25101310e-01 1.85009196e-01
-7.96685100e-01 -1.89664349e-01 2.67483741e-01 6.91014111e-01
6.15098536e-01 -1.65153158e+00 -5.12182713e-01 5.12651622e-01
4.70798790e-01 -5.73904157e-01 7.00131476e-01 5.62257946e-01
-1.02778584e-01 8.90008986e-01 -6.63206875e-01 -3.58072281e-01
-1.17643392e+00 4.88974959e-01 2.06269503e-01 -7.31998384e-01
-1.02887310e-01 1.13672400e+00 6.80141568e-01 -5.27570128e-01
1.50973454e-01 -4.12757844e-01 -2.27810711e-01 2.43253782e-01
6.12311065e-01 3.27874392e-01 1.83364645e-01 -8.96620214e-01
-6.09839201e-01 1.19982827e+00 -2.22632140e-01 3.17463547e-01
1.20040810e+00 -1.03976004e-01 -2.47597009e-01 2.61074364e-01
1.19834876e+00 2.40366198e-02 -1.41797733e+00 -4.31964636e-01
-7.35413581e-02 -7.49591053e-01 -1.41422987e-01 -1.08642960e+00
-7.06284463e-01 1.08900845e+00 6.12561285e-01 1.43972889e-01
1.26463938e+00 1.91469878e-01 5.49565315e-01 1.10803634e-01
5.73177338e-01 -6.81698680e-01 1.52662009e-01 8.43348563e-01
1.11743927e+00 -1.18881834e+00 -5.03820814e-02 -7.30937302e-01
-1.39199868e-01 1.19563091e+00 6.24825835e-01 -2.08649293e-01
9.12223339e-01 3.56302738e-01 5.73596843e-02 -3.47394198e-02
-4.54881430e-01 -4.74077582e-01 9.80758250e-01 3.84201586e-01
5.60510337e-01 3.48612815e-01 -2.52066076e-01 9.09380317e-01
1.96561329e-02 2.48322323e-01 -4.41965848e-01 8.16633999e-01
-3.86337370e-01 -1.40803826e+00 -4.56727207e-01 3.15185398e-01
-4.77254301e-01 -2.85863459e-01 -3.63790333e-01 4.80286956e-01
5.05629838e-01 7.99130976e-01 8.38127211e-02 -4.80237246e-01
4.55436319e-01 7.76644051e-02 1.05141413e+00 -4.34895277e-01
-8.33792269e-01 -1.50904134e-01 -2.80447811e-01 -9.13740516e-01
-3.50136399e-01 -1.19960673e-01 -7.76256025e-01 -2.99599916e-01
-6.19525239e-02 -7.01702759e-02 4.29625332e-01 1.05086780e+00
5.08573532e-01 5.16811848e-01 5.04738748e-01 -1.23741853e+00
-3.09808046e-01 -8.64236176e-01 -8.94208550e-01 3.30431461e-01
3.11023563e-01 -1.10917604e+00 -4.96367849e-02 -2.21452162e-01] | [9.593878746032715, 2.021925210952759] |
41ea5cd3-61d0-4f42-8a46-944608755fa7 | using-structured-events-to-predict-stock | null | null | https://aclanthology.org/D14-1148 | https://aclanthology.org/D14-1148.pdf | Using Structured Events to Predict Stock Price Movement: An Empirical Investigation | null | ['Yue Zhang', 'Ting Liu', 'Junwen Duan', 'Xiao Ding'] | 2014-10-01 | null | null | null | emnlp-2014-10 | ['stock-market-prediction', 'stock-prediction'] | ['time-series', 'time-series'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.203945636749268, 3.779231309890747] |
d5147de8-929c-4bb4-9765-44f58fd9b758 | from-examples-to-rules-neural-guided-rule | 2202.00475 | null | https://arxiv.org/abs/2202.00475v1 | https://arxiv.org/pdf/2202.00475v1.pdf | From Examples to Rules: Neural Guided Rule Synthesis for Information Extraction | While deep learning approaches to information extraction have had many successes, they can be difficult to augment or maintain as needs shift. Rule-based methods, on the other hand, can be more easily modified. However, crafting rules requires expertise in linguistics and the domain of interest, making it infeasible for most users. Here we attempt to combine the advantages of these two directions while mitigating their drawbacks. We adapt recent advances from the adjacent field of program synthesis to information extraction, synthesizing rules from provided examples. We use a transformer-based architecture to guide an enumerative search, and show that this reduces the number of steps that need to be explored before a rule is found. Further, we show that without training the synthesis algorithm on the specific domain, our synthesized rules achieve state-of-the-art performance on the 1-shot scenario of a task that focuses on few-shot learning for relation classification, and competitive performance in the 5-shot scenario. | ['Mihai Surdeanu', 'Rebecca Sharp', 'George C. G. Barbosa', 'Marco A. Valenzuela-Escarcega', 'Robert Vacareanu'] | 2022-01-16 | null | https://aclanthology.org/2022.lrec-1.665 | https://aclanthology.org/2022.lrec-1.665.pdf | lrec-2022-6 | ['enumerative-search', 'relation-classification'] | ['computer-code', 'natural-language-processing'] | [ 4.72419024e-01 2.44887948e-01 -4.07226175e-01 -3.46379042e-01
-8.27836931e-01 -4.80092347e-01 4.91700709e-01 4.68484342e-01
-2.26333663e-01 5.45637131e-01 6.79479837e-02 -8.40327024e-01
-4.57577333e-02 -9.50889409e-01 -6.58378601e-01 -1.59673095e-01
2.27389395e-01 3.47766221e-01 4.95992631e-01 -5.11254370e-01
2.33308792e-01 2.77817369e-01 -1.89664042e+00 4.03713942e-01
8.49397004e-01 6.80271089e-01 2.53109299e-02 6.82092071e-01
-4.32790011e-01 1.16065836e+00 -4.42308068e-01 -4.79970574e-01
-2.64071953e-02 -4.69667315e-01 -9.64243770e-01 -7.10614473e-02
2.63846964e-01 -4.54532385e-01 -3.61939490e-01 8.79046440e-01
2.60150522e-01 2.89069563e-01 5.18840790e-01 -1.05094385e+00
-5.52190483e-01 9.83032644e-01 -3.00681770e-01 1.82404578e-01
4.68255579e-01 6.88557699e-02 1.20869505e+00 -8.42878997e-01
8.35357845e-01 8.63476217e-01 6.10131800e-01 5.66317558e-01
-1.18004906e+00 -5.40387034e-01 1.53679416e-01 3.43508899e-01
-1.01542068e+00 -8.54241848e-01 6.45569146e-01 -4.37159508e-01
1.72744691e+00 2.95878630e-02 6.42005563e-01 7.60853112e-01
-1.02071881e-01 9.43125367e-01 5.48365295e-01 -1.02704465e+00
2.23888829e-01 1.18869813e-02 4.66430038e-01 9.76008654e-01
4.48452622e-01 -6.88422918e-02 -5.55831313e-01 -2.60611717e-02
3.61252546e-01 -2.71786153e-01 -2.15273112e-01 -5.84261894e-01
-8.16594005e-01 9.97054577e-01 2.58445982e-02 5.98077834e-01
-6.83898479e-02 -6.16001934e-02 5.20591855e-01 5.75038612e-01
2.05158219e-01 8.76957655e-01 -5.78813910e-01 -4.00450349e-01
-1.20813990e+00 4.43263859e-01 1.25839150e+00 1.19910514e+00
6.45460308e-01 -3.26875076e-02 -1.08732574e-01 7.93439150e-01
9.22753066e-02 -1.15555100e-01 3.94930929e-01 -8.79352808e-01
4.63573843e-01 7.07601845e-01 7.16024190e-02 -8.25722218e-01
-3.88891101e-01 -2.18315318e-01 -1.38133109e-01 3.10455173e-01
7.48266876e-01 -2.36193776e-01 -8.93283427e-01 1.54437411e+00
1.09794132e-01 -3.34201187e-01 8.06507766e-02 3.17647785e-01
7.70814538e-01 6.33142948e-01 7.62818754e-02 -4.28867370e-01
1.43787587e+00 -1.02778125e+00 -6.42861903e-01 -5.88655114e-01
1.06764591e+00 -5.10810256e-01 1.14650631e+00 5.00340044e-01
-1.25987411e+00 -2.50208139e-01 -1.48788810e+00 -2.75786310e-01
-6.33698761e-01 -2.94265598e-01 9.13625777e-01 5.95258534e-01
-7.61759460e-01 7.58901775e-01 -8.71210575e-01 -5.29344261e-01
4.97134298e-01 2.67449588e-01 -5.99619672e-02 -8.22429582e-02
-1.20740283e+00 1.24341059e+00 6.05101347e-01 -3.54020327e-01
-4.77289706e-01 -9.24287379e-01 -1.07948267e+00 4.19840366e-01
9.20070052e-01 -3.97643030e-01 1.81633282e+00 -5.81444085e-01
-1.38498855e+00 7.90814638e-01 -2.66328067e-01 -4.75775570e-01
1.67321697e-01 -1.02707259e-01 -2.48024002e-01 -1.19848132e-01
-1.37106016e-01 3.56980652e-01 4.99810755e-01 -8.09073627e-01
-9.28857327e-01 -1.07728548e-01 3.94339740e-01 2.34147869e-02
-5.84939301e-01 3.02284539e-01 -3.59565347e-01 -4.04821783e-01
-1.96733207e-01 -6.94587529e-01 -7.48152286e-02 -9.66381952e-02
-1.42188057e-01 -4.86508667e-01 7.42238700e-01 -6.06755912e-01
1.61643338e+00 -1.96880579e+00 3.17500941e-02 7.43962824e-02
1.81280136e-01 5.57527423e-01 9.03534368e-02 4.58770037e-01
-7.43770367e-03 1.67871431e-01 -3.15861017e-01 1.11029390e-02
1.85489699e-01 1.42563492e-01 -2.66402543e-01 -2.95021385e-02
4.60194260e-01 8.64495039e-01 -1.07266688e+00 -6.36276066e-01
1.18276417e-01 7.30291307e-02 -6.68325603e-01 1.33964300e-01
-5.31211019e-01 -2.87218750e-01 -2.70869464e-01 3.92290115e-01
9.38819945e-02 -2.94626713e-01 5.39732277e-01 -3.17617282e-02
5.80710471e-02 7.23943770e-01 -1.11530054e+00 1.85993993e+00
-8.45858693e-01 6.84218347e-01 -2.64927328e-01 -1.09305954e+00
8.07242334e-01 4.19661552e-01 2.37400368e-01 -4.21930611e-01
5.46685793e-03 3.54372442e-01 3.59337240e-01 -6.59500897e-01
5.00866532e-01 -2.95556992e-01 -2.47342631e-01 7.92888463e-01
3.89475018e-01 -2.34327629e-01 7.60502279e-01 1.66923255e-01
1.38236356e+00 4.11362737e-01 9.20826674e-01 1.63452968e-01
2.53921002e-01 5.19988239e-01 5.03177226e-01 7.88494647e-01
3.43102738e-02 1.50447726e-01 6.58867836e-01 -5.21757722e-01
-1.02380979e+00 -5.19038856e-01 9.32252035e-02 1.49083197e+00
-3.99358332e-01 -6.46727920e-01 -6.95553660e-01 -8.72601807e-01
-1.43404841e-01 1.18303382e+00 -2.73814112e-01 -2.09031001e-01
-8.23974848e-01 -5.72852612e-01 5.89740574e-01 5.95745265e-01
1.63239524e-01 -1.13741195e+00 -1.25787473e+00 4.96745586e-01
-6.82602152e-02 -1.03485906e+00 -1.55880913e-01 5.95907032e-01
-6.70747280e-01 -1.12630606e+00 -2.94521809e-01 -8.73713613e-01
3.65152240e-01 3.75736281e-02 1.41555500e+00 3.51646960e-01
-1.32773668e-01 1.51243731e-01 -5.83179653e-01 -6.76134706e-01
-6.30683601e-01 4.16125953e-01 -4.82635736e-01 -6.21290386e-01
9.48286176e-01 -5.62272429e-01 7.04206601e-02 -1.28699824e-01
-7.22394347e-01 1.27342016e-01 3.93467188e-01 9.28836405e-01
4.72321808e-02 2.10122436e-01 6.09393120e-01 -1.31095982e+00
6.89621568e-01 -2.30085298e-01 -4.77759540e-01 5.05211890e-01
-8.23876858e-01 2.65513510e-01 4.95191306e-01 -2.89501309e-01
-1.10296190e+00 2.86862075e-01 6.99308962e-02 -8.74870569e-02
-1.84461236e-01 7.72510290e-01 -2.08229318e-01 5.04704658e-03
1.01248419e+00 -5.62562272e-02 -2.10578337e-01 -2.81611860e-01
5.48682451e-01 6.57828271e-01 3.79820287e-01 -6.61667168e-01
5.87171376e-01 -7.27270320e-02 -2.14491472e-01 -5.89501858e-01
-8.78336430e-01 -7.89577141e-02 -5.78212142e-01 1.31780624e-01
6.19756162e-01 -4.49507028e-01 -3.37743551e-01 8.17862619e-03
-1.41329038e+00 -4.94631559e-01 -4.58217084e-01 8.47217962e-02
-7.49380469e-01 1.50642857e-01 -6.81961358e-01 -6.54235065e-01
-3.70587379e-01 -1.07629812e+00 6.66754901e-01 1.52094886e-01
-8.70897174e-01 -7.72941113e-01 -5.10759987e-02 8.73425826e-02
3.53916317e-01 1.43961878e-02 1.53957045e+00 -1.13906348e+00
-3.28901023e-01 -2.76281774e-01 -1.98663697e-02 -1.39060169e-02
1.48135737e-01 5.07091396e-02 -9.08057690e-01 1.12079836e-01
-9.87436622e-03 -4.85817969e-01 5.88323951e-01 9.09215808e-02
9.18818414e-01 -2.95766562e-01 -4.96592611e-01 3.17751229e-01
1.13700223e+00 6.11395240e-01 4.83387381e-01 4.13701713e-01
5.44413567e-01 7.72604048e-01 6.02956057e-01 2.22740129e-01
2.87129611e-01 7.18266726e-01 -1.28443614e-01 1.19733676e-01
-2.23275125e-01 -2.40398288e-01 1.55594796e-01 6.16586089e-01
-7.95318857e-02 1.35317016e-02 -1.29910243e+00 7.09645331e-01
-1.97890282e+00 -9.74642992e-01 3.95004779e-01 2.01529717e+00
1.26225936e+00 5.39307475e-01 2.28327066e-01 4.95381594e-01
4.66339022e-01 1.08340904e-01 -5.00629544e-01 -6.87701344e-01
3.37571621e-01 6.05992973e-01 1.21251591e-01 5.44233143e-01
-1.04062998e+00 1.23304284e+00 6.66345072e+00 7.36490071e-01
-8.88645709e-01 1.37603357e-02 2.93746471e-01 -2.76980758e-01
-3.83839965e-01 2.34524652e-01 -8.16439986e-01 -4.93795238e-02
1.12568724e+00 -3.93987238e-01 5.50113678e-01 9.46011305e-01
-2.53308356e-01 -7.59282857e-02 -1.68708146e+00 7.78352678e-01
1.77327514e-01 -1.50239301e+00 -1.87991589e-01 -2.64423490e-01
4.88534778e-01 -2.50475466e-01 -3.75058591e-01 7.03837931e-01
5.27803123e-01 -1.08426940e+00 6.45010114e-01 1.97109073e-01
5.65716863e-01 -7.50719905e-01 4.29723769e-01 3.94768536e-01
-9.97524202e-01 -1.67620480e-01 -1.66848883e-01 -4.57627833e-01
4.22820495e-03 2.95176119e-01 -8.45364332e-01 4.35945243e-01
2.45209709e-01 5.53192437e-01 -5.28916717e-01 9.35536623e-01
-5.12051940e-01 4.20029879e-01 -1.50919110e-01 -4.17895019e-01
3.22863646e-03 3.54639262e-01 3.48094702e-01 1.33041847e+00
2.90286928e-01 2.61216640e-01 1.76382840e-01 7.96154201e-01
-4.45160493e-02 6.57833740e-02 -9.03478920e-01 -2.82230258e-01
6.24938250e-01 1.07541132e+00 -8.51433873e-01 -6.58695638e-01
-7.20066130e-01 4.43962127e-01 5.17939329e-01 2.44594067e-01
-5.73797524e-01 -1.16665137e+00 2.29391471e-01 -3.54377367e-02
7.11124837e-01 -1.15605488e-01 -4.50726658e-01 -1.17225218e+00
3.23019065e-02 -1.41948402e+00 4.65998054e-01 -5.57883799e-01
-7.47091591e-01 4.04276639e-01 2.71568537e-01 -8.75249982e-01
-7.56124556e-01 -6.25353098e-01 -5.01925766e-01 5.70026338e-01
-1.15367341e+00 -9.18221772e-01 1.02318734e-01 2.78685957e-01
7.81824470e-01 -1.15095481e-01 1.06261778e+00 2.63736278e-01
-4.27215785e-01 7.07412124e-01 -4.91266906e-01 2.05772161e-01
3.96786273e-01 -1.07693493e+00 4.66835260e-01 1.14340723e+00
4.63719398e-01 7.47379839e-01 9.61478889e-01 -5.16285419e-01
-1.60355437e+00 -6.53923035e-01 1.22060990e+00 -4.19839948e-01
8.83955956e-01 -3.85630459e-01 -9.06582117e-01 8.80546749e-01
1.23219959e-01 -1.67916194e-01 6.13339663e-01 5.75159073e-01
-7.03689754e-01 1.02969639e-01 -8.72541130e-01 7.90069163e-01
1.08821154e+00 -6.61777496e-01 -1.00072443e+00 3.25291514e-01
7.69232929e-01 -4.42875773e-01 -9.77112055e-01 2.58066148e-01
6.01259947e-01 -6.69366241e-01 5.32596827e-01 -1.00948346e+00
6.84941351e-01 -1.52451217e-01 -2.00747639e-01 -1.17548764e+00
-2.33186066e-01 -8.20617378e-01 -6.39758468e-01 1.23554218e+00
9.21676815e-01 -1.65790513e-01 7.42814302e-01 8.02282274e-01
-2.67600894e-01 -9.42221820e-01 -5.28183520e-01 -7.89819360e-01
2.20915347e-01 -7.19995916e-01 7.12307811e-01 9.06241536e-01
9.02218103e-01 8.85704637e-01 -1.96899369e-01 -4.28286403e-01
1.30472451e-01 2.44426355e-01 6.87491775e-01 -1.35717511e+00
-5.31402647e-01 -6.33649349e-01 -8.57678801e-02 -8.20196629e-01
1.86986357e-01 -1.02074146e+00 4.52749163e-01 -1.67111290e+00
1.27556279e-01 -2.83892930e-01 -2.34559268e-01 1.01031661e+00
-9.66682434e-02 -2.46047765e-01 3.36018473e-01 -2.09289556e-03
-4.69072461e-01 -1.38477432e-02 8.02818000e-01 -2.39368454e-01
-4.58273053e-01 -1.69659719e-01 -1.05215776e+00 8.34574044e-01
5.66218913e-01 -4.63502288e-01 -5.86833239e-01 -4.03522611e-01
5.03688514e-01 2.63055973e-02 -1.72841430e-01 -9.95981932e-01
4.47630823e-01 -1.32812679e-01 9.57109593e-03 -2.29576945e-01
1.34961486e-01 -6.19522810e-01 -2.45492235e-01 4.28593218e-01
-5.90897083e-01 -1.02364227e-01 3.14620167e-01 2.57098049e-01
-8.74433741e-02 -7.42766082e-01 5.70320606e-01 -1.82836995e-01
-8.04710090e-01 8.97668488e-03 -5.50116897e-01 1.32056996e-01
9.18129921e-01 -1.73527151e-01 -4.02391523e-01 -3.62112582e-01
-5.67160606e-01 1.90224443e-02 4.81851965e-01 2.56913275e-01
4.74449575e-01 -9.86393034e-01 -3.56205374e-01 5.66419549e-02
2.87415087e-01 -1.21583842e-01 -3.03780496e-01 7.50616550e-01
-3.18989158e-01 5.03248751e-01 -1.41517848e-01 -1.66720942e-01
-1.21049082e+00 9.02487218e-01 1.60058647e-01 -5.76665819e-01
-8.22902262e-01 6.24361515e-01 -2.87161410e-01 -3.21486264e-01
4.65282351e-01 -4.65129375e-01 -3.72580945e-01 2.89769679e-01
7.65380919e-01 1.94566607e-01 4.57154870e-01 -2.90404689e-02
-2.88598686e-01 3.42080265e-01 -5.24970531e-01 -1.34996042e-01
1.50546670e+00 3.74994010e-01 6.93243891e-02 5.25803089e-01
7.43466735e-01 -2.31311753e-01 -9.24112260e-01 -3.80940497e-01
5.57767332e-01 -1.41283870e-01 5.77957742e-02 -6.21213794e-01
-6.70959234e-01 8.79803658e-01 3.75315957e-02 4.56939548e-01
1.07359171e+00 -1.11362383e-01 9.49163139e-01 8.92727196e-01
3.32035273e-01 -1.12439764e+00 -1.73445448e-01 8.16812873e-01
3.79654557e-01 -1.08816624e+00 9.57704112e-02 -5.18234074e-01
-3.72354358e-01 1.42220652e+00 5.17086864e-01 -1.73202902e-02
5.51589370e-01 5.94158232e-01 -2.30821297e-01 -3.23510885e-01
-1.04802811e+00 -2.23589689e-01 8.52717608e-02 6.09682262e-01
7.92294562e-01 -2.19407097e-01 -2.93136686e-01 5.89382887e-01
-2.58099318e-01 3.45435679e-01 4.97736573e-01 1.57459915e+00
-6.85212433e-01 -1.25238812e+00 -2.58699749e-02 7.12654591e-01
-4.18677092e-01 -3.92577201e-01 -2.60063708e-01 9.50859666e-01
4.86289933e-02 1.11741281e+00 -4.21613194e-02 -3.78381729e-01
4.53363210e-01 6.55862391e-01 7.52410591e-01 -1.18080735e+00
-5.59897602e-01 -9.08065140e-02 7.51175344e-01 -3.66319776e-01
-3.48123193e-01 -7.22497225e-01 -1.10444701e+00 -2.26952732e-01
-3.09026957e-01 1.37053639e-01 2.84701437e-01 1.29192901e+00
2.20327303e-01 6.74581707e-01 -1.72404735e-03 -5.86478710e-01
-6.16693258e-01 -8.86619031e-01 -1.06115535e-01 6.90643415e-02
1.46666571e-01 -6.23004079e-01 1.38804996e-02 1.80342525e-01] | [9.57179069519043, 7.779696941375732] |
3138f430-8e27-47cf-90df-2a38b0e441d6 | learning-reconstructability-for-drone-aerial | 2209.10174 | null | https://arxiv.org/abs/2209.10174v1 | https://arxiv.org/pdf/2209.10174v1.pdf | Learning Reconstructability for Drone Aerial Path Planning | We introduce the first learning-based reconstructability predictor to improve view and path planning for large-scale 3D urban scene acquisition using unmanned drones. In contrast to previous heuristic approaches, our method learns a model that explicitly predicts how well a 3D urban scene will be reconstructed from a set of viewpoints. To make such a model trainable and simultaneously applicable to drone path planning, we simulate the proxy-based 3D scene reconstruction during training to set up the prediction. Specifically, the neural network we design is trained to predict the scene reconstructability as a function of the proxy geometry, a set of viewpoints, and optionally a series of scene images acquired in flight. To reconstruct a new urban scene, we first build the 3D scene proxy, then rely on the predicted reconstruction quality and uncertainty measures by our network, based off of the proxy geometry, to guide the drone path planning. We demonstrate that our data-driven reconstructability predictions are more closely correlated to the true reconstruction quality than prior heuristic measures. Further, our learned predictor can be easily integrated into existing path planners to yield improvements. Finally, we devise a new iterative view planning framework, based on the learned reconstructability, and show superior performance of the new planner when reconstructing both synthetic and real scenes. | ['Hui Huang', 'Hao Zhang', 'Chi-Wing Fu', 'Ke Xie', 'Yue Hu', 'Liqiang Lin', 'Yilin Liu'] | 2022-09-21 | null | null | null | null | ['3d-scene-reconstruction'] | ['computer-vision'] | [ 8.68155658e-02 4.39638972e-01 -1.63144141e-01 -5.62967956e-01
-6.33788347e-01 -8.45175564e-01 4.76339340e-01 -2.08237931e-01
8.85188729e-02 7.07640350e-01 3.48979622e-01 -2.31549934e-01
-3.41625512e-01 -1.21660960e+00 -9.36939716e-01 -2.24915043e-01
-6.21719584e-02 9.93161261e-01 1.51633680e-01 -4.43479687e-01
2.16317058e-01 8.34813714e-01 -1.41494060e+00 -1.18900083e-01
8.25170040e-01 8.38803709e-01 4.45304781e-01 7.39976943e-01
6.53005362e-01 5.70097446e-01 1.19480467e-03 1.74975824e-02
8.78339112e-01 -2.71306157e-01 -8.59612167e-01 4.70121205e-01
2.19121650e-01 -7.91814685e-01 -5.34020245e-01 6.47481561e-01
9.63337123e-02 4.14485216e-01 4.46088135e-01 -1.04264140e+00
1.23059146e-01 2.17552125e-01 6.90868720e-02 -2.89323270e-01
6.16460323e-01 4.75731194e-01 9.87360835e-01 -6.99718058e-01
1.00544608e+00 1.01761985e+00 5.78912556e-01 3.26640338e-01
-1.08633447e+00 -2.21379846e-01 2.54715174e-01 1.14780150e-01
-1.11348355e+00 -3.41215283e-01 8.84476304e-01 -3.76857549e-01
9.17846203e-01 -3.98892350e-02 1.25464392e+00 7.35180259e-01
4.56621468e-01 5.29344678e-01 7.80492663e-01 2.32480429e-02
2.66385615e-01 -2.46483609e-01 -6.06510878e-01 1.28023481e+00
1.03911420e-03 7.87173450e-01 -3.93580735e-01 1.34095117e-01
9.82019663e-01 -3.74560542e-02 -7.22326636e-01 -1.06458569e+00
-1.57463503e+00 7.65168726e-01 7.46806026e-01 -5.26112676e-01
-4.74540114e-01 2.79882431e-01 -4.57586311e-02 1.79846078e-01
2.74816960e-01 1.00148511e+00 -6.76040053e-01 -1.93943139e-02
-7.42795944e-01 5.79177737e-01 8.28309417e-01 1.17403126e+00
9.64937270e-01 1.84824422e-01 2.20278665e-01 3.34365100e-01
2.99786359e-01 5.84474266e-01 -3.15338731e-01 -1.63969672e+00
5.83429694e-01 4.73704576e-01 3.51646006e-01 -9.52057540e-01
-4.27047968e-01 -3.62862259e-01 -3.69566619e-01 6.02545917e-01
5.22996373e-02 -1.20163120e-01 -8.42318475e-01 1.59739864e+00
4.79430526e-01 1.19620606e-01 2.68486161e-02 1.09014165e+00
2.48226717e-01 6.85594618e-01 -8.23203266e-01 -1.23210877e-01
4.01178658e-01 -1.02412009e+00 -1.75092593e-01 -3.44660342e-01
4.31709230e-01 -4.20701593e-01 8.91703844e-01 6.42090559e-01
-1.14612842e+00 -5.05323410e-01 -1.27596295e+00 1.16966717e-01
7.55004510e-02 -4.80366312e-02 4.82070714e-01 1.20939359e-01
-1.15022862e+00 9.49367046e-01 -9.30671215e-01 -3.59695435e-01
2.45446712e-01 2.91951358e-01 -3.46508145e-01 -3.81431848e-01
-6.57458246e-01 1.10548496e+00 5.82124054e-01 -6.62121251e-02
-1.76953197e+00 -5.29062450e-01 -1.01836765e+00 -1.61208838e-01
7.45380521e-01 -1.25381434e+00 1.20860410e+00 -5.03270268e-01
-1.84847891e+00 5.60464859e-01 1.83865294e-01 -3.67256999e-01
4.61507887e-01 -1.35700032e-01 -2.95384508e-02 1.31669790e-01
9.12627578e-02 1.07932329e+00 6.12668574e-01 -1.78968859e+00
-7.79777229e-01 -9.17326361e-02 6.79895699e-01 7.58255005e-01
5.92610121e-01 -1.04191196e+00 -4.55834061e-01 -6.24785982e-02
4.83214915e-01 -1.08115387e+00 -5.98042250e-01 3.15645576e-01
-3.94544125e-01 4.46866393e-01 5.82362056e-01 -3.34294349e-01
6.01250827e-01 -1.56123316e+00 6.17056549e-01 2.68440574e-01
2.14569643e-01 -2.42807224e-01 -3.02602768e-01 5.26063681e-01
2.25765467e-01 -1.57656193e-01 -4.52001601e-01 -1.76782906e-01
-2.06767410e-01 5.83402693e-01 -4.13541436e-01 4.34518874e-01
-1.34260848e-01 6.81663871e-01 -1.15919507e+00 -1.36420742e-01
5.08796275e-01 1.83279961e-01 -1.14202321e+00 5.70684433e-01
-6.70020998e-01 8.64656687e-01 -5.73822558e-01 5.89777589e-01
4.35590535e-01 -2.06730748e-03 1.41646117e-01 -1.23453736e-01
-3.43019724e-01 6.00031316e-01 -7.44350433e-01 2.29394722e+00
-7.39347875e-01 5.62759578e-01 3.72647792e-02 -6.25295103e-01
9.58617926e-01 3.16854715e-02 6.53449297e-01 -2.19303593e-01
3.63144167e-02 -1.14075474e-01 -4.02486026e-01 -4.23159599e-01
7.56084800e-01 -3.04078132e-01 -1.47937894e-01 2.79540658e-01
1.15318865e-01 -1.21355915e+00 -2.20839873e-01 -2.84679532e-02
1.08689487e+00 7.17123926e-01 6.12445951e-01 -4.19523418e-02
3.30236733e-01 4.82722312e-01 6.07689440e-01 2.96472400e-01
-5.77224381e-02 6.31806135e-01 3.56305599e-01 -8.80023658e-01
-1.27895260e+00 -1.27597976e+00 1.05201781e-01 3.51365000e-01
5.69027960e-01 -3.94513249e-01 -6.17163062e-01 -8.56023967e-01
-6.20644838e-02 9.95019555e-01 -3.31389636e-01 -7.21816942e-02
-6.69651270e-01 -6.82200789e-02 1.15273900e-01 2.79976547e-01
5.46657085e-01 -6.79452419e-01 -1.02830684e+00 2.31537431e-01
-3.74727666e-01 -1.08022189e+00 -3.18855524e-01 1.40194327e-01
-9.68572617e-01 -1.28288531e+00 2.71828651e-01 -3.71443033e-01
5.64546406e-01 5.42214751e-01 1.22202373e+00 1.50541991e-01
2.13139951e-01 7.90524244e-01 -2.71209717e-01 -3.36572565e-02
-4.76182997e-01 -1.30221263e-01 2.10179523e-01 -4.78708088e-01
-6.10525131e-01 -9.08916593e-01 -4.65123892e-01 4.39012825e-01
-5.88155389e-01 5.56787729e-01 4.27555233e-01 8.39960396e-01
1.01409161e+00 2.72395611e-01 -1.72453389e-01 -5.20299077e-01
2.06882536e-01 -3.63974601e-01 -1.00374925e+00 -5.40245650e-03
-6.33487821e-01 1.97920650e-01 8.11439693e-01 3.85427810e-02
-1.08511996e+00 4.31378037e-01 -1.87344387e-01 -6.27005398e-01
-2.48640075e-01 6.71305478e-01 -3.26135546e-01 -2.10248873e-01
6.34211779e-01 -1.59667749e-02 -2.70333201e-01 -2.85536237e-02
5.74742675e-01 -1.78778455e-01 4.63841379e-01 -6.20099604e-01
1.13435459e+00 6.30736351e-01 2.93893933e-01 -4.18771058e-01
-1.03662777e+00 -4.99739237e-02 -1.06461835e+00 -4.80021298e-01
9.19920623e-01 -9.70526457e-01 -6.39853895e-01 -1.25156835e-01
-1.11779201e+00 -7.91810453e-01 -5.49288869e-01 6.69164360e-01
-1.34535766e+00 9.45832878e-02 -4.82264571e-02 -5.02173066e-01
1.68440610e-01 -1.35198987e+00 1.16865504e+00 -3.30991954e-01
3.71834226e-02 -1.04326510e+00 2.70270944e-01 2.91040659e-01
-8.18611160e-02 7.86984324e-01 8.77932668e-01 8.11847299e-03
-1.52249372e+00 8.98867920e-02 1.31930187e-01 -5.72875813e-02
6.80918843e-02 -1.10344775e-01 -6.16712153e-01 -3.57991874e-01
1.46125123e-01 -3.84858042e-01 5.38931072e-01 4.75856155e-01
1.03165627e+00 -4.45769250e-01 -4.88828748e-01 1.10164964e+00
1.61182928e+00 1.74253613e-01 3.35436970e-01 3.71046513e-01
5.86098254e-01 4.91787046e-01 9.59588528e-01 6.48353755e-01
6.77980304e-01 6.81464612e-01 1.18612075e+00 3.37398142e-01
7.85883665e-02 -9.83322263e-01 2.12211505e-01 7.98000395e-01
-2.99846590e-01 -6.24393463e-01 -1.01667356e+00 3.87795210e-01
-1.63582230e+00 -7.74021506e-01 2.92387694e-01 2.30892038e+00
1.06200248e-01 2.88786650e-01 -1.16114624e-01 -1.24179736e-01
6.16252534e-02 5.14446795e-01 -7.40869761e-01 -9.17620882e-02
4.29772317e-01 -1.96492106e-01 8.01292717e-01 1.00011897e+00
-8.00664425e-01 1.21061790e+00 6.75926161e+00 1.59234643e-01
-9.27702665e-01 -2.20410481e-01 1.83814272e-01 -1.50511101e-01
-8.93157125e-01 5.06086171e-01 -6.26472294e-01 -2.21417293e-01
6.91679657e-01 -8.16272199e-02 1.18025386e+00 1.03947186e+00
3.50784481e-01 -1.45343974e-01 -1.36587214e+00 6.48495495e-01
9.62297097e-02 -1.59331667e+00 2.65219659e-01 3.14490706e-01
8.10689688e-01 1.74798504e-01 -4.03399058e-02 1.70329183e-01
9.34583962e-01 -9.24359500e-01 9.18970585e-01 6.62795782e-01
7.54848242e-01 -9.50806677e-01 4.20340836e-01 9.62916017e-01
-1.23467600e+00 -1.67471126e-01 -3.76466334e-01 -1.05652042e-01
6.27856672e-01 3.24859828e-01 -1.41935909e+00 8.55505586e-01
2.98761964e-01 1.06174457e+00 4.09370894e-03 9.69091713e-01
-5.42258441e-01 1.82440758e-01 -2.57099718e-01 3.54710162e-01
3.87737364e-01 -3.27431828e-01 9.93464828e-01 2.73526847e-01
6.45341158e-01 4.63513106e-01 5.66461205e-01 9.94388521e-01
7.91399106e-02 -4.39533800e-01 -1.24662495e+00 3.75634849e-01
3.90906692e-01 9.73142684e-01 -4.68653321e-01 -1.44857839e-01
-5.20325452e-03 7.79171884e-01 4.37268257e-01 4.29276317e-01
-1.01555312e+00 3.38587105e-01 7.84586549e-01 2.67094493e-01
2.12585285e-01 -7.65922248e-01 -1.47354797e-01 -1.18816221e+00
-2.50647008e-01 -6.96862876e-01 -1.56479016e-01 -1.15045118e+00
-7.28703380e-01 8.40469956e-01 3.47601861e-01 -1.80773652e+00
-5.19101560e-01 -4.91911560e-01 -3.63378346e-01 4.23466861e-01
-1.71187627e+00 -1.12338018e+00 -5.58063507e-01 4.16878998e-01
9.06983793e-01 -1.40798211e-01 7.37203479e-01 -3.90301675e-01
-8.65318924e-02 -2.50513107e-01 -4.30573314e-01 -3.11298102e-01
3.78221273e-01 -1.14720893e+00 5.19120157e-01 8.79408896e-01
5.23251481e-02 1.46951258e-01 7.40835845e-01 -7.51202524e-01
-1.42774415e+00 -1.24895906e+00 2.05011129e-01 -7.25244522e-01
2.62703121e-01 -5.11552170e-02 -2.89028525e-01 1.12199879e+00
-6.62881806e-02 6.90305009e-02 1.87368095e-01 -1.48491576e-01
8.31510797e-02 -1.05083011e-01 -1.23069537e+00 6.13681018e-01
1.60615134e+00 -2.62005001e-01 -5.70030272e-01 2.82818854e-01
9.47587311e-01 -1.01751602e+00 -8.64182889e-01 7.09053755e-01
5.16330242e-01 -1.33036673e+00 1.05606055e+00 -4.66448486e-01
5.36131680e-01 -3.52502197e-01 -4.94226009e-01 -1.87832880e+00
-4.63442683e-01 -6.09922886e-01 9.65594202e-02 3.62366468e-01
4.20807600e-01 -3.42903197e-01 9.74299371e-01 3.22946906e-01
-8.27629387e-01 -8.87095034e-01 -9.72569466e-01 -5.95806897e-01
-2.24132821e-01 -5.68559587e-01 8.51606786e-01 6.44266069e-01
-3.48642498e-01 1.79056361e-01 -4.99489009e-01 7.01253176e-01
5.42548776e-01 3.97616327e-01 1.14888227e+00 -1.19563162e+00
-3.97183448e-01 -4.09514457e-02 -1.64556637e-01 -1.78672540e+00
1.84567541e-01 -8.20400238e-01 5.68784297e-01 -1.66168964e+00
-4.44646180e-01 -6.50965273e-01 4.04517740e-01 2.38639608e-01
2.90889412e-01 -4.36221480e-01 1.71400577e-01 1.87625125e-01
-4.55021143e-01 8.88305187e-01 1.75920367e+00 1.69363022e-01
-4.33409542e-01 1.77790686e-01 -2.28453696e-01 1.04317439e+00
6.24612331e-01 -4.29181993e-01 -9.10831928e-01 -7.25573003e-01
3.61686110e-01 9.52781200e-01 3.24000984e-01 -1.35928428e+00
1.14567213e-01 -7.50924647e-01 3.44619066e-01 -9.35217917e-01
8.26783299e-01 -1.15004826e+00 2.05687687e-01 4.00534928e-01
-2.88344353e-01 2.59954661e-01 3.48877981e-02 8.83329213e-01
-1.52770272e-02 1.55968875e-01 6.88250601e-01 -5.51489592e-01
-6.55016541e-01 7.99702048e-01 -3.24872047e-01 -1.03096068e-01
1.05403399e+00 -3.78164053e-01 9.89822373e-02 -5.98316431e-01
-5.64100266e-01 3.81597877e-01 9.54898059e-01 4.76657338e-02
1.17456710e+00 -1.34430075e+00 -4.37221259e-01 4.24019486e-01
3.99785005e-02 5.81255138e-01 -4.24383245e-02 4.54088271e-01
-9.13761139e-01 4.43413913e-01 -4.87554312e-01 -6.97658181e-01
-6.37719274e-01 6.70134962e-01 5.36098182e-01 -4.56710339e-01
-6.62562966e-01 6.11554682e-01 3.15979004e-01 -9.86273289e-01
-1.99130937e-01 -7.74565876e-01 1.35717034e-01 -5.92152119e-01
-1.18730329e-01 3.08654368e-01 -2.39735425e-01 -4.44118649e-01
2.14139372e-02 7.18261182e-01 5.39192975e-01 -4.85863358e-01
1.55099511e+00 -4.44193065e-01 2.54262030e-01 2.58125544e-01
7.52432883e-01 1.82751194e-01 -2.11738825e+00 9.38418582e-02
-4.94003296e-01 -7.59024441e-01 1.26548648e-01 -7.15329945e-01
-1.00249386e+00 6.55532598e-01 8.07122514e-02 -2.46889338e-01
1.07509911e+00 -2.59347707e-01 9.17352319e-01 8.00512195e-01
1.09108233e+00 -8.34155381e-01 3.16301316e-01 6.75780296e-01
1.18257475e+00 -1.20656586e+00 2.84918517e-01 -4.47759002e-01
-7.77335823e-01 1.30472314e+00 6.13211334e-01 -3.52239192e-01
5.84534645e-01 1.12152755e-01 -1.16007693e-01 -3.57045114e-01
-9.77920651e-01 -1.08915143e-01 4.14715946e-01 7.85704374e-01
-3.50529701e-01 1.55905291e-01 4.91924495e-01 -7.97894690e-03
-7.14065731e-01 -4.18687314e-02 7.19518661e-01 6.79371119e-01
-8.23195934e-01 -9.62540865e-01 -1.46626472e-01 1.50020882e-01
6.36182547e-01 1.55946419e-01 -2.48070613e-01 9.97676730e-01
3.67165416e-01 6.64564013e-01 -1.56058565e-01 -8.81833017e-01
4.98841435e-01 -3.91236097e-01 6.18018568e-01 -8.82192850e-01
-5.97179197e-02 -3.09403628e-01 4.59425420e-01 -1.04118681e+00
-2.58741945e-01 -5.82822204e-01 -1.20126092e+00 -3.60948801e-01
-8.83849785e-02 -4.44651172e-02 3.78885269e-01 9.66241598e-01
1.49728402e-01 3.69964808e-01 8.87411714e-01 -1.26364255e+00
-1.89975157e-01 -3.47249031e-01 -2.59411871e-01 -2.06154466e-01
5.87381065e-01 -9.25721288e-01 -1.59407496e-01 -1.81486253e-02] | [8.636751174926758, -2.8139805793762207] |
09e4bbbb-4e07-49b1-a0e5-9155cc1c6fd5 | impact-of-physical-activity-on-sleepa-deep | 1607.07034 | null | http://arxiv.org/abs/1607.07034v1 | http://arxiv.org/pdf/1607.07034v1.pdf | Impact of Physical Activity on Sleep:A Deep Learning Based Exploration | The importance of sleep is paramount for maintaining physical, emotional and
mental wellbeing. Though the relationship between sleep and physical activity
is known to be important, it is not yet fully understood. The explosion in
popularity of actigraphy and wearable devices, provides a unique opportunity to
understand this relationship. Leveraging this information source requires new
tools to be developed to facilitate data-driven research for sleep and activity
patient-recommendations.
In this paper we explore the use of deep learning to build sleep quality
prediction models based on actigraphy data. We first use deep learning as a
pure model building device by performing human activity recognition (HAR) on
raw sensor data, and using deep learning to build sleep prediction models. We
compare the deep learning models with those build using classical approaches,
i.e. logistic regression, support vector machines, random forest and adaboost.
Secondly, we employ the advantage of deep learning with its ability to handle
high dimensional datasets. We explore several deep learning models on the raw
wearable sensor output without performing HAR or any other feature extraction.
Our results show that using a convolutional neural network on the raw
wearables output improves the predictive value of sleep quality from physical
activity, by an additional 8% compared to state-of-the-art non-deep learning
approaches, which itself shows a 15% improvement over current practice.
Moreover, utilizing deep learning on raw data eliminates the need for data
pre-processing and simplifies the overall workflow to analyze actigraphy data
for sleep and physical activity research. | ['Luis Fernandez-Luque', 'Ferda Ofli', 'Ahmed Elmagarmid', 'Shahrad Taheri', 'Jaideep Srivastava', 'Shafiq Joty', 'Aarti Sathyanarayana', 'Teresa Arora'] | 2016-07-24 | null | null | null | null | ['sleep-quality-prediction', 'sleep-quality-prediction-1'] | ['medical', 'medical'] | [-2.63800509e-02 -1.42432541e-01 -4.08799022e-01 -4.98839021e-01
-1.81988701e-01 -2.21253559e-02 1.60805881e-01 2.95177937e-01
-7.60589480e-01 9.06698585e-01 5.84615767e-01 -3.89068991e-01
-2.04799309e-01 -8.95147622e-01 -2.96313316e-01 -6.62072122e-01
-1.26771510e-01 2.49124914e-01 -2.59807587e-01 -1.25863040e-02
-2.97636539e-01 4.11331624e-01 -1.62921858e+00 -1.12078197e-01
6.61127090e-01 1.30698025e+00 -2.32055590e-01 7.62023211e-01
-1.63708515e-02 6.85324669e-01 -5.59759498e-01 1.66317523e-01
1.83761209e-01 -4.46876645e-01 -4.39166665e-01 -5.91660179e-02
2.16149151e-01 -4.70238656e-01 -1.40015213e-02 1.63946882e-01
7.81407893e-01 3.08458716e-01 1.53357014e-01 -1.11334288e+00
-2.66098469e-01 -8.88941884e-02 3.96530032e-02 6.62390530e-01
5.36601484e-01 3.50536644e-01 5.52368641e-01 -4.03580874e-01
-7.98525289e-02 4.88908619e-01 1.15415382e+00 4.44773495e-01
-1.44332254e+00 -7.20170140e-01 -5.45188308e-01 3.65977317e-01
-1.20681214e+00 -6.44616842e-01 5.24616897e-01 -2.63113916e-01
1.36096430e+00 2.64198065e-01 1.55223668e+00 1.19076967e+00
4.89180356e-01 3.56953770e-01 1.34642327e+00 -2.90738434e-01
5.39116621e-01 -3.40064205e-02 2.56268919e-01 7.14137971e-01
3.82215619e-01 1.57643110e-01 -6.98140621e-01 4.84378226e-02
3.24036151e-01 7.35628724e-01 1.61557928e-01 3.68324034e-02
-6.77539825e-01 6.35445833e-01 3.95805389e-01 4.38147277e-01
-6.92451000e-01 1.88709617e-01 1.81536227e-01 2.76766688e-01
5.93668640e-01 2.85118490e-01 -4.96601492e-01 -6.55238092e-01
-1.62208033e+00 -7.92237930e-03 9.54817295e-01 2.89035644e-02
8.17943037e-01 -1.58953723e-02 1.32863577e-02 7.34908342e-01
4.19865429e-01 5.51723719e-01 8.98502529e-01 -9.43010747e-01
-5.90967946e-02 1.05673790e+00 4.93949093e-02 -8.26430023e-01
-1.19736874e+00 -4.26675469e-01 -9.27649617e-01 2.62261540e-01
3.75138730e-01 -1.41802818e-01 -5.92732906e-01 1.22253180e+00
2.82620668e-01 1.91180289e-01 -2.18152910e-01 5.93109488e-01
8.54610443e-01 2.22358301e-01 1.86867937e-01 -2.60242790e-01
1.47341514e+00 -8.45735848e-01 -7.96824574e-01 -3.71098548e-01
5.56583107e-01 -2.42731020e-01 1.21684849e+00 6.07911229e-01
-9.23774421e-01 -5.88067591e-01 -1.40578198e+00 -4.19990659e-01
-7.77423620e-01 1.51539564e-01 7.70210862e-01 1.16940093e+00
-1.28914940e+00 8.86471808e-01 -1.70302260e+00 -6.71800852e-01
8.32318485e-01 1.01623356e+00 -4.75243926e-01 5.06011173e-02
-7.62916148e-01 1.01057005e+00 1.42522082e-02 9.37386826e-02
-5.49418390e-01 -6.32805347e-01 -7.50103772e-01 2.34513909e-01
6.61905250e-03 -8.91047418e-01 9.10764575e-01 -6.78336501e-01
-1.69091535e+00 7.17818081e-01 -1.07609920e-01 -7.62798250e-01
9.98453796e-02 -6.11854315e-01 -7.63812363e-01 -4.43173945e-02
-1.81039497e-01 1.94198728e-01 5.39935231e-01 -3.16114694e-01
-3.14646780e-01 -6.22833788e-01 -1.10900901e-01 -2.92733878e-01
-6.19149745e-01 -2.65267551e-01 1.10745989e-01 -4.27167676e-02
-2.70299703e-01 -9.97274816e-01 -1.73907802e-01 2.77555227e-01
5.99090941e-02 -1.24502391e-01 5.85495472e-01 -7.47850180e-01
1.41902947e+00 -1.89579296e+00 -4.76577312e-01 3.77625562e-02
5.31379580e-01 5.13488054e-01 3.30492914e-01 4.16981995e-01
1.57870322e-01 -1.41442418e-01 -9.96331125e-02 -7.30505645e-01
-7.77670220e-02 7.12678075e-01 4.59877878e-01 7.69277871e-01
-8.79477151e-03 1.09326708e+00 -8.34639430e-01 -3.15746456e-01
6.70142770e-01 6.86532140e-01 -4.92745161e-01 3.71232867e-01
2.54078746e-01 3.64262402e-01 -4.51996513e-02 4.77950305e-01
1.92685813e-01 -3.30716699e-01 1.39781192e-01 -5.42360321e-02
-3.97275269e-01 6.45710051e-01 -6.86038733e-01 1.87985444e+00
-8.17472935e-01 6.40061915e-01 -2.23578408e-01 -1.08377647e+00
1.08131182e+00 1.01968095e-01 7.67216623e-01 -1.12965584e+00
2.57933766e-01 -5.13898358e-02 -7.02376291e-02 -7.70384550e-01
7.73526579e-02 -6.03516221e-01 8.98351446e-02 7.25116134e-01
2.47230530e-01 5.79733253e-01 -1.02780856e-01 -3.50013107e-01
1.55288124e+00 1.44353718e-01 7.58253038e-01 -2.58590817e-01
3.66511613e-01 -2.82519817e-01 5.29463649e-01 4.36772585e-01
-3.60469192e-01 2.52811581e-01 1.14620693e-01 -9.18470263e-01
-6.20717525e-01 -1.01076698e+00 2.66235173e-02 1.16113484e+00
-3.83392692e-01 -6.98644817e-01 -3.82337540e-01 -4.83211815e-01
-5.73434420e-02 4.55381602e-01 -1.01214004e+00 -3.89005929e-01
-4.28922564e-01 -1.00004447e+00 5.31108797e-01 7.57872462e-01
6.81636155e-01 -9.73337948e-01 -1.10180104e+00 4.29402739e-01
-1.97005435e-03 -7.90334463e-01 2.53544688e-01 6.28009558e-01
-1.22752345e+00 -1.08658481e+00 -2.61108309e-01 -6.19133264e-02
1.12340897e-01 2.22300850e-02 1.16191328e+00 2.32498571e-01
-4.06149954e-01 4.11447406e-01 -1.46575376e-01 -4.97420728e-01
-2.56980620e-02 3.80631596e-01 4.13717449e-01 -1.03083789e-01
9.70884442e-01 -1.36431980e+00 -1.27594006e+00 -1.09056858e-02
-6.81225002e-01 -1.74679369e-01 7.88916945e-01 4.37246442e-01
4.77001458e-01 -1.79350212e-01 4.67628390e-01 -3.95356625e-01
4.60499585e-01 -7.87248075e-01 -7.00261071e-02 -2.40902632e-01
-1.25513637e+00 5.28260209e-02 5.38759291e-01 -2.10557599e-02
-5.31982660e-01 4.71987687e-02 -6.16252482e-01 5.24343140e-02
-4.43241298e-01 2.79424697e-01 8.30357000e-02 2.06786841e-01
7.97097325e-01 2.36835092e-01 3.87837082e-01 -8.05372596e-01
-4.84033264e-02 7.10684836e-01 3.18188190e-01 -6.40460625e-02
2.47555912e-01 7.08953679e-01 2.64063001e-01 -1.10530818e+00
-1.02662241e+00 -6.94698155e-01 -8.04404378e-01 -1.19167589e-01
1.17379975e+00 -8.67181301e-01 -9.54440534e-01 2.12768748e-01
-3.83551627e-01 -7.92303324e-01 -4.00737435e-01 4.89391446e-01
-4.96033311e-01 5.78809381e-02 -2.17076480e-01 -9.17898118e-01
-7.43329644e-01 -5.76152682e-01 9.87405121e-01 3.19341898e-01
-6.70704067e-01 -1.22941220e+00 6.97443008e-01 7.72493720e-01
7.48825908e-01 5.68392038e-01 2.97671854e-01 -6.80901825e-01
-9.36899930e-02 -1.96216136e-01 1.97604164e-01 5.67167997e-01
4.56267148e-01 -5.04191399e-01 -1.10631609e+00 -1.52787179e-01
2.69985169e-01 -4.02856290e-01 5.74728072e-01 3.45446616e-01
8.99506390e-01 -3.95107985e-01 -7.59771019e-02 6.65273607e-01
1.35031915e+00 -9.21690986e-02 9.87278461e-01 6.45539939e-01
4.08772290e-01 3.77639867e-02 6.41023591e-02 5.65039694e-01
6.46283865e-01 7.07536459e-01 3.91582549e-01 -4.13136631e-01
-1.45540223e-01 9.66286957e-02 3.81337821e-01 5.20538151e-01
-4.54026729e-01 3.61265451e-01 -7.86318123e-01 3.47424567e-01
-1.72110450e+00 -9.32115316e-01 -4.90242913e-02 2.16168332e+00
5.06503284e-01 1.28326461e-01 8.23084414e-01 4.66218650e-01
-3.20580542e-01 6.61112517e-02 -5.22835970e-01 -5.71003377e-01
2.48047426e-01 8.84612918e-01 4.14779276e-01 8.29466805e-02
-8.62577021e-01 3.75283331e-01 6.39188242e+00 1.42426044e-01
-1.28039491e+00 5.39802551e-01 3.07350904e-01 -8.26300979e-01
4.96268749e-01 -3.74136895e-01 -5.76312363e-01 7.11244941e-01
1.88164163e+00 2.53628403e-01 7.10709274e-01 9.32007730e-01
9.58893895e-01 -5.91177166e-01 -9.94041979e-01 1.01965988e+00
-1.85968485e-02 -1.13904572e+00 -8.84736121e-01 5.21396458e-01
2.71428734e-01 3.82517487e-01 -3.89354855e-01 4.69851136e-01
7.13116601e-02 -1.16719282e+00 4.97429073e-02 9.20691729e-01
4.38375145e-01 -5.48358679e-01 8.45592439e-01 1.94507092e-01
-8.81622612e-01 -3.30354422e-01 -1.69501513e-01 -9.33908761e-01
-2.56869756e-02 8.53437662e-01 -7.97296941e-01 1.33682743e-01
1.02888370e+00 8.21562111e-01 -9.58304703e-01 9.16063130e-01
3.55386036e-03 1.02726495e+00 -5.80109358e-01 -1.37245476e-01
3.41698639e-02 -1.39512613e-01 -2.86081761e-01 1.08836150e+00
3.67992282e-01 -5.29131852e-03 -4.84213941e-02 6.73958123e-01
1.98033601e-01 -1.12417839e-01 -5.41814864e-01 -1.49679808e-02
-1.92893356e-01 1.42588055e+00 -7.18705297e-01 -2.39563540e-01
-6.97251856e-01 8.17955256e-01 2.69199014e-01 -2.33246997e-01
-5.79196274e-01 5.07081635e-02 9.42919731e-01 6.64093614e-01
2.00709403e-01 -5.75366318e-01 -5.83442271e-01 -9.11576867e-01
-1.14525057e-01 -5.71335793e-01 3.29651892e-01 -5.71434319e-01
-1.03201842e+00 1.36039704e-01 -2.60780931e-01 -9.50216234e-01
-1.03929609e-01 -4.60416406e-01 -8.60423207e-01 5.01174629e-01
-1.26973116e+00 -1.01312327e+00 -6.88938558e-01 7.30992436e-01
1.78940967e-01 1.60295233e-01 1.01092291e+00 5.04580259e-01
-7.23451197e-01 2.78293043e-01 8.54339376e-02 -3.34147587e-02
3.25291693e-01 -1.27424586e+00 1.09323196e-01 5.44808447e-01
1.51260674e-01 8.31733584e-01 4.52056974e-01 -3.20104718e-01
-1.46098673e+00 -9.84921932e-01 1.00895572e+00 -7.13195503e-01
4.72520471e-01 -4.10307676e-01 -6.26876056e-01 4.91164356e-01
7.60176703e-02 1.33282185e-01 1.72461581e+00 5.36532462e-01
2.66786307e-01 -7.89958715e-01 -1.12296402e+00 1.90490022e-01
9.41884696e-01 -3.66534978e-01 -6.79014683e-01 3.04733645e-02
1.15860738e-01 7.12350160e-02 -1.15762854e+00 6.26560897e-02
1.07436562e+00 -1.43110788e+00 1.04315937e+00 -5.71250260e-01
4.28358972e-01 -1.15198381e-01 1.35185286e-01 -1.07303607e+00
-3.97024930e-01 -3.79101217e-01 -6.43713236e-01 9.24297810e-01
2.64859535e-02 -4.35823649e-01 1.10691011e+00 9.89067018e-01
-2.76188225e-01 -9.54210520e-01 -1.10327148e+00 -6.83953047e-01
-3.49562466e-01 -6.78952456e-01 6.37581766e-01 6.18016362e-01
1.08149767e-01 5.08371532e-01 -5.11726677e-01 -2.41838902e-01
1.87131435e-01 -1.22414671e-01 9.17709112e-01 -1.45601702e+00
-3.12185407e-01 -1.23822346e-01 -6.89375043e-01 -5.47845542e-01
-2.00451553e-01 -6.71620429e-01 -3.33613217e-01 -2.00282168e+00
-1.05793715e-01 -2.25066066e-01 -7.81623065e-01 9.95414853e-01
1.51532099e-01 7.16534138e-01 -2.43789166e-01 -1.30288556e-01
-7.43002117e-01 5.80333889e-01 7.15598762e-01 6.80851415e-02
-6.85949087e-01 3.04002821e-01 -7.56264448e-01 6.08101189e-01
1.12906969e+00 -5.18736780e-01 -3.43982399e-01 -4.43044491e-02
3.64375293e-01 -1.62852958e-01 5.64948618e-01 -1.71496701e+00
9.14093778e-02 1.84136957e-01 9.02368188e-01 -2.90898442e-01
6.46189272e-01 -1.05884647e+00 2.40047500e-01 8.10776889e-01
4.24374305e-02 5.48771024e-02 2.59496897e-01 5.00887394e-01
3.91468734e-01 2.93344080e-01 5.14007986e-01 -1.37447059e-01
-4.35952395e-01 4.31356207e-02 -7.14075685e-01 -3.08881730e-01
6.29250824e-01 -5.41477859e-01 -7.02638105e-02 -3.30017328e-01
-1.20499349e+00 -4.31519747e-02 2.88564652e-01 2.98489720e-01
3.65885824e-01 -1.29919124e+00 -1.56430423e-01 4.62766409e-01
1.86229572e-02 -4.14056718e-01 3.42351049e-01 1.38990545e+00
-3.27475548e-01 3.44771296e-01 -6.72321260e-01 -5.15226066e-01
-1.25019777e+00 5.22389054e-01 3.47655952e-01 -4.24902588e-01
-7.02166915e-01 2.31763199e-01 -7.07034528e-01 -1.16031520e-01
-5.37859648e-03 -6.39653623e-01 -2.41137952e-01 2.97870878e-02
6.54971063e-01 8.95873249e-01 5.63557625e-01 -1.91533461e-01
-5.85797369e-01 1.31230146e-01 4.53223258e-01 2.47973502e-01
1.84647763e+00 -1.93166271e-01 -6.49837926e-02 7.10265338e-01
1.15272641e+00 -1.75908193e-01 -1.09177077e+00 2.01492652e-01
1.01961359e-01 -1.99906275e-01 3.86967957e-01 -8.73014927e-01
-9.22544062e-01 9.80199993e-01 1.31526816e+00 3.81838918e-01
1.24853420e+00 -1.52985632e-01 1.09764636e+00 5.40174544e-01
1.09871633e-01 -1.10689664e+00 1.20454960e-01 -1.45220431e-02
4.34926927e-01 -1.26781607e+00 4.45431679e-01 4.13405567e-01
-9.93623808e-02 1.08017373e+00 4.24344420e-01 -2.20417589e-01
9.62009847e-01 1.09621316e-01 8.30977038e-02 -3.83254409e-01
-4.95960921e-01 -4.75459278e-01 1.48927853e-01 7.29398131e-01
5.30802310e-01 1.11615263e-01 -5.25110245e-01 8.42330396e-01
-5.63777208e-01 8.46219540e-01 2.92409509e-01 9.72736120e-01
-5.69257319e-01 -1.25470269e+00 -1.44329265e-01 1.01719916e+00
-7.31642544e-01 -1.38852792e-02 -1.48436636e-01 6.38073862e-01
6.16393507e-01 1.20591462e+00 1.62922591e-01 -5.63504636e-01
3.16597134e-01 4.82381105e-01 2.50905931e-01 -6.22979581e-01
-7.88974822e-01 -3.59096587e-01 2.70856082e-01 -8.72652471e-01
-7.29803741e-01 -6.42759562e-01 -8.52803528e-01 -3.86416078e-01
2.14943752e-01 -2.14081883e-01 8.63896787e-01 1.30130661e+00
6.60397708e-01 5.85529566e-01 3.21293622e-01 -1.02405083e+00
1.12980329e-01 -9.65417206e-01 -5.78906894e-01 1.23932049e-01
6.03869557e-01 -7.66240776e-01 -7.96523318e-02 -6.87101530e-03] | [13.60063648223877, 3.384312391281128] |
5bcc73a7-b88f-4416-8a79-19c7c2cb5519 | xastnn-improved-code-representations-for | 2303.07104 | null | https://arxiv.org/abs/2303.07104v1 | https://arxiv.org/pdf/2303.07104v1.pdf | xASTNN: Improved Code Representations for Industrial Practice | The application of deep learning techniques in software engineering becomes increasingly popular. One key problem is developing high-quality and easy-to-use source code representations for code-related tasks. The research community has acquired impressive results in recent years. However, due to the deployment difficulties and performance bottlenecks, seldom these approaches are applied to the industry. In this paper, we present xASTNN, an eXtreme Abstract Syntax Tree (AST)-based Neural Network for source code representation, aiming to push this technique to industrial practice. The proposed xASTNN has three advantages. First, xASTNN is completely based on widely-used ASTs and does not require complicated data pre-processing, making it applicable to various programming languages and practical scenarios. Second, three closely-related designs are proposed to guarantee the effectiveness of xASTNN, including statement subtree sequence for code naturalness, gated recursive unit for syntactical information, and gated recurrent unit for sequential information. Third, a dynamic batching algorithm is introduced to significantly reduce the time complexity of xASTNN. Two code comprehension downstream tasks, code classification and code clone detection, are adopted for evaluation. The results demonstrate that our xASTNN can improve the state-of-the-art while being faster than the baselines. | ['Hongyu Zhang', 'Xi Cheng', 'Yang Chen', 'Xibin Zhao', 'Min Zhou', 'Zhiwei Xu'] | 2023-03-13 | null | null | null | null | ['code-classification'] | ['computer-code'] | [ 1.58785388e-01 -3.42370361e-01 -4.08280015e-01 -3.72368366e-01
-5.02697825e-01 -5.86908385e-02 -1.39407665e-02 1.04110986e-01
-3.29274349e-02 -6.31906316e-02 1.15439156e-02 -8.47093582e-01
2.16894269e-01 -6.08317196e-01 -4.73231405e-01 -3.66928875e-01
-1.34182237e-02 -3.48236442e-01 3.20759326e-01 -2.12726906e-01
4.74279374e-01 -1.92720164e-02 -1.58191204e+00 5.77950656e-01
1.19473577e+00 8.33025396e-01 7.15210974e-01 4.36017632e-01
-7.26218343e-01 1.27214146e+00 -6.63360536e-01 -3.65245700e-01
-1.10740937e-01 -5.70222497e-01 -6.83106661e-01 -3.32364857e-01
1.17457055e-01 -1.76902518e-01 -1.12555534e-01 1.16987276e+00
3.23678613e-01 -2.81003982e-01 2.04563916e-01 -1.09841597e+00
-8.26918542e-01 1.10434628e+00 -9.01757479e-01 9.30931121e-02
2.13161066e-01 -1.01609811e-01 1.15011775e+00 -7.43047833e-01
2.52917767e-01 1.06333375e+00 8.21724415e-01 4.36450899e-01
-8.86815548e-01 -8.16446006e-01 2.75785357e-01 2.17917293e-01
-1.10131621e+00 -2.15281427e-01 8.58002186e-01 -6.50923193e-01
1.22770500e+00 8.31761733e-02 3.31041574e-01 1.18616188e+00
5.89818776e-01 1.13800848e+00 5.03741205e-01 -5.29048681e-01
2.66584396e-01 -1.64244562e-01 5.69813430e-01 9.93508279e-01
2.52359718e-01 -2.10842893e-01 -1.81682214e-01 -3.11270267e-01
6.01938963e-01 4.08412755e-01 -2.62387395e-01 -3.92518103e-01
-1.09982026e+00 1.01977432e+00 6.59417629e-01 6.53086722e-01
-1.44474491e-01 4.60386753e-01 1.03428900e+00 4.55151320e-01
1.98075205e-01 3.70354027e-01 -5.76225281e-01 -5.94872832e-01
-9.73774493e-01 8.02798271e-02 5.29985428e-01 1.32786262e+00
4.91637468e-01 5.14752388e-01 -7.10023493e-02 1.11079895e+00
4.67626899e-01 1.82382122e-01 1.09781516e+00 -4.80550259e-01
7.51894951e-01 1.23045290e+00 -6.29923582e-01 -1.04770601e+00
-2.86823183e-01 -6.81902707e-01 -9.10691321e-01 1.51686072e-01
-6.19007275e-02 -3.14409956e-02 -8.08998883e-01 1.40055192e+00
-2.47536495e-01 3.45345698e-02 -1.12860858e-01 5.28780758e-01
8.41724396e-01 8.63426149e-01 -9.84735191e-02 4.21181098e-02
1.26829863e+00 -1.30628586e+00 -5.54127872e-01 -3.92116666e-01
1.06515396e+00 -6.13774657e-01 1.34505689e+00 4.37241107e-01
-8.13861251e-01 -7.83726335e-01 -1.20825911e+00 -1.26367241e-01
-9.21256393e-02 5.15957117e-01 9.75873888e-01 7.32790709e-01
-9.79788482e-01 3.77768785e-01 -1.09921110e+00 -2.81472534e-01
4.30516332e-01 1.94293246e-01 1.19096294e-01 1.10475220e-01
-7.42190838e-01 3.67298305e-01 4.33511555e-01 2.00276926e-01
-7.35394299e-01 -5.80666006e-01 -1.06478870e+00 4.44965005e-01
3.08585376e-01 -3.51679116e-01 1.63992274e+00 -1.18944454e+00
-1.60055864e+00 4.77591932e-01 -5.78195825e-02 -3.08625638e-01
6.59341216e-02 -5.11782408e-01 -4.32725459e-01 -4.18740928e-01
7.77602345e-02 2.71851774e-02 8.01819146e-01 -8.26419115e-01
-4.53759879e-01 -1.47525877e-01 -6.40717298e-02 -3.43699723e-01
-6.77589715e-01 3.40749234e-01 -5.32621026e-01 -9.26005661e-01
2.88198702e-02 -7.74306178e-01 -2.96927363e-01 -1.52774826e-01
-3.75082791e-01 -4.07316506e-01 8.36733341e-01 -6.44627571e-01
1.66455007e+00 -2.49435091e+00 1.90316051e-01 -9.38280001e-02
3.38485599e-01 4.84488517e-01 -2.90558934e-01 5.31368434e-01
-2.32656568e-01 -4.11872938e-02 -4.67965811e-01 4.67650779e-02
5.33148460e-02 -9.94341075e-02 -6.45116806e-01 1.81444302e-01
2.11040363e-01 9.70893919e-01 -7.02691317e-01 -2.46558920e-01
-1.10925160e-01 1.40167162e-01 -8.42389047e-01 4.52287197e-01
-4.08884734e-01 -9.64376405e-02 -7.75728524e-01 6.79912031e-01
4.89629716e-01 -4.03012842e-01 -6.15185574e-02 2.78617412e-01
-3.13558787e-01 4.53473449e-01 -7.87202120e-01 2.16987991e+00
-9.41547453e-01 6.76488519e-01 -1.46779627e-01 -9.95506525e-01
1.20961452e+00 5.89704439e-02 -3.94163243e-02 -6.69485390e-01
3.51874024e-01 4.17471349e-01 2.13893011e-01 -6.95451140e-01
3.15161139e-01 3.59271675e-01 -3.15076947e-01 6.12835109e-01
-4.82227467e-02 1.12281673e-01 5.32294028e-02 7.95951188e-02
1.30827773e+00 3.41011733e-01 4.19994891e-01 -3.65562767e-01
6.42365098e-01 -2.49632090e-01 5.93411803e-01 5.87195814e-01
1.02453388e-01 3.55283320e-01 7.77679324e-01 -5.33510327e-01
-7.39865184e-01 -6.43579721e-01 9.45041478e-02 1.45831645e+00
-4.70260009e-02 -7.91278899e-01 -9.86693084e-01 -8.87122989e-01
-1.61291122e-01 7.52848983e-01 -5.15556991e-01 -3.32577348e-01
-7.86922574e-01 -5.02133906e-01 7.80354857e-01 9.17940557e-01
7.35583365e-01 -1.24334598e+00 -1.02317417e+00 3.74709219e-01
1.07926801e-01 -5.03898382e-01 -5.51362395e-01 2.13054702e-01
-1.00885546e+00 -9.52598512e-01 -6.72698736e-01 -1.08363080e+00
4.55835462e-01 2.69739151e-01 1.11749518e+00 6.10972762e-01
-1.89932078e-01 -2.72293270e-01 -6.02608919e-01 -2.64052153e-01
-6.27662778e-01 5.32057583e-01 -5.35269916e-01 -2.99028933e-01
4.99466956e-01 -5.63910782e-01 -2.74724245e-01 4.92609143e-02
-9.06468987e-01 1.57236010e-01 1.05158818e+00 9.31804240e-01
-4.44357805e-02 -6.21153675e-02 6.01168275e-01 -1.18323970e+00
6.84203684e-01 -4.11091506e-01 -6.86018169e-01 2.31671378e-01
-6.77932441e-01 2.12925121e-01 9.08455431e-01 -2.72196800e-01
-1.21740949e+00 -1.82728335e-01 -5.02775431e-01 2.18486413e-03
-9.69890971e-03 1.09672475e+00 -1.21924594e-01 8.91692042e-02
6.73084736e-01 3.91225368e-01 -1.72570482e-01 -7.08280921e-01
1.45604476e-01 1.07529926e+00 3.30161065e-01 -7.05289483e-01
5.22102356e-01 -1.07168421e-01 -5.82145691e-01 -6.22833848e-01
-5.45790732e-01 -3.18227202e-01 -4.42502528e-01 2.50446141e-01
5.83970845e-01 -8.23999107e-01 -4.38009083e-01 7.80202687e-01
-1.36284721e+00 -2.66589999e-01 2.61553288e-01 2.96945125e-01
-3.49805713e-01 5.61006844e-01 -9.19881403e-01 -4.98536229e-01
-5.38299322e-01 -1.53009832e+00 1.00771296e+00 5.49029075e-02
-1.09507591e-01 -7.32292950e-01 -3.10300197e-03 -5.50890565e-02
7.31241763e-01 8.60366672e-02 1.39167118e+00 -5.74191809e-01
-6.40062451e-01 -4.02425341e-02 -3.31830382e-01 4.69035298e-01
1.42049193e-01 2.21408635e-01 -8.68237793e-01 -4.28004235e-01
2.78202742e-01 -3.69228601e-01 9.36749995e-01 1.26622081e-01
1.61303723e+00 -2.24812776e-01 -4.59112018e-01 6.97612226e-01
1.35387599e+00 4.50268477e-01 5.18196762e-01 4.31642056e-01
8.53886962e-01 3.33803117e-01 3.05947959e-01 3.60685050e-01
2.48582751e-01 5.52690804e-01 4.52845603e-01 5.06982170e-02
7.05630779e-02 -1.60281792e-01 6.76078796e-01 1.59869909e+00
3.72236818e-01 -7.56696537e-02 -1.28353846e+00 4.86581326e-01
-1.82504880e+00 -5.59469640e-01 -3.50760281e-01 1.93868077e+00
8.34364772e-01 3.13781142e-01 -1.45920783e-01 1.09002113e-01
6.51850402e-01 1.72152504e-01 -5.94344318e-01 -5.13721645e-01
4.29615796e-01 2.27702841e-01 -1.05919287e-01 -1.09641336e-01
-9.89589095e-01 7.21220732e-01 5.35601616e+00 8.32059920e-01
-1.35423195e+00 1.53486997e-01 3.91025484e-01 4.09207165e-01
-2.21340731e-01 -6.61224946e-02 -5.56717813e-01 5.39654553e-01
1.03903162e+00 -2.36070156e-01 1.55140430e-01 1.55178559e+00
-2.83617198e-01 3.19889992e-01 -1.32283294e+00 9.11631584e-01
6.61931857e-02 -1.29893339e+00 -7.40380287e-02 -3.31937909e-01
5.32014489e-01 1.81714624e-01 1.79221369e-02 7.79518425e-01
2.45451719e-01 -9.16095257e-01 7.84157932e-01 -1.05893590e-01
8.32790852e-01 -8.36029887e-01 9.55077767e-01 2.71267265e-01
-1.39496171e+00 -4.62743849e-01 -4.77355093e-01 -1.61983624e-01
-2.36801982e-01 6.77753270e-01 -4.54261690e-01 6.26439154e-01
7.31908321e-01 1.06220794e+00 -8.96996081e-01 1.00032365e+00
-2.66048223e-01 8.15809488e-01 2.38391414e-01 -5.02424300e-01
4.03413743e-01 1.59900442e-01 -2.95673888e-02 1.57326245e+00
4.67956513e-01 -4.47512478e-01 1.43802196e-01 1.29609871e+00
-1.29757091e-01 1.18029624e-01 -5.54877043e-01 -1.77261949e-01
3.26583415e-01 1.04668868e+00 -6.72947705e-01 -2.06965253e-01
-8.14973772e-01 8.46842766e-01 4.47790891e-01 2.97331572e-01
-9.53596175e-01 -1.10019958e+00 4.69862938e-01 -1.27726421e-01
4.80085790e-01 -1.44606143e-01 -3.89996588e-01 -1.33711243e+00
2.97602117e-01 -1.23727095e+00 2.18385160e-01 -5.69245100e-01
-9.66194093e-01 1.09667003e+00 -2.17682600e-01 -1.17761993e+00
-2.51661360e-01 -8.04205418e-01 -9.04294908e-01 6.48432255e-01
-1.46755970e+00 -9.88960683e-01 -3.86716902e-01 1.12935834e-01
9.07295585e-01 -5.29744327e-01 7.83277631e-01 4.23655361e-01
-8.56448948e-01 7.97410309e-01 1.85379833e-01 4.13669348e-01
3.24244261e-01 -1.21023691e+00 1.00449860e+00 1.11107206e+00
-3.48250344e-02 1.25835729e+00 3.30381870e-01 -4.32325125e-01
-1.68661225e+00 -1.31061840e+00 6.11809552e-01 -1.82020798e-01
7.55076349e-01 -7.24424720e-01 -1.19805706e+00 7.48313546e-01
3.37346911e-01 -1.20630018e-01 5.88073611e-01 2.56111652e-01
-7.50698268e-01 -1.36934549e-01 -5.81295788e-01 5.85484505e-01
9.04582679e-01 -5.35391629e-01 -7.27777183e-01 3.75373922e-02
8.80857587e-01 -5.44442594e-01 -5.38009226e-01 3.25648755e-01
3.19094300e-01 -1.12790751e+00 4.94110256e-01 -4.03750330e-01
8.65612864e-01 -2.01465681e-01 -6.89683156e-03 -1.11268699e+00
-3.27042699e-01 -6.19427681e-01 -1.04233779e-01 1.33618307e+00
3.86326283e-01 -4.84510422e-01 6.32518947e-01 1.00816011e-01
-5.95620871e-01 -7.92998791e-01 -7.29997337e-01 -7.64478445e-01
1.74288169e-01 -4.28690225e-01 7.25874841e-01 9.13924634e-01
2.23427728e-01 5.28433621e-01 -3.72811407e-02 -1.31481186e-01
5.86691611e-02 6.43531501e-01 7.75448322e-01 -1.26509809e+00
-5.84045768e-01 -6.92344785e-01 -4.24346685e-01 -1.42215836e+00
3.31289560e-01 -9.97961164e-01 1.38010755e-01 -1.25750291e+00
3.55517626e-01 -3.65478277e-01 -2.78992742e-01 7.06877887e-01
-2.89688617e-01 -3.89635503e-01 1.03667945e-01 1.66705206e-01
-4.22547340e-01 7.09804654e-01 7.30103493e-01 -3.07718724e-01
-1.32738456e-01 7.15008155e-02 -7.20262468e-01 7.40278602e-01
6.34444296e-01 -5.04814923e-01 -6.02219760e-01 -8.43530893e-01
3.28506887e-01 4.01850045e-02 -5.89230917e-02 -1.00884783e+00
1.40024185e-01 1.97072953e-01 -1.49828687e-01 -4.86888319e-01
-3.58742952e-01 -6.49062097e-01 -1.57907829e-01 8.10584307e-01
-4.88501698e-01 5.52702308e-01 3.39264393e-01 4.20066148e-01
-2.85002887e-01 -7.58648932e-01 6.77066624e-01 -3.44981849e-02
-8.89437675e-01 5.86528108e-02 -4.37849551e-01 7.82896578e-02
7.96980977e-01 -1.76370088e-02 -2.29800299e-01 4.30822372e-02
2.91632682e-01 -8.64237174e-02 4.24586475e-01 8.64216447e-01
7.18402207e-01 -1.15504098e+00 -7.01491952e-01 5.55324912e-01
5.04449010e-01 -7.30975764e-03 4.90752235e-02 5.23017585e-01
-7.66425669e-01 4.48426932e-01 -1.52868718e-01 -6.84521139e-01
-9.80071127e-01 7.68188000e-01 5.41216247e-02 -1.49639070e-01
-9.28184927e-01 9.16868567e-01 5.08138657e-01 -6.60322011e-01
3.42173964e-01 -9.14762259e-01 -7.83552602e-02 -5.19993961e-01
7.63371289e-01 1.45949528e-01 2.44034752e-01 -1.54436277e-02
-2.78171867e-01 4.60391104e-01 -5.46093404e-01 5.88557363e-01
1.47970760e+00 2.57298201e-01 -5.20259976e-01 6.24833167e-01
1.22197235e+00 -4.64385673e-02 -9.29868817e-01 -1.88794851e-01
5.36600769e-01 -3.00020278e-01 1.21427506e-01 -6.58868372e-01
-1.27782869e+00 1.20702767e+00 4.06392634e-01 2.87181050e-01
1.09039056e+00 -2.52415091e-01 9.32441235e-01 7.03529716e-01
3.90199482e-01 -6.50146127e-01 1.25400096e-01 6.75381303e-01
8.50026548e-01 -8.16236436e-01 -2.57584274e-01 -2.91039616e-01
-3.47101182e-01 1.48713624e+00 9.79840875e-01 -1.90752465e-02
4.10512060e-01 6.31731987e-01 1.70226619e-01 -2.18221292e-01
-8.14438522e-01 4.61193651e-01 1.33336661e-03 3.42099309e-01
1.04312027e+00 -4.26093191e-01 -2.27536917e-01 6.90650582e-01
-4.95862961e-02 -3.18974815e-02 7.19736159e-01 1.10704637e+00
-4.17852044e-01 -1.19508982e+00 -1.54672740e-02 6.82150066e-01
-4.53850418e-01 -3.50822598e-01 -1.37631699e-01 7.10732162e-01
-2.03316659e-01 6.30820453e-01 -2.30649590e-01 -4.72842038e-01
2.43293941e-01 -1.31324664e-01 3.98111418e-02 -9.76586521e-01
-8.45869005e-01 -1.35484070e-01 -3.56687754e-01 -5.31627655e-01
-1.36055708e-01 -4.56249356e-01 -1.28010046e+00 -1.34564251e-01
-6.76961899e-01 2.92748123e-01 5.39707661e-01 6.23421073e-01
7.14941978e-01 9.45710838e-01 5.31677008e-01 -4.84675884e-01
-7.96527445e-01 -1.02418196e+00 -2.64664769e-01 2.92203575e-02
4.86754596e-01 -4.90452081e-01 -5.02832606e-02 7.90623669e-03] | [7.547199726104736, 7.919604778289795] |
e65959a6-649e-44b9-8e13-d6e85149c7f0 | regularity-learning-via-explicit-distribution | 2112.03649 | null | https://arxiv.org/abs/2112.03649v2 | https://arxiv.org/pdf/2112.03649v2.pdf | Regularity Learning via Explicit Distribution Modeling for Skeletal Video Anomaly Detection | Anomaly detection in surveillance videos is challenging and important for ensuring public security. Different from pixel-based anomaly detection methods, pose-based methods utilize highly-structured skeleton data, which decreases the computational burden and also avoids the negative impact of background noise. However, unlike pixel-based methods, which could directly exploit explicit motion features such as optical flow, pose-based methods suffer from the lack of alternative dynamic representation. In this paper, a novel Motion Embedder (ME) is proposed to provide a pose motion representation from the probability perspective. Furthermore, a novel task-specific Spatial-Temporal Transformer (STT) is deployed for self-supervised pose sequence reconstruction. These two modules are then integrated into a unified framework for pose regularity learning, which is referred to as Motion Prior Regularity Learner (MoPRL). MoPRL achieves the state-of-the-art performance by an average improvement of 4.7% AUC on several challenging datasets. Extensive experiments validate the versatility of each proposed module. | ['Wei Wu', 'Cewu Lu', 'Weihao Gan', 'Dongliang Wang', 'Haisheng Su', 'Andong Deng', 'Haoshu Fang', 'Zhongyin Zhao', 'Shoubin Yu'] | 2021-12-07 | null | null | null | null | ['anomaly-detection-in-surveillance-videos', 'anomaly-detection-in-surveillance-videos'] | ['computer-vision', 'methodology'] | [ 3.39673549e-01 -1.87416777e-01 -1.71894088e-01 -1.02129221e-01
-6.80427730e-01 -2.63485044e-01 6.75246596e-01 8.23212266e-02
-3.87865573e-01 3.59127790e-01 1.40615597e-01 -1.91953987e-01
7.87844136e-02 -6.08038783e-01 -7.35983849e-01 -8.97457838e-01
6.34732246e-02 -2.34027505e-01 6.64484024e-01 1.74373239e-01
5.99840609e-03 2.45226666e-01 -1.35684073e+00 -9.63861775e-03
9.65720177e-01 1.15455592e+00 -1.15009949e-01 4.87931281e-01
1.28850743e-01 7.39231884e-01 -2.53908962e-01 -3.35687339e-01
3.46685290e-01 -5.00281572e-01 -3.10689121e-01 3.00338417e-01
8.11466336e-01 -5.81283212e-01 -5.83873212e-01 9.82122123e-01
3.04812014e-01 1.81176320e-01 4.18374121e-01 -1.11082089e+00
9.15988013e-02 -7.83248469e-02 -9.98962879e-01 3.80796075e-01
6.52645528e-01 3.08940113e-01 7.88784087e-01 -7.15412021e-01
5.93134224e-01 1.15572405e+00 7.67683685e-01 5.56144834e-01
-1.03797531e+00 -3.67169112e-01 5.51201522e-01 4.05366153e-01
-1.27161336e+00 -3.22794676e-01 1.03590560e+00 -3.09312880e-01
5.85013449e-01 1.80669829e-01 9.38671708e-01 1.34610271e+00
1.87740192e-01 1.10261011e+00 7.37183094e-01 -3.31437774e-02
7.97274560e-02 -4.90660638e-01 -2.17287242e-01 1.11612129e+00
6.00861371e-01 5.35169803e-02 -5.91943204e-01 -2.83783585e-01
6.20003998e-01 3.19964141e-01 -4.88784045e-01 -7.82725692e-01
-1.27578783e+00 5.74166119e-01 1.71732292e-01 -1.18667282e-01
-4.14204180e-01 2.39967778e-01 3.94185305e-01 -7.79543146e-02
3.49162281e-01 -1.44320717e-02 -2.35560954e-01 -2.17900842e-01
-1.03311205e+00 1.15109921e-01 5.02994478e-01 6.23590350e-01
5.88631392e-01 3.05415154e-01 -3.50393653e-01 2.95764416e-01
6.62700772e-01 4.77165639e-01 3.36903244e-01 -8.54658306e-01
3.76757383e-01 7.03868866e-01 1.16282292e-01 -1.45698464e+00
-4.45812672e-01 -5.59387743e-01 -7.93683350e-01 -5.57029694e-02
4.44837898e-01 -3.73704359e-02 -8.44472170e-01 1.80098915e+00
6.91093385e-01 8.55566442e-01 -1.20640211e-01 9.41876352e-01
6.16344750e-01 5.15498042e-01 2.54866779e-01 -3.81668329e-01
1.13730609e+00 -1.08907521e+00 -6.73566222e-01 -2.91714162e-01
5.20370245e-01 -5.65385878e-01 6.48990452e-01 3.92502367e-01
-8.03096294e-01 -3.41526210e-01 -1.03588402e+00 3.67748201e-01
1.38907865e-01 -5.60427606e-02 4.79044408e-01 6.93121910e-01
-6.67027116e-01 2.67461658e-01 -1.47323501e+00 -4.31331903e-01
6.57758236e-01 2.29630750e-02 -3.77613902e-01 -2.40684263e-02
-7.96310365e-01 6.44582987e-01 1.59454554e-01 2.95041442e-01
-1.22107685e+00 -5.85642576e-01 -1.24240327e+00 -3.19333315e-01
5.71110070e-01 -8.81189108e-01 9.07151699e-01 -6.64707720e-01
-1.52274883e+00 5.95718741e-01 -5.32907844e-01 -5.36818981e-01
9.13055360e-01 -6.48598075e-01 -3.07893097e-01 4.97447938e-01
1.55369937e-02 5.70542336e-01 1.16808689e+00 -1.11469257e+00
-7.25430965e-01 -2.02540711e-01 -1.52571782e-01 1.20647840e-01
-5.35298944e-01 -1.85570702e-01 -7.49461591e-01 -9.73794281e-01
3.80697012e-01 -8.05237591e-01 -3.38656723e-01 5.14253914e-01
-2.40580365e-01 2.83486601e-02 1.18193245e+00 -8.09090078e-01
1.48988414e+00 -1.98628247e+00 2.13049561e-01 1.42086953e-01
2.18472108e-01 4.81421262e-01 -4.47656699e-02 1.56175084e-02
1.60845861e-01 -3.56717110e-01 -6.43231928e-01 -3.85746241e-01
-2.18782246e-01 2.38255769e-01 -1.83041263e-02 9.15112793e-01
4.78096873e-01 7.79019356e-01 -1.16539156e+00 -6.41904950e-01
5.42567909e-01 4.40028250e-01 -6.71213150e-01 1.01323262e-01
-3.06187242e-01 8.23103428e-01 -6.54728770e-01 8.58301997e-01
7.31429279e-01 -1.58958063e-01 -3.23991254e-02 -2.16947600e-01
2.27651577e-02 -2.64421515e-02 -1.11230338e+00 2.07933307e+00
-2.50757858e-02 3.93354297e-01 -3.51639539e-02 -9.71133232e-01
7.07064450e-01 3.54720473e-01 8.81652355e-01 -3.46399039e-01
-2.58186515e-02 1.07276805e-01 -1.68518975e-01 -6.00736141e-01
1.03158243e-01 2.04760849e-01 1.60958365e-01 2.38776356e-01
-1.49211600e-01 4.45110589e-01 1.36066198e-01 1.73633790e-03
1.45641160e+00 6.41721129e-01 4.75580662e-01 1.32740382e-02
9.64552939e-01 -1.81206912e-01 1.20548511e+00 4.58715588e-01
-6.33554220e-01 6.93841100e-01 3.98813546e-01 -6.70184493e-01
-5.56364059e-01 -1.20681906e+00 -6.32416755e-02 5.46994567e-01
4.45777118e-01 -5.50693929e-01 -6.58960640e-01 -1.04564834e+00
-2.17814907e-01 2.38914862e-01 -4.29332644e-01 -1.32250085e-01
-8.20617914e-01 -9.30910289e-01 5.38609922e-01 5.05590618e-01
8.62146854e-01 -6.90171838e-01 -9.96547401e-01 1.35329813e-01
-6.87867880e-01 -1.52872920e+00 -5.69897771e-01 -5.47327340e-01
-1.01580691e+00 -1.24450409e+00 -6.76280081e-01 -3.53447437e-01
7.64792323e-01 4.74890262e-01 7.50682354e-01 6.44267723e-02
-4.43952441e-01 7.60118246e-01 -4.84756172e-01 -1.53089002e-01
-4.38225456e-02 -1.27829149e-01 1.44021079e-01 5.24473906e-01
3.44458193e-01 -6.63973927e-01 -1.13571692e+00 3.52993071e-01
-9.26143408e-01 -1.70460671e-01 7.51254022e-01 8.78212869e-01
7.66730070e-01 -1.99671775e-01 2.62283593e-01 -4.83863056e-01
-6.09955341e-02 -3.70880425e-01 -5.73487997e-01 1.52820751e-01
-2.79036164e-01 -1.08175154e-03 4.47509617e-01 -4.09451336e-01
-1.01051986e+00 4.72195327e-01 5.66465259e-02 -6.78006411e-01
-2.34752417e-01 1.67593375e-01 -3.46208245e-01 -4.54503894e-02
4.62720186e-01 5.50426900e-01 1.07408978e-01 -3.03021729e-01
9.52730775e-02 -5.99235184e-02 7.77599454e-01 -4.09105599e-01
1.19017088e+00 9.12704408e-01 2.38407955e-01 -9.45080996e-01
-9.43799853e-01 -6.29530311e-01 -5.80588400e-01 -3.30386221e-01
1.00600815e+00 -8.93518627e-01 -4.46814746e-01 6.42113626e-01
-1.14626276e+00 9.15259346e-02 -8.89895037e-02 5.75820863e-01
-5.16898930e-01 1.05175066e+00 -2.98352838e-01 -9.25711036e-01
-1.97270438e-01 -1.03895557e+00 1.08617496e+00 1.33600608e-01
-1.30996019e-01 -1.05289793e+00 4.54078838e-02 3.35807949e-01
7.78030604e-02 8.91865373e-01 2.53400356e-01 -3.00729573e-01
-8.66306901e-01 -3.50220740e-01 -8.17871466e-02 3.11126530e-01
2.42594108e-01 -1.65862918e-01 -9.53827381e-01 -4.14343238e-01
9.15394798e-02 3.68038565e-02 1.08667350e+00 4.19188559e-01
1.13876903e+00 -2.78253973e-01 -4.31066662e-01 9.33446288e-01
1.12184322e+00 -8.87825340e-02 6.96011543e-01 4.14759547e-01
1.04982996e+00 5.09408116e-01 7.53303587e-01 6.30857885e-01
3.39366168e-01 7.50852525e-01 6.36932850e-01 3.56845371e-02
-4.79892753e-02 -3.44180822e-01 8.13843668e-01 3.94989192e-01
-1.57535136e-01 -5.48817404e-02 -7.64155865e-01 3.80321622e-01
-2.09388208e+00 -1.06236577e+00 -6.48402497e-02 2.12127090e+00
4.32534367e-01 1.44852892e-01 2.94482112e-01 2.30167434e-01
6.03504419e-01 6.38916492e-01 -5.89537740e-01 2.55706310e-01
-4.50981557e-02 -2.24219903e-01 4.04926389e-01 2.02424109e-01
-1.62175262e+00 7.61121690e-01 5.21168232e+00 7.49713778e-01
-8.69908214e-01 3.76961865e-02 3.26207519e-01 1.33087948e-01
-9.36822221e-02 -6.94767162e-02 -6.15601063e-01 5.88790357e-01
5.56151807e-01 2.71905571e-01 -2.87554227e-02 7.17751861e-01
4.27970737e-01 -1.47681087e-01 -8.93778622e-01 9.36746538e-01
2.42556676e-01 -8.60296547e-01 2.22588524e-01 -1.05570599e-01
5.87707996e-01 -3.04687560e-01 1.44114345e-02 -2.24764407e-01
-1.66104525e-01 -6.16719961e-01 5.63838124e-01 6.02116168e-01
4.82424617e-01 -7.60274172e-01 6.36472940e-01 2.10806862e-01
-1.47064090e+00 -1.16926543e-01 -2.44551629e-01 2.26104170e-01
4.42823917e-01 6.21547580e-01 -3.56030554e-01 8.81001592e-01
8.06412637e-01 1.23549712e+00 -5.35790205e-01 1.32628953e+00
-3.42588067e-01 7.94514596e-01 -3.56505901e-01 4.65924501e-01
4.09871459e-01 -1.79648325e-02 1.15534091e+00 1.20421076e+00
3.44861418e-01 -1.29233032e-01 4.86834288e-01 4.75685865e-01
4.36941713e-01 -8.86422172e-02 -6.22103214e-01 2.51918644e-01
2.18133599e-01 1.26279414e+00 -6.78211451e-01 3.61952148e-02
-7.47103095e-01 1.19170403e+00 -1.26894504e-01 2.63176650e-01
-1.02550185e+00 5.52285537e-02 7.30229557e-01 2.00970322e-01
5.94662905e-01 -3.61632407e-01 2.08514154e-01 -1.33223510e+00
4.73718613e-01 -9.30461347e-01 5.77394664e-01 3.18763331e-02
-1.19789398e+00 2.44465083e-01 1.42741293e-01 -2.01062107e+00
-2.15085328e-01 -6.09669328e-01 -8.57784927e-01 2.54489064e-01
-1.63117111e+00 -1.36322439e+00 -7.11837590e-01 6.02254152e-01
5.84832728e-01 -1.45570502e-01 6.71708882e-01 3.44692737e-01
-9.36304271e-01 6.82771504e-01 -3.22762638e-01 2.87090510e-01
5.58608174e-01 -9.47178304e-01 2.46607453e-01 1.55617428e+00
1.18327007e-01 3.57241511e-01 5.35098493e-01 -7.16045082e-01
-1.51217198e+00 -1.45347154e+00 3.06568027e-01 -5.55871427e-01
6.14942014e-01 1.53829670e-03 -7.96316803e-01 4.15863991e-01
-2.00106576e-01 4.90197778e-01 6.54916525e-01 -4.74611998e-01
-4.36160356e-01 -1.36282265e-01 -9.69514012e-01 5.71823299e-01
1.38149714e+00 -1.32316500e-01 -4.42973703e-01 -1.78475231e-02
5.64121783e-01 -4.56048250e-01 -4.51449543e-01 7.45821774e-01
5.48953116e-01 -1.10747123e+00 1.21071756e+00 -3.92454028e-01
2.01542258e-01 -7.73103297e-01 -1.24527410e-01 -8.45934868e-01
-1.08972803e-01 -9.49814081e-01 -9.43675697e-01 9.99630809e-01
6.19038055e-03 -5.92026830e-01 8.93076956e-01 1.08670577e-01
-2.29750052e-01 -8.04089427e-01 -1.01017773e+00 -1.00339329e+00
-5.24100125e-01 -6.28972828e-01 1.23625316e-01 7.41922736e-01
-3.64007950e-01 -6.39998242e-02 -6.38487339e-01 4.90623474e-01
9.62737143e-01 -1.32982954e-01 7.61831999e-01 -1.16728747e+00
-3.71854872e-01 -3.85033220e-01 -9.67187583e-01 -1.35094523e+00
1.19594648e-01 -6.30948961e-01 9.21737626e-02 -1.40391791e+00
1.29796863e-01 1.80630498e-02 -3.88794839e-01 2.12826714e-01
-5.79749167e-01 3.09867084e-01 -4.41113822e-02 8.83311406e-02
-1.07847381e+00 8.28903258e-01 1.20733595e+00 -6.00152612e-02
-2.07237884e-01 9.11212862e-02 -3.98104072e-01 1.23123443e+00
6.34820342e-01 -3.72614324e-01 -5.71112692e-01 -3.83559465e-01
-1.88686728e-01 -5.59150726e-02 7.06349254e-01 -1.41166353e+00
1.64409488e-01 -1.20510526e-01 4.76760536e-01 -7.32609630e-01
3.81629765e-01 -8.86250317e-01 -2.02184707e-01 8.10900033e-01
1.50870308e-01 2.15612590e-01 1.96804211e-01 1.31001318e+00
-2.39888415e-01 1.32287413e-01 7.28208125e-01 1.45344406e-01
-8.21835756e-01 6.37426674e-01 -2.94936329e-01 5.06192334e-02
1.17536676e+00 -3.89209241e-01 -1.16867311e-01 -4.17167664e-01
-4.62735862e-01 2.85955369e-01 3.14017802e-01 3.36604297e-01
8.42531860e-01 -1.35060728e+00 -7.22420275e-01 1.89122841e-01
2.66099393e-01 2.55438805e-01 3.70568633e-01 1.26498699e+00
-6.15200639e-01 2.15748340e-01 -7.89238513e-02 -1.01443672e+00
-1.09950387e+00 4.10602301e-01 2.82620788e-01 -2.94842213e-01
-1.01707685e+00 6.36155605e-01 3.03284705e-01 8.82610530e-02
3.63346428e-01 -1.51874140e-01 -2.46753976e-01 -1.53006658e-01
7.01592386e-01 5.40417433e-01 -2.36845970e-01 -7.65610099e-01
-6.25303388e-01 7.84622729e-01 1.98277421e-02 -7.44926557e-02
1.10461831e+00 -7.33531415e-02 7.19517022e-02 1.04660392e-01
9.28253174e-01 -7.19837025e-02 -1.80201614e+00 -3.37526619e-01
1.56234086e-01 -6.64903522e-01 -2.15392828e-01 -4.12973315e-02
-1.17403638e+00 8.73772919e-01 5.96509874e-01 -2.41915315e-01
1.25800061e+00 -2.52093017e-01 1.12567008e+00 5.31584024e-01
2.94021904e-01 -7.78653622e-01 4.94745046e-01 3.01116198e-01
6.19708717e-01 -1.34060895e+00 1.48688644e-01 -5.79529285e-01
-4.95439589e-01 9.26883519e-01 8.84516239e-01 -8.04097801e-02
5.94609618e-01 -5.35448678e-02 -3.97434831e-02 -5.13890125e-02
-4.93367612e-01 -2.63068974e-01 5.57068825e-01 7.04924941e-01
2.80684024e-01 -3.08538496e-01 -3.06788832e-01 3.42156142e-01
1.88289091e-01 -3.38237882e-01 6.39415011e-02 1.12694812e+00
-3.68846118e-01 -9.83843625e-01 -3.92515063e-01 2.71787643e-01
-4.11702603e-01 3.51491332e-01 -3.99667807e-02 5.97692728e-01
1.32045180e-01 8.95040631e-01 -7.11744651e-02 -4.41073716e-01
1.93969205e-01 8.10763892e-03 4.80898410e-01 -3.13433766e-01
-6.77534640e-02 2.01396212e-01 -1.04053579e-01 -9.69963491e-01
-8.14341426e-01 -9.33234453e-01 -1.08477080e+00 6.97147846e-02
-1.81741342e-02 -2.27051407e-01 8.68416801e-02 8.88454616e-01
2.45576888e-01 4.32917476e-01 6.18532896e-01 -8.07831168e-01
-3.72930527e-01 -6.55868769e-01 7.61890411e-02 4.54046845e-01
5.97094715e-01 -8.79367113e-01 -2.85809517e-01 1.56853721e-01] | [8.188711166381836, 1.1622964143753052] |
a8a7ccf8-d968-4673-8ab6-274f9adef1af | va-learning-as-a-more-efficient-alternative | 2305.18161 | null | https://arxiv.org/abs/2305.18161v1 | https://arxiv.org/pdf/2305.18161v1.pdf | VA-learning as a more efficient alternative to Q-learning | In reinforcement learning, the advantage function is critical for policy improvement, but is often extracted from a learned Q-function. A natural question is: Why not learn the advantage function directly? In this work, we introduce VA-learning, which directly learns advantage function and value function using bootstrapping, without explicit reference to Q-functions. VA-learning learns off-policy and enjoys similar theoretical guarantees as Q-learning. Thanks to the direct learning of advantage function and value function, VA-learning improves the sample efficiency over Q-learning both in tabular implementations and deep RL agents on Atari-57 games. We also identify a close connection between VA-learning and the dueling architecture, which partially explains why a simple architectural change to DQN agents tends to improve performance. | ['Michal Valko', 'Mark Rowland', 'Rémi Munos', 'Yunhao Tang'] | 2023-05-29 | null | null | null | null | ['q-learning'] | ['methodology'] | [-4.60919172e-01 2.63514549e-01 -5.52756727e-01 -1.17236659e-01
-9.64914858e-01 -9.02079940e-01 5.75889826e-01 3.05155776e-02
-7.67767072e-01 1.13860548e+00 1.41298771e-01 -7.96837389e-01
-3.81453037e-01 -8.16048861e-01 -8.32313597e-01 -6.90059006e-01
-4.53749448e-01 5.95867515e-01 2.26179823e-01 -6.14934087e-01
1.94410622e-01 2.15001971e-01 -1.39454627e+00 -4.17135190e-03
9.07491982e-01 9.53168213e-01 -2.23016459e-02 8.51560950e-01
-1.69847682e-01 1.13258350e+00 -8.52632046e-01 -3.48688900e-01
7.95451403e-01 -8.11962843e-01 -7.96758473e-01 -3.62201005e-01
2.70575792e-01 -7.77199805e-01 -3.72525930e-01 8.19405794e-01
6.28548205e-01 2.15785950e-01 4.47905213e-01 -1.33501256e+00
-1.58219948e-01 8.20233226e-01 -4.06843185e-01 3.92990261e-01
7.51833478e-03 7.02381313e-01 1.44916975e+00 -1.80964723e-01
3.93587232e-01 1.50747919e+00 5.20090401e-01 5.67633450e-01
-1.23591018e+00 -5.15600979e-01 1.00566015e-01 3.19280386e-01
-7.69872308e-01 -8.71815756e-02 3.92036736e-01 -1.08332433e-01
1.09634364e+00 -1.33771673e-01 9.95228291e-01 8.64742637e-01
4.68833596e-01 1.06654751e+00 1.40808523e+00 -2.41528854e-01
6.46202326e-01 -1.82008356e-01 -3.84735107e-01 5.65416515e-01
7.90083706e-02 9.68327761e-01 -2.85097152e-01 -2.61170745e-01
1.04618239e+00 -2.70562708e-01 7.98641145e-02 -9.88646090e-01
-8.52977455e-01 1.08639181e+00 4.17768359e-01 -8.22748914e-02
-4.27808404e-01 9.25734222e-01 7.22900033e-01 1.00996232e+00
3.37897427e-02 8.25519800e-01 -9.39304590e-01 -8.21080625e-01
-4.48206395e-01 6.94993913e-01 7.63192117e-01 5.91256797e-01
7.49142051e-01 5.17810822e-01 -4.32001173e-01 4.45882261e-01
3.63567285e-02 8.13243151e-01 4.68731493e-01 -1.61194229e+00
3.05857331e-01 2.77901173e-01 3.85128438e-01 -5.37242591e-02
-5.44088185e-01 -5.71776807e-01 -1.57411277e-01 1.05173099e+00
8.32303107e-01 -5.86326659e-01 -6.34755552e-01 2.00764608e+00
2.52728969e-01 -1.20360449e-01 2.85869718e-01 7.34719396e-01
1.91246364e-02 2.96452224e-01 -8.94743651e-02 -3.74654830e-01
9.17975247e-01 -8.64273190e-01 -4.81237411e-01 -1.17118150e-01
7.29102254e-01 -1.99209854e-01 1.41873145e+00 5.12864053e-01
-1.31471193e+00 -3.17565352e-01 -9.54260111e-01 2.86885828e-01
-6.33856133e-02 -3.34227562e-01 8.39044273e-01 7.14310765e-01
-1.11242652e+00 9.31186914e-01 -7.26561546e-01 -9.07357857e-02
4.62567359e-01 6.33356571e-01 2.04453960e-01 4.02049989e-01
-1.19114220e+00 1.12106752e+00 4.54795122e-01 -6.96845889e-01
-1.26951230e+00 -5.96849859e-01 -4.94983017e-01 1.10767089e-01
8.85045826e-01 -7.49728441e-01 1.95280874e+00 -1.21727443e+00
-2.33386159e+00 1.12413235e-01 4.31500226e-01 -1.07972145e+00
7.65517116e-01 -2.45255232e-01 3.13481800e-02 5.67973927e-02
-2.98188865e-01 6.48804247e-01 1.16724098e+00 -9.94541585e-01
-8.82142842e-01 -1.75825298e-01 6.56835914e-01 3.88278037e-01
2.39327904e-02 -4.64121848e-01 2.66127110e-01 -3.40633601e-01
-7.03906000e-01 -7.86820173e-01 -5.03576994e-01 -7.87675083e-02
3.95351827e-01 -3.75277877e-01 4.13249373e-01 -2.86530823e-01
1.00243330e+00 -1.70579231e+00 -1.42194163e-02 2.27259651e-01
2.62707919e-01 3.28532457e-01 -5.79781771e-01 3.95836443e-01
2.55302638e-02 -2.17890993e-01 -8.78160372e-02 4.94364798e-01
3.48103940e-01 6.44172907e-01 -4.20795172e-01 4.60612565e-01
2.42151637e-02 1.28265762e+00 -1.27621388e+00 -1.02457121e-01
2.79125035e-01 -4.75442819e-02 -1.02417994e+00 2.65729994e-01
-6.87628508e-01 2.22066671e-01 -4.29531544e-01 1.57747239e-01
4.34770703e-01 6.51120245e-02 4.99445826e-01 5.04987121e-01
-1.01120822e-01 4.78239328e-01 -1.05458307e+00 1.65532160e+00
-4.72993284e-01 3.77398938e-01 7.95256197e-02 -1.08808553e+00
7.31455386e-01 7.15048313e-02 6.19856834e-01 -1.37689555e+00
2.62500018e-01 2.32068494e-01 3.96752447e-01 -2.52282351e-01
3.66983533e-01 -1.76287696e-01 -5.97408712e-02 7.07154691e-01
2.03051612e-01 -2.95925647e-01 3.01903814e-01 4.48422395e-02
1.20122886e+00 4.39281046e-01 5.64199865e-01 -5.01055419e-01
2.85855949e-01 1.32853389e-01 5.09119987e-01 1.12478220e+00
-3.10377389e-01 -1.87653497e-01 1.13960242e+00 -5.49453020e-01
-1.11022592e+00 -1.19404840e+00 1.43032730e-01 1.62979841e+00
-1.00168124e-01 -5.53324163e-01 -6.78551137e-01 -1.24724114e+00
4.61868197e-01 7.34361351e-01 -8.11899424e-01 -2.72899181e-01
-7.52455890e-01 -3.97245973e-01 4.17564243e-01 6.00090921e-01
2.96711892e-01 -1.08360958e+00 -1.04219413e+00 4.29361403e-01
4.70195740e-01 -3.12989444e-01 -3.83214891e-01 5.72156727e-01
-1.04818082e+00 -1.15722549e+00 -4.61164743e-01 -1.65035680e-01
3.99615988e-02 8.99737403e-02 1.45656490e+00 -5.38878478e-02
1.91415071e-01 4.75951105e-01 -2.18595833e-01 -4.84762073e-01
-6.49830341e-01 2.56317496e-01 1.29691869e-01 -7.20674455e-01
5.84652275e-02 -6.31933749e-01 -7.11660624e-01 1.92095906e-01
-6.20955467e-01 -5.08516133e-01 6.51519299e-01 1.21550107e+00
3.70335579e-01 -1.75296023e-01 8.01744401e-01 -6.79353237e-01
9.15264964e-01 -2.51849949e-01 -1.27655625e+00 8.96798000e-02
-1.01012766e+00 8.29169631e-01 9.48651373e-01 -3.64013374e-01
-6.34625018e-01 -1.98991135e-01 -3.19300771e-01 -2.83211023e-01
3.57323378e-01 1.04473777e-01 2.22733870e-01 3.55641991e-02
8.43575716e-01 3.16129625e-02 5.45838833e-01 -2.78538197e-01
8.03912401e-01 1.44713730e-01 3.23942035e-01 -8.49221110e-01
6.62344873e-01 1.70323312e-01 -3.56757566e-02 -2.16643468e-01
-5.76125264e-01 3.08489110e-02 -4.73249368e-02 2.55803633e-02
3.29199016e-01 -7.93985486e-01 -1.43555450e+00 6.79603918e-03
-6.93893254e-01 -1.17640388e+00 -1.15299666e+00 4.36781257e-01
-1.20658326e+00 -2.61981227e-03 -5.50678074e-01 -7.80684352e-01
-3.55670989e-01 -1.04663718e+00 5.39879024e-01 3.07015210e-01
2.24074215e-01 -9.42838848e-01 4.74333465e-01 -2.54019350e-01
6.08324230e-01 -2.51304716e-01 8.82207334e-01 -4.72119778e-01
-4.23232347e-01 5.66115797e-01 -3.77414748e-02 3.14983428e-01
5.38982544e-03 -3.15847486e-01 -7.83733606e-01 -7.71138549e-01
-2.04425842e-01 -6.74188495e-01 8.01697373e-01 6.33399129e-01
9.74911690e-01 -5.01007915e-01 3.37535173e-01 6.87380254e-01
1.31476116e+00 4.55634356e-01 5.14053166e-01 7.41255760e-01
1.74697995e-01 1.01638950e-01 6.90507829e-01 9.14551735e-01
3.24485093e-01 6.02675557e-01 7.29181826e-01 2.32948974e-01
9.76740569e-02 -3.99839461e-01 8.36010158e-01 3.19373995e-01
-7.30154589e-02 2.06571922e-01 -6.31157994e-01 1.53210700e-01
-1.96591055e+00 -1.12238431e+00 3.54815364e-01 2.21627522e+00
1.09704149e+00 3.88233066e-01 8.95864308e-01 -2.11040899e-01
1.30863010e-03 1.03316650e-01 -1.19205225e+00 -7.47539043e-01
4.34724502e-02 4.99383122e-01 9.07224357e-01 7.08153665e-01
-6.31264150e-01 1.06577933e+00 7.94233227e+00 1.08064640e+00
-1.07320917e+00 2.01460615e-01 2.89796859e-01 -4.30299968e-01
-4.66165215e-01 3.16487849e-02 -4.66358304e-01 2.49322861e-01
1.10714376e+00 -4.73359734e-01 1.03516030e+00 1.04126048e+00
1.69447407e-01 -4.24024463e-02 -1.04418039e+00 7.24507928e-01
-5.65782487e-01 -1.38524222e+00 -2.00933918e-01 3.23128700e-01
7.96989977e-01 3.53932738e-01 2.18891174e-01 9.98174489e-01
1.17959666e+00 -1.07024086e+00 6.87200606e-01 5.35412244e-02
7.10190177e-01 -1.21213531e+00 6.83049500e-01 3.31838012e-01
-8.77069056e-01 -5.12651920e-01 -5.67151248e-01 -4.19963628e-01
-3.10272157e-01 -7.50875026e-02 -7.88800001e-01 1.78952128e-01
3.95521551e-01 4.44452643e-01 -3.10026348e-01 8.88652861e-01
-5.98813355e-01 7.25227892e-01 -1.80902824e-01 -9.77271795e-02
6.98773861e-01 -3.67817104e-01 3.29871178e-01 8.19896877e-01
-2.00615805e-02 -7.83471018e-02 2.00336859e-01 7.47400165e-01
1.74601730e-02 -1.81411412e-02 -4.91572827e-01 -8.35516304e-02
2.80936480e-01 1.02202702e+00 -3.63509983e-01 -2.41553053e-01
-2.23563239e-01 5.29650629e-01 4.79002088e-01 3.72127563e-01
-8.19596112e-01 -1.74541056e-01 1.08852935e+00 -1.48135414e-02
6.27637506e-01 -3.34442526e-01 5.98303229e-02 -7.60196209e-01
-3.70019436e-01 -1.43006623e+00 4.51051682e-01 -2.39932820e-01
-1.01737690e+00 2.73427963e-01 -9.76866926e-04 -1.23636925e+00
-1.06159556e+00 -6.75889194e-01 -3.68298620e-01 4.87961292e-01
-1.63558745e+00 -3.83397281e-01 2.58525491e-01 7.27630615e-01
3.95749301e-01 -4.72930282e-01 6.04649723e-01 -2.54528731e-01
-2.30008185e-01 8.13573301e-01 6.83937669e-01 -1.61533266e-01
5.92462599e-01 -1.78060198e+00 6.97337866e-01 2.11197600e-01
1.64034292e-01 2.11602718e-01 8.48721087e-01 -1.59126356e-01
-1.67913544e+00 -5.65946281e-01 -2.30876774e-01 -5.59783280e-01
9.35673058e-01 -4.90036868e-02 -5.40910423e-01 4.30859417e-01
5.23924172e-01 6.77472726e-02 2.97541529e-01 9.68398452e-02
-3.87171417e-01 -3.86615694e-01 -1.03262925e+00 5.00969470e-01
9.69096661e-01 -4.58014339e-01 -5.12740731e-01 -1.18499465e-01
7.57079542e-01 -4.91640866e-01 -7.69569278e-01 -7.16542751e-02
5.73654890e-01 -1.19918025e+00 9.87527728e-01 -9.73151624e-01
2.27249339e-01 -7.71908164e-02 -2.11004410e-02 -1.90752554e+00
-1.96761087e-01 -1.03127229e+00 -5.56942463e-01 3.21462721e-01
6.35523051e-02 -8.47204447e-01 9.12114084e-01 1.05930373e-01
1.66761503e-01 -8.69708002e-01 -1.07024240e+00 -1.27107680e+00
8.83230507e-01 -3.75350356e-01 8.29804540e-01 6.02948546e-01
9.17125270e-02 4.12189126e-01 -3.75590831e-01 -4.29675937e-01
7.20691144e-01 1.77128538e-01 9.77443397e-01 -8.33221614e-01
-9.21535850e-01 -8.46106172e-01 6.15440272e-02 -1.33954704e+00
1.51780248e-01 -6.80349410e-01 -9.07604322e-02 -1.17371070e+00
-4.01670709e-02 -4.27620173e-01 -4.91056055e-01 5.31818986e-01
-5.34770899e-02 -2.87426174e-01 4.74782884e-01 -7.40442276e-02
-9.36860859e-01 9.16806579e-01 1.58527815e+00 -5.72525002e-02
-5.18528759e-01 3.38068552e-04 -7.16204882e-01 5.33185124e-01
1.07657862e+00 -5.78268588e-01 -6.80852294e-01 -3.17767054e-01
4.86150682e-01 5.22622168e-01 1.31204814e-01 -7.74906039e-01
-2.01345727e-01 -6.73817039e-01 3.47044766e-02 -1.88603342e-01
-1.75298918e-02 -5.43233633e-01 -5.64619601e-01 8.71095121e-01
-4.86090600e-01 4.52732682e-01 3.75678152e-01 4.45152134e-01
1.24567784e-01 -2.39336744e-01 6.79649234e-01 -2.30080798e-01
-7.95266092e-01 1.85731530e-01 -6.93610251e-01 6.68163300e-01
7.38775432e-01 2.75487956e-02 -4.06301111e-01 -6.49841905e-01
-2.60491133e-01 5.14254153e-01 3.75984013e-01 2.29406849e-01
3.18306357e-01 -1.27194560e+00 -7.13964283e-01 1.27639949e-01
-1.40221968e-01 -3.63501281e-01 -2.17747480e-01 5.87729394e-01
-2.68127620e-01 2.78471917e-01 -6.18825614e-01 -2.88206190e-01
-7.83407271e-01 5.56556642e-01 6.56140208e-01 -6.24931097e-01
-5.63662469e-01 4.54925716e-01 2.86755115e-02 -5.81393719e-01
3.22962821e-01 -5.09401262e-01 1.43840149e-01 -1.49324434e-02
6.39159799e-01 4.73718852e-01 -2.30280757e-01 1.51667178e-01
-1.76395684e-01 1.08409896e-01 4.74506505e-02 -4.67086792e-01
1.19654107e+00 1.32567152e-01 3.16805035e-01 2.02445313e-01
8.08811963e-01 -1.13079056e-01 -2.07301641e+00 -2.16509670e-01
1.79081038e-01 -3.68795395e-01 5.59821203e-02 -9.97997165e-01
-1.16304791e+00 7.48153627e-01 8.24966788e-01 1.47828594e-01
9.20345545e-01 -2.40926325e-01 6.67787492e-01 8.84546459e-01
7.39200056e-01 -1.42786658e+00 4.37197655e-01 8.72217536e-01
6.73792422e-01 -1.28692329e+00 3.46726067e-02 6.82218134e-01
-8.10551584e-01 1.03031278e+00 6.14001215e-01 -3.90536696e-01
2.38097906e-01 4.77339149e-01 2.16919899e-01 5.47775514e-02
-1.23593926e+00 -6.69365525e-01 -2.07556739e-01 6.99909985e-01
1.08119905e-01 1.96742699e-01 -2.64960736e-01 3.01945329e-01
-3.85228157e-01 5.34004234e-02 3.82014304e-01 9.33034182e-01
-6.92098439e-01 -1.58182621e+00 -1.95491299e-01 4.29271430e-01
-3.57724935e-01 1.09332740e-01 -1.28567696e-01 1.02919567e+00
-8.20180923e-02 7.13954687e-01 1.73852146e-01 -2.41942838e-01
2.88645238e-01 -1.47587731e-01 9.71353114e-01 -2.47224674e-01
-1.02981150e+00 -5.06841317e-02 -1.96113884e-02 -9.93137360e-01
-8.29326138e-02 -4.87417966e-01 -1.46999979e+00 -6.38924539e-01
3.24495323e-02 4.77547318e-01 4.34657633e-01 9.20510352e-01
1.57252192e-01 5.66546440e-01 6.08790040e-01 -4.05279398e-01
-1.60197747e+00 -5.59417129e-01 -6.64903820e-01 1.33462876e-01
5.67825973e-01 -7.60976970e-01 -1.24178901e-01 -7.41882682e-01] | [4.069138050079346, 2.0927541255950928] |
0b9d7383-5f0f-4c6a-9269-7ac4145760be | building-low-resource-ner-models-using-non-1 | null | null | https://aclanthology.org/2021.dash-1.11 | https://aclanthology.org/2021.dash-1.11.pdf | Building Low-Resource NER Models Using Non-Speaker Annotations | In low-resource natural language processing (NLP), the key problems are a lack of target language training data, and a lack of native speakers to create it. Cross-lingual methods have had notable success in addressing these concerns, but in certain common circumstances, such as insufficient pre-training corpora or languages far from the source language, their performance suffers. In this work we propose a complementary approach to building low-resource Named Entity Recognition (NER) models using “non-speaker” (NS) annotations, provided by annotators with no prior experience in the target language. We recruit 30 participants in a carefully controlled annotation experiment with Indonesian, Russian, and Hindi. We show that use of NS annotators produces results that are consistently on par or better than cross-lingual methods built on modern contextual representations, and have the potential to outperform with additional effort. We conclude with observations of common annotation patterns and recommended implementation practices, and motivate how NS annotations can be used in addition to prior methods for improved performance. | ['Dan Roth', 'Stephen Mayhew', 'Francesca Marini', 'Tatiana Tsygankova'] | null | null | null | null | naacl-dash-2021-6 | ['low-resource-named-entity-recognition'] | ['natural-language-processing'] | [ 1.76736355e-01 2.11390123e-01 -1.79398671e-01 -7.72594988e-01
-1.46224391e+00 -8.41202080e-01 5.50923645e-01 3.21500301e-01
-9.24842119e-01 8.35299730e-01 7.21647620e-01 -4.54209805e-01
2.89690971e-01 -1.12022012e-01 -2.82155097e-01 -2.27167606e-01
3.99281800e-01 5.24863183e-01 -4.46227342e-02 -2.67556638e-01
4.99885641e-02 3.13016295e-01 -9.74590302e-01 1.74134254e-01
7.90704429e-01 2.13973299e-01 1.65218189e-01 3.07595015e-01
-4.96006638e-01 8.51763666e-01 -6.92885160e-01 -5.43810904e-01
8.08132216e-02 -2.27944344e-01 -1.12651527e+00 -2.22037166e-01
4.97016013e-01 2.68844604e-01 2.28508472e-01 8.19456339e-01
7.14355409e-01 3.32408458e-01 3.33954990e-01 -5.99814057e-01
-6.98378623e-01 9.74481165e-01 -2.47679815e-01 2.63751984e-01
4.66797054e-01 -1.14792123e-01 9.79799569e-01 -8.69331777e-01
8.59623075e-01 1.12237561e+00 1.06796670e+00 8.73493731e-01
-1.42652559e+00 -4.99175072e-01 1.07375145e-01 -4.01563972e-01
-1.47320855e+00 -1.02324247e+00 3.59017164e-01 -2.98487365e-01
1.17341232e+00 4.53941673e-02 -1.67913288e-01 1.12170875e+00
-3.15389097e-01 4.54715908e-01 1.08541703e+00 -1.01644206e+00
1.67073663e-02 6.51134372e-01 1.13157198e-01 3.24982226e-01
9.85854045e-02 -3.33676457e-01 -6.38647497e-01 -2.69711077e-01
3.99045348e-01 -6.04525626e-01 -1.98771074e-01 -1.72633864e-02
-1.20019865e+00 7.61562049e-01 -3.21280919e-02 9.85247850e-01
-3.12921524e-01 -2.49356672e-01 6.59728587e-01 1.48993820e-01
5.47458589e-01 8.56884539e-01 -1.05128181e+00 -3.51301163e-01
-9.60039735e-01 -1.62648275e-01 1.16343379e+00 1.03149915e+00
6.70364976e-01 1.48264542e-01 9.30131823e-02 1.28115666e+00
1.70151934e-01 3.73255670e-01 8.85204852e-01 -7.73604631e-01
6.51285589e-01 3.84495884e-01 3.64227891e-01 -6.47898972e-01
-3.88409317e-01 -8.66051912e-02 -2.59764224e-01 -1.47647932e-01
6.53087199e-01 -4.66908604e-01 -7.05288649e-01 1.89730620e+00
2.07274541e-01 -2.90384352e-01 5.14425516e-01 3.96400154e-01
4.82988089e-01 5.54384530e-01 9.04065132e-01 -2.95027763e-01
1.58675504e+00 -7.09621310e-01 -6.71665490e-01 -6.29323721e-01
1.06114817e+00 -1.15653431e+00 1.26179004e+00 -1.20218284e-02
-9.41636026e-01 -4.82528806e-01 -6.24978900e-01 -2.93476641e-01
-6.01085246e-01 3.24897349e-01 5.42111576e-01 1.14793324e+00
-1.04624724e+00 4.54460651e-01 -7.26645470e-01 -8.52431178e-01
-1.83572620e-01 6.84213713e-02 -6.65984869e-01 -6.74589798e-02
-1.14553154e+00 1.26157081e+00 4.80314046e-01 1.71260148e-01
-3.72713923e-01 -6.86272740e-01 -1.12302542e+00 -2.34623984e-01
3.44147265e-01 9.43288729e-02 1.55172169e+00 -1.00251687e+00
-1.26768029e+00 1.17300105e+00 -3.57526004e-01 -1.68839917e-01
2.58795768e-01 -5.91477275e-01 -6.93194747e-01 -3.31597984e-01
5.22989273e-01 4.61658180e-01 1.64078489e-01 -1.08404946e+00
-6.73292458e-01 -4.12410140e-01 -1.59783438e-01 3.63531917e-01
-3.65465939e-01 9.56899107e-01 -1.71964318e-01 -4.94605213e-01
-2.07814679e-01 -9.53903079e-01 -4.29403961e-01 -6.93511248e-01
-1.38934642e-01 -5.10539174e-01 4.65857714e-01 -1.01883841e+00
1.11154163e+00 -2.14606810e+00 -4.04138058e-01 4.06409577e-02
-3.10250401e-01 5.87922513e-01 -1.09899759e-01 5.90604842e-01
-7.35752210e-02 6.03597224e-01 -9.40033644e-02 -4.47767794e-01
3.53573076e-02 3.79624158e-01 -1.34564728e-01 2.89236546e-01
3.41476232e-01 5.72710752e-01 -9.28531051e-01 -6.58389747e-01
-9.20939967e-02 5.96357644e-01 -3.14314514e-02 -4.34166752e-02
5.90322688e-02 5.66563785e-01 -2.63067216e-01 4.10394669e-01
1.49056852e-01 1.68959171e-01 5.95312178e-01 1.23782143e-01
-6.78541899e-01 9.92801368e-01 -1.30572486e+00 1.69532418e+00
-8.83512795e-01 5.20135045e-01 3.74391526e-01 -7.02200234e-01
8.88325691e-01 7.25615263e-01 -6.12692870e-02 -4.87645358e-01
-4.87625748e-02 5.81282079e-01 -3.09422845e-03 -4.56906736e-01
6.88042939e-01 -3.69507134e-01 -4.01823640e-01 5.91361165e-01
2.30056316e-01 2.41695214e-02 2.75812268e-01 -1.19887576e-01
1.01284432e+00 3.08808774e-01 7.64117181e-01 -4.85831738e-01
5.36100507e-01 3.55562091e-01 8.95492673e-01 7.46087492e-01
-4.42738533e-01 2.35911608e-01 1.97220698e-01 -2.52021521e-01
-1.19919026e+00 -5.99070489e-01 -3.42294842e-01 1.69868588e+00
-5.32624900e-01 -4.74140644e-01 -7.30660021e-01 -6.95010424e-01
-6.39910817e-01 1.05430675e+00 -3.49470466e-01 3.52413923e-01
-1.12332511e+00 -4.73519504e-01 1.12423575e+00 7.05096483e-01
2.08224311e-01 -1.38616180e+00 -3.89117837e-01 5.44141710e-01
-3.50522697e-01 -1.30505204e+00 -3.82205278e-01 3.45554024e-01
-5.35549223e-01 -6.28353119e-01 -5.68771601e-01 -1.19539630e+00
4.55530941e-01 -1.09685346e-01 1.24683523e+00 -7.02267587e-02
-1.70226279e-03 4.22691524e-01 -4.66276377e-01 -4.86590862e-01
-7.02871859e-01 4.27778840e-01 1.60341293e-01 -4.21835244e-01
9.11490083e-01 -2.50394225e-01 9.79072899e-02 3.56919587e-01
-7.58260131e-01 -4.05686378e-01 5.42266190e-01 6.67164743e-01
3.62159252e-01 -2.06841290e-01 7.04383552e-01 -1.37386847e+00
5.32495081e-01 -3.16466868e-01 -2.92223901e-01 4.00462449e-01
-4.07570094e-01 6.87255114e-02 6.38354659e-01 -3.11207980e-01
-1.63149571e+00 3.35273415e-01 -4.41130579e-01 2.48296514e-01
-6.85567558e-01 6.41466975e-01 -3.55608523e-01 8.96650031e-02
1.07138085e+00 -1.83322459e-01 -3.68752003e-01 -9.04260635e-01
3.76522273e-01 7.52467036e-01 4.88450944e-01 -8.79988432e-01
5.97880244e-01 -1.17801242e-01 -6.88021660e-01 -1.03001487e+00
-1.06496644e+00 -6.57395005e-01 -9.43081200e-01 3.06067616e-01
8.64391088e-01 -1.00400901e+00 -6.43351525e-02 -4.44760397e-02
-1.19323659e+00 -5.01586080e-01 -3.26251388e-01 6.20943367e-01
-1.78002253e-01 1.85443491e-01 -6.10785067e-01 -8.47133219e-01
-2.65647620e-01 -7.04896033e-01 8.39104891e-01 2.78386593e-01
-7.26328731e-01 -1.33656013e+00 2.11366430e-01 4.50916499e-01
5.25019526e-01 -4.09325920e-02 8.50419462e-01 -1.40788019e+00
2.30862036e-01 -3.50534886e-01 2.81044282e-02 4.84140694e-01
1.55893937e-01 -1.79473504e-01 -1.05622065e+00 3.74981202e-02
-2.76577156e-02 -5.56047201e-01 8.55236501e-02 -6.23198971e-02
2.92547762e-01 -2.62346417e-01 -2.66216457e-01 -9.53073055e-02
1.40158188e+00 1.34319037e-01 3.78429741e-01 3.08573425e-01
6.42116249e-01 1.00402617e+00 3.86053592e-01 -5.26183937e-03
4.42216694e-01 5.14018416e-01 -6.20045364e-01 3.52260247e-02
-2.22237920e-03 -3.61532360e-01 4.62652743e-01 9.73919213e-01
2.32400913e-02 -1.17391743e-01 -1.39520657e+00 9.04270589e-01
-1.52806485e+00 -8.23173583e-01 -3.61888460e-03 2.23608375e+00
1.17323613e+00 -6.59444332e-02 -3.28427479e-02 -2.96739787e-01
9.70082223e-01 -4.71321531e-02 -1.14546649e-01 -5.75608432e-01
-1.78176433e-01 3.39558810e-01 4.61723179e-01 5.40346980e-01
-1.14845955e+00 1.32135665e+00 6.55785465e+00 6.11688375e-01
-9.54299808e-01 4.73706245e-01 3.69294435e-01 4.00210738e-01
-8.53590444e-02 2.26273566e-01 -1.27804625e+00 8.21096748e-02
1.57748711e+00 -1.59949198e-01 2.00176463e-01 1.02858829e+00
3.69363115e-03 1.09314673e-01 -1.08627558e+00 6.42336786e-01
2.83945203e-01 -7.95084596e-01 -5.24760246e-01 -1.58120394e-01
5.86535990e-01 4.22494382e-01 -5.53047061e-01 6.53223634e-01
7.70165861e-01 -8.81378829e-01 6.79820418e-01 -5.64800762e-02
7.37719774e-01 -5.07150233e-01 9.63593423e-01 4.10375297e-01
-1.18543959e+00 2.57654637e-01 -4.18938279e-01 8.73186737e-02
4.76187259e-01 8.97357091e-02 -9.42348063e-01 4.41344708e-01
5.86729705e-01 1.81157559e-01 -5.55135667e-01 7.47653961e-01
-5.21885574e-01 9.86101270e-01 -5.10835767e-01 -1.14838853e-01
2.97454268e-01 1.92746386e-01 3.12733293e-01 1.67949295e+00
1.27869308e-01 1.65431827e-01 4.64536816e-01 2.51319379e-01
-2.13959798e-01 8.34072232e-01 -7.12002516e-01 -2.39129692e-01
5.84641516e-01 1.38869810e+00 -7.46612668e-01 -3.93870562e-01
-7.37108111e-01 8.18257689e-01 6.58743024e-01 2.54811287e-01
-2.71494985e-01 -4.53667611e-01 4.63599056e-01 2.73004472e-02
5.25567718e-02 -4.12040263e-01 -3.08574677e-01 -1.18220294e+00
4.33213934e-02 -1.04259622e+00 5.91624379e-01 -5.33308864e-01
-1.47331440e+00 8.83521497e-01 -1.18067950e-01 -7.08801150e-01
-4.00737762e-01 -7.61669040e-01 -3.61361295e-01 1.09933817e+00
-1.24939024e+00 -1.23343837e+00 2.30107203e-01 3.35593641e-01
6.09217465e-01 -1.85992736e-02 1.30079889e+00 4.89156067e-01
-5.28259516e-01 5.98514736e-01 -4.78284955e-02 5.76812506e-01
1.09927583e+00 -1.18061495e+00 3.99060041e-01 9.67210114e-01
5.25242150e-01 1.05604386e+00 6.68595850e-01 -6.46228731e-01
-8.75738263e-01 -9.28141534e-01 1.72161400e+00 -7.83628881e-01
7.86550522e-01 -4.07891721e-01 -1.02592099e+00 1.01430142e+00
5.00045776e-01 -1.39486685e-01 1.28177750e+00 8.42389524e-01
-4.42531437e-01 2.56644130e-01 -1.13026786e+00 5.00957072e-01
8.24202836e-01 -8.14117730e-01 -1.01675928e+00 3.37886304e-01
6.36497736e-01 -2.55954802e-01 -9.65968668e-01 1.38954716e-02
4.21838641e-01 -2.91333199e-01 4.85467464e-01 -8.21434438e-01
-1.25660107e-01 -2.63843417e-01 -1.68957815e-01 -1.31786931e+00
-1.42479315e-01 -6.81171775e-01 8.96822870e-01 2.01128221e+00
8.91615212e-01 -5.42297781e-01 4.14993823e-01 1.26426518e+00
-3.00390929e-01 6.83536753e-02 -9.21224177e-01 -8.28784943e-01
2.49808282e-01 -5.78085303e-01 3.72836709e-01 1.35319018e+00
1.40011802e-01 7.74739504e-01 -2.27472916e-01 1.61052316e-01
1.84337988e-01 -5.64997554e-01 5.97796082e-01 -1.38842630e+00
2.77799089e-02 -1.09750398e-01 -2.77943105e-01 -5.14523208e-01
5.94576418e-01 -8.32736731e-01 6.34928644e-01 -1.36200011e+00
-1.60310641e-01 -9.05253768e-01 -3.07948291e-01 1.04406178e+00
-4.80650477e-02 2.62727797e-01 7.57368281e-02 2.59044975e-01
-4.87161309e-01 9.58703458e-02 3.92849475e-01 4.16978061e-01
-4.47389573e-01 -2.44569689e-01 -1.01573730e+00 9.67496157e-01
8.40985894e-01 -8.17311525e-01 -1.66622177e-02 -7.82635748e-01
2.34856367e-01 -2.61543781e-01 -1.56434029e-01 -1.10204327e+00
2.27246180e-01 -8.49315971e-02 3.87023419e-01 2.00920980e-02
3.86175066e-02 -6.94399476e-01 -1.45086711e-02 -8.96513984e-02
-6.00662708e-01 3.01506639e-01 4.10182744e-01 1.74613655e-01
-1.90865815e-01 -7.73980260e-01 6.75808549e-01 -3.62492681e-01
-8.38791192e-01 -2.44742498e-01 -5.23592055e-01 5.51792979e-01
7.01654136e-01 2.80283280e-02 -2.04781860e-01 -8.35897028e-02
-7.52282977e-01 -9.80028808e-02 5.14976263e-01 4.25575763e-01
-1.62754864e-01 -1.09300721e+00 -8.01049232e-01 -1.23097491e-03
1.56958818e-01 -2.93505162e-01 -6.66235313e-02 5.64439714e-01
-4.14886922e-01 5.56690753e-01 -5.50769307e-02 -2.53970399e-02
-1.04180443e+00 2.66303182e-01 6.58841431e-02 -3.42237025e-01
-4.88459796e-01 7.70265162e-01 -1.10363834e-01 -9.05052066e-01
1.25336617e-01 2.04221815e-01 -3.61651927e-01 1.94316074e-01
6.74820960e-01 1.24357276e-01 2.53262162e-01 -1.20973456e+00
-5.73607564e-01 1.19591311e-01 -3.29051375e-01 -6.29723907e-01
1.30773497e+00 -3.31934780e-01 6.16306514e-02 6.95169091e-01
8.54931295e-01 7.87297606e-01 -5.99364042e-01 -5.29742122e-01
6.96699560e-01 -1.74794286e-01 -2.65523970e-01 -8.59029651e-01
-4.09142137e-01 6.61240935e-01 3.02225769e-01 2.37737566e-01
6.48587465e-01 1.83225706e-01 6.35269582e-01 6.89872324e-01
5.00213861e-01 -1.47019494e+00 -5.79278529e-01 7.46606231e-01
5.08678854e-01 -1.30262518e+00 -2.52994210e-01 -3.37888658e-01
-9.40959334e-01 9.64330792e-01 6.15886390e-01 1.67698860e-01
5.49136817e-01 2.26847872e-01 7.09356427e-01 -1.32379219e-01
-4.69811141e-01 -2.65032560e-01 -3.92328948e-02 7.53904343e-01
1.22134101e+00 9.47435275e-02 -4.65423912e-01 6.27300858e-01
-3.72049272e-01 -2.77981907e-01 3.81434232e-01 1.19466209e+00
-2.85732776e-01 -1.54089153e+00 -3.36215675e-01 7.35209063e-02
-1.10103476e+00 -5.56611001e-01 -3.50975931e-01 9.49824870e-01
1.07735105e-01 1.18019783e+00 -2.49425266e-02 2.97569782e-01
4.89062369e-01 8.24003339e-01 3.25884223e-02 -1.15816998e+00
-9.07776773e-01 1.19088463e-01 8.18403542e-01 -2.25898251e-01
-6.04869962e-01 -9.71396983e-01 -1.06325030e+00 1.05558701e-01
-4.11130548e-01 6.99474573e-01 8.69912803e-01 1.09469271e+00
2.57902950e-01 -9.09176618e-02 1.22074455e-01 -6.52433336e-01
-4.53809530e-01 -1.21561229e+00 -4.15433884e-01 3.45074594e-01
-1.86888874e-01 -2.50475436e-01 -2.35028446e-01 4.46739197e-01] | [10.109796524047852, 9.796167373657227] |
75fd769f-a245-414f-b9b7-9f0a469a822c | robust-human-identity-anonymization-using | 2301.04243 | null | https://arxiv.org/abs/2301.04243v1 | https://arxiv.org/pdf/2301.04243v1.pdf | Robust Human Identity Anonymization using Pose Estimation | Many outdoor autonomous mobile platforms require more human identity anonymized data to power their data-driven algorithms. The human identity anonymization should be robust so that less manual intervention is needed, which remains a challenge for current face detection and anonymization systems. In this paper, we propose to use the skeleton generated from the state-of-the-art human pose estimation model to help localize human heads. We develop criteria to evaluate the performance and compare it with the face detection approach. We demonstrate that the proposed algorithm can reduce missed faces and thus better protect the identity information for the pedestrians. We also develop a confidence-based fusion method to further improve the performance. | ['Henrik I. Christensen', 'David Paz', 'Jing-Yan Liao', 'Hengyuan Zhang'] | 2023-01-10 | null | null | null | null | ['face-detection'] | ['computer-vision'] | [ 6.39252886e-02 1.83430091e-01 2.36744717e-01 -7.36837983e-01
-5.09722471e-01 -5.83862066e-01 3.96987498e-01 -3.39535117e-01
-6.90888822e-01 8.64192486e-01 1.06549338e-02 1.18663646e-01
2.97882378e-01 -8.00672054e-01 -6.02954566e-01 -6.83351398e-01
-3.22161615e-02 3.94398063e-01 4.52189386e-01 2.22157198e-03
-9.68976703e-04 8.40030968e-01 -1.88594103e+00 6.61596423e-03
8.22985768e-01 9.25271690e-01 -3.28486770e-01 5.48781455e-01
4.36969250e-01 1.48314714e-01 -6.22381508e-01 -8.64703894e-01
5.82817137e-01 -8.82793311e-03 -4.14743006e-01 -2.32296392e-01
6.74471498e-01 -4.88089770e-01 -1.77567691e-01 1.17418814e+00
9.08806026e-01 1.14523247e-02 5.30200481e-01 -1.63377559e+00
-8.27663988e-02 2.25357682e-01 -9.24018204e-01 -2.18231082e-01
5.83755553e-01 -8.33644345e-02 1.23092905e-01 -9.00334597e-01
5.92432022e-01 1.38619626e+00 9.40313518e-01 7.42751479e-01
-8.29612732e-01 -1.13304627e+00 -5.78023158e-02 3.74368489e-01
-2.02495098e+00 -1.01457047e+00 5.60116947e-01 -2.77266651e-01
2.47261018e-01 1.58942476e-01 5.86877406e-01 1.03631055e+00
-5.68027087e-02 4.94825393e-01 9.15046573e-01 -4.73309517e-01
1.95140436e-01 3.15711111e-01 -1.70084029e-01 1.16087341e+00
1.03617537e+00 -4.23612036e-02 -6.17132068e-01 -3.32259953e-01
3.90040547e-01 -2.48169601e-01 -4.01915014e-02 -5.72375178e-01
-5.31486809e-01 7.45797932e-01 1.78182796e-01 -2.37815067e-01
-2.19477102e-01 -9.81473923e-03 2.74101049e-01 -4.05787885e-01
3.51609200e-01 -1.33760303e-01 1.50999725e-01 2.49181777e-01
-1.19060540e+00 4.85562325e-01 7.68001795e-01 1.10812128e+00
7.73739696e-01 -1.72875568e-01 -2.46748999e-02 4.37963963e-01
6.37927592e-01 8.74630153e-01 2.44765267e-01 -1.02376366e+00
3.03715527e-01 5.45106351e-01 1.82368070e-01 -1.24777257e+00
-3.26558709e-01 1.01297848e-01 -5.64176917e-01 3.49715263e-01
4.67948288e-01 -3.96931499e-01 -7.22418308e-01 1.53691792e+00
7.05979109e-01 2.72124950e-02 -1.77638829e-01 8.24647367e-01
6.59493804e-01 3.75780575e-02 6.27657026e-02 -3.33971269e-02
1.46553934e+00 -5.07558823e-01 -6.24675214e-01 -1.31048083e-01
3.65375698e-01 -7.60330141e-01 5.48089802e-01 9.62477997e-02
-7.16981590e-01 -4.06409979e-01 -1.22284150e+00 6.52965084e-02
-4.18542743e-01 3.57005835e-01 4.03257251e-01 1.66962075e+00
-8.56697738e-01 1.85560331e-01 -7.67609537e-01 -9.37826335e-01
6.81604385e-01 8.50612819e-01 -7.67750680e-01 7.95128122e-02
-1.10184264e+00 7.87959814e-01 3.11487228e-01 4.47282707e-03
-6.44398868e-01 -2.02129737e-01 -9.82486010e-01 -3.59552145e-01
2.18171477e-01 -6.64288759e-01 9.76556718e-01 -4.77233768e-01
-1.45560908e+00 1.20998597e+00 -4.78738695e-01 -7.58999109e-01
8.40672076e-01 -3.35486263e-01 -3.71902585e-01 3.20890754e-01
3.95711094e-01 9.13594902e-01 9.62745786e-01 -1.43502796e+00
-9.23133850e-01 -7.14799047e-01 -3.60832393e-01 1.18198857e-01
-3.25471163e-01 1.28804162e-01 -6.22801244e-01 -3.60690624e-01
-2.37676105e-03 -1.21554661e+00 -1.26427077e-02 2.46688411e-01
-4.45999056e-01 1.82982266e-01 1.26493716e+00 -8.23335350e-01
1.03168130e+00 -1.96774793e+00 -3.36810857e-01 5.46843827e-01
1.56495973e-01 4.37954575e-01 3.33101928e-01 -4.68845516e-02
2.61779487e-01 1.48406014e-01 -1.47268534e-01 -7.33267486e-01
-1.62598029e-01 8.40250328e-02 1.50343791e-01 9.85459507e-01
-1.93272773e-02 5.26762009e-01 -4.05987561e-01 -8.70943010e-01
3.11292142e-01 5.35590470e-01 -3.58954996e-01 9.86037627e-02
4.59272563e-01 2.81475812e-01 -3.33660185e-01 9.26287532e-01
1.17141759e+00 5.76023459e-01 -4.15926799e-03 -2.90261120e-01
-6.51719794e-02 -4.37436432e-01 -1.50766492e+00 1.22451365e+00
-8.62175226e-02 3.36123973e-01 4.90730613e-01 -4.30147499e-01
9.80359077e-01 1.72548324e-01 5.34359157e-01 -1.49315760e-01
4.80115563e-01 4.49588634e-02 -1.37524500e-01 -3.55034739e-01
5.99939108e-01 1.13516741e-01 -5.37806414e-02 1.51853308e-01
-2.87624985e-01 3.76756638e-01 -1.72832489e-01 9.82368961e-02
8.43096793e-01 1.88689053e-01 5.35551488e-01 -2.86886811e-01
9.68759358e-01 -2.32773751e-01 7.90986657e-01 5.23987949e-01
-8.52253020e-01 5.18086433e-01 -7.02907443e-02 -2.03265160e-01
-8.95077348e-01 -9.64860916e-01 -1.80135518e-01 8.51706624e-01
3.37894887e-01 -1.86305657e-01 -1.43401277e+00 -8.22329462e-01
2.95487881e-01 3.40629131e-01 -4.76525784e-01 -2.43938640e-01
-5.88787615e-01 -7.60856092e-01 1.12217796e+00 3.53118122e-01
9.26467121e-01 -4.63810384e-01 -5.98005474e-01 -2.05402806e-01
-2.09453255e-01 -1.17721331e+00 -3.29389751e-01 -3.89935791e-01
-1.75417289e-01 -1.19154489e+00 -6.14880681e-01 -5.74262023e-01
7.96150863e-01 3.58369499e-01 3.92574877e-01 -9.03315917e-02
-2.81654358e-01 4.71869916e-01 -3.25509489e-01 -6.95912540e-01
-2.25806966e-01 1.69250876e-01 7.92261958e-01 4.75518137e-01
5.63684285e-01 -3.36539894e-01 -6.86150074e-01 6.58572435e-01
-3.10113817e-01 -4.71473634e-01 3.18886489e-01 3.67377549e-01
5.07316351e-01 1.75479114e-01 4.16141182e-01 -8.07320416e-01
4.04636413e-01 -2.79998183e-01 -5.47090352e-01 2.87176788e-01
-4.25312161e-01 -5.47474623e-02 3.97512525e-01 -2.08131045e-01
-1.22081447e+00 8.95933986e-01 -1.41696259e-01 -2.57685035e-01
-2.49912426e-01 -4.38571423e-01 -1.01462173e+00 -9.10799921e-01
5.89216352e-01 -2.25274786e-02 2.33511224e-01 -3.97474706e-01
5.02099216e-01 1.01316237e+00 8.53064358e-01 -5.28849185e-01
1.01725721e+00 8.72731328e-01 1.78800285e-01 -1.12121260e+00
-1.78642645e-01 -5.32688618e-01 -9.55215037e-01 -4.63762045e-01
8.02569926e-01 -1.02127481e+00 -1.10358238e+00 5.04034221e-01
-1.25937843e+00 5.70496619e-01 9.73596126e-02 2.95968294e-01
-5.04042566e-01 7.11942077e-01 -4.49335761e-02 -1.27552319e+00
-4.89848137e-01 -9.52273786e-01 1.21244478e+00 4.05973345e-01
-2.80538082e-01 -3.90092880e-01 -6.84003457e-02 2.96133280e-01
3.67600517e-03 5.73724210e-01 1.03706501e-01 -5.42562485e-01
-3.74431878e-01 -5.26353300e-01 -1.73206240e-01 3.79144438e-02
1.01633601e-01 1.53524995e-01 -1.23597169e+00 -4.51400608e-01
-3.93567421e-02 1.03204608e-01 6.19418383e-01 3.04645777e-01
8.66819143e-01 -3.69124174e-01 -7.78007627e-01 7.91762114e-01
9.48531330e-01 4.91899960e-02 6.48512244e-01 2.83617377e-01
8.43199909e-01 9.64370370e-01 7.04192281e-01 5.03111184e-01
4.66384053e-01 8.11352968e-01 2.54691809e-01 -1.48967290e-02
-1.05421945e-01 -3.93399417e-01 4.27608609e-01 9.24443230e-02
-1.17175996e-01 -1.51470020e-01 -8.14659894e-01 2.53806621e-01
-1.92007947e+00 -1.01133859e+00 -1.64615124e-01 2.26216221e+00
4.62707579e-01 -3.70883942e-02 5.54472268e-01 1.38380617e-01
1.33339548e+00 -2.57253915e-01 -3.26367676e-01 -8.66437778e-02
-4.93915193e-02 9.99487862e-02 1.12862766e+00 6.02496743e-01
-1.48279476e+00 1.16108906e+00 6.88806343e+00 6.01987720e-01
-8.04469824e-01 1.03756145e-01 3.98689657e-01 -4.31541540e-03
4.23718452e-01 -2.41320968e-01 -1.30452263e+00 5.11775017e-01
6.99491918e-01 -3.57649624e-01 8.47555846e-02 1.12631667e+00
3.11691135e-01 -2.61948258e-01 -8.36845160e-01 1.11455917e+00
3.81669223e-01 -6.81082428e-01 -9.82243270e-02 2.70014346e-01
3.66163105e-01 -5.49869001e-01 -2.05928311e-02 -2.58683801e-01
2.79094577e-01 -8.45996737e-01 8.00331056e-01 6.64775372e-01
8.86287987e-01 -1.04566979e+00 8.86863053e-01 1.88696876e-01
-1.62352002e+00 9.84295681e-02 -4.70377833e-01 1.81079552e-01
2.69110441e-01 3.11680675e-01 -1.21377409e+00 5.07441640e-01
6.45057678e-01 2.54207492e-01 -9.78779376e-01 1.09441817e+00
-1.28662840e-01 2.49657243e-01 -6.95629418e-01 5.35730459e-02
-5.41630983e-01 1.28161103e-01 6.62897706e-01 1.02550054e+00
4.46579993e-01 1.81332994e-02 1.99766472e-01 4.15061623e-01
-7.04574138e-02 1.43274009e-01 -8.59941483e-01 4.58870351e-01
9.01311874e-01 1.20142496e+00 -7.13018179e-01 -1.26754353e-02
-2.57211953e-01 1.12798929e+00 8.90638754e-02 3.43577191e-03
-8.61432970e-01 -4.20591563e-01 9.37243223e-01 6.01764798e-01
8.11827332e-02 -2.88056970e-01 -2.65454680e-01 -8.76035988e-01
1.12667464e-01 -6.58691525e-01 3.72871161e-01 -2.73451298e-01
-1.08041978e+00 4.35277641e-01 4.40072566e-02 -1.16515219e+00
-3.45149666e-01 -5.49790084e-01 -3.19145083e-01 7.54663885e-01
-8.88814330e-01 -1.57889390e+00 -5.35497010e-01 8.05430949e-01
-8.78406391e-02 -4.35950994e-01 7.67150223e-01 5.04397690e-01
-6.82489514e-01 1.24348378e+00 -1.48019448e-01 3.39663029e-01
9.94024277e-01 -6.24955237e-01 4.45035994e-01 1.27918470e+00
-2.44281769e-01 7.69647181e-01 7.22787201e-01 -1.17971301e+00
-1.23152351e+00 -1.27052259e+00 4.08045709e-01 -6.09963715e-01
2.45385338e-02 -5.33111334e-01 -5.20629525e-01 6.08043671e-01
-3.40175658e-01 3.30306068e-02 7.43206680e-01 -1.92461178e-01
-1.97662130e-01 -2.69411623e-01 -1.91939235e+00 3.32874745e-01
1.12561584e+00 -4.18707669e-01 -3.60999882e-01 -1.56160176e-01
3.71328264e-01 -1.77259788e-01 -4.76863295e-01 5.02615631e-01
9.35998321e-01 -9.70166564e-01 1.04707682e+00 -1.00415751e-01
-5.27996302e-01 -7.37230122e-01 -3.46224606e-01 -6.14036441e-01
-1.83133230e-01 -5.67656100e-01 -1.60596102e-01 1.62794173e+00
-3.97349820e-02 -6.16448998e-01 1.46507716e+00 9.03738797e-01
5.14619291e-01 -5.89984283e-02 -1.21053171e+00 -8.83934796e-01
-3.92479748e-01 7.73086026e-03 7.03399837e-01 5.58088303e-01
-3.23951393e-01 -1.38624564e-01 -5.38254440e-01 6.66208386e-01
1.20710003e+00 -5.92172325e-01 1.18508410e+00 -1.29845810e+00
4.83351082e-01 5.17826080e-02 -1.02977979e+00 -3.05054784e-01
3.27900201e-01 -5.94279408e-01 1.41262775e-02 -1.04794085e+00
1.85031131e-01 -1.71851873e-01 1.96274340e-01 4.50132012e-01
-7.23098889e-02 5.05607724e-01 4.48129550e-02 -8.53643473e-03
-5.15009522e-01 5.95208526e-01 3.88622165e-01 2.10928172e-02
6.73556775e-02 2.09878549e-01 -7.72439480e-01 9.57434595e-01
7.47389734e-01 -5.90840578e-01 -1.33322164e-01 1.15102209e-01
-4.68726635e-01 -5.52611291e-01 5.77959836e-01 -1.62619507e+00
4.62193936e-01 2.13003814e-01 8.22186589e-01 -5.88040948e-01
4.10095960e-01 -1.06576192e+00 1.75785840e-01 6.38045669e-01
3.51025999e-01 -7.36806216e-03 2.21572712e-01 5.50846279e-01
1.49173975e-01 -1.88905343e-01 1.16130042e+00 -3.13406549e-02
-6.37705863e-01 3.61280978e-01 -1.82949841e-01 -2.98413694e-01
1.56582987e+00 -4.88291562e-01 -1.41102120e-01 -4.10411775e-01
-4.34535056e-01 1.54449835e-01 9.98100519e-01 2.64830917e-01
5.88165522e-01 -1.41420472e+00 -6.10543728e-01 3.35853189e-01
1.65757611e-01 -3.56988072e-01 -8.38143378e-02 2.49933273e-01
-7.05521762e-01 1.50967106e-01 -5.00082433e-01 -4.16745186e-01
-1.83418798e+00 4.80504334e-01 4.29004043e-01 3.55738103e-01
-1.72942400e-01 8.04700375e-01 -1.46079853e-01 -4.23583597e-01
2.29262859e-01 4.18141484e-01 -2.06974611e-01 8.53642523e-02
7.36930847e-01 8.52844477e-01 -2.19884031e-02 -1.32637751e+00
-9.15988266e-01 7.33314753e-01 -4.32925634e-02 -2.96725214e-01
7.05494940e-01 -3.28754604e-01 6.39869273e-03 -3.59001398e-01
9.94539618e-01 3.71590674e-01 -1.10882807e+00 1.29274815e-01
-7.07177445e-02 -7.98055649e-01 -2.09255815e-01 -3.09941709e-01
-8.15234601e-01 4.82116014e-01 1.26594174e+00 -2.23954991e-01
7.97888458e-01 -2.85275012e-01 6.62749648e-01 2.44664997e-01
8.33606660e-01 -1.22036898e+00 -6.13133967e-01 2.76588723e-02
5.49768746e-01 -1.37494874e+00 4.84532148e-01 -9.34404373e-01
-3.23179156e-01 7.62501419e-01 8.43996286e-01 2.06063554e-01
6.59257531e-01 4.72221494e-01 -7.36459866e-02 1.83293447e-01
2.43394345e-01 -3.99131268e-01 1.78363144e-01 1.14658332e+00
-5.91539452e-03 1.70405596e-01 -4.28065300e-01 8.60204577e-01
-6.39549732e-01 -1.12021305e-01 1.62867099e-01 9.56226647e-01
-6.14547431e-01 -1.33884788e+00 -9.62216079e-01 2.12712869e-01
-3.60929042e-01 3.73769820e-01 -9.06335473e-01 6.20097160e-01
4.85057652e-01 1.05889666e+00 -3.04864496e-01 -6.23222053e-01
4.40528899e-01 1.90185830e-01 4.97744709e-01 -3.64342988e-01
-2.19994903e-01 -4.21112180e-01 1.64471090e-01 -6.72739267e-01
-5.05449116e-01 -8.62735450e-01 -1.17821324e+00 -7.50820458e-01
-2.15246618e-01 -1.69865742e-01 7.27939487e-01 5.98126829e-01
4.79658544e-01 -2.01613396e-01 6.13999963e-01 -1.04607630e+00
-2.35003993e-01 -5.16090512e-01 -5.61404705e-01 3.28505397e-01
-6.19146414e-02 -1.09230411e+00 -9.67291296e-02 1.14341438e-01] | [12.871909141540527, 0.7117544412612915] |
62f2fc82-0d66-442c-925e-5e81472036b8 | improving-sequence-to-sequence-semantic | null | null | https://aclanthology.org/2020.intexsempar-1.3 | https://aclanthology.org/2020.intexsempar-1.3.pdf | Improving Sequence-to-Sequence Semantic Parser for Task Oriented Dialog | Task Oriented Parsing (TOP) attempts to map utterances to compositional requests, including multiple intents and their slots. Previous work focus on a tree-based hierarchical meaning representation, and applying constituency parsing techniques to address TOP. In this paper, we propose a new format of meaning representation that is more compact and amenable to sequence-to-sequence (seq-to-seq) models. A simple copy-augmented seq-to-seq parser is built and evaluated over a public TOP dataset, resulting in 3.44% improvement over prior best seq-to-seq parser (exact match accuracy), which is also comparable to constituency parsers’ performance. | ['Chaoting Xuan'] | null | null | null | null | emnlp-intexsempar-2020-11 | ['constituency-parsing'] | ['natural-language-processing'] | [ 7.84006417e-01 7.66261280e-01 -4.03709352e-01 -9.20541227e-01
-1.26500225e+00 -7.02891827e-01 3.62713039e-01 3.46149147e-01
-1.14755402e-03 6.51167989e-01 1.22233653e+00 -5.19395828e-01
2.67513037e-01 -6.79239213e-01 -3.45760971e-01 -6.51604459e-02
1.27039522e-01 5.49934804e-01 2.50691354e-01 -6.90671980e-01
2.97554731e-01 -3.16669136e-01 -1.35617137e+00 1.08226633e+00
2.56091505e-01 6.53821349e-01 3.51393104e-01 1.06854844e+00
-9.55799639e-01 1.25091374e+00 -6.64125800e-01 -2.41834491e-01
-1.86384365e-01 -4.97145176e-01 -1.44014323e+00 -4.00181383e-01
5.47775865e-01 -5.72606586e-02 3.30148727e-01 6.78057015e-01
1.55261174e-01 -1.27334416e-01 1.42999753e-01 -8.09331775e-01
-5.61850667e-01 1.13380694e+00 -6.09832108e-02 -1.77283715e-02
1.20834041e+00 -3.81956428e-01 1.68782198e+00 -4.73397732e-01
9.60090995e-01 1.69744623e+00 6.51238382e-01 9.00331676e-01
-1.30011582e+00 -4.70373690e-01 2.96088099e-01 -4.51993525e-01
-7.19665706e-01 -4.79092717e-01 5.08779705e-01 -2.73464024e-01
1.79458022e+00 4.22645181e-01 5.26494622e-01 1.20114315e+00
4.57228571e-01 8.71036112e-01 1.31979144e+00 -7.79439688e-01
2.80800194e-01 -4.98922020e-01 7.44462550e-01 5.25834441e-01
-2.88936887e-02 -2.25647792e-01 -8.69612753e-01 -3.68427515e-01
4.53048378e-01 -4.81088847e-01 4.71124560e-01 1.91283867e-01
-1.19606435e+00 1.01759970e+00 6.44813254e-02 5.34475327e-01
-3.97488713e-01 1.78310961e-01 7.13589728e-01 3.46255869e-01
3.95376563e-01 5.16856432e-01 -8.80304575e-01 -6.27473950e-01
-4.64989871e-01 5.83415329e-01 1.13445473e+00 1.41287076e+00
5.19308269e-01 -1.25182167e-01 -4.17693615e-01 8.22250724e-01
2.05679059e-01 7.15065002e-02 5.90084672e-01 -1.05873334e+00
7.35616863e-01 7.09540665e-01 -1.77375674e-01 -4.37206566e-01
-8.09751391e-01 8.73359814e-02 -3.25772911e-01 -6.59949899e-01
1.74286142e-01 -1.47273928e-01 -8.78597856e-01 1.99966681e+00
2.91565984e-01 -3.50216240e-01 6.11146152e-01 4.70946759e-01
1.11690950e+00 7.80409992e-01 9.09391940e-01 -4.46042642e-02
2.11050558e+00 -6.78062975e-01 -7.47001112e-01 -6.01965964e-01
1.06575441e+00 -7.35532522e-01 1.15172267e+00 -4.59227003e-02
-1.12809479e+00 -4.38919067e-01 -7.02992320e-01 -3.77805918e-01
-8.09821263e-02 -4.18346971e-01 9.86666977e-01 6.34329200e-01
-1.17019308e+00 1.99438363e-01 -5.09347916e-01 -7.18748093e-01
-6.29401058e-02 1.59322515e-01 -4.37393248e-01 1.83818579e-01
-1.18593204e+00 7.03509688e-01 1.02415419e+00 -7.15232849e-01
-4.06921178e-01 -6.09220922e-01 -1.25515747e+00 3.19159850e-02
1.81556895e-01 -8.94331634e-01 1.85357869e+00 -6.22962952e-01
-1.62688768e+00 8.65102172e-01 -8.32677662e-01 -4.02987272e-01
-4.61524576e-01 -2.95360804e-01 -2.99602032e-01 -1.09287404e-01
5.97249568e-01 1.02483130e+00 3.39091241e-01 -9.96526897e-01
-1.05536127e+00 -2.88065493e-01 5.06907701e-01 2.75236100e-01
4.96413037e-02 7.99461246e-01 -2.49322429e-02 -7.12294102e-01
6.00697994e-01 -8.23352158e-01 -1.53916702e-01 -9.86257851e-01
-3.43150973e-01 -6.28456533e-01 2.73273915e-01 -7.07287133e-01
1.46764994e+00 -1.80851698e+00 -1.36736900e-01 -5.34592927e-01
-1.04228355e-01 -2.45723918e-01 -2.94518471e-01 1.07092714e+00
-2.06944674e-01 3.89174044e-01 -4.20018792e-01 -3.89129013e-01
1.13937996e-01 8.12213480e-01 -3.57921332e-01 -4.07893956e-01
5.37120998e-01 1.00585461e+00 -9.41904068e-01 -7.51292467e-01
3.64000760e-02 -1.02484673e-01 -9.20884073e-01 3.35394114e-01
-4.18606132e-01 2.55186856e-01 -6.24843538e-01 8.10617566e-01
3.00593406e-01 2.13288665e-01 7.88478553e-01 2.31774062e-01
-2.25788429e-01 1.37617171e+00 -4.65207875e-01 2.00462413e+00
-7.21024454e-01 2.95993060e-01 4.51372080e-02 -8.40220988e-01
1.00219369e+00 6.23585761e-01 1.49768993e-01 -6.49473369e-01
-7.65449032e-02 -1.69232883e-03 4.68095094e-02 -4.03282225e-01
8.79378915e-01 -7.60424256e-01 -1.14677024e+00 2.29824752e-01
2.59620994e-01 -3.07735503e-01 3.15524131e-01 1.44501701e-01
1.52183747e+00 4.27485853e-01 8.48694980e-01 -6.36950374e-01
4.94510978e-01 4.79364067e-01 7.22763419e-01 5.89876950e-01
-2.69934945e-02 3.75689536e-01 7.83717394e-01 -6.34991109e-01
-1.07643008e+00 -7.70980835e-01 -4.00159732e-02 1.91236448e+00
-2.36228570e-01 -8.56450796e-01 -9.47208941e-01 -8.01136434e-01
-4.50746447e-01 1.25258291e+00 -5.09834290e-01 4.18819755e-01
-1.17291760e+00 -3.75296801e-01 8.53340328e-01 7.18088388e-01
2.07034096e-01 -1.35380852e+00 -6.00992441e-01 8.08551371e-01
-5.88715374e-01 -1.45437503e+00 -4.95145060e-02 4.78477180e-01
-8.91019642e-01 -7.58555174e-01 -3.73554081e-02 -9.99100208e-01
6.11922145e-02 -1.82117924e-01 1.51148820e+00 -1.50277406e-01
2.16357231e-01 -7.55148232e-02 -9.41908002e-01 -5.47828197e-01
-1.04835582e+00 3.40375483e-01 -3.32808405e-01 -6.20682597e-01
5.34316778e-01 -3.74972373e-01 -1.95789728e-02 -1.19393960e-01
-7.45431662e-01 4.71530586e-01 6.19273067e-01 7.80956388e-01
3.18507224e-01 -4.69786644e-01 5.70615590e-01 -1.19100845e+00
7.73962915e-01 -3.19303393e-01 -3.48906107e-02 1.29315674e-01
-3.30619305e-01 3.69464010e-01 6.44894958e-01 1.54531315e-01
-1.42305362e+00 2.70337313e-01 -9.49220240e-01 5.10494053e-01
-6.65599108e-01 4.42963719e-01 -1.26658514e-01 5.33399463e-01
3.58169377e-01 1.48545876e-01 -3.16331029e-01 -5.09512246e-01
4.34043646e-01 6.52888179e-01 5.63428760e-01 -1.07982194e+00
1.55960813e-01 -2.03236826e-02 -3.58803533e-02 -7.32729673e-01
-1.18902171e+00 -8.18779945e-01 -7.99652755e-01 3.92319739e-01
1.11570871e+00 -7.39746511e-01 -3.23270142e-01 -1.42631844e-01
-1.42795980e+00 -3.23322594e-01 -2.99124599e-01 1.33155644e-01
-8.01795840e-01 3.71521801e-01 -9.13947225e-01 -7.73842812e-01
-6.61475480e-01 -6.73621535e-01 1.46840775e+00 -1.37985781e-01
-8.19983423e-01 -8.24802399e-01 1.23811439e-01 4.98227924e-01
2.95848429e-01 1.85736179e-01 1.28451169e+00 -8.46358955e-01
6.02216944e-02 1.70586139e-01 -2.79609442e-01 -1.17620379e-01
9.08560231e-02 -4.51325327e-01 -7.97440350e-01 7.30692446e-02
2.18862221e-02 -3.31065565e-01 4.37633395e-01 2.71544844e-01
4.09777671e-01 -5.99247992e-01 -8.54432583e-02 1.51722178e-01
1.23526168e+00 3.16882014e-01 5.57537258e-01 5.85883498e-01
3.98915470e-01 9.71596420e-01 9.13310349e-01 1.86914325e-01
9.83505905e-01 6.88153386e-01 1.66215912e-01 1.93112195e-01
-3.10881257e-01 -8.10177505e-01 4.57240045e-01 9.97834384e-01
3.43195349e-01 -9.61888582e-02 -9.13586438e-01 6.16214573e-01
-1.80034244e+00 -7.49300241e-01 -3.03939193e-01 1.49928546e+00
9.29039299e-01 2.16046825e-01 1.43294260e-01 -2.35652566e-01
5.43320179e-01 2.52791673e-01 5.20793125e-02 -1.28560567e+00
1.87252965e-02 4.50067639e-01 3.69335651e-01 5.37616014e-01
-1.05647397e+00 1.60860312e+00 7.77492619e+00 5.18377364e-01
-5.35267413e-01 3.37515682e-01 2.64497608e-01 5.06500125e-01
-5.09871483e-01 4.65672761e-01 -1.07293916e+00 7.17121810e-02
1.28328598e+00 1.31002545e-01 5.43478094e-02 9.96335268e-01
-3.65929082e-02 3.38349156e-02 -1.10576057e+00 5.46757817e-01
-2.11199552e-01 -1.26638067e+00 1.46896258e-01 -1.85028270e-01
3.07860434e-01 1.39427884e-02 -7.73194969e-01 7.48555481e-01
6.32395327e-01 -5.80700219e-01 1.08547246e+00 -1.49336740e-01
5.65953076e-01 -5.22874415e-01 8.73616755e-01 5.37949383e-01
-1.74271858e+00 -9.00698379e-02 -3.75543654e-01 -6.13286972e-01
5.46076596e-01 4.78317812e-02 -1.09615040e+00 7.39969790e-01
6.23712242e-01 3.48492622e-01 -2.05682904e-01 -4.83354414e-03
-4.11221147e-01 9.67443883e-01 -9.51830745e-02 -3.73801261e-01
4.52730089e-01 2.54476279e-01 5.53043664e-01 1.91307795e+00
1.58887997e-01 5.93172014e-01 4.73018616e-01 4.40759987e-01
7.74174556e-02 4.17909831e-01 -5.55435181e-01 -5.50679006e-02
5.37826419e-01 1.03244829e+00 -7.41300583e-01 -8.33975673e-01
-5.24944365e-01 8.29061210e-01 4.49385941e-01 -2.93499321e-01
-3.11610579e-01 -1.11711703e-01 8.26356411e-01 -1.12172753e-01
1.90082520e-01 -2.10364029e-01 -4.16703582e-01 -9.07229006e-01
-1.13113008e-01 -9.30606186e-01 6.95433319e-01 -6.31917715e-01
-1.14708757e+00 7.35849977e-01 5.79546571e-01 -6.92177534e-01
-5.69974661e-01 -5.17102480e-01 -5.73916733e-01 7.58767784e-01
-1.21435368e+00 -1.55057812e+00 2.65148610e-01 1.17965542e-01
1.12635803e+00 5.30867651e-02 1.38158715e+00 -2.04626963e-01
-2.55546570e-01 4.10215408e-01 -7.39596725e-01 2.49297872e-01
1.10612221e-01 -1.52879870e+00 1.34258711e+00 7.97497213e-01
8.29043761e-02 8.85572076e-01 7.45492935e-01 -7.08169639e-01
-1.42704952e+00 -9.30998981e-01 1.70176375e+00 -6.64583802e-01
6.80868804e-01 -6.86201632e-01 -7.01224685e-01 9.54137325e-01
4.54878747e-01 -5.95078766e-01 8.53551567e-01 6.43568695e-01
-5.11051595e-01 5.54440200e-01 -1.27752388e+00 4.17811155e-01
1.39476562e+00 -5.12957871e-01 -1.45855796e+00 -5.93200885e-03
1.27509296e+00 -5.48246443e-01 -1.07344007e+00 4.39272285e-01
7.59114087e-01 -9.11589742e-01 5.62913597e-01 -9.17519510e-01
4.87119406e-01 5.36659025e-02 -7.15715706e-01 -8.69636059e-01
-5.18037438e-01 -7.06586599e-01 2.87446052e-01 1.44281387e+00
5.74476063e-01 -6.49689317e-01 8.03474784e-01 5.33601999e-01
-7.51749516e-01 -5.06757498e-01 -1.07392013e+00 -6.89667881e-01
1.95250213e-01 -9.74902928e-01 8.87074888e-01 7.71824598e-01
6.39879167e-01 1.04611063e+00 -5.25452420e-02 1.25291916e-02
3.13792378e-01 2.82609224e-01 5.06564021e-01 -1.02515554e+00
-4.14706796e-01 -2.92767435e-01 -3.15403521e-01 -1.36404276e+00
3.11822355e-01 -9.66667056e-01 3.76145989e-01 -1.97483552e+00
2.11990654e-01 -5.19366026e-01 4.96898592e-02 9.90817368e-01
-2.83170551e-01 -5.07911444e-02 3.00465554e-01 5.42474426e-02
-4.59194064e-01 1.75315395e-01 9.14316535e-01 3.73052731e-02
-1.63966820e-01 -2.76564598e-01 -1.04576719e+00 7.91113555e-01
1.08164608e+00 -7.90963769e-01 -1.69014350e-01 -3.34515065e-01
2.71804243e-01 3.68414193e-01 -3.60398203e-01 -6.35038197e-01
-1.23244844e-01 -3.09148133e-01 -2.94012278e-01 -7.45987713e-01
3.28564703e-01 -2.80550003e-01 -1.43776909e-01 4.80648786e-01
-6.71764553e-01 3.22648108e-01 4.04119819e-01 8.79818872e-02
-2.43425563e-01 -5.32840550e-01 2.29923055e-01 -5.24346769e-01
-1.04747903e+00 -4.29719537e-01 -7.37287819e-01 1.92134961e-01
6.67305291e-01 -6.65227771e-02 -3.05967331e-01 -2.58432060e-01
-6.16577506e-01 -3.03430706e-02 1.59837633e-01 6.91316843e-01
5.42325437e-01 -1.02453017e+00 -8.07245731e-01 1.41565353e-01
5.21771550e-01 -1.93910033e-01 -7.06737563e-02 2.70926833e-01
-4.59769487e-01 9.86994505e-01 -9.37554389e-02 -4.68955189e-01
-1.56395376e+00 3.99307460e-01 -2.13410512e-01 -4.18938249e-01
-8.20995986e-01 8.67818534e-01 3.83034587e-01 -9.90389705e-01
-1.97748378e-01 -8.97577524e-01 -1.80987656e-01 -2.79172987e-01
5.54542959e-01 -1.19850347e-02 -9.31195617e-02 -9.17558074e-01
-5.59705615e-01 5.41312218e-01 4.10669297e-02 -2.35038519e-01
9.81079102e-01 -1.37043506e-01 -3.72633696e-01 2.72578388e-01
1.04586530e+00 -7.92797282e-02 -6.50791883e-01 -1.48195922e-01
7.39750862e-01 -1.10998392e-01 -2.91232675e-01 -6.62413061e-01
-2.35962146e-03 4.81217682e-01 4.47132923e-02 6.09883010e-01
9.33503568e-01 5.18717885e-01 1.07734728e+00 4.75930810e-01
6.16914451e-01 -9.17246342e-01 -3.24355841e-01 1.11031485e+00
7.86054671e-01 -8.84843528e-01 -3.62922460e-01 -8.11628520e-01
-7.66020477e-01 1.05828846e+00 3.00300717e-01 1.62157044e-01
2.48749062e-01 6.73772395e-01 2.68336773e-01 -2.97276616e-01
-1.17768252e+00 -4.60887372e-01 -1.22481875e-01 6.44615948e-01
1.04740584e+00 4.74557549e-01 -6.39308989e-01 5.56908727e-01
-8.12753797e-01 -2.04523027e-01 1.37449592e-01 1.38953841e+00
-7.53306687e-01 -1.73469460e+00 -1.82835117e-01 2.68656760e-01
-8.74672115e-01 -5.61027229e-01 -6.46457195e-01 8.05037796e-01
-1.09226303e-02 1.38925552e+00 1.45228341e-01 -3.89917731e-01
6.33002996e-01 5.64303696e-01 3.32856268e-01 -1.39199507e+00
-9.61210668e-01 1.17648669e-01 1.14912677e+00 -7.14095056e-01
-4.96587396e-01 -8.19982588e-01 -1.82950044e+00 -7.22950324e-02
5.66694373e-03 1.12803519e-01 6.49974287e-01 9.83835697e-01
2.36451685e-01 5.23803353e-01 4.65344548e-01 -6.16392136e-01
-2.44426891e-01 -1.19164479e+00 -1.32364184e-01 3.82298708e-01
-3.99460383e-02 -1.20895118e-01 2.79202759e-01 1.06344894e-02] | [10.435626029968262, 9.39526081085205] |
7e01d2c5-a173-4385-9f26-93b53fac7b98 | fault-detection-in-induction-motors-using | 2306.09365 | null | https://arxiv.org/abs/2306.09365v1 | https://arxiv.org/pdf/2306.09365v1.pdf | Fault Detection in Induction Motors using Functional Dimensionality Reduction Methods | The implementation of strategies for fault detection and diagnosis on rotating electrical machines is crucial for the reliability and safety of modern industrial systems. The contribution of this work is a methodology that combines conventional strategy of Motor Current Signature Analysis with functional dimensionality reduction methods, namely Functional Principal Components Analysis and Functional Diffusion Maps, for detecting and classifying fault conditions in induction motors. The results obtained from the proposed scheme are very encouraging, revealing a potential use in the future not only for real-time detection of the presence of a fault in an induction motor, but also in the identification of a greater number of types of faults present through an offline analysis. | ['Ángela Fernández', 'Carlos M. Alaíz', 'José M. Bossio', 'María Barroso'] | 2023-06-14 | null | null | null | null | ['dimensionality-reduction', 'fault-detection'] | ['methodology', 'miscellaneous'] | [ 2.75176793e-01 -3.69293004e-01 2.23407283e-01 2.76031613e-01
1.34529889e-01 -3.72003168e-01 3.94406855e-01 9.56603289e-02
1.78377166e-01 4.83161330e-01 -3.35184515e-01 -5.63935399e-01
-1.05696833e+00 -4.97687191e-01 -3.58539820e-02 -1.12018239e+00
-3.32762092e-01 4.80797023e-01 7.64080808e-02 -1.55324399e-01
6.00499749e-01 1.38217771e+00 -1.77038133e+00 -4.39428724e-03
5.37599385e-01 8.98404121e-01 6.55547976e-01 4.09336239e-01
3.98408383e-01 8.41420531e-01 -7.72098005e-01 6.47084177e-01
-1.31769150e-01 -6.11482203e-01 -9.62555766e-01 5.03954411e-01
-7.04056621e-01 -3.16209323e-03 -2.92577624e-01 1.02469730e+00
2.69373983e-01 3.28015506e-01 8.85914028e-01 -9.35018182e-01
2.00709570e-02 1.02974795e-01 -1.94101870e-01 5.77180743e-01
1.54786840e-01 5.38707003e-02 4.29631680e-01 -7.45862067e-01
4.08290863e-01 7.90279984e-01 5.87292202e-02 -1.08913139e-01
-6.70629621e-01 -1.55499533e-01 -3.62731904e-01 9.10382867e-01
-1.07972491e+00 -1.64966732e-01 9.41653192e-01 -7.13293672e-01
9.34814453e-01 3.66287917e-01 4.13134634e-01 3.93691331e-01
7.74434924e-01 4.67604727e-01 1.08354521e+00 -5.33511102e-01
3.67723405e-01 -5.23142330e-02 4.99019682e-01 2.63275683e-01
4.92027044e-01 8.55020881e-02 -1.70261234e-01 4.74109799e-02
4.22151536e-01 5.58634102e-02 -4.71335620e-01 -4.22702372e-01
-9.87926781e-01 7.03697622e-01 -1.47229269e-01 1.19915664e+00
-9.53338146e-01 -3.18594068e-01 6.96768403e-01 5.63638449e-01
3.75993043e-01 6.71069443e-01 -4.52835768e-01 -5.09598017e-01
-6.03978455e-01 -8.73471424e-03 4.24141347e-01 2.56616652e-01
2.45347604e-01 4.15651411e-01 3.57767701e-01 4.00816858e-01
1.21252611e-01 3.75836730e-01 1.02000654e+00 -6.51576102e-01
-3.63039337e-02 9.40993071e-01 -1.83454901e-02 -8.71923864e-01
-6.57077491e-01 -2.80916482e-01 -4.21151221e-01 4.88769174e-01
-1.58717126e-01 -2.21940875e-01 -4.06269222e-01 7.72426426e-01
1.62749797e-01 -1.62966982e-01 1.61736831e-01 7.36938179e-01
3.49702612e-02 4.57983732e-01 -4.48911011e-01 -3.89211535e-01
1.12112236e+00 -2.20278084e-01 -1.09620965e+00 2.76647836e-01
8.44374418e-01 -7.13990211e-01 2.56395757e-01 8.93437743e-01
-7.24716425e-01 -5.24967551e-01 -1.29188395e+00 8.21363926e-01
-4.90561396e-01 5.07890463e-01 5.17316759e-01 4.67680126e-01
-7.13919699e-01 1.01098537e+00 -8.20152104e-01 -3.23404729e-01
-5.88952117e-02 3.18996876e-01 -6.94888830e-01 -1.50334075e-01
-1.17565596e+00 1.50826073e+00 5.09831786e-01 5.06704569e-01
-6.30979240e-01 -2.89960206e-01 -6.67674243e-01 -1.01437032e-01
8.73280875e-03 2.44675249e-01 7.60550797e-01 -6.56576574e-01
-1.33934200e+00 2.75207222e-01 -4.98199984e-02 -4.06402260e-01
1.69561118e-01 -1.03805751e-01 -7.33158231e-01 6.31894171e-01
-1.88608423e-01 -5.46127558e-01 8.70923936e-01 -8.55289400e-01
-6.38302386e-01 -7.05884874e-01 -4.54789311e-01 1.72337368e-02
-2.08181739e-01 1.02537140e-01 5.88234603e-01 -2.44334072e-01
3.58525395e-01 -7.23490775e-01 1.15556791e-01 -7.71132946e-01
-9.42994505e-02 -6.25656426e-01 1.69979346e+00 -9.31926787e-01
8.09295714e-01 -1.92873633e+00 4.20368284e-01 6.21116102e-01
-1.73848167e-01 5.30854464e-01 2.97522604e-01 8.32372129e-01
-6.07977688e-01 -5.71745932e-01 -2.07632810e-01 5.11901617e-01
-3.50431949e-01 2.15267196e-01 -1.32838994e-01 1.00165963e+00
5.39757669e-01 4.11950409e-01 -6.15102291e-01 2.85254598e-01
8.54629159e-01 2.63181806e-01 4.71335709e-01 3.07442080e-02
4.13755238e-01 4.08444583e-01 -4.96325433e-01 2.78007865e-01
2.97954410e-01 3.30325067e-01 1.57096252e-01 -2.73169637e-01
-3.63287389e-01 1.19773164e-01 -1.31377220e+00 1.00894463e+00
-3.06139797e-01 8.14530730e-01 3.13763246e-02 -1.85506988e+00
1.03837693e+00 8.57114136e-01 9.74571884e-01 -6.06180727e-01
3.73264462e-01 1.77813113e-01 1.97138265e-01 -9.33729053e-01
2.63301581e-01 9.69316512e-02 1.71382606e-01 4.55865115e-01
2.92101175e-01 1.06142104e-01 6.38644159e-01 -1.90891057e-01
1.18322301e+00 -1.37304217e-01 -1.98269375e-02 -5.14822721e-01
8.65614235e-01 -4.66364138e-02 8.51951316e-02 -1.00481436e-01
3.36554348e-02 -9.64734331e-02 5.23637176e-01 -7.63574019e-02
-7.01796532e-01 -4.19656664e-01 -5.63249886e-01 1.28924385e-01
6.01355396e-02 3.08902562e-01 -5.69323480e-01 -3.94335777e-01
1.73199296e-01 7.26422489e-01 -3.41681391e-01 -6.14207387e-01
-4.53728855e-01 -1.00583971e+00 -5.13579929e-03 2.60381669e-01
3.49066406e-02 -1.12277448e+00 -8.86196911e-01 4.30924177e-01
1.99495703e-01 -6.73736155e-01 5.60279667e-01 5.30016661e-01
-1.20058024e+00 -1.57410860e+00 -4.95357364e-01 -7.35234022e-01
7.61572123e-01 2.57365465e-01 2.14915738e-01 1.05467238e-01
-7.37507164e-01 3.58319461e-01 -6.08174264e-01 -1.19265988e-01
-7.78958023e-01 -3.13438147e-01 2.75378257e-01 -2.87669264e-02
3.02290112e-01 -3.36921543e-01 -3.35570097e-01 6.27821743e-01
-9.21925783e-01 -6.42745376e-01 7.85997808e-01 5.49551010e-01
-9.02929157e-02 1.28429627e+00 1.03518176e+00 -5.73173702e-01
8.53220582e-01 -7.83625007e-01 -5.46941459e-01 -1.36658520e-01
-1.10770595e+00 -2.43235648e-01 5.68834722e-01 8.49781372e-03
-9.13761437e-01 -1.48729831e-01 -2.70942271e-01 -1.62087888e-01
-6.23205781e-01 6.37842774e-01 -2.09613517e-01 -1.86007991e-01
4.52985853e-01 1.85113862e-01 6.43328130e-01 -7.80487418e-01
-1.71473883e-02 7.90597200e-01 5.70774019e-01 1.06983326e-01
5.05674362e-01 1.39939398e-01 3.75148028e-01 -1.13088453e+00
3.64354014e-01 -1.11875594e+00 -7.25067556e-01 -4.70710963e-01
4.30872500e-01 -4.63325530e-01 -4.03956980e-01 9.11639392e-01
-9.15821970e-01 1.13038972e-01 -3.58946949e-01 6.57757461e-01
-3.63817006e-01 6.38935566e-01 -6.58725679e-01 -8.39995980e-01
-1.75091967e-01 -1.00751328e+00 4.19807285e-01 -6.01973943e-02
-2.39066198e-01 -1.14020252e+00 -1.06034465e-01 1.58431426e-01
3.40701222e-01 8.85416120e-02 1.02289486e+00 -5.58742583e-01
1.36758953e-01 -9.26280379e-01 2.76604563e-01 9.87871349e-01
7.07798958e-01 8.88413489e-02 -6.76898658e-01 -4.15847033e-01
6.69798970e-01 2.19049841e-01 4.28745061e-01 1.50049463e-01
6.12717628e-01 2.00855404e-01 -4.90577817e-01 -2.96909660e-01
1.33370149e+00 9.22655404e-01 7.44048655e-01 2.53166527e-01
4.77602392e-01 8.23881209e-01 9.44633961e-01 2.99077153e-01
-4.78728473e-01 5.23631215e-01 3.05537462e-01 -4.10137698e-02
-2.83668358e-02 5.48294842e-01 2.80317992e-01 8.62335861e-01
-1.96389332e-02 -4.19223793e-02 -6.36438549e-01 6.70419931e-01
-1.61087894e+00 -1.01886964e+00 -5.09963453e-01 1.86148548e+00
9.25796404e-02 9.78480652e-02 -5.84940277e-02 1.39919496e+00
8.34178627e-01 -4.44083661e-01 -3.40886772e-01 -5.86783230e-01
1.10973651e-02 1.83037564e-01 3.82003039e-01 1.68470174e-01
-8.63956571e-01 -7.56671727e-02 6.72944450e+00 7.48158038e-01
-1.35891759e+00 -1.35927647e-01 9.76045243e-03 4.07303780e-01
3.16563576e-01 -2.47744054e-01 -2.10169286e-01 6.09396338e-01
1.27876759e+00 4.62232344e-02 1.81007519e-01 5.82547486e-01
5.28377175e-01 -6.41385138e-01 -5.36143959e-01 5.34380615e-01
8.90353248e-02 -8.33970904e-01 -3.56444240e-01 1.31520882e-01
5.53128839e-01 -2.12124363e-01 -2.43162841e-01 -3.08994353e-01
-5.37018299e-01 -4.32304770e-01 4.50158328e-01 5.02259552e-01
3.12664539e-01 -1.12400436e+00 1.16447091e+00 1.83959052e-01
-7.84015000e-01 -4.62164640e-01 -8.91667455e-02 -3.09968174e-01
2.62379825e-01 9.47219491e-01 -7.35791504e-01 8.89350712e-01
3.04997236e-01 7.68102229e-01 -2.46471405e-01 1.01565754e+00
-2.25393891e-01 6.79943264e-01 9.19055939e-02 1.30999029e-01
-8.49488080e-02 -3.36844981e-01 5.29965580e-01 8.68424237e-01
3.25801641e-01 -4.14652199e-01 -4.57414478e-01 5.69641769e-01
7.11124063e-01 -1.08432367e-01 -9.51343417e-01 -3.04058313e-01
3.47424120e-01 1.36506033e+00 -1.00380325e+00 -1.30849645e-01
-1.87521324e-01 8.72198939e-01 -4.08599049e-01 1.83853418e-01
-1.91722706e-01 -9.85890746e-01 4.98926014e-01 6.33839220e-02
4.42474931e-01 -3.09863001e-01 -1.02327131e-01 -3.11878979e-01
1.38267269e-02 -6.97582901e-01 2.07941502e-01 -6.47295415e-01
-9.30733919e-01 2.97482729e-01 4.06689018e-01 -1.09707475e+00
-6.95618331e-01 -1.15609264e+00 -1.00682104e+00 1.17022634e+00
-1.08372712e+00 -4.18621123e-01 2.79255450e-01 3.59137595e-01
4.75543261e-01 -4.82730836e-01 8.45396042e-01 2.65496701e-01
-6.90814734e-01 -4.61959839e-01 5.58393121e-01 -2.21412957e-01
-9.18649212e-02 -1.15210497e+00 6.21323027e-02 1.06857789e+00
-3.30776870e-01 1.72430083e-01 8.70868742e-01 -5.43728411e-01
-1.60990918e+00 -5.56284070e-01 6.22989714e-01 3.00058052e-02
7.28859425e-01 2.53147572e-01 -9.07216072e-01 1.64112095e-02
2.02418640e-01 -3.86061728e-01 2.17475370e-01 -2.73489594e-01
8.44209254e-01 -7.82823637e-02 -1.25975883e+00 -5.49977757e-02
1.15740657e-01 -6.34175658e-01 -7.47730970e-01 3.15527678e-01
-1.77681372e-01 2.28584170e-01 -1.17097569e+00 5.98488271e-01
1.10855974e-01 -6.10519111e-01 5.55773020e-01 -5.16847037e-02
-1.57573581e-01 -4.24015045e-01 6.22699618e-01 -1.55995202e+00
-4.64539945e-01 -3.93047690e-01 -1.07195929e-01 8.08358133e-01
-6.15855195e-02 -8.06405246e-01 3.09935480e-01 -7.46571645e-02
-4.23663586e-01 -6.80002093e-01 -1.17472959e+00 -6.92868233e-01
-2.46202245e-01 -2.87923902e-01 9.85355079e-02 9.34588730e-01
6.66102231e-01 1.11053057e-01 2.71724492e-01 2.10790008e-01
1.99463859e-01 -1.40576020e-01 -1.13903411e-01 -1.42551732e+00
1.81015313e-01 -5.25665164e-01 -1.12849438e+00 3.07585038e-02
2.58043945e-01 -5.00972033e-01 2.02002794e-01 -1.62069559e+00
-5.20205796e-01 -1.22969113e-01 -5.20141006e-01 1.65121928e-01
1.53266698e-01 1.02596208e-01 -3.83794725e-01 2.71796882e-01
1.16743498e-01 2.19658628e-01 8.87100399e-01 5.98907471e-02
2.09341466e-01 3.20748121e-01 -1.02782257e-01 4.75362003e-01
9.28347826e-01 -2.05174968e-01 -7.07392335e-01 3.05796295e-01
-1.22122876e-01 4.67078499e-02 2.31729701e-01 -1.33345318e+00
-2.03029111e-01 2.11269096e-01 2.14527711e-01 -7.05551982e-01
-1.82834685e-01 -9.58195329e-01 3.03192437e-01 8.84309769e-01
3.60638767e-01 5.03957272e-01 2.13812739e-01 2.31814682e-01
-4.81336534e-01 -7.98548758e-01 4.60311085e-01 1.05171070e-01
-1.16351545e+00 -5.79070687e-01 -1.26050580e+00 -8.95437360e-01
1.37118065e+00 -2.42553011e-01 -2.42604926e-01 2.30872765e-01
-5.62710702e-01 -2.05365658e-01 1.95802510e-01 5.98260283e-01
7.17423201e-01 -1.00991416e+00 -3.50122064e-01 4.86987263e-01
-1.25552669e-01 -4.05796975e-01 4.32586998e-01 1.18230343e+00
-6.65888071e-01 7.10492432e-01 -5.90753675e-01 -3.22123975e-01
-1.12742436e+00 7.00335085e-01 2.88625807e-01 8.18664432e-02
-7.05288887e-01 2.82819510e-01 -4.46945250e-01 3.89506549e-01
-5.15073119e-03 -9.69630107e-02 -7.99813092e-01 2.76538610e-01
4.62961555e-01 1.00929916e+00 9.06195998e-01 -9.00116205e-01
-3.02085638e-01 3.02814871e-01 2.22224146e-01 9.07849446e-02
1.02986443e+00 -4.24676351e-02 -5.14490724e-01 7.41835833e-01
1.00138533e+00 -5.64578474e-01 -8.22600901e-01 3.63214999e-01
3.57896864e-01 -5.66421263e-02 6.07681990e-01 -7.59462655e-01
-8.86664450e-01 6.92961752e-01 1.08164656e+00 6.71141088e-01
1.05646729e+00 -3.03263903e-01 2.82923818e-01 3.67148608e-01
3.34277183e-01 -1.30887890e+00 -1.72494024e-01 8.76261294e-02
7.41905034e-01 -6.09712660e-01 1.83089748e-02 -5.72898649e-02
-4.24882025e-01 1.46890628e+00 -4.82413620e-02 -1.85338601e-01
7.72873878e-01 3.37663800e-01 -7.07413582e-03 -5.28404057e-01
-5.07915676e-01 -1.32053778e-01 2.26549909e-01 7.38092005e-01
4.19693410e-01 1.07120499e-01 -7.81063378e-01 -6.04775771e-02
5.23269892e-01 -7.90451020e-02 3.87782454e-01 1.47661936e+00
-5.80222607e-01 -1.01467359e+00 -6.97407961e-01 4.85194474e-01
-3.63703370e-01 5.52660882e-01 -5.12002148e-02 6.83249652e-01
1.16274595e-01 1.20020628e+00 1.04807682e-01 -4.80899006e-01
5.35025418e-01 3.27844799e-01 5.03894866e-01 -2.56589830e-01
-1.35832325e-01 -2.62327492e-01 -1.09118177e-03 -2.58254617e-01
-3.28764528e-01 -1.04062498e+00 -1.32111299e+00 1.11045733e-01
-8.17114890e-01 5.99258006e-01 1.33333957e+00 1.25951135e+00
1.70904383e-01 1.01568031e+00 1.15718603e+00 -8.17634404e-01
-6.53964341e-01 -1.49486601e+00 -1.19727182e+00 3.66613746e-01
5.96142113e-02 -1.04171789e+00 -6.13870978e-01 -6.88689277e-02] | [6.676941871643066, 2.414522171020508] |
9f4886e6-033a-46a9-86bc-a5a549c2b81f | autonomous-sputter-synthesis-of-thin-film | 2305.11122 | null | https://arxiv.org/abs/2305.11122v2 | https://arxiv.org/pdf/2305.11122v2.pdf | Autonomous sputter synthesis of thin film nitrides with composition controlled by Bayesian optimization of optical plasma emission | Autonomous experimentation has emerged as an efficient approach to accelerate the pace of materials discovery. Although instruments for autonomous synthesis have become popular in molecular and polymer science, solution processing of hybrid materials and nanoparticles, examples of autonomous tools for physical vapour deposition are scarce yet important for the semiconductor industry. Here, we report the design and implementation of an autonomous instrument for sputter deposition of thin films with controlled composition, leveraging a highly automated sputtering reactor custom-controlled by Python, optical emission spectroscopy (OES), and Bayesian optimization algorithm. We modeled film composition, measured by x-ray fluorescence, as a linear function of emission lines monitored during the co-sputtering from elemental Zn and Ti targets in N$_2$ atmosphere. A Bayesian control algorithm, informed by OES, navigates the space of sputtering power to fabricate films with user-defined composition, by minimizing the absolute error between desired and measured emission signals. We validated our approach by autonomously fabricating Zn$_x$Ti$_{1-x}$N$_y$ films with deviations from the targeted cation composition within relative 3.5 %, even for 15 nm thin films, demonstrating that the proposed approach can reliably synthesize thin films with specific composition and minimal human interference. Moreover, the proposed method can be extended to more difficult synthesis experiments where plasma intensity depends non-linearly on pressure, or the elemental sticking coefficients strongly depend on the substrate temperature. | ['Andriy Zakutayev', 'Rebecca W. Smaha', 'John S. Mangum', 'Sage R. Bauers', 'Stephen Schaefer', 'Kendal Johnson', 'Kevin R. Talley', 'Davi M. Febba'] | 2023-05-18 | null | null | null | null | ['bayesian-optimization'] | ['methodology'] | [ 5.12310505e-01 5.96836880e-02 -3.54632661e-02 -3.39422256e-01
-2.24692330e-01 -3.30461472e-01 3.57218295e-01 8.73749107e-02
-5.29099345e-01 1.01848233e+00 -8.59431684e-01 -1.40675351e-01
-1.76847398e-01 -8.37313771e-01 -6.96206272e-01 -1.13031638e+00
3.46794188e-01 1.19328105e+00 4.22267497e-01 -7.50968456e-02
2.88440168e-01 9.07382488e-01 -1.70656705e+00 6.09128326e-02
8.22162569e-01 1.07255423e+00 6.12549543e-01 5.03327906e-01
-1.47815198e-01 -2.73179673e-02 -6.59465551e-01 6.38314486e-02
1.75668746e-01 -2.87574291e-01 -8.30135345e-02 -1.46284804e-01
1.08639058e-02 3.62370968e-01 3.99183929e-01 1.11890209e+00
1.41336545e-01 -1.83049649e-01 1.07172775e+00 -6.19956613e-01
-4.37950939e-01 8.79514813e-01 -3.62814844e-01 -1.67839736e-01
8.89283270e-02 5.76604366e-01 5.98608136e-01 -5.66064596e-01
7.19904423e-01 8.54565859e-01 1.02408074e-01 8.29384089e-01
-1.41477418e+00 -9.18283939e-01 -1.25430763e-01 -1.28536716e-01
-1.33331406e+00 -4.47864473e-01 5.56674182e-01 -7.27515519e-01
1.12724805e+00 4.19776082e-01 7.70547271e-01 1.03104639e+00
5.19720912e-01 -3.03548247e-01 1.29318595e+00 -6.38600826e-01
6.15046382e-01 8.50061178e-01 -2.75519282e-01 4.71641421e-01
3.92926782e-01 2.41652757e-01 -2.86506712e-01 -3.03962324e-02
5.16427159e-01 -2.14139819e-01 -1.12012446e-01 -1.98867664e-01
-7.55220234e-01 5.87908089e-01 1.29777595e-01 4.46790189e-01
-6.57586694e-01 1.52661502e-01 -7.99469128e-02 1.32508233e-01
3.50037038e-01 1.04787207e+00 -3.20733637e-01 2.17349697e-02
-9.29586947e-01 3.31335187e-01 9.35917974e-01 8.65913153e-01
7.52871037e-01 1.50398806e-01 7.42562860e-02 5.93109667e-01
6.35799587e-01 1.06544447e+00 7.18719140e-02 -6.39001966e-01
-1.98886976e-01 4.16781873e-01 5.19293725e-01 -2.72470653e-01
-2.45512843e-01 2.33850762e-01 -3.49536836e-01 5.37178576e-01
3.79710615e-01 -2.62019008e-01 -8.58169496e-01 1.32002640e+00
8.16728055e-01 -6.04998708e-01 6.28318591e-03 9.28261936e-01
3.48565549e-01 8.52239132e-01 -3.14243399e-02 -1.00276804e+00
1.25129807e+00 -4.50549096e-01 -7.70130992e-01 2.97721386e-01
-2.79740214e-01 -7.19443977e-01 1.06957901e+00 7.63833880e-01
-1.33730626e+00 -2.13609770e-01 -1.36098504e+00 4.04968590e-01
-2.92987615e-01 -8.07881132e-02 6.68434799e-01 1.05862045e+00
-5.44042766e-01 1.06866539e+00 -1.01907361e+00 -1.16733797e-01
3.61168236e-01 6.72349632e-01 8.14342648e-02 2.62946039e-01
-8.46213162e-01 6.28782749e-01 4.32005599e-02 1.08917607e-02
-9.52330291e-01 -1.09722292e+00 -1.77567959e-01 -2.30026454e-01
3.12507212e-01 -6.39776409e-01 1.09948039e+00 -5.19807816e-01
-2.79147243e+00 6.99626327e-01 -1.81139246e-01 -3.71049851e-01
4.58310992e-01 -6.02265857e-02 -4.62234616e-01 1.08838119e-01
7.83983991e-02 2.78294474e-01 1.19169736e+00 -1.07904625e+00
-3.22940409e-01 -2.89143264e-01 -5.69506109e-01 -2.53633797e-01
-1.17784135e-01 3.74236077e-01 3.42536151e-01 3.80781919e-01
-1.18257534e-02 -7.38442421e-01 -4.98862594e-01 -1.03358716e-01
-4.90296930e-01 -3.92205656e-01 6.59547865e-01 1.52404398e-01
6.30434871e-01 -1.77672124e+00 3.28424722e-01 6.21701360e-01
-3.98832560e-02 6.20843992e-02 5.33322871e-01 5.69784999e-01
2.18965083e-01 -2.76671699e-03 -4.89053279e-01 3.76389250e-02
1.92539275e-01 -2.39359558e-01 -1.20626949e-01 6.87080801e-01
-1.91923343e-02 3.25382203e-01 -7.98824072e-01 -8.79974738e-02
3.42675000e-01 3.36062610e-01 -4.18576598e-01 2.09953114e-01
-1.00857413e+00 4.52379555e-01 -2.71329015e-01 6.66826963e-01
5.35364211e-01 -1.17004342e-01 2.90873975e-01 -4.91329879e-02
-1.05613387e+00 2.32764676e-01 -9.44894373e-01 1.44344544e+00
-2.89856076e-01 8.57855231e-02 7.41789699e-01 -4.06967431e-01
1.25582659e+00 3.14081609e-01 5.42283356e-01 -7.21159279e-01
4.82942909e-01 5.24278164e-01 3.75616014e-01 -5.12619972e-01
3.81712228e-01 -9.16593909e-01 3.06692004e-01 4.24670488e-01
-1.21701799e-01 -1.12009048e+00 2.55831420e-01 -2.29953378e-01
9.51744795e-01 -9.84941423e-02 -3.50840203e-02 -8.28515470e-01
5.31961858e-01 9.29734707e-02 1.18625917e-01 8.21228564e-01
1.68802470e-01 1.63367435e-01 1.94113061e-01 -2.25562736e-01
-1.55342484e+00 -1.00488949e+00 -6.87757432e-01 6.19156957e-01
3.09773237e-01 -3.56685102e-01 -8.21536541e-01 1.16001755e-01
5.83355408e-03 6.92009509e-01 -2.35146582e-01 -3.86578850e-02
-3.42765927e-01 -9.10688758e-01 -4.46171314e-02 -2.01282263e-01
-1.13912232e-01 -1.32045400e+00 -6.92238986e-01 4.09480989e-01
8.93116772e-01 -1.11418808e+00 4.21165794e-01 2.47651160e-01
-2.97429591e-01 -8.76129568e-01 -2.77828544e-01 1.21966064e-01
5.37644982e-01 -2.90363282e-01 7.68888533e-01 -1.00572057e-01
-1.16783690e+00 3.82129610e-01 -5.19611984e-02 -9.69712555e-01
-9.16904211e-01 7.59279951e-02 6.08038783e-01 -2.68401243e-02
5.73365577e-02 -9.40838099e-01 -6.94590449e-01 3.25879186e-01
-5.62467694e-01 -2.71555223e-03 3.99860471e-01 1.80597514e-01
8.84168863e-01 2.29608670e-01 5.04825830e-01 -1.13304293e+00
5.22343278e-01 -2.22629756e-01 -1.24747205e+00 -7.42048351e-03
-6.52417958e-01 -1.42095581e-01 7.48690009e-01 -3.32414359e-01
-9.03259933e-01 -1.75270915e-01 -2.23490730e-01 -3.27663064e-01
-4.18825984e-01 1.04655541e-01 -3.24624926e-01 -8.22997317e-02
1.03215969e+00 -8.92925858e-02 1.32687882e-01 -3.02162636e-02
1.32370159e-01 6.64359093e-01 1.00450680e-01 -1.00911915e+00
6.82353377e-01 6.67780936e-01 2.72011757e-01 -1.44657457e+00
-4.07084376e-01 4.20558155e-02 -4.09161270e-01 -3.52757752e-01
1.10231102e+00 -4.29768920e-01 -1.35111701e+00 1.41239598e-01
-8.22569311e-01 -5.40605605e-01 -4.85840678e-01 4.65815514e-01
-4.18592960e-01 -2.20878813e-02 -4.38942432e-01 -1.27671003e+00
-7.75300264e-01 -1.45145535e+00 8.30491364e-01 2.43112072e-01
-3.97795886e-01 -4.66596365e-01 4.63007949e-02 1.48234323e-01
7.49703169e-01 3.86504501e-01 1.04308283e+00 -7.13879988e-02
-7.70499527e-01 -3.17077756e-01 2.44869813e-01 4.76704650e-02
4.29064840e-01 7.46483147e-01 -7.01035917e-01 -1.15929969e-01
3.70906413e-01 -1.03812143e-01 4.57884371e-01 4.47383344e-01
8.60980749e-01 1.01400189e-01 -5.98069906e-01 2.52000332e-01
1.28602302e+00 4.46688414e-01 3.33365142e-01 1.88363001e-01
6.30175889e-01 3.75246048e-01 5.47322810e-01 6.21377289e-01
-4.73034948e-01 7.18189538e-01 5.16462862e-01 5.13207614e-01
2.81348199e-01 3.14068377e-01 3.41667533e-01 4.41404164e-01
-2.27789059e-01 -1.13348193e-01 -6.11381769e-01 -5.11297919e-02
-1.24822307e+00 -7.37508833e-01 -4.42409098e-01 2.32726622e+00
1.06918240e+00 2.39647642e-01 1.74689800e-01 -7.96424001e-02
7.71943688e-01 -3.69572878e-01 -7.31847286e-01 -5.60245752e-01
2.55001396e-01 1.13592231e+00 6.77312791e-01 5.85642159e-01
-4.38471437e-01 1.04930425e+00 6.45263004e+00 4.27035451e-01
-1.69124937e+00 -2.16876213e-02 -2.97982335e-01 -5.40014863e-01
-5.53726494e-01 3.72687951e-02 -1.19234431e+00 7.69678831e-01
1.10188007e+00 -2.18093656e-02 5.23322225e-01 7.76677668e-01
2.57477254e-01 -2.05686569e-01 -1.23422062e+00 5.96969366e-01
-4.59826887e-01 -1.45592284e+00 -2.46862337e-01 6.45684376e-02
5.71497917e-01 -1.08850472e-01 8.16569403e-02 -2.70976871e-01
1.80672452e-01 -7.47456253e-01 9.13283646e-01 5.44810057e-01
8.77228856e-01 -6.00231409e-01 -1.70615003e-01 3.45426261e-01
-3.49765301e-01 -6.94297906e-03 -6.02677166e-01 -5.94430640e-02
4.99225467e-01 1.17944884e+00 -1.23683512e+00 2.75294602e-01
8.49929154e-01 2.25332752e-01 4.68480140e-01 3.53764534e-01
-2.50078440e-01 4.30088669e-01 -8.19899023e-01 -5.83340287e-01
-1.82601511e-01 -8.38600993e-01 6.92815363e-01 7.85040617e-01
3.89596492e-01 4.52573180e-01 -1.71154857e-01 1.55951416e+00
1.80780783e-01 7.91193321e-02 -2.96287864e-01 -2.81102329e-01
6.73491716e-01 1.28542376e+00 -9.38394606e-01 -1.24985345e-01
3.06970119e-01 3.09590966e-01 5.76256588e-02 -1.77409686e-02
-8.77872229e-01 -2.18906015e-01 8.54467392e-01 6.92040980e-01
3.66336942e-01 -3.72545123e-01 -1.53475672e-01 -6.80180430e-01
-2.52974153e-01 -5.64515769e-01 -2.79985458e-01 -4.24512774e-01
-1.08998990e+00 4.92136300e-01 5.73016256e-02 -2.78272748e-01
1.48060396e-01 -1.00183892e+00 -4.45948452e-01 7.54851460e-01
-1.25580716e+00 -2.48533770e-01 -8.27119276e-02 2.55728573e-01
2.30848178e-01 -1.76550746e-01 8.58959913e-01 1.12811975e-01
-7.86580384e-01 1.70179814e-01 8.71074125e-02 -9.92466629e-01
5.65810859e-01 -1.02062643e+00 -6.96612895e-02 4.85207498e-01
-1.72074124e-01 5.05558610e-01 1.24792862e+00 -8.00435007e-01
-1.81502128e+00 -7.32755363e-01 -7.65764043e-02 -3.44138861e-01
7.28039861e-01 -7.40567327e-01 -5.17825425e-01 1.44658625e-01
2.01702490e-01 -2.42128819e-01 6.58384442e-01 -2.35541090e-01
1.17741391e-01 -1.94596335e-01 -1.53004289e+00 5.37816584e-01
8.56262505e-01 -1.21222250e-01 1.51147604e-01 8.96206558e-01
7.03689456e-01 -5.65162539e-01 -1.07966924e+00 6.32064104e-01
4.12335545e-01 -9.23258543e-01 7.72211611e-01 -3.84393364e-01
3.56861860e-01 -2.96722293e-01 -2.29543313e-01 -8.35882366e-01
-1.65188953e-01 -7.33720243e-01 2.18542352e-01 1.01013756e+00
7.76310861e-01 -6.20845735e-01 8.58021140e-01 1.04110670e+00
-4.82123494e-01 -5.05137086e-01 -1.01223695e+00 -5.38409710e-01
-1.05316684e-01 -3.29134226e-01 5.69171727e-01 4.35374945e-01
1.22641183e-01 6.17105141e-02 -1.01997532e-01 5.86871624e-01
6.10931873e-01 1.33509055e-01 7.18742669e-01 -1.37803638e+00
-3.50573719e-01 -3.46488684e-01 -8.63660127e-02 -3.39828372e-01
3.04948419e-01 -8.14431071e-01 3.31778646e-01 -1.01695645e+00
-4.51654196e-01 -1.03976393e+00 5.59675656e-02 -5.02735786e-02
4.85386580e-01 -1.81619704e-01 -1.04581803e-01 5.39154820e-02
-2.22256079e-01 5.20540118e-01 1.13576484e+00 -8.64836276e-02
-3.91823351e-01 8.44369382e-02 -2.56109893e-01 3.39847237e-01
5.42597711e-01 -8.26522291e-01 -2.47627780e-01 -3.53883067e-03
4.93698269e-01 4.21181927e-03 -2.35086814e-01 -1.28965962e+00
2.82199144e-01 -4.70559984e-01 -7.41549060e-02 2.13801432e-02
6.35965765e-01 -1.11504626e+00 6.35681927e-01 3.88167948e-01
-3.46638225e-02 -6.11232102e-01 3.25922109e-02 2.68366843e-01
4.10618722e-01 -6.47191823e-01 9.78985071e-01 -4.44182158e-01
3.25404227e-01 3.45853090e-01 -6.59411669e-01 -7.53390729e-01
1.29619300e+00 -9.83484462e-02 -1.07861251e-01 4.50884402e-01
-6.58402383e-01 -8.50191712e-02 7.68376291e-01 -8.34812820e-02
3.36880863e-01 -7.49716163e-01 -1.99568912e-01 3.83695781e-01
-3.45526040e-01 3.99042934e-01 3.93480211e-01 7.30469465e-01
-8.32128942e-01 2.31775016e-01 -2.61680484e-02 -1.12735605e+00
-9.99703825e-01 4.79241461e-01 4.28106576e-01 1.46293119e-01
-4.39155012e-01 1.01852250e+00 -7.08850101e-02 -1.08107962e-01
-2.74541140e-01 -4.03823197e-01 2.29768366e-01 -3.20310920e-01
4.88996744e-01 2.12650508e-01 3.96498948e-01 8.75160918e-02
-1.23368435e-01 3.79529923e-01 -4.17201042e-01 -2.33709127e-01
1.48567355e+00 3.13373566e-01 -4.31734145e-01 5.87187767e-01
4.26900148e-01 1.63567007e-01 -1.08973253e+00 1.84424669e-01
-3.80586207e-01 -2.39020959e-01 -2.85989225e-01 -5.08932054e-01
-6.99974597e-01 4.02606457e-01 3.58416587e-01 4.08449203e-01
3.95359993e-01 1.86087728e-01 3.04292649e-01 3.61676961e-01
5.69476962e-01 -1.10748506e+00 -2.07230583e-01 1.48042470e-01
7.12778568e-01 -7.93971419e-01 3.31572652e-01 -6.74287379e-01
-5.79827204e-02 1.17783952e+00 7.03402102e-01 -2.42472082e-01
6.52493536e-01 5.79070508e-01 -2.35484373e-02 -7.84680128e-01
-8.06098402e-01 4.39228803e-01 -4.38626647e-01 4.21200067e-01
2.63757944e-01 2.86031067e-01 -2.15971991e-01 4.06425118e-01
-5.13250589e-01 -1.80596367e-01 4.72651899e-01 8.24094415e-01
-6.42942250e-01 -1.25965822e+00 -3.85602713e-01 3.66348505e-01
-2.42543310e-01 1.08134024e-01 -2.83885270e-01 5.27115285e-01
5.88953197e-02 6.38434291e-01 1.33171692e-01 1.54812664e-01
4.62395489e-01 1.53569788e-01 8.94162118e-01 -8.51626039e-01
-3.32420558e-01 -9.50185359e-02 -1.82004482e-01 -3.27287048e-01
-5.35049319e-01 -7.73807585e-01 -1.32678056e+00 4.77146264e-03
-5.91625392e-01 5.89010060e-01 1.32979023e+00 9.32169259e-01
1.90286666e-01 5.05939066e-01 6.51464522e-01 -1.20616961e+00
-2.55286306e-01 -7.86065876e-01 -1.01084232e+00 -2.11544514e-01
-1.34250239e-01 -8.99016261e-01 -5.54869533e-01 -5.92295527e-01] | [5.39114236831665, 4.942857265472412] |
9ca49e48-2542-4b2e-8f99-74f1f9f87850 | sub-policy-adaptation-for-hierarchical-1 | null | null | https://openreview.net/forum?id=rkfJB9Sj2V | https://openreview.net/pdf?id=rkfJB9Sj2V | Sub-policy Adaptation for Hierarchical Reinforcement Learning | Hierarchical Reinforcement Learning is a promising approach to long-horizon decision-making problems with sparse rewards. Unfortunately, most methods still decouple the lower-level skill acquisition process and the training of a higher level that controls the skills in a new task. Treating the skills as fixed can lead to significant sub-optimality in the transfer setting. In this work, we propose a novel algorithm to discover a set of skills, and continuously adapt them along with the higher level even when training on a new task. Our main contributions are two-fold. First, we derive a new hierarchical policy gradient, as well as an unbiased latent-dependent baseline. We introduce Hierarchical Proximal Policy Optimization (HiPPO), an on-policy method to efficiently train all levels of the hierarchy simultaneously. Second, we propose a method of training time-abstractions that improves the robustness of the obtained skills to environment changes. Code and results are available at sites.google.com/view/hippo-rl. | ['Anonymous'] | 2019-05-16 | null | null | null | icml-workshop-amtl-2019-6 | ['hierarchical-reinforcement-learning'] | ['methodology'] | [ 1.99879751e-01 1.73469204e-02 -2.51122266e-01 -1.11813851e-01
-9.99460757e-01 -7.22258151e-01 3.79412919e-01 5.96939810e-02
-6.78394437e-01 1.07396722e+00 1.59654185e-01 -3.23859096e-01
-3.91179889e-01 -5.52127481e-01 -9.46541667e-01 -9.24984217e-01
-1.84489995e-01 6.07183456e-01 3.28644067e-01 -3.45236391e-01
3.57338995e-01 4.04497653e-01 -1.40705788e+00 -3.30661871e-02
1.07999992e+00 8.85074914e-01 4.84673560e-01 7.51040697e-01
3.79128963e-01 9.95035172e-01 -2.44011134e-01 1.35862064e-02
6.13543332e-01 -3.79395783e-01 -9.84341562e-01 7.87328854e-02
3.24380606e-01 -4.98534858e-01 -2.60587156e-01 1.06698108e+00
5.50146580e-01 6.15838706e-01 3.75803292e-01 -9.19968367e-01
-2.77303010e-01 5.89574039e-01 -2.91652590e-01 1.48094460e-01
-1.49216084e-03 4.00615782e-01 9.35847461e-01 -4.46226716e-01
4.09271240e-01 1.23060441e+00 4.44686204e-01 5.85181713e-01
-1.25113308e+00 -5.07573843e-01 6.90053880e-01 7.07461834e-02
-8.42204154e-01 -2.54669487e-01 5.03641903e-01 -6.59254193e-01
6.54266179e-01 -1.95898175e-01 7.71272480e-01 1.25393999e+00
2.35239655e-01 9.14511740e-01 1.76833785e+00 -4.07076716e-01
3.86176288e-01 -8.38330239e-02 -4.20047417e-02 8.78137350e-01
-1.35722116e-01 7.69332409e-01 -6.49378598e-01 -6.96556717e-02
8.97794902e-01 -1.74608052e-01 -2.35899203e-02 -7.67483473e-01
-1.12824047e+00 8.18069696e-01 4.16205615e-01 1.26847222e-01
-7.01609910e-01 3.81924927e-01 2.60848075e-01 5.69454730e-01
2.79333591e-01 8.43137324e-01 -6.60673797e-01 -3.04986417e-01
-7.48983622e-01 5.61286271e-01 5.88196278e-01 7.87466884e-01
7.66334116e-01 1.12941861e-01 -6.94756031e-01 5.78214288e-01
-1.27661183e-01 3.32874238e-01 4.35650587e-01 -1.36987925e+00
4.78317797e-01 -1.57476049e-02 6.15965545e-01 -4.81415540e-01
-3.93480092e-01 -7.33836353e-01 -3.02700728e-01 4.68612373e-01
5.12614965e-01 -5.32915831e-01 -8.28693688e-01 1.99077940e+00
5.52522242e-01 1.79906234e-01 -6.87821060e-02 8.20536017e-01
-2.19953045e-01 4.45066780e-01 4.75922599e-02 -2.61421770e-01
9.87045527e-01 -1.39902401e+00 -5.27710438e-01 -4.47826862e-01
3.25371087e-01 -3.55812192e-01 1.36614311e+00 5.92012286e-01
-1.43213010e+00 -7.09472835e-01 -8.65726769e-01 6.63834810e-02
-1.24824822e-01 -1.14154583e-02 7.54507780e-01 1.29171014e-01
-1.21511769e+00 1.13922846e+00 -9.56515551e-01 1.74716045e-03
2.86160469e-01 4.65972990e-01 2.41743103e-01 2.02422947e-01
-1.35694289e+00 9.96643841e-01 5.49761057e-01 -1.40209869e-01
-1.72681296e+00 -4.74960685e-01 -6.48219824e-01 2.06620812e-01
1.07845640e+00 -6.37288451e-01 1.92686391e+00 -9.69064772e-01
-2.11541319e+00 3.89531493e-01 8.43388289e-02 -4.91838396e-01
6.56071723e-01 -5.92186153e-01 3.56729507e-01 -3.99285071e-02
2.63579935e-01 6.39529824e-01 1.05060971e+00 -1.19826233e+00
-1.03374720e+00 -3.72291297e-01 4.17689979e-01 6.75014138e-01
-3.19886327e-01 -2.65313983e-01 -2.55785197e-01 -4.92804199e-01
-2.83068091e-01 -1.10818553e+00 -5.56830347e-01 -4.80870277e-01
-4.06411327e-02 -2.68523604e-01 -1.41123636e-02 -6.66883051e-01
1.09997308e+00 -1.81084514e+00 7.96095490e-01 -8.23028106e-03
2.20768340e-03 1.55183300e-01 -1.92301095e-01 3.02173913e-01
2.04780385e-01 -3.19557339e-01 -2.36834809e-01 -4.03564572e-01
1.20159827e-01 3.00006121e-01 -4.02132720e-01 1.86234608e-01
-6.67628199e-02 9.74477649e-01 -1.35673618e+00 -2.58955479e-01
-7.10864458e-03 -6.07418083e-02 -6.69296324e-01 3.48445207e-01
-4.48294073e-01 1.02167976e+00 -7.54654706e-01 4.08450246e-01
3.29435617e-02 -1.62233979e-01 2.62550175e-01 5.04372239e-01
-3.43262494e-01 3.71889800e-01 -9.81838405e-01 1.88743985e+00
-7.08160579e-01 -1.00123674e-01 1.64569482e-01 -1.01890624e+00
5.30414999e-01 2.42871746e-01 5.67246258e-01 -6.95285141e-01
8.76167193e-02 1.14163384e-01 -5.85145280e-02 -4.16599959e-01
5.87037623e-01 -8.06084722e-02 -2.05051631e-01 1.54333353e-01
2.26856992e-01 -1.07870519e-01 2.82542467e-01 -2.30110227e-03
9.06211913e-01 8.28048766e-01 3.53939086e-01 -2.37739310e-01
4.03809071e-01 3.52401286e-02 6.87604249e-01 1.19697273e+00
-2.92784184e-01 -1.48119792e-01 7.17425525e-01 -2.75006413e-01
-9.03487146e-01 -9.34324801e-01 1.97879285e-01 1.67159247e+00
-5.60618527e-02 -3.01199015e-02 -7.10206985e-01 -7.79095054e-01
2.39606693e-01 6.21354520e-01 -7.74229407e-01 -8.40760991e-02
-6.90539539e-01 -3.74387473e-01 1.67475283e-01 5.02356708e-01
3.77334058e-01 -1.33011544e+00 -8.78710210e-01 2.44412065e-01
-5.91963418e-02 -9.00571883e-01 -5.55475116e-01 5.55100203e-01
-1.13193917e+00 -7.69168139e-01 -8.33173394e-01 -4.77205724e-01
5.48729599e-01 -5.54877780e-02 1.13308358e+00 -1.26555860e-01
2.47424155e-01 4.65454012e-01 -2.24965632e-01 -3.06642234e-01
-2.76965827e-01 4.50927973e-01 3.22118491e-01 -1.51790127e-01
-1.81207448e-01 -6.55264735e-01 -4.82969522e-01 6.80814236e-02
-4.03605282e-01 1.33963361e-01 7.85382092e-01 1.03073239e+00
7.03712583e-01 -8.99325311e-02 5.40207088e-01 -9.12689388e-01
8.51768374e-01 -3.04301888e-01 -1.26102245e+00 1.54767454e-01
-9.36693847e-01 3.99929076e-01 7.09125519e-01 -4.93093282e-01
-1.07079339e+00 3.28788787e-01 6.00969866e-02 -5.05642295e-01
-6.71357214e-02 3.91066968e-01 1.54440239e-01 3.67250927e-02
7.50348568e-01 2.35028908e-01 -2.76333958e-01 -5.00246227e-01
4.13444012e-01 1.67387411e-01 4.60291684e-01 -1.23262763e+00
7.59720325e-01 1.32382005e-01 2.32896022e-02 -2.96430379e-01
-1.38166416e+00 -5.58377802e-01 -8.03886294e-01 -1.92889571e-01
6.77502871e-01 -9.57951725e-01 -8.86667311e-01 3.53163660e-01
-7.22694874e-01 -1.30776441e+00 -5.70311904e-01 3.61481905e-01
-1.13340855e+00 6.29267618e-02 -5.07875264e-01 -8.14067245e-01
-1.75506085e-01 -1.24911177e+00 1.02146828e+00 2.98850060e-01
1.70375675e-01 -9.01582956e-01 3.69782150e-01 3.33070546e-01
2.59860188e-01 8.23781863e-02 5.55551350e-01 -2.31304407e-01
-5.23665905e-01 3.93179417e-01 2.19341069e-01 2.98412412e-01
-1.84457272e-01 -6.85542285e-01 -8.00832033e-01 -8.45057130e-01
1.52460590e-01 -8.86233628e-01 9.14421916e-01 5.20595789e-01
1.18804431e+00 -4.30818230e-01 -3.61885168e-02 6.69657886e-01
1.19999778e+00 3.53004634e-01 1.88783541e-01 7.08368778e-01
6.08002186e-01 6.24247193e-01 9.73824382e-01 4.88961518e-01
1.87166899e-01 8.27274561e-01 3.71345252e-01 2.06386775e-01
2.18579516e-01 -4.99874979e-01 6.78766489e-01 6.34832323e-01
-3.32706273e-01 3.00900072e-01 -6.09155774e-01 5.78207195e-01
-2.16520858e+00 -8.94252121e-01 4.09276277e-01 2.35755682e+00
1.10163403e+00 1.89960077e-01 4.77137417e-01 -2.93908954e-01
3.32809597e-01 6.02462701e-03 -1.11835051e+00 -2.79575527e-01
5.06122410e-01 4.01614070e-01 7.44087398e-01 9.98739719e-01
-1.16992033e+00 1.40786481e+00 6.35233355e+00 6.43412888e-01
-9.54231918e-01 3.64520997e-01 2.77506799e-01 -3.47648710e-01
1.33930802e-01 1.22609856e-02 -9.96736944e-01 2.41177827e-01
9.03094471e-01 -1.63561821e-01 1.07564628e+00 9.94062543e-01
3.22581559e-01 -1.06448926e-01 -1.09261167e+00 4.62712020e-01
-4.43915576e-01 -8.72105956e-01 -3.99365336e-01 1.36070207e-01
1.10273623e+00 1.25044242e-01 2.32724845e-01 8.80624831e-01
8.50987852e-01 -8.03259194e-01 9.08957362e-01 4.89280730e-01
5.40208936e-01 -8.44948232e-01 3.04240495e-01 6.71339393e-01
-8.12937617e-01 -4.54019934e-01 -3.86587620e-01 -3.51172894e-01
-3.04805655e-02 1.13119997e-01 -5.14347970e-01 5.06642044e-01
5.62537193e-01 5.58866024e-01 -5.20466685e-01 7.70142615e-01
-8.51752818e-01 5.47741354e-01 4.75565679e-02 8.17157701e-02
6.02228284e-01 -2.63599753e-01 4.79289681e-01 8.66534293e-01
1.47351816e-01 3.25950794e-02 7.81471550e-01 6.61590099e-01
1.13310471e-01 -2.19147000e-03 -5.09286702e-01 -2.65778508e-02
3.20339352e-01 1.11504424e+00 -3.98152620e-01 -3.26174885e-01
-1.62652373e-01 1.14661252e+00 1.01660740e+00 4.84933853e-01
-7.36505210e-01 -3.07587273e-02 5.36109269e-01 -2.13638261e-01
3.88681740e-01 -4.16393906e-01 -6.22266494e-02 -1.16632199e+00
-3.79744638e-03 -1.19840157e+00 4.16412652e-01 -3.77729297e-01
-8.79147470e-01 2.67979980e-01 7.49801919e-02 -1.06347466e+00
-4.19755101e-01 -5.36744535e-01 -1.12972423e-01 8.33314478e-01
-1.75592160e+00 -6.91248834e-01 3.66354994e-02 6.85506880e-01
6.66108429e-01 -1.45937428e-01 6.91679537e-01 1.91588998e-02
-4.01349753e-01 4.35246885e-01 3.59197319e-01 -2.85007685e-01
5.78199804e-01 -1.67784774e+00 2.32668191e-01 8.46072972e-01
-2.97146570e-02 5.05436718e-01 6.10561609e-01 -5.97810686e-01
-1.02867579e+00 -8.38175595e-01 4.22450811e-01 -4.32029754e-01
8.95381093e-01 -2.91513503e-01 -7.40759730e-01 8.47302437e-01
8.45764056e-02 -5.09215772e-01 1.17852516e-01 5.14416754e-01
1.21281795e-01 -2.05137506e-02 -6.40730917e-01 6.66591763e-01
9.89945948e-01 -3.82814527e-01 -6.33457243e-01 6.22352183e-01
8.25929821e-01 -8.38867366e-01 -9.08238590e-01 1.19674332e-01
4.38382447e-01 -8.67291510e-01 9.59015369e-01 -8.51491451e-01
3.52393031e-01 -3.45695019e-02 1.42332196e-01 -1.66324055e+00
-6.41891420e-01 -1.15531909e+00 -4.01448131e-01 6.37381792e-01
2.17341900e-01 -5.87663949e-01 7.43888259e-01 1.48157045e-01
-2.60920018e-01 -1.01940227e+00 -6.49593949e-01 -1.09333611e+00
3.89148742e-01 -3.54491882e-02 2.51162648e-01 6.84365690e-01
4.13307436e-02 5.55428207e-01 -8.46722662e-01 3.13952476e-01
6.72435403e-01 2.20354512e-01 6.61430240e-01 -1.12245727e+00
-9.23014462e-01 -3.45944017e-01 3.78146112e-01 -1.39721715e+00
2.45658368e-01 -7.15445638e-01 4.60417181e-01 -1.35527599e+00
2.45103482e-02 -5.29141247e-01 -7.29681492e-01 6.02192938e-01
-3.57256830e-01 -6.23492658e-01 3.47944260e-01 3.28994572e-01
-6.93611562e-01 7.80349970e-01 1.79819608e+00 2.15287223e-01
-4.76951063e-01 3.52632314e-01 -7.95792401e-01 6.96112990e-01
1.18993843e+00 -7.22806513e-01 -6.77072108e-01 -4.77692336e-01
2.07341328e-01 3.70408535e-01 1.65367335e-01 -1.03420961e+00
8.48393440e-02 -5.35461485e-01 2.63773978e-01 -1.66147470e-01
3.42427760e-01 -5.49316645e-01 -4.75970894e-01 7.82143056e-01
-7.57274032e-01 1.80873051e-01 1.21186279e-01 6.49830699e-01
1.25974745e-01 -4.59396362e-01 9.81829643e-01 -4.06884134e-01
-6.86215222e-01 4.23243791e-01 -3.67843330e-01 1.85040146e-01
1.00105095e+00 2.83164829e-01 -7.06075504e-02 -3.70585263e-01
-8.94722581e-01 8.09679508e-01 3.06627542e-01 3.85992259e-01
2.06941679e-01 -1.10242438e+00 -6.12783849e-01 -2.48266637e-01
-2.20931217e-01 1.13123223e-01 2.72806622e-02 8.32308054e-01
-6.77074939e-02 6.14212275e-01 -5.19227564e-01 -2.31905356e-01
-8.93199623e-01 8.77068222e-01 4.76643354e-01 -9.27689135e-01
-4.44536328e-01 8.67559731e-01 2.87202626e-01 -5.36635280e-01
5.18179178e-01 -2.73544073e-01 -2.03929707e-01 -9.48897228e-02
3.46527368e-01 3.78795654e-01 -2.31463283e-01 -1.15759932e-01
8.68332535e-02 3.62624258e-01 -9.28404555e-02 -5.89471340e-01
1.16609323e+00 -9.51851681e-02 1.92223683e-01 7.08721578e-01
5.69704056e-01 -3.08882326e-01 -1.99705899e+00 -4.42887664e-01
2.29476467e-01 -4.69467431e-01 1.05797611e-01 -9.59953189e-01
-6.06189609e-01 8.20213497e-01 5.56519270e-01 -4.19303142e-02
1.08405375e+00 -3.50100636e-01 5.48683226e-01 8.26524734e-01
6.52621508e-01 -1.49776959e+00 4.19572771e-01 9.44301248e-01
8.87869895e-01 -1.07931912e+00 1.60720628e-02 1.61558062e-01
-8.74739468e-01 6.65154934e-01 7.73218393e-01 -1.66231036e-01
1.28944471e-01 -7.90799335e-02 -8.58444273e-02 -1.36432990e-01
-9.22974706e-01 -6.03356838e-01 8.82547870e-02 5.32070696e-01
2.07683384e-01 1.80282995e-01 -1.49701014e-01 1.12536341e-01
-2.36893520e-01 2.10897908e-01 2.06750005e-01 1.09949172e+00
-6.87073529e-01 -1.31845188e+00 -2.89067775e-01 1.98103100e-01
-4.38789189e-01 -4.76857051e-02 -1.40800342e-01 4.72663134e-01
3.54187153e-02 6.06274784e-01 -3.27947855e-01 -1.01136610e-01
4.67111468e-01 2.00832114e-01 9.31947351e-01 -7.90978909e-01
-6.78866446e-01 1.25644669e-01 -8.41152072e-02 -8.32830906e-01
-2.03890830e-01 -9.53615606e-01 -1.05917847e+00 5.90690784e-02
1.42011821e-01 3.18908215e-01 3.16633016e-01 9.07869577e-01
1.61419272e-01 8.84533882e-01 7.36311257e-01 -1.09768605e+00
-1.32535112e+00 -8.61899614e-01 -5.62141538e-01 2.54491538e-01
5.43997228e-01 -9.48393047e-01 -1.15032300e-01 -4.38686013e-02] | [4.11032772064209, 1.7204103469848633] |
518cc747-78fb-4973-a28e-8e444c5f72f1 | discovering-and-explaining-the-non-causality | 2304.00668 | null | https://arxiv.org/abs/2304.00668v4 | https://arxiv.org/pdf/2304.00668v4.pdf | Discovering and Explaining the Non-Causality of Deep Learning in SAR ATR | In recent years, deep learning has been widely used in SAR ATR and achieved excellent performance on the MSTAR dataset. However, due to constrained imaging conditions, MSTAR has data biases such as background correlation, i.e., background clutter properties have a spurious correlation with target classes. Deep learning can overfit clutter to reduce training errors. Therefore, the degree of overfitting for clutter reflects the non-causality of deep learning in SAR ATR. Existing methods only qualitatively analyze this phenomenon. In this paper, we quantify the contributions of different regions to target recognition based on the Shapley value. The Shapley value of clutter measures the degree of overfitting. Moreover, we explain how data bias and model bias contribute to non-causality. Concisely, data bias leads to comparable signal-to-clutter ratios and clutter textures in training and test sets. And various model structures have different degrees of overfitting for these biases. The experimental results of various models under standard operating conditions on the MSTAR dataset support our conclusions. Our code is available at https://github.com/waterdisappear/Data-Bias-in-MSTAR. | ['Yongxiang Liu', 'Wenpeng Zhang', 'Li Liu', 'Wei Yang', 'Weijie Li'] | 2023-04-03 | null | null | null | null | ['selection-bias'] | ['natural-language-processing'] | [-1.17618069e-01 -4.85329688e-01 -5.21126240e-02 -3.67683351e-01
-8.21026504e-01 -4.55987990e-01 2.70855010e-01 -3.55925977e-01
2.37243809e-02 6.89483523e-01 2.97092736e-01 -2.95525730e-01
-1.19678594e-01 -7.64332175e-01 -4.78428602e-01 -1.18021512e+00
-2.80249447e-01 3.85812633e-02 4.64106873e-02 -1.72584996e-01
-2.74696261e-01 7.10313737e-01 -8.89472008e-01 3.32349062e-01
8.20544243e-01 1.23979509e+00 1.00685861e-02 1.79119289e-01
9.58747938e-02 8.02380204e-01 -7.75972426e-01 -1.41340494e-01
5.85906088e-01 -3.65974724e-01 -6.93948641e-02 -4.24736589e-02
7.98113883e-01 -4.06622201e-01 -8.40141833e-01 1.43805873e+00
6.41691923e-01 -2.90443987e-01 7.16322601e-01 -1.05310869e+00
-5.96682727e-01 5.97442150e-01 -9.29979861e-01 6.57379568e-01
-5.10610223e-01 3.86802524e-01 8.25742841e-01 -1.07776821e+00
-1.32912517e-01 1.19778931e+00 9.30459976e-01 -3.84985544e-02
-1.35614491e+00 -9.39806640e-01 2.02768117e-01 -7.64517486e-02
-1.44872820e+00 -2.59463012e-01 8.21267068e-01 -6.33316994e-01
2.36267254e-01 4.04188007e-01 5.93956411e-01 1.21802175e+00
5.76693296e-01 6.40665352e-01 1.40245676e+00 -6.34335876e-02
1.28787115e-01 -3.03947300e-01 5.20984113e-01 3.69178772e-01
8.80645096e-01 6.58074260e-01 -2.26945832e-01 -1.08613268e-01
9.60185409e-01 1.76380202e-02 -4.64928687e-01 -2.10249454e-01
-1.02975261e+00 8.89361143e-01 1.03956771e+00 1.46862596e-01
-9.72121134e-02 2.17567220e-01 1.48913562e-01 2.06064969e-01
2.85320878e-01 3.69085342e-01 -3.01650673e-01 9.15781558e-01
-7.90377140e-01 1.07156716e-01 3.03130418e-01 7.06650257e-01
4.05458301e-01 9.29860532e-01 -4.90708113e-01 8.38912189e-01
2.78750479e-01 1.26152992e+00 7.59212440e-03 -6.15896225e-01
3.55952501e-01 -7.39585161e-02 2.76526302e-01 -1.24056506e+00
-5.41847467e-01 -1.61163175e+00 -1.34029233e+00 4.01051283e-01
5.91399252e-01 -3.25967461e-01 -1.16333497e+00 1.62532163e+00
-7.72910118e-02 -3.95705737e-02 2.47898504e-01 1.31226289e+00
9.52403605e-01 4.15250570e-01 5.73902652e-02 -2.32348576e-01
1.29206705e+00 -5.66083789e-01 -9.08328056e-01 -7.52627373e-01
1.86310902e-01 -8.39430511e-01 9.81105447e-01 3.08999896e-01
-4.58736390e-01 -6.57054901e-01 -1.22421205e+00 7.11580396e-01
8.22521076e-02 2.75534362e-01 6.48993492e-01 6.27451301e-01
-3.01185638e-01 4.14117634e-01 -6.44815207e-01 2.40931101e-02
6.18618309e-01 -5.71823306e-02 2.31401876e-01 -1.19898841e-01
-1.23507071e+00 8.58231187e-01 1.60870537e-01 7.06516981e-01
-1.15586376e+00 -6.73598707e-01 -4.34363365e-01 -2.57816147e-02
3.82625043e-01 -4.80744272e-01 1.09205651e+00 -1.13100576e+00
-6.22034669e-01 4.43947941e-01 4.05533463e-01 -5.08245051e-01
4.61005718e-01 -5.04314125e-01 -8.66017818e-01 -1.98144406e-01
-1.54997138e-02 4.71011214e-02 8.23154032e-01 -1.78644454e+00
-1.35726318e-01 -4.45757896e-01 -2.53705025e-01 -1.73878655e-01
2.72877932e-01 -1.33137241e-01 1.10963978e-01 -9.44674015e-01
5.99730790e-01 -6.99995339e-01 -2.19360918e-01 -1.04478009e-01
-3.45204622e-01 6.26602769e-01 5.60768008e-01 -5.43402195e-01
1.02314270e+00 -2.33423471e+00 -5.69328427e-01 2.94614941e-01
3.43136400e-01 2.57920530e-02 -3.06104273e-01 2.07207710e-01
-2.91585505e-01 -8.19149613e-02 -3.13523859e-01 8.45667779e-01
-2.15369955e-01 1.11370608e-01 -7.77637661e-01 6.78362191e-01
4.06524926e-01 6.90720677e-01 -7.72677600e-01 -2.53772825e-01
-6.85188472e-02 2.51929045e-01 -1.23112626e-01 -8.88990331e-03
-3.68858539e-02 3.48552257e-01 -5.35618186e-01 9.83814180e-01
1.30051994e+00 -2.11363867e-01 5.74210882e-02 -7.96786726e-01
-2.73536861e-01 -2.21802384e-01 -9.85494196e-01 7.29835749e-01
1.18063472e-01 8.24677050e-01 1.52620047e-01 -7.72741854e-01
1.03375483e+00 -2.57747442e-01 2.94139475e-01 -9.59099114e-01
2.32872993e-01 3.46368164e-01 6.05107546e-01 -1.83793932e-01
2.08815187e-01 -3.78184378e-01 1.47898406e-01 -1.71322405e-01
-2.90011615e-01 1.18143775e-01 -5.64351737e-01 -1.46898702e-01
9.16519821e-01 -3.23547959e-01 1.23712815e-01 -5.98659813e-01
-3.85621041e-02 4.80640501e-01 7.31722653e-01 1.06456685e+00
-2.51087219e-01 4.30557907e-01 2.57012486e-01 -6.99676096e-01
-6.91243052e-01 -1.40471733e+00 -5.50651431e-01 4.05805618e-01
1.09487966e-01 2.20328957e-01 -1.15184054e-01 -3.00809801e-01
4.07069288e-02 6.53438389e-01 -4.62106675e-01 -2.88800806e-01
-6.57544196e-01 -1.52275193e+00 6.47207379e-01 6.96296930e-01
7.14303970e-01 -6.32090926e-01 -5.82082450e-01 5.92168123e-02
-3.06380093e-01 -1.08587301e+00 -1.21676989e-01 5.67668080e-01
-9.96569693e-01 -9.42117989e-01 -6.10856950e-01 -3.60623360e-01
4.98801500e-01 7.78064013e-01 1.32581794e+00 -1.07835308e-01
-5.27346313e-01 3.22715677e-02 -2.16029093e-01 -7.21905768e-01
-1.41516790e-01 -5.62783659e-01 1.03954300e-01 -1.88974723e-01
2.88044602e-01 -2.06796050e-01 -7.52336502e-01 5.43645442e-01
-7.42061496e-01 -2.81485260e-01 9.85788882e-01 1.02387047e+00
6.87168777e-01 2.05887601e-01 1.86954454e-01 -4.18601692e-01
1.79394111e-01 -3.40173870e-01 -1.04385126e+00 2.14321036e-02
-3.59512448e-01 4.95808125e-02 1.63908273e-01 -4.03609693e-01
-9.16451037e-01 -1.02310352e-01 2.43763283e-01 -6.30719960e-01
1.74803026e-02 8.33248138e-01 -3.34574491e-01 -1.46646038e-01
8.62503946e-01 1.12013020e-01 -3.03295553e-01 -2.31361762e-01
-1.10688500e-01 3.05161119e-01 5.44403195e-01 -5.05231738e-01
1.33752489e+00 5.36259711e-01 9.69439521e-02 -9.83346283e-01
-1.43151355e+00 -9.49939042e-02 -2.95705914e-01 -4.92872119e-01
5.07791042e-01 -1.39924860e+00 -8.26254413e-02 5.65070808e-01
-8.70557010e-01 -5.39762378e-01 -1.51627928e-01 7.67119408e-01
5.75196482e-02 2.90992618e-01 -2.23501608e-01 -1.17689002e+00
-2.62470514e-01 -7.05659270e-01 6.42757297e-01 1.27754420e-01
2.22457573e-01 -5.42175233e-01 -1.97234645e-01 9.88185555e-02
5.72757125e-01 5.10864377e-01 8.17793071e-01 -3.04030657e-01
-7.73773074e-01 2.24343129e-02 -5.26854038e-01 3.03633839e-01
3.30662504e-02 -1.10225119e-01 -1.30545616e+00 -5.23247123e-01
2.84194499e-01 -4.23342511e-02 1.15246403e+00 9.28685009e-01
1.11611903e+00 -3.22399348e-01 -2.23087713e-01 6.15992069e-01
1.70760000e+00 3.24850708e-01 9.66083407e-01 4.18918997e-01
5.01075029e-01 4.62335289e-01 8.20927918e-01 3.57873559e-01
-5.27075648e-01 6.03803515e-01 7.91304648e-01 -5.97562432e-01
-2.15977818e-01 1.98750511e-01 4.41179097e-01 4.99097198e-01
-3.36330943e-02 -1.80776790e-01 -1.22971117e+00 1.92723349e-01
-1.50103998e+00 -1.05135798e+00 -7.64402866e-01 2.30058527e+00
4.34385240e-01 2.65302420e-01 -1.65944993e-01 -2.46666178e-01
5.96975803e-01 3.35706443e-01 -6.58457994e-01 4.57823068e-01
-7.71364927e-01 2.26347297e-02 1.07839906e+00 4.57494766e-01
-1.16632104e+00 8.01309466e-01 6.41584492e+00 6.99775457e-01
-1.26690185e+00 8.80548358e-02 6.75901353e-01 -7.48792142e-02
-2.25762427e-01 -1.28976539e-01 -7.07909465e-01 2.56354034e-01
5.48591852e-01 1.62544087e-01 4.73547056e-02 6.31056011e-01
3.28839153e-01 1.68006308e-03 -6.30523801e-01 8.18817735e-01
-1.75557867e-01 -1.06563771e+00 5.50382808e-02 2.01958001e-01
3.19373667e-01 3.77162009e-01 1.69192493e-01 3.70045483e-01
5.18553197e-01 -9.79862809e-01 8.00429106e-01 5.57361007e-01
6.89747512e-01 -4.41859931e-01 1.04000962e+00 -6.42740279e-02
-9.85272586e-01 -1.57450110e-01 -6.72950029e-01 1.34632140e-01
-1.83313161e-01 1.14903450e+00 -3.39943707e-01 6.43801332e-01
9.53337550e-01 2.32412413e-01 -4.59940046e-01 1.04403913e+00
-7.18717128e-02 7.55484641e-01 -8.78285393e-02 4.07294095e-01
1.90587908e-01 -3.01622987e-01 7.21976757e-01 1.24478793e+00
4.45819020e-01 2.55283952e-01 4.00091141e-01 1.03778625e+00
2.60918111e-01 -4.13055629e-01 -5.29167056e-01 -2.92341579e-02
5.29341698e-01 1.11480618e+00 -5.20869374e-01 7.07427179e-03
-4.56534773e-01 3.07450712e-01 -2.85110235e-01 6.06545329e-01
-1.18399537e+00 1.85506977e-02 8.03432822e-01 3.41378450e-01
2.92347550e-01 -4.27744538e-01 -4.56582129e-01 -1.04859710e+00
-1.48569524e-01 -1.07743800e+00 4.33363229e-01 -9.85657871e-01
-1.76320922e+00 6.81331754e-01 1.52288098e-02 -1.65090835e+00
7.75490403e-01 -8.86620581e-01 -7.29067862e-01 9.23871636e-01
-1.55221593e+00 -1.04850674e+00 -9.44614768e-01 4.26023960e-01
2.95137584e-01 -3.86117250e-01 5.19890904e-01 1.56992331e-01
-6.76484346e-01 3.26393723e-01 1.47292063e-01 5.94962835e-01
9.27305162e-01 -9.01995361e-01 1.86649665e-01 1.15502977e+00
-9.41444468e-03 4.28566784e-01 1.04703295e+00 -8.02551389e-01
-1.37748408e+00 -1.25031817e+00 -6.81319162e-02 -1.79944083e-01
9.93569970e-01 -1.36998057e-01 -1.00100029e+00 3.91350836e-01
1.31031200e-01 2.31819168e-01 7.22057700e-01 2.55286306e-01
-6.63671136e-01 -5.92495620e-01 -7.91241407e-01 6.41635954e-01
9.49324906e-01 -9.72665101e-02 -4.00213480e-01 4.56876099e-01
5.36455095e-01 -3.98072302e-01 -7.13230371e-01 8.57741535e-01
4.49648052e-01 -9.67870355e-01 1.17198396e+00 -5.52409947e-01
3.16130042e-01 -6.90570891e-01 -7.62651563e-01 -1.27507806e+00
-1.10602772e+00 4.41412292e-02 6.92096278e-02 8.28871608e-01
3.94435823e-01 -7.21423745e-01 4.25913870e-01 -4.20186110e-02
-5.13019979e-01 -2.49279663e-01 -7.47389793e-01 -1.35599053e+00
2.23959923e-01 -2.50450134e-01 3.92643392e-01 1.10842717e+00
-8.89378905e-01 3.54594350e-01 -3.00393462e-01 1.13347781e+00
1.17943239e+00 3.93260449e-01 6.22990549e-01 -1.39656246e+00
-2.39652827e-01 -2.98790514e-01 2.03545000e-02 -6.18400931e-01
-2.69193351e-01 -7.39449024e-01 4.08123322e-02 -1.31633270e+00
2.20311716e-01 -6.20282769e-01 -5.40015221e-01 4.79080200e-01
-7.47073144e-02 1.33676007e-01 1.76652834e-01 6.82485402e-01
2.99323238e-02 5.15971541e-01 1.27171278e+00 -5.95522106e-01
1.96516499e-01 -1.99473187e-01 -7.38941133e-01 7.28356838e-01
1.36728001e+00 -6.23790681e-01 1.09425366e-01 -7.46070027e-01
1.33418053e-01 -1.83047708e-02 8.57155383e-01 -1.53059959e+00
-2.65312970e-01 -3.85669291e-01 1.08353221e+00 -9.53944266e-01
7.44803920e-02 -1.10500741e+00 3.80424738e-01 8.07667613e-01
8.61232951e-02 -2.64100343e-01 5.79733014e-01 6.78529084e-01
-3.15287054e-01 1.01190537e-01 1.11372745e+00 -1.40434816e-01
-5.52124023e-01 2.05161288e-01 -4.84513223e-01 9.38736126e-02
4.58727986e-01 -1.88432097e-01 -7.79111683e-01 -2.03047022e-01
-4.42363262e-01 5.75262085e-02 -4.44301683e-03 9.16445702e-02
3.97382736e-01 -1.44529676e+00 -9.63384271e-01 5.54556549e-02
3.07364464e-01 -3.26835006e-01 2.80870944e-01 1.14331663e+00
-2.02397734e-01 2.52821267e-01 -2.52454758e-01 -8.96072209e-01
-1.10100019e+00 3.03920269e-01 8.14055204e-01 -1.12117350e-01
-4.60606396e-01 5.40261030e-01 7.42301106e-01 4.60949726e-02
6.25228286e-02 -5.04179120e-01 1.68747738e-01 3.93728316e-02
4.85465109e-01 1.40210345e-01 -1.28214061e-02 -3.11600059e-01
-4.52667773e-01 4.99766499e-01 3.12462226e-02 2.30218127e-01
1.12717688e+00 3.60357434e-01 1.38914898e-01 3.78203243e-01
4.25542176e-01 2.21373141e-01 -1.39882994e+00 -3.20554107e-01
-1.63252577e-01 -6.53883755e-01 5.49339235e-01 -1.14054406e+00
-1.34453905e+00 8.22594821e-01 1.19339466e+00 3.82532738e-02
1.14888477e+00 -1.10037073e-01 1.60363749e-01 5.18129945e-01
3.18275362e-01 -8.15306962e-01 3.65226835e-01 7.98599184e-01
1.36938190e+00 -1.36087084e+00 3.08327734e-01 -5.16085684e-01
-8.03681970e-01 1.03545058e+00 7.78748751e-01 -1.19584359e-01
5.24952233e-01 7.38060415e-01 3.86691034e-01 -5.26021838e-01
-1.52666405e-01 -3.63885254e-01 9.36264694e-02 7.39423096e-01
3.76940548e-01 3.64033461e-01 1.27269104e-01 7.43704259e-01
-4.29343395e-02 -3.77444744e-01 4.08453137e-01 6.15659237e-01
-7.90627539e-01 -4.72852439e-01 -9.19096053e-01 4.54170078e-01
-2.15200469e-01 -2.94652641e-01 -5.73294818e-01 1.04704630e+00
-1.41226247e-01 8.33701491e-01 -1.04235987e-04 -3.54262531e-01
4.78349477e-01 -3.50766748e-01 4.95842993e-01 -3.06976110e-01
-3.22072834e-01 6.15252018e-01 2.25407943e-01 -3.36376369e-01
-2.33388796e-01 -4.96481836e-01 -9.27362382e-01 -1.20810695e-01
-5.73064506e-01 2.22582873e-02 1.52546093e-01 3.88275117e-01
5.99002577e-02 9.04159009e-01 4.77430373e-01 -4.69156623e-01
-8.01619649e-01 -9.97897863e-01 -8.00924003e-01 -1.75304458e-01
4.67717350e-01 -8.97818983e-01 -7.16092288e-01 -2.77811170e-01] | [8.281984329223633, -0.9763543009757996] |
973a35fa-b320-45bd-86f4-f80766109ab7 | towards-standardizing-korean-grammatical | 2210.14389 | null | https://arxiv.org/abs/2210.14389v3 | https://arxiv.org/pdf/2210.14389v3.pdf | Towards standardizing Korean Grammatical Error Correction: Datasets and Annotation | Research on Korean grammatical error correction (GEC) is limited, compared to other major languages such as English. We attribute this problematic circumstance to the lack of a carefully designed evaluation benchmark for Korean GEC. In this work, we collect three datasets from different sources (Kor-Lang8, Kor-Native, and Kor-Learner) that covers a wide range of Korean grammatical errors. Considering the nature of Korean grammar, We then define 14 error types for Korean and provide KAGAS (Korean Automatic Grammatical error Annotation System), which can automatically annotate error types from parallel corpora. We use KAGAS on our datasets to make an evaluation benchmark for Korean, and present baseline models trained from our datasets. We show that the model trained with our datasets significantly outperforms the currently used statistical Korean GEC system (Hanspell) on a wider range of error types, demonstrating the diversity and usefulness of the datasets. The implementations and datasets are open-sourced. | ['Gyutae Kim', 'Alice Oh', 'Minjoon Seo', 'Kihyo Park', 'Junhee Cho', 'Gyuwan Kim', 'Sungjoon Park', 'Soyoung Yoon'] | 2022-10-25 | null | null | null | null | ['grammatical-error-correction'] | ['natural-language-processing'] | [-4.53065425e-01 -3.10653657e-01 1.55514464e-01 -4.72394228e-01
-1.22805929e+00 -5.10558546e-01 -2.56932080e-01 4.20216203e-01
-7.77786791e-01 1.03534567e+00 3.82031947e-01 -4.98822331e-01
3.57922703e-01 -5.07955074e-01 -6.34839296e-01 -3.80559713e-02
1.84839979e-01 3.54122370e-01 3.06619667e-02 -4.38974649e-01
3.94010216e-01 -3.17540854e-01 -8.35544944e-01 2.06523821e-01
1.81382763e+00 2.88385808e-01 3.51389527e-01 6.96597219e-01
1.78379938e-01 4.20925170e-01 -7.87266195e-01 -9.90379632e-01
-3.15747440e-01 -5.79126894e-01 -1.10380507e+00 -6.90751672e-01
4.37007457e-01 -1.66465133e-01 3.32718104e-01 1.17890799e+00
5.70018828e-01 -5.17046675e-02 4.55407202e-01 -1.02521014e+00
-1.00743532e+00 1.12045395e+00 1.35775521e-01 -6.82960898e-02
4.37043190e-01 -2.30567500e-01 1.04302192e+00 -1.20953906e+00
9.84026670e-01 8.69189858e-01 1.26442492e+00 1.12656391e+00
-7.99502790e-01 -5.38093984e-01 7.53447860e-02 8.92107561e-02
-1.39636958e+00 -4.09562975e-01 6.21805012e-01 -4.06362601e-02
1.58777237e+00 1.66543543e-01 3.41093630e-01 1.38626921e+00
1.05856836e-01 6.86942160e-01 9.29709375e-01 -9.96438861e-01
4.23902972e-03 -2.22682431e-02 5.71214080e-01 8.41649473e-01
5.34717023e-01 -1.62132040e-01 -3.11698735e-01 1.70026347e-02
-5.96048310e-02 -5.46913326e-01 -6.76671267e-01 5.34250975e-01
-1.13027203e+00 6.27090454e-01 7.60953650e-02 3.75999838e-01
2.70124346e-01 3.11082810e-01 7.83754826e-01 5.14000356e-01
4.05095756e-01 3.93778145e-01 -1.01728642e+00 -6.02895260e-01
-4.09102499e-01 3.45490456e-01 8.79117846e-01 1.59852350e+00
3.01165551e-01 2.83514351e-01 3.84861439e-01 1.18526351e+00
5.43186009e-01 6.82284892e-01 6.07350171e-01 -7.25164175e-01
1.11096263e+00 4.87772167e-01 -1.30012572e-01 -7.77735889e-01
-5.43788970e-01 -3.27937901e-02 -6.23714864e-01 -3.73436511e-01
6.47626162e-01 -1.35783792e-01 -5.17646551e-01 1.90473139e+00
-5.08716181e-02 -6.37779176e-01 3.87688011e-01 6.78834677e-01
1.06348753e+00 5.00342190e-01 4.78245288e-01 -2.09163073e-02
1.02424014e+00 -1.00950634e+00 -8.71517837e-01 -1.77613974e-01
1.83606052e+00 -7.88924217e-01 1.57010388e+00 3.24335992e-01
-7.72107005e-01 -2.73647696e-01 -9.30444300e-01 -5.79917967e-01
-6.07175946e-01 5.16968668e-01 5.77009320e-01 8.63867283e-01
-7.82287061e-01 4.99960721e-01 -8.84472430e-01 -6.52548671e-01
-4.33886409e-01 -1.03122830e-01 -5.39881527e-01 -1.86985567e-01
-1.31839883e+00 9.17228162e-01 8.26845050e-01 2.57636428e-01
-3.02600443e-01 -6.27048135e-01 -1.11947024e+00 -4.23338741e-01
1.32744655e-01 -4.12346840e-01 1.31104422e+00 -7.61301100e-01
-1.10991168e+00 9.74274933e-01 -7.07152411e-02 -4.32937033e-02
1.86669812e-01 -4.43586558e-01 -8.99998486e-01 -5.48532724e-01
2.96171918e-03 3.56065184e-01 -9.74119529e-02 -1.30753613e+00
-6.39650166e-01 -4.80276942e-01 -2.98575580e-01 -2.34467220e-02
-1.48687541e-01 6.05631411e-01 -4.14929688e-01 -1.14552510e+00
-8.42084555e-05 -1.01743054e+00 9.91411507e-03 -8.43324065e-01
-4.32314098e-01 -3.85318875e-01 3.23826075e-01 -1.42095566e+00
2.19395781e+00 -2.06901550e+00 6.52795285e-02 -1.16412461e-01
-2.68919051e-01 2.42785648e-01 -3.50138605e-01 5.56930721e-01
-8.22200105e-02 8.39197397e-01 -5.99374354e-01 -7.32549608e-01
6.67265952e-02 4.57155824e-01 5.38395755e-02 1.94831640e-02
5.53869791e-02 1.01899529e+00 -1.11663365e+00 -4.80650038e-01
-4.89507228e-01 -1.46929130e-01 -9.97771263e-01 1.43982582e-02
-1.38751507e-01 1.78746760e-01 -5.90098044e-03 1.18609881e+00
6.59191668e-01 5.23119330e-01 5.77088654e-01 2.32857853e-01
2.30981894e-02 8.15827370e-01 -9.83729124e-01 1.87379467e+00
-4.82268959e-01 2.09729914e-02 -2.86151916e-01 -4.12221432e-01
8.54990363e-01 2.92101383e-01 -3.71543020e-01 -3.98027211e-01
-1.27850994e-01 1.06867671e+00 1.51374400e-01 -3.27996820e-01
1.18704879e+00 2.47269586e-01 -9.41493213e-01 3.62021685e-01
6.05541050e-01 -1.63949981e-01 4.15992469e-01 2.95144022e-01
1.03577721e+00 7.09680021e-02 4.79377866e-01 -3.52557808e-01
3.61307979e-01 1.91768870e-01 1.06462491e+00 6.47378504e-01
-3.68765444e-01 6.44181252e-01 1.92562789e-01 -2.09342241e-01
-8.88488472e-01 -9.17469621e-01 -2.89576966e-02 8.04121912e-01
-2.84938753e-01 -1.21765423e+00 -9.37912881e-01 -1.50400281e+00
-2.10390940e-01 8.10937643e-01 -2.92391032e-01 1.49875423e-02
-1.04312980e+00 -8.15337598e-01 1.08052123e+00 6.97007656e-01
3.02181274e-01 -1.34229553e+00 1.96060553e-01 3.70538324e-01
-6.64172709e-01 -1.11059546e+00 -8.65846694e-01 1.27265096e-01
-8.82254481e-01 -1.28442252e+00 8.72966945e-02 -1.15092731e+00
6.16477251e-01 -3.64175618e-01 1.64007914e+00 5.67626297e-01
1.77597240e-01 2.45319963e-01 -9.24134374e-01 -4.56185728e-01
-9.36538160e-01 3.75746459e-01 1.64485142e-01 -1.04859042e+00
5.58828473e-01 -1.13871999e-01 1.90140992e-01 -1.00223124e-02
-4.45284545e-01 -1.74248472e-01 -4.36431058e-02 9.79703605e-01
4.01516616e-01 -1.83764741e-01 7.13014245e-01 -1.38701963e+00
6.31658137e-01 -3.93481940e-01 -5.60301840e-01 4.89386082e-01
-7.95767725e-01 -8.32229480e-02 6.92689419e-01 -3.19859223e-03
-1.05846715e+00 -5.05467117e-01 -6.69654429e-01 4.88828003e-01
7.65588274e-03 9.47567523e-01 -5.26152551e-01 1.36042371e-01
3.62308532e-01 1.88031048e-02 -6.86465859e-01 -8.29240084e-01
1.73458546e-01 8.40822995e-01 6.75020933e-01 -1.13303554e+00
2.11441174e-01 -5.43687224e-01 -6.58947825e-01 -3.58087063e-01
-6.51889861e-01 -3.64044309e-02 -7.00353026e-01 5.39290383e-02
5.82816601e-01 -9.79628205e-01 -2.31438398e-01 9.22764778e-01
-1.58225119e+00 -4.49947238e-01 1.19608536e-01 4.62197006e-01
-2.55822361e-01 4.88956302e-01 -1.27719617e+00 -4.63021427e-01
-5.36744833e-01 -9.48970854e-01 9.50692713e-01 -1.64843351e-01
-4.60136682e-01 -1.30348587e+00 2.39521891e-01 2.52796501e-01
2.01618195e-01 3.22184339e-02 1.31824028e+00 -4.67476726e-01
-3.06090057e-01 1.05927832e-01 1.55291453e-01 6.47514999e-01
-1.55270770e-01 2.48637602e-01 -3.36280048e-01 -2.05495968e-01
-4.63872582e-01 -5.35989106e-01 8.79689038e-01 -6.73773140e-02
1.13409507e+00 -3.64250958e-01 9.52682868e-02 8.22520912e-01
1.71533024e+00 1.54801965e-01 7.42166102e-01 4.25161064e-01
8.73247921e-01 3.53930831e-01 7.42772341e-01 1.06008336e-01
1.10859597e+00 5.87841809e-01 2.00521648e-01 3.77891183e-01
-1.00629032e-01 -6.98330879e-01 6.99185073e-01 2.04714489e+00
-3.88098240e-01 -6.31707013e-01 -1.10969162e+00 7.65455604e-01
-1.70140946e+00 -3.98301512e-01 -6.75736904e-01 2.10974121e+00
1.37414241e+00 -3.13131124e-01 -2.98328787e-01 4.33970578e-02
3.46917152e-01 -3.57054442e-01 4.74688709e-01 -9.45373476e-01
-3.02430093e-01 2.29061633e-01 2.09850937e-01 6.34850204e-01
-1.04193115e+00 1.43209183e+00 6.54973173e+00 7.21523464e-01
-6.31337941e-01 4.54121113e-01 1.82341963e-01 5.35620987e-01
-4.91094142e-01 8.15581083e-02 -1.31111050e+00 7.73147345e-01
1.24383926e+00 1.37436703e-01 2.43867114e-01 6.37719333e-01
-4.59903553e-02 -1.29239514e-01 -1.00871372e+00 6.96931422e-01
2.72580832e-01 -8.31249118e-01 -3.89759876e-02 -2.25095823e-01
8.62573743e-01 1.50381299e-02 -4.88299549e-01 8.98129821e-01
6.59582019e-01 -7.73308516e-01 8.63381684e-01 3.72805327e-01
9.26725388e-01 -8.59269440e-01 1.01541150e+00 2.10854605e-01
-7.26013958e-01 5.21675088e-02 -5.80045104e-01 2.38557644e-02
1.59201086e-01 6.30022764e-01 -3.50324154e-01 8.14686239e-01
9.64402020e-01 8.08158159e-01 -9.36942399e-01 6.61639273e-01
-7.75733769e-01 1.23984063e+00 -9.67157930e-02 -6.28190488e-02
-2.20722005e-01 -1.82765216e-01 4.70972866e-01 1.76050353e+00
8.17808032e-01 3.81788462e-02 7.37389848e-02 6.56839013e-01
-4.43712145e-01 4.71384823e-01 -3.77280235e-01 -4.23376188e-02
7.66086876e-01 9.98270333e-01 -8.80006254e-02 -2.05248117e-01
-3.92395020e-01 1.18720627e+00 9.80420053e-01 1.68264061e-01
-8.33275855e-01 -7.23755896e-01 5.53620040e-01 -3.96758556e-01
-5.46925813e-02 -6.19520605e-01 -3.81731331e-01 -1.64134014e+00
2.31492549e-01 -1.19874883e+00 6.55925751e-01 -7.15366006e-01
-1.36623394e+00 5.65627992e-01 -1.85835928e-01 -9.58521664e-01
-1.94464892e-01 -9.01389182e-01 -3.83915484e-01 8.49209011e-01
-1.42943454e+00 -1.37101960e+00 -2.05245972e-01 4.25136656e-01
4.32846576e-01 -1.72437634e-02 1.24665487e+00 6.36558771e-01
-8.13900650e-01 1.25804448e+00 1.15032934e-01 6.53096616e-01
1.13782263e+00 -1.65056598e+00 5.43251812e-01 1.23606765e+00
2.77752280e-02 8.59959602e-01 8.47709477e-02 -9.68156159e-01
-1.16436982e+00 -1.23317170e+00 2.02853394e+00 -9.71276343e-01
6.50414288e-01 -6.68595016e-01 -1.11472178e+00 9.92780685e-01
2.11618487e-02 8.42232183e-02 7.25956023e-01 5.41897595e-01
-4.88245845e-01 9.55171511e-02 -9.29397166e-01 5.44163287e-01
1.48193109e+00 -3.98712814e-01 -5.45130134e-01 -9.84632075e-02
9.00608003e-01 -7.28926539e-01 -1.07726371e+00 4.55444962e-01
1.83318973e-01 -6.10356450e-01 2.04408690e-01 -9.20866370e-01
6.85814679e-01 -3.22473854e-01 -4.59899455e-01 -1.97591376e+00
-2.16579184e-01 -2.44746193e-01 4.97255474e-02 1.85107219e+00
8.10513973e-01 -5.97139597e-01 6.01912811e-02 3.32303971e-01
-9.34273303e-01 -4.63538796e-01 -7.97331333e-01 -8.70266080e-01
5.97549736e-01 -8.77796590e-01 4.95758563e-01 1.31912446e+00
3.39635164e-01 -4.90166023e-02 -2.24897742e-01 2.64210224e-01
2.43796811e-01 -2.97256976e-01 5.42571127e-01 -8.02834213e-01
-1.22884035e-01 -5.42039871e-02 -2.48605862e-01 -4.62505698e-01
4.19254512e-01 -1.27428496e+00 2.67206162e-01 -1.02959812e+00
1.52121603e-01 -1.10026741e+00 -6.37646541e-02 8.87358665e-01
-6.75033391e-01 3.28317434e-01 1.30904987e-01 -2.14014351e-02
-4.75033760e-01 5.54315805e-01 6.86366856e-01 1.35771096e-01
9.17900447e-03 -5.76829851e-01 -7.37492681e-01 7.15808570e-01
7.40775883e-01 -7.39067972e-01 3.20314556e-01 -1.02308500e+00
7.09784508e-01 -2.63861954e-01 4.65364102e-03 -7.75025070e-01
-9.65059027e-02 -2.89460905e-02 3.42664942e-02 -3.13391447e-01
-5.73993325e-01 -4.57550704e-01 2.83408370e-02 2.20537797e-01
-7.70233572e-02 7.06955135e-01 1.97700441e-01 8.40762407e-02
-4.30230051e-01 -5.87143481e-01 5.22575557e-01 -7.96014294e-02
-9.93361950e-01 1.29515901e-02 -3.42509717e-01 6.27151132e-01
4.28828865e-01 2.75977422e-02 -6.49328649e-01 6.40986115e-02
-5.16015589e-01 2.68590987e-01 5.85661113e-01 6.18597746e-01
4.84604895e-01 -1.51219583e+00 -8.16248298e-01 2.31748655e-01
7.13716328e-01 -2.32392937e-01 1.39687033e-02 7.56196380e-01
-6.86347067e-01 2.42283285e-01 1.84533939e-01 -9.34131444e-02
-1.24185884e+00 2.12582186e-01 2.66912282e-01 -6.06526494e-01
-2.88113534e-01 6.86723769e-01 -6.29417181e-01 -1.42372119e+00
-9.87297818e-02 -7.60894775e-01 8.92208219e-02 -5.41924298e-01
2.66602218e-01 4.23184752e-01 5.67200065e-01 -5.62926471e-01
-4.76832360e-01 2.60551065e-01 5.01171798e-02 2.84876406e-01
1.09739327e+00 -2.79583573e-01 -5.07424772e-01 4.02863264e-01
9.56183672e-01 7.88056970e-01 -1.84875920e-01 2.64662169e-02
5.64875126e-01 -2.86087126e-01 -4.16842192e-01 -1.34141481e+00
-7.02476323e-01 5.06616354e-01 1.38842419e-01 -5.52669287e-01
1.09147251e+00 -1.47170216e-01 9.32230353e-01 4.08787519e-01
7.61850476e-01 -1.55576694e+00 -3.62701893e-01 1.49834895e+00
1.00296354e+00 -1.33521557e+00 -5.76403260e-01 -7.42384911e-01
-6.85348272e-01 1.13017452e+00 1.09500074e+00 7.57695884e-02
5.14285326e-01 3.69532943e-01 2.81012356e-01 1.44966081e-01
-7.13045895e-01 3.00874189e-02 1.42882302e-01 7.12018967e-01
1.08109093e+00 3.25187653e-01 -1.23281276e+00 1.32220936e+00
-5.52208483e-01 -1.25195876e-01 8.39753270e-01 8.87959301e-01
3.78915891e-02 -1.63817382e+00 -7.00901747e-02 5.69170527e-02
-6.48091674e-01 -6.39683545e-01 -2.12196171e-01 1.18982244e+00
4.36843842e-01 8.40450943e-01 -1.50142401e-01 -2.24014848e-01
4.01625365e-01 2.81835586e-01 7.53221631e-01 -8.74035776e-01
-7.97807217e-01 -4.62770015e-01 6.88118994e-01 -4.69290584e-01
-1.10425174e-01 -6.46568716e-01 -1.44053912e+00 -3.85538131e-01
-5.03433704e-01 2.18339175e-01 6.25095367e-01 7.15588450e-01
2.49263570e-01 1.15752369e-01 7.03863800e-02 -8.03269371e-02
-4.53541040e-01 -1.35522270e+00 -7.60224819e-01 5.57966352e-01
-2.42397115e-01 -1.46083117e-01 -3.83461535e-01 5.61996847e-02] | [11.060962677001953, 10.733016967773438] |
d063a46c-1785-443b-849b-31fb3b23e5fc | dgfont-robust-deformable-generative-networks | 2212.14742 | null | https://arxiv.org/abs/2212.14742v1 | https://arxiv.org/pdf/2212.14742v1.pdf | DGFont++: Robust Deformable Generative Networks for Unsupervised Font Generation | Automatic font generation without human experts is a practical and significant problem, especially for some languages that consist of a large number of characters. Existing methods for font generation are often in supervised learning. They require a large number of paired data, which are labor-intensive and expensive to collect. In contrast, common unsupervised image-to-image translation methods are not applicable to font generation, as they often define style as the set of textures and colors. In this work, we propose a robust deformable generative network for unsupervised font generation (abbreviated as DGFont++). We introduce a feature deformation skip connection (FDSC) to learn local patterns and geometric transformations between fonts. The FDSC predicts pairs of displacement maps and employs the predicted maps to apply deformable convolution to the low-level content feature maps. The outputs of FDSC are fed into a mixer to generate final results. Moreover, we introduce contrastive self-supervised learning to learn a robust style representation for fonts by understanding the similarity and dissimilarities of fonts. To distinguish different styles, we train our model with a multi-task discriminator, which ensures that each style can be discriminated independently. In addition to adversarial loss, another two reconstruction losses are adopted to constrain the domain-invariant characteristics between generated images and content images. Taking advantage of FDSC and the adopted loss functions, our model is able to maintain spatial information and generates high-quality character images in an unsupervised manner. Experiments demonstrate that our model is able to generate character images of higher quality than state-of-the-art methods. | ['Yue Lu', 'Li Sun', 'Yangchen Xie', 'Xinyuan Chen'] | 2022-12-30 | null | null | null | null | ['unsupervised-image-to-image-translation'] | ['computer-vision'] | [ 4.94743854e-01 -2.37539366e-01 2.81062573e-01 -3.88797224e-01
-4.19996023e-01 -9.73137021e-01 6.64598227e-01 -3.04958344e-01
-1.41557142e-01 6.64554179e-01 -1.94128662e-01 5.54347038e-03
2.32168317e-01 -1.17943418e+00 -9.97482717e-01 -8.51793349e-01
5.51346600e-01 4.23773319e-01 1.28237456e-01 -3.39984506e-01
1.86139882e-01 6.31030798e-01 -1.51699853e+00 3.63841027e-01
1.25209498e+00 9.75979149e-01 3.89847934e-01 5.77231765e-01
-3.69503349e-01 4.18414474e-01 -6.78520203e-01 -5.86490929e-01
4.70222354e-01 -8.01541448e-01 -4.86592144e-01 3.89619470e-01
6.13764703e-01 -3.19906592e-01 -1.20316848e-01 1.19408989e+00
5.51654518e-01 2.49600485e-02 1.03093100e+00 -9.86025035e-01
-1.20508790e+00 3.14620733e-01 -3.21997941e-01 -3.83864969e-01
1.66779533e-01 5.00905931e-01 7.21897185e-01 -8.14833581e-01
6.93740249e-01 1.17613089e+00 2.88378626e-01 7.02687562e-01
-1.54699838e+00 -6.10774815e-01 1.29824206e-01 -3.17400515e-01
-1.15723038e+00 -1.18521951e-01 1.16832721e+00 -5.62055528e-01
1.85356751e-01 3.94418627e-01 6.40050232e-01 1.46453583e+00
5.64235970e-02 7.08712161e-01 1.49976408e+00 -4.59949195e-01
2.42731065e-01 4.00571167e-01 -5.06925583e-01 6.39393091e-01
-4.85102348e-02 3.11027486e-02 -4.22116280e-01 1.95203006e-01
1.20662439e+00 -1.57028958e-01 -2.72727668e-01 -5.03802776e-01
-1.17753077e+00 8.07334423e-01 5.10071158e-01 1.99456573e-01
-1.35670617e-01 -9.47660506e-02 1.53754860e-01 3.80291283e-01
3.82103294e-01 4.89385247e-01 9.75134894e-02 1.17023669e-01
-8.70429575e-01 3.21992159e-01 5.70983291e-01 1.04498422e+00
7.76638865e-01 3.13757151e-01 -4.32996482e-01 9.19103146e-01
9.99047086e-02 9.34999287e-01 6.03327513e-01 -5.10429025e-01
5.85146964e-01 5.52723229e-01 1.25193298e-01 -1.13736546e+00
1.09555483e-01 -1.75545514e-01 -1.11083257e+00 7.05557883e-01
3.99796814e-01 -3.20008248e-02 -9.88770485e-01 1.64942849e+00
-3.06133833e-03 -1.78277418e-01 2.02176552e-02 9.85003114e-01
5.26118875e-01 6.26320720e-01 -1.91841796e-01 1.30011603e-01
1.05823576e+00 -7.95316577e-01 -5.79445720e-01 -1.00494601e-01
5.16467541e-02 -1.04834652e+00 1.55377960e+00 3.03846270e-01
-1.18713093e+00 -9.69023705e-01 -1.02793181e+00 -1.09423697e-01
-5.27965724e-01 4.45360214e-01 4.12418067e-01 6.99414134e-01
-8.32120061e-01 7.39329576e-01 -4.75608230e-01 -1.20544568e-01
4.73253399e-01 1.99370280e-01 -6.58010170e-02 2.10466743e-01
-1.08297300e+00 6.56661510e-01 3.45258653e-01 7.79467821e-02
-7.46821582e-01 -3.86882126e-01 -7.66705394e-01 -1.82138637e-01
-1.27369389e-01 -6.49828374e-01 6.24552071e-01 -1.53459060e+00
-2.06947827e+00 8.99469376e-01 1.55357286e-01 -1.48078352e-01
1.11368728e+00 2.80912034e-04 -4.00497973e-01 4.18101028e-02
1.70548782e-02 9.09992337e-01 1.45985925e+00 -1.42655778e+00
-3.00935030e-01 3.81623283e-02 -1.68865904e-01 2.45887831e-01
-7.21531510e-01 -1.52553543e-01 -3.28154415e-01 -1.15827811e+00
-5.27398065e-02 -8.31201732e-01 9.66574475e-02 3.84434253e-01
-6.54694557e-01 2.69852340e-01 9.37024295e-01 -6.21765256e-01
8.28339696e-01 -2.04509020e+00 5.40130794e-01 3.20813060e-01
-5.30315051e-03 2.70506829e-01 -2.73947269e-01 1.53267264e-01
1.05230145e-01 -5.67973871e-03 -5.00208437e-01 -4.93674576e-01
1.27354236e-02 1.79631919e-01 -5.33620298e-01 1.34702861e-01
5.83040416e-01 1.00312054e+00 -6.84702098e-01 -3.69321197e-01
3.19560707e-01 5.64829707e-01 -3.41257840e-01 4.19680029e-01
-5.12565911e-01 7.80436456e-01 -4.51537311e-01 4.50270802e-01
8.90773714e-01 -1.13091439e-01 8.94843508e-03 -1.73410267e-01
-3.31111364e-02 -1.80485040e-01 -1.08588028e+00 1.72495508e+00
-5.67471743e-01 4.44963962e-01 -3.24224979e-01 -7.47625828e-01
1.41783917e+00 -9.61526670e-03 -9.68801305e-02 -5.60403228e-01
2.06855655e-01 3.67882103e-01 -2.68316209e-01 -2.38640547e-01
3.70386243e-01 -3.96016287e-03 -8.75292495e-02 3.71155769e-01
-9.34826732e-02 -6.58710599e-01 1.85484514e-01 -1.40558332e-01
4.94224370e-01 4.87846494e-01 -3.55730355e-01 -2.41073936e-01
7.58422017e-01 -3.73559505e-01 3.98514390e-01 4.87498671e-01
4.38377589e-01 1.16534746e+00 3.90490323e-01 -4.40393597e-01
-1.41596413e+00 -1.38285661e+00 1.84303951e-02 6.04600251e-01
2.48891532e-01 2.79475451e-01 -9.83723938e-01 -7.68621325e-01
-5.49861453e-02 4.77345794e-01 -5.88651597e-01 -5.98252900e-02
-7.07900107e-01 -5.40082633e-01 4.89767969e-01 5.28004706e-01
7.18895078e-01 -1.31588423e+00 -3.18967402e-01 1.95563156e-02
-4.00218852e-02 -1.07814610e+00 -8.24315131e-01 3.96819748e-02
-7.67060280e-01 -7.34653652e-01 -1.26812840e+00 -1.07109249e+00
1.11655486e+00 -2.55566984e-01 9.93381202e-01 -2.82700956e-01
-2.24398077e-01 -3.49305533e-02 -2.52728701e-01 -1.12680666e-01
-8.53753090e-01 9.08481702e-02 -7.01121539e-02 6.32371962e-01
-1.59775466e-01 -7.20418036e-01 -6.81479216e-01 3.40681314e-01
-1.23777878e+00 3.34693015e-01 6.77104473e-01 1.02432048e+00
7.90833652e-01 -1.43625274e-01 3.69326144e-01 -9.59317863e-01
7.85775244e-01 7.36841112e-02 -8.74072611e-01 2.52265930e-01
-4.28694010e-01 3.90096933e-01 1.29742658e+00 -8.62346113e-01
-1.11107039e+00 3.09633642e-01 6.07502796e-02 -5.33471406e-01
-2.97173858e-01 -1.17511548e-01 -5.15371501e-01 -2.62503266e-01
6.96201861e-01 6.45406902e-01 1.29595160e-01 -4.61115986e-01
5.38472593e-01 5.41745365e-01 5.54542422e-01 -7.78964579e-01
1.36414385e+00 5.03654122e-01 -2.73406953e-02 -6.09036922e-01
-3.99537951e-01 4.04183656e-01 -8.46995234e-01 -2.41276965e-01
1.00576568e+00 -6.77895546e-01 -4.00361627e-01 9.01868224e-01
-1.10046053e+00 -4.93590117e-01 -5.35805643e-01 9.74718630e-02
-7.53461301e-01 4.17225868e-01 -5.61559856e-01 -5.38348794e-01
-3.78647000e-01 -1.26064444e+00 1.14943016e+00 3.17814708e-01
3.30370478e-02 -1.00664902e+00 -1.20471776e-01 1.03650033e-01
5.43262243e-01 5.75480700e-01 1.08276010e+00 -4.01318856e-02
-6.88567042e-01 -3.92842665e-03 -2.09001914e-01 8.12619030e-01
4.16975975e-01 3.26824188e-02 -8.69396567e-01 -2.74312019e-01
-3.20850700e-01 -5.47541976e-01 6.65814936e-01 2.74425112e-02
1.39158118e+00 -3.24942201e-01 1.27272114e-01 8.08328152e-01
1.40345716e+00 1.70369223e-01 8.09081256e-01 2.06080884e-01
9.96533573e-01 5.95175743e-01 2.86466032e-01 2.49241292e-01
-2.67819520e-02 7.43986547e-01 1.95346221e-01 -3.69055748e-01
-3.21870029e-01 -4.48776871e-01 4.21310365e-01 7.88217783e-01
-1.84665591e-01 -3.45611960e-01 -3.55407774e-01 2.50324517e-01
-1.55065334e+00 -7.68347502e-01 4.36371416e-02 2.42937875e+00
1.12707734e+00 2.61360049e-01 8.06521922e-02 1.56525318e-02
9.00786698e-01 1.69585034e-01 -7.11396694e-01 -5.99655867e-01
-5.92415333e-01 4.91002828e-01 3.28918189e-01 2.50339061e-01
-1.09504092e+00 1.12800086e+00 5.25259924e+00 9.12594497e-01
-1.49005175e+00 -2.72422493e-01 8.50074470e-01 1.58459857e-01
-5.85370064e-01 -2.38401175e-01 -6.14843369e-01 8.66345286e-01
3.93136173e-01 2.07983762e-01 6.53143108e-01 6.73520505e-01
1.53141007e-01 3.87473166e-01 -9.34599400e-01 9.27207708e-01
2.86359400e-01 -1.21616840e+00 3.87009442e-01 -1.19166814e-01
1.04135478e+00 -7.07005322e-01 5.28339148e-01 -1.47422507e-01
2.15581119e-01 -1.12435198e+00 1.01826286e+00 7.85418272e-01
1.18228924e+00 -7.27444112e-01 2.96283424e-01 1.41767740e-01
-1.10312808e+00 1.40663743e-01 -5.16797602e-01 2.71198392e-01
1.34466678e-01 5.62845767e-01 -2.30829403e-01 5.78616798e-01
4.76232976e-01 4.92179215e-01 -6.15459800e-01 6.13310218e-01
-4.59245622e-01 2.31024727e-01 -7.22258613e-02 -1.35873914e-01
8.00143480e-02 -6.13226533e-01 3.71081382e-01 1.02615345e+00
5.22440493e-01 -4.85246569e-01 9.18378234e-02 1.52546084e+00
-2.47156873e-01 1.35237157e-01 -5.36344886e-01 -1.20405152e-01
1.97417870e-01 1.20913517e+00 -6.70762658e-01 -3.04345638e-01
-1.19870476e-01 1.76049519e+00 1.68610036e-01 3.37834924e-01
-8.90762687e-01 -6.01773858e-01 4.16078478e-01 1.24679223e-01
4.29795593e-01 -1.35571197e-01 -3.75464618e-01 -1.26888990e+00
3.13035160e-01 -1.03315699e+00 -2.08695412e-01 -7.96148181e-01
-1.52918768e+00 8.99106622e-01 -3.43615025e-01 -1.73530436e+00
-1.69409275e-01 -8.34154725e-01 -7.57911801e-01 1.25679755e+00
-1.43669474e+00 -1.53718853e+00 -5.05205512e-01 6.26873672e-01
4.59062338e-01 -3.54353279e-01 8.92034352e-01 1.12196833e-01
-3.12040389e-01 9.12764013e-01 2.80849189e-01 4.98478681e-01
8.95874619e-01 -1.38007152e+00 5.84932685e-01 8.85315895e-01
4.83742431e-02 3.82927358e-01 4.93279576e-01 -6.00710154e-01
-1.16725802e+00 -1.33930945e+00 5.00293374e-01 -1.42983839e-01
3.93727899e-01 -7.76300371e-01 -7.98028886e-01 2.67060339e-01
1.96026728e-01 8.52468014e-02 3.35809261e-01 -6.27656639e-01
-5.51192522e-01 -2.31980607e-01 -1.20358062e+00 7.78587341e-01
1.03340530e+00 -4.79423463e-01 -2.63291746e-01 3.70519996e-01
4.45507050e-01 -4.39686894e-01 -8.19040656e-01 1.25984997e-01
4.98008281e-01 -9.75232840e-01 8.41793656e-01 -2.91597635e-01
7.57673144e-01 -3.68788451e-01 1.56294212e-01 -1.36725485e+00
-2.03004211e-01 -6.91497147e-01 2.33237743e-01 1.53078687e+00
4.34644073e-01 -6.86821759e-01 7.35605478e-01 3.33096236e-01
8.58139768e-02 -4.94522274e-01 -4.95080024e-01 -1.02763927e+00
4.30887669e-01 1.93762913e-01 7.65929341e-01 8.64283085e-01
-6.59557045e-01 1.28650591e-01 -5.54453671e-01 -2.58398801e-02
6.84905648e-01 5.38209915e-01 7.19643950e-01 -1.10538960e+00
-5.41829407e-01 -6.05013907e-01 -3.24643999e-01 -9.91073191e-01
2.52154708e-01 -8.18862498e-01 -1.78158004e-02 -1.13996470e+00
-7.19936341e-02 -6.07695580e-01 4.45683068e-03 2.06225336e-01
-2.04915747e-01 5.16252458e-01 2.95949876e-01 2.33049005e-01
-5.23775518e-02 8.11157346e-01 1.75016499e+00 -3.94062251e-01
-3.03688318e-01 4.22271714e-02 -4.04444277e-01 4.86118495e-01
9.12853181e-01 -9.24921706e-02 -4.42876458e-01 -5.26657343e-01
4.91207205e-02 -3.05232167e-01 3.63513470e-01 -1.06024432e+00
-1.88345119e-01 -1.10461943e-01 8.17633867e-01 -2.05303967e-01
3.01776290e-01 -5.81755280e-01 8.47991332e-02 2.57604837e-01
-4.39947814e-01 -3.21471617e-02 9.93421953e-03 2.71490186e-01
-2.85752803e-01 -2.33205497e-01 1.05024028e+00 -1.26747459e-01
-2.97659934e-01 3.33234608e-01 -1.79709718e-01 6.00355230e-02
8.75770807e-01 -2.69397676e-01 -6.86411187e-02 -2.37566695e-01
-4.60094601e-01 -2.47640684e-01 7.85312116e-01 5.44677436e-01
6.99973106e-01 -1.57568038e+00 -7.40970314e-01 4.89421606e-01
1.35853112e-01 6.47672638e-02 2.02418774e-01 1.84924468e-01
-6.92380786e-01 -7.31370151e-02 -5.48819959e-01 -4.53577429e-01
-1.01354373e+00 5.87363243e-01 2.30300814e-01 -1.61983907e-01
-6.44431233e-01 6.88056946e-01 2.63412565e-01 -5.47370911e-01
1.37246661e-02 -3.55792403e-01 -5.64461872e-02 -9.10323188e-02
3.10415149e-01 -2.19990239e-02 -1.41513690e-01 -5.54410458e-01
1.92156389e-01 9.90023673e-01 -2.35257726e-02 -1.85680360e-01
1.08086824e+00 1.63450792e-01 -4.21718024e-02 2.23499969e-01
1.13059759e+00 1.83961824e-01 -1.65848756e+00 1.09341862e-02
-4.55029458e-01 -5.62426329e-01 -4.07384604e-01 -7.85875261e-01
-1.21320343e+00 1.01648438e+00 9.35649574e-01 3.76484282e-02
1.32462776e+00 -4.30932254e-01 8.21432412e-01 -1.95728289e-03
2.67574728e-01 -1.21076834e+00 3.70499194e-01 2.42556185e-01
1.09311426e+00 -1.12669885e+00 -3.51886064e-01 -4.66101259e-01
-8.28925848e-01 1.27768886e+00 6.05374157e-01 -3.51410985e-01
1.13169782e-01 2.95469880e-01 2.59000003e-01 3.36501777e-01
-3.89585420e-02 9.78182419e-04 4.18602109e-01 7.50136435e-01
2.93437392e-01 7.64660686e-02 -2.19034955e-01 2.63557076e-01
-3.38605076e-01 -2.18882963e-01 3.35004956e-01 6.25359476e-01
-2.02893373e-02 -1.60530329e+00 -4.44863051e-01 2.07976103e-01
-9.20013711e-02 -1.69554189e-01 -4.84870136e-01 4.18460727e-01
3.04347426e-01 5.37954569e-01 1.32516533e-01 -4.19662237e-01
3.19564521e-01 7.36804232e-02 5.69180310e-01 -3.42912346e-01
-6.05105877e-01 1.58343948e-02 -4.08964157e-01 -2.77738452e-01
-1.25127628e-01 -4.94724870e-01 -8.96026313e-01 -1.61729693e-01
-8.71965587e-02 -1.89770997e-01 5.22577465e-01 5.12647331e-01
3.29144448e-01 6.52974069e-01 1.05352402e+00 -8.61729324e-01
-6.19348645e-01 -7.44002581e-01 -5.86459041e-01 9.05737400e-01
-1.20100789e-01 -5.18167973e-01 -2.76832700e-01 4.47069734e-01] | [11.711909294128418, -0.4040853679180145] |
28f6af16-773d-4fdf-a3e1-deb12ee143ee | audio-driven-talking-face-generation-with | 2304.08945 | null | https://arxiv.org/abs/2304.08945v1 | https://arxiv.org/pdf/2304.08945v1.pdf | Audio-Driven Talking Face Generation with Diverse yet Realistic Facial Animations | Audio-driven talking face generation, which aims to synthesize talking faces with realistic facial animations (including accurate lip movements, vivid facial expression details and natural head poses) corresponding to the audio, has achieved rapid progress in recent years. However, most existing work focuses on generating lip movements only without handling the closely correlated facial expressions, which degrades the realism of the generated faces greatly. This paper presents DIRFA, a novel method that can generate talking faces with diverse yet realistic facial animations from the same driving audio. To accommodate fair variation of plausible facial animations for the same audio, we design a transformer-based probabilistic mapping network that can model the variational facial animation distribution conditioned upon the input audio and autoregressively convert the audio signals into a facial animation sequence. In addition, we introduce a temporally-biased mask into the mapping network, which allows to model the temporal dependency of facial animations and produce temporally smooth facial animation sequence. With the generated facial animation sequence and a source image, photo-realistic talking faces can be synthesized with a generic generation network. Extensive experiments show that DIRFA can generate talking faces with realistic facial animations effectively. | ['Shijian Lu', 'Xiaoqin Zhang', 'Jiahui Zhang', 'Fangneng Zhan', 'Yingchen Yu', 'Rongliang Wu'] | 2023-04-18 | null | null | null | null | ['talking-face-generation', 'face-generation'] | ['computer-vision', 'computer-vision'] | [ 7.24309832e-02 3.20428193e-01 2.76421398e-01 -5.63994050e-01
-8.21187437e-01 -2.85540700e-01 6.02323174e-01 -1.26963878e+00
3.39423031e-01 5.99212408e-01 4.09532070e-01 3.67949188e-01
3.98961335e-01 -5.05521715e-01 -6.33528233e-01 -9.32323873e-01
-4.05815467e-02 2.32424274e-01 -1.90777496e-01 -2.10167289e-01
-3.26498151e-01 5.94866931e-01 -2.06264400e+00 4.41519052e-01
4.37371165e-01 8.78148854e-01 1.86510548e-01 8.82909894e-01
-1.44692555e-01 7.74166763e-01 -7.97905803e-01 -4.88662392e-01
1.34882201e-02 -6.88182652e-01 -3.92614543e-01 2.41582021e-01
3.98457736e-01 -5.77209949e-01 -3.37273836e-01 9.26277399e-01
6.69933736e-01 1.56038806e-01 8.51238191e-01 -1.73978114e+00
-7.00857341e-01 4.75062490e-01 -6.88767672e-01 -4.37314630e-01
4.96186376e-01 2.08345890e-01 3.13800901e-01 -9.79794204e-01
7.24050105e-01 2.13983154e+00 4.51658964e-01 1.32181263e+00
-1.20062995e+00 -1.22037804e+00 4.43260521e-02 1.10426232e-01
-1.66366529e+00 -9.48585868e-01 1.13237953e+00 -2.33755499e-01
1.14797957e-01 2.52180129e-01 7.61272371e-01 1.42081058e+00
9.96692181e-02 5.89920402e-01 7.52089083e-01 -2.94937342e-01
-4.67217080e-02 3.64256501e-02 -8.73427033e-01 6.52095079e-01
-7.24303722e-01 1.99917570e-01 -5.61126590e-01 -2.36334175e-01
8.65748346e-01 -4.58637953e-01 -3.19943666e-01 1.03659645e-01
-8.95833671e-01 7.42782116e-01 -9.96153355e-02 -7.88279846e-02
-5.38658023e-01 3.95494282e-01 2.57585287e-01 -5.27047366e-02
5.02321839e-01 -1.87111884e-01 2.31794968e-01 -1.33049771e-01
-1.15567994e+00 5.69750965e-01 5.87352514e-01 9.28101122e-01
5.30785918e-01 7.94406354e-01 -3.58457118e-01 9.59570110e-01
5.87163508e-01 9.17903066e-01 2.44079888e-01 -1.58236313e+00
-8.72090980e-02 -2.54150569e-01 8.91137198e-02 -1.08089435e+00
4.75949831e-02 2.54101008e-01 -9.28270102e-01 3.40868473e-01
1.83708131e-01 -5.38103342e-01 -7.74283826e-01 2.43201303e+00
5.88529348e-01 6.19087577e-01 7.96860084e-02 9.14180517e-01
9.69060242e-01 1.20670629e+00 1.31893784e-01 -6.56542122e-01
1.20321620e+00 -4.86874908e-01 -1.34144211e+00 2.62058079e-01
-1.56760171e-01 -9.16354656e-01 1.03363740e+00 2.32777312e-01
-1.36426997e+00 -7.04672635e-01 -6.14923120e-01 -5.36307544e-02
3.10064107e-01 -1.90263484e-02 2.62253910e-01 6.10788286e-01
-1.46298504e+00 1.78610489e-01 -5.31030178e-01 2.03719229e-01
2.53849506e-01 1.94376707e-01 -3.07951361e-01 1.26540795e-01
-1.41547930e+00 4.93878216e-01 -1.77648991e-01 3.05480957e-01
-1.28965664e+00 -8.46342921e-01 -1.11126590e+00 5.13235740e-02
-8.73442367e-02 -6.11378968e-01 1.46018136e+00 -1.57521760e+00
-2.29554248e+00 5.32284737e-01 -7.42009759e-01 1.02942437e-01
5.21889687e-01 2.01880544e-01 -4.87467021e-01 2.49436215e-01
-4.69103195e-02 1.48695326e+00 1.45675814e+00 -1.41934013e+00
-1.47306651e-01 -2.92473473e-03 -4.96639431e-01 2.25490540e-01
-7.55368248e-02 4.72409487e-01 -5.60775876e-01 -7.49608636e-01
-4.71193999e-01 -1.06030166e+00 3.37380692e-02 4.84257370e-01
-2.88770705e-01 -1.73425421e-01 1.31771767e+00 -5.39519668e-01
7.54905164e-01 -2.40478063e+00 1.99722171e-01 6.43614773e-03
-1.70878395e-01 1.42532103e-02 -4.14240122e-01 1.44630998e-01
-3.93398374e-01 -4.53942753e-02 -8.08445439e-02 -7.45928168e-01
-4.10993695e-02 2.63686895e-01 -6.12890899e-01 3.15036565e-01
3.30603361e-01 7.24412620e-01 -6.98307872e-01 -7.61795521e-01
8.21429268e-02 1.28595054e+00 -6.49333596e-01 4.35637712e-01
-3.38924825e-01 7.28771031e-01 -2.46740922e-01 3.65531206e-01
8.44603360e-01 2.83891499e-01 4.44514714e-02 -1.64787322e-01
4.70173135e-02 -2.65928000e-01 -1.15223908e+00 1.40699041e+00
-5.38176239e-01 8.36400688e-01 6.11177385e-01 -3.49137068e-01
1.17931509e+00 8.24770570e-01 4.44394439e-01 -2.27425620e-01
3.02260697e-01 -1.68343648e-01 -2.54909277e-01 -5.65688312e-01
2.08526582e-01 -6.43383384e-01 6.13003373e-02 3.62797290e-01
2.41708066e-02 -5.28269827e-01 -1.99741960e-01 -8.14255923e-02
2.07525790e-01 3.00293356e-01 -3.13263863e-01 -5.13962172e-02
6.76423311e-01 -6.70639694e-01 6.19028389e-01 -4.26060967e-02
-8.81902575e-02 8.33109081e-01 4.50093389e-01 -2.00242400e-01
-9.40891683e-01 -1.14152253e+00 7.45750368e-02 1.02213478e+00
-1.88063998e-02 -2.32355207e-01 -1.22619474e+00 -3.53464410e-02
-3.83151084e-01 6.61584318e-01 -5.86243451e-01 -1.38267756e-01
-5.82590520e-01 -3.06838304e-01 8.55503261e-01 2.69665688e-01
4.52656627e-01 -1.39361787e+00 -4.48057763e-02 9.69742686e-02
-5.94921470e-01 -1.03744137e+00 -8.24969351e-01 -8.14993560e-01
-2.84639329e-01 -6.04475617e-01 -1.20971489e+00 -1.05501688e+00
4.43943977e-01 2.59786323e-02 7.74712622e-01 -3.87221873e-01
-2.48804942e-01 2.06963122e-01 1.83699220e-01 -4.85493213e-01
-9.55170035e-01 -7.50302196e-01 3.75321537e-01 6.04781449e-01
-2.24699557e-01 -7.25416720e-01 -3.53970408e-01 3.16421151e-01
-9.50581431e-01 3.47737074e-01 -1.13361523e-01 6.62828147e-01
4.85935569e-01 2.62089190e-03 8.19094121e-01 -3.29252154e-01
6.64814293e-01 -3.80332947e-01 -4.89650398e-01 -6.85755834e-02
2.09097058e-01 -1.13587588e-01 8.44410658e-01 -9.25125241e-01
-1.69777048e+00 1.27008304e-01 -5.92516184e-01 -1.10606551e+00
-2.41761684e-01 -1.45165995e-01 -5.61463356e-01 8.91360715e-02
5.24460435e-01 2.72717237e-01 5.66337705e-01 -1.32563502e-01
7.35719621e-01 7.55472600e-01 9.32442427e-01 -6.55214369e-01
7.34737158e-01 6.69419944e-01 1.45772159e-01 -1.12855136e+00
-2.76114523e-01 3.43779236e-01 -1.65433794e-01 -6.93899274e-01
8.61413896e-01 -1.06193054e+00 -1.13303721e+00 8.31040621e-01
-1.44467831e+00 -3.80964339e-01 -1.02010190e-01 3.81221563e-01
-1.02708483e+00 3.01711112e-01 -6.98236525e-01 -1.12052894e+00
-2.47059897e-01 -1.43726456e+00 1.32327700e+00 2.81974792e-01
-4.24402565e-01 -5.49765527e-01 6.07489944e-02 -9.88715142e-02
3.82955819e-01 5.97693145e-01 6.07575774e-01 5.10475099e-01
-4.40403163e-01 2.73578435e-01 9.03964490e-02 3.52570504e-01
4.73662645e-01 7.70512938e-01 -1.25668013e+00 1.22104473e-01
-1.83225777e-02 -2.48464793e-01 2.54638702e-01 8.79866421e-01
1.21014595e+00 -6.59569800e-01 -1.15239963e-01 7.53845632e-01
7.29212165e-01 4.13472563e-01 7.07000852e-01 -7.07887471e-01
5.46920598e-01 1.13613260e+00 3.89647365e-01 4.11674023e-01
1.95279554e-01 8.70351315e-01 3.62454623e-01 -1.00662909e-01
-4.18702364e-01 -5.44394195e-01 6.95701718e-01 5.88275492e-01
1.31070390e-02 -2.19385237e-01 -3.18666011e-01 4.76446331e-01
-1.42383969e+00 -1.31236172e+00 2.19115540e-01 1.66826057e+00
9.88775730e-01 -5.23319423e-01 2.66916722e-01 -6.94895536e-03
1.18761897e+00 1.80936411e-01 -5.14212251e-01 -5.73195815e-01
-1.61123350e-02 2.77378917e-01 -3.93511623e-01 7.66478360e-01
-7.02050865e-01 9.48355258e-01 6.91407299e+00 1.07819271e+00
-1.53239977e+00 -2.80127730e-02 7.29442656e-01 -3.19135755e-01
-6.22152448e-01 -4.66090232e-01 -6.42660499e-01 6.61283672e-01
1.13596642e+00 -3.78221631e-01 3.48844886e-01 8.11169028e-01
9.01777208e-01 4.92202371e-01 -8.69131207e-01 1.17936897e+00
1.05313718e-01 -1.37639093e+00 5.85525393e-01 -1.67655706e-01
6.82615340e-01 -7.15508223e-01 5.94770908e-01 -1.74097344e-03
3.29104006e-01 -1.32489765e+00 8.96145642e-01 6.87708735e-01
1.52708697e+00 -1.06631982e+00 1.53309345e-01 3.03575993e-01
-1.33064842e+00 2.53390104e-01 -4.88705598e-02 2.21411422e-01
7.09975719e-01 3.00924052e-02 -7.13779330e-01 1.03376418e-01
6.70987248e-01 5.06394982e-01 2.90086925e-01 4.56776917e-01
-7.72260949e-02 5.88027954e-01 -3.26518506e-01 1.06700763e-01
8.86821654e-03 -2.18380228e-01 6.00982547e-01 1.06562948e+00
6.53822780e-01 2.04573944e-01 -1.09187834e-01 1.14988363e+00
-7.07671717e-02 8.78085047e-02 -8.32552731e-01 1.50616318e-01
5.39787889e-01 1.15127110e+00 -1.73946649e-01 -1.98615775e-01
1.28755495e-02 7.44994044e-01 -2.70239949e-01 5.15863597e-01
-1.15356219e+00 -7.24438354e-02 1.16674590e+00 2.27845967e-01
-6.66096881e-02 3.69039327e-02 3.04211199e-01 -7.59913087e-01
-2.55737931e-01 -9.52009141e-01 -1.56707570e-01 -1.23454428e+00
-9.20234323e-01 9.29008126e-01 1.23115368e-01 -1.07023466e+00
-8.43367994e-01 -1.05821379e-01 -8.28745306e-01 1.16962099e+00
-1.28908634e+00 -1.12987936e+00 -3.59307736e-01 9.59319592e-01
8.02102864e-01 -6.89652702e-03 7.97299564e-01 3.08840424e-01
-3.77101928e-01 8.10488641e-01 -3.12931031e-01 -8.47620070e-02
9.09316599e-01 -5.31198263e-01 3.92898053e-01 3.55912328e-01
-2.48540923e-01 3.56675059e-01 9.95628715e-01 -3.62165600e-01
-1.00594604e+00 -1.31368923e+00 6.48305953e-01 -3.53345275e-02
1.93718016e-01 -4.04983133e-01 -1.00826192e+00 4.63227153e-01
3.21884811e-01 -6.62328601e-02 5.73208094e-01 -7.15834022e-01
-1.63911939e-01 -1.45085052e-01 -1.37824285e+00 8.99568796e-01
8.95111740e-01 -5.37780523e-01 -1.43554345e-01 2.10300937e-01
7.23306000e-01 -3.62616897e-01 -5.58329463e-01 2.79436201e-01
7.33135700e-01 -8.42555046e-01 9.10655320e-01 -4.41482484e-01
4.17465776e-01 -4.58260864e-01 1.09996796e-01 -1.24252784e+00
-9.38947275e-02 -1.33724201e+00 1.49088293e-01 1.52173233e+00
6.23546951e-02 -2.64556348e-01 6.91345692e-01 4.58039612e-01
1.26052096e-01 -4.29632902e-01 -8.92865837e-01 -5.83586037e-01
5.34278043e-02 -3.86591196e-01 9.03752744e-01 6.55270755e-01
-2.19929874e-01 5.14027588e-02 -8.84490848e-01 1.73165530e-01
5.56256711e-01 -9.82410237e-02 8.61474991e-01 -9.58726585e-01
3.08561623e-02 -3.29127222e-01 -1.90258935e-01 -9.01201904e-01
8.53375256e-01 -5.02385378e-01 4.64231163e-01 -9.93693769e-01
-8.95739645e-02 -1.83202982e-01 6.26687944e-01 1.28375411e-01
-7.39112198e-02 3.59000564e-01 2.28084430e-01 -5.30743934e-02
1.83874890e-01 1.09626615e+00 1.59975588e+00 4.33958285e-02
-1.17710277e-01 1.99123789e-02 -4.32001293e-01 9.37394500e-01
4.47265446e-01 -2.51425624e-01 -8.36330593e-01 -9.83288661e-02
-3.75931442e-01 6.90780461e-01 3.27825069e-01 -7.56461263e-01
-5.56706227e-02 -3.27161521e-01 3.35038662e-01 -2.89665878e-01
1.10722101e+00 -5.05744338e-01 7.92252183e-01 1.12232119e-01
-5.32548845e-01 3.29153314e-02 3.28458458e-01 3.30189586e-01
-4.11850095e-01 2.27284119e-01 1.23664105e+00 6.09000698e-02
-1.95628360e-01 6.21120036e-01 -6.88408077e-01 -5.78341335e-02
1.03387773e+00 -1.40809909e-01 1.94247618e-01 -1.29931569e+00
-9.36219871e-01 -3.66196707e-02 3.29154611e-01 5.34697771e-01
7.45561004e-01 -1.66993511e+00 -1.02726305e+00 4.97661084e-01
-4.01140958e-01 -1.64190643e-02 5.89084685e-01 2.05714583e-01
-3.46474618e-01 -2.18316051e-03 -3.96118969e-01 -6.63496971e-01
-1.61683035e+00 4.62691277e-01 5.43024123e-01 4.63535815e-01
-4.88335520e-01 8.99468720e-01 7.42127776e-01 -9.25515071e-02
1.79301456e-01 1.52687684e-01 -3.30193698e-01 -1.63052394e-03
7.88198292e-01 1.13286279e-01 -5.01175940e-01 -1.19086468e+00
-1.08141311e-01 7.59140849e-01 3.38718772e-01 -4.95745748e-01
9.26303089e-01 -1.95699334e-01 1.18412331e-01 2.81559199e-01
1.29835021e+00 1.49078086e-01 -1.73990822e+00 1.90402120e-01
-9.28148627e-01 -4.83327121e-01 -1.26668736e-01 -1.55931339e-01
-1.40338135e+00 1.01812351e+00 3.68901134e-01 -1.99623361e-01
1.21553588e+00 -6.97353780e-02 7.96956778e-01 -3.01805168e-01
2.54714400e-01 -5.76212227e-01 1.86017409e-01 1.56789228e-01
1.16160142e+00 -6.54823899e-01 -6.95880949e-01 -7.11655438e-01
-7.72920132e-01 1.09799790e+00 5.54979742e-01 -4.09896635e-02
7.12298930e-01 7.26778924e-01 3.32872957e-01 1.54914647e-01
-8.70876372e-01 3.36186230e-01 1.14242867e-01 1.01933837e+00
1.79600254e-01 -9.78323370e-02 2.27270707e-01 3.67764205e-01
-5.79298735e-01 1.64858848e-01 6.22473836e-01 1.16398387e-01
-2.33254775e-01 -7.37971187e-01 -7.35098779e-01 -1.67188764e-01
-4.77350771e-01 1.44736143e-02 -2.05254987e-01 5.71958840e-01
2.27002222e-02 9.20113266e-01 2.99111664e-01 -1.33522645e-01
4.41083275e-02 3.21749002e-01 4.85397667e-01 -3.32485020e-01
-5.32222651e-02 3.31218988e-01 -5.31818308e-02 -5.35454452e-01
-4.19667453e-01 -5.61838448e-01 -1.32281208e+00 -6.58019006e-01
3.69143412e-02 2.21803352e-01 4.50511247e-01 5.98118544e-01
4.45610553e-01 3.90548885e-01 9.38165545e-01 -1.43507028e+00
-2.02946678e-01 -8.09717715e-01 -6.61057830e-01 3.19822729e-01
5.78524411e-01 -7.86691427e-01 -5.04412591e-01 6.00874245e-01] | [13.209757804870605, -0.44384264945983887] |
9eec67ed-0b75-4272-8d5f-cd9899565c40 | minding-language-models-lack-of-theory-of | 2306.00924 | null | https://arxiv.org/abs/2306.00924v1 | https://arxiv.org/pdf/2306.00924v1.pdf | Minding Language Models' (Lack of) Theory of Mind: A Plug-and-Play Multi-Character Belief Tracker | Theory of Mind (ToM)$\unicode{x2014}$the ability to reason about the mental states of other people$\unicode{x2014}$is a key element of our social intelligence. Yet, despite their ever more impressive performance, large-scale neural language models still lack basic theory of mind capabilities out-of-the-box. We posit that simply scaling up models will not imbue them with theory of mind due to the inherently symbolic and implicit nature of the phenomenon, and instead investigate an alternative: can we design a decoding-time algorithm that enhances theory of mind of off-the-shelf neural language models without explicit supervision? We present SymbolicToM, a plug-and-play approach to reason about the belief states of multiple characters in reading comprehension tasks via explicit symbolic representation. More concretely, our approach tracks each entity's beliefs, their estimation of other entities' beliefs, and higher-order levels of reasoning, all through graphical representations, allowing for more precise and interpretable reasoning than previous approaches. Empirical results on the well-known ToMi benchmark (Le et al., 2019) demonstrate that SymbolicToM dramatically enhances off-the-shelf neural networks' theory of mind in a zero-shot setting while showing robust out-of-distribution performance compared to supervised baselines. Our work also reveals spurious patterns in existing theory of mind benchmarks, emphasizing the importance of out-of-distribution evaluation and methods that do not overfit a particular dataset. | ['Yulia Tsvetkov', 'Yejin Choi', 'Alane Suhr', 'Peter West', 'Sachin Kumar', 'Melanie Sclar'] | 2023-06-01 | null | null | null | null | ['reading-comprehension'] | ['natural-language-processing'] | [ 2.70093471e-01 6.41073763e-01 -7.83988535e-02 -5.23573637e-01
-3.55093271e-01 -2.51311690e-01 9.51239467e-01 3.85230571e-01
-4.33798701e-01 5.61791360e-01 2.93794423e-01 -5.95711887e-01
-2.49285355e-01 -9.69182849e-01 -9.27114964e-01 -2.51858830e-01
2.17574775e-01 1.04818606e+00 -4.61272709e-02 -3.85533690e-01
1.05644628e-01 -1.26992151e-01 -1.63670886e+00 2.47156724e-01
8.13794374e-01 7.12024212e-01 -7.77195916e-02 4.91715938e-01
-1.00414813e-01 1.52044773e+00 -3.49105239e-01 -8.96990180e-01
-3.59086156e-01 -4.50631231e-01 -1.03135240e+00 -2.52343625e-01
6.35473669e-01 -3.18878829e-01 -2.40395010e-01 1.30206883e+00
7.09017664e-02 6.46895766e-02 8.31862867e-01 -1.04271770e+00
-1.06315362e+00 1.18627346e+00 -5.64051211e-01 2.66295850e-01
5.46816468e-01 3.15033108e-01 1.23091948e+00 -4.45817441e-01
4.65015382e-01 1.57104766e+00 7.18765259e-01 5.34187853e-01
-1.65046918e+00 -6.20901406e-01 3.02023828e-01 2.87858516e-01
-1.20008576e+00 -5.35097778e-01 4.61241186e-01 -3.33982557e-01
1.29021215e+00 8.20617154e-02 5.44453919e-01 1.32063961e+00
6.62166334e-04 9.30019379e-01 1.11904633e+00 -3.79783928e-01
2.94452548e-01 -3.98287363e-02 5.82368076e-01 1.16399229e+00
4.45998460e-01 -3.15184087e-01 -8.86952937e-01 -1.46819456e-02
6.65520787e-01 -1.46681815e-01 8.26656520e-02 -1.87356964e-01
-1.45970964e+00 6.87017977e-01 1.46750897e-01 1.31363720e-01
-4.86603975e-01 5.92540443e-01 1.92615911e-01 1.92799464e-01
4.55795348e-01 5.23211598e-01 -4.22592849e-01 -5.05847335e-01
-1.01530373e+00 4.09457386e-01 8.27666998e-01 5.57335436e-01
5.94932735e-01 -1.91159546e-01 8.97063781e-03 6.45093858e-01
2.62353033e-01 5.55463493e-01 4.70440894e-01 -1.17011654e+00
3.79065871e-01 6.42696559e-01 -1.53691560e-01 -8.94712865e-01
-5.34689903e-01 -4.63719577e-01 -7.94469774e-01 -2.30589579e-03
7.31390119e-01 -2.08337426e-01 -6.01747632e-01 2.23424745e+00
-5.23350611e-02 2.11452603e-01 9.27405506e-02 5.65516651e-01
6.06521308e-01 2.92174399e-01 3.92156765e-02 2.14438498e-01
1.49156499e+00 -6.99954569e-01 -3.53635192e-01 -7.57627666e-01
6.47872329e-01 -9.23641771e-02 1.21656036e+00 6.24812365e-01
-1.50089395e+00 -2.64844000e-01 -9.84304190e-01 -3.57361525e-01
-3.22578073e-01 -1.51663244e-01 1.11458707e+00 5.29715836e-01
-1.15980864e+00 5.84995747e-01 -1.05015969e+00 -1.81131706e-01
6.54507279e-01 3.87006372e-01 -1.94921136e-01 1.26558572e-01
-1.37569094e+00 1.38708174e+00 2.86071867e-01 -3.05908054e-01
-7.69217432e-01 -6.84319258e-01 -9.30657983e-01 2.89166480e-01
5.54592669e-01 -7.69230723e-01 1.35813093e+00 -9.73648369e-01
-1.29143333e+00 1.22052956e+00 -2.83773839e-01 -7.33438432e-01
3.24572533e-01 -2.47107655e-01 -1.32488236e-01 1.40247300e-01
1.22738957e-01 7.49916553e-01 5.80374539e-01 -9.30290878e-01
-2.02856869e-01 -3.40809882e-01 4.74886030e-01 5.63913845e-02
-1.33662105e-01 -7.01297000e-02 -2.20298573e-01 -1.21470243e-01
1.59586981e-01 -6.66713238e-01 2.50101089e-01 -7.06052184e-02
-5.07566214e-01 -5.16840518e-01 -1.01576895e-01 -4.84316915e-01
8.97357643e-01 -1.81810188e+00 4.04273570e-01 3.77995968e-02
6.65287316e-01 -5.31626977e-02 2.69935671e-02 2.48851806e-01
-2.56816834e-01 1.39194801e-01 -5.39052626e-03 -7.63489187e-01
5.31367600e-01 1.91351309e-01 -3.20116341e-01 4.98010814e-01
2.18372151e-01 1.04399014e+00 -9.92066741e-01 -6.68395996e-01
7.84825310e-02 3.42485547e-01 -7.33146012e-01 -2.79733896e-01
-7.02933788e-01 -2.69092321e-01 -2.71837533e-01 4.07278448e-01
2.12429017e-01 -9.11395729e-01 3.18647087e-01 2.24396706e-01
2.85513163e-01 6.97294891e-01 -9.07345831e-01 1.72883570e+00
-1.86570719e-01 6.09680653e-01 -2.49341324e-01 -9.47557449e-01
4.51826453e-01 1.62687600e-01 4.29290645e-02 -7.53006995e-01
4.24984604e-01 -1.28199607e-01 4.53355670e-01 -2.57784694e-01
2.92222202e-01 -2.98213542e-01 -2.39977017e-02 9.83047366e-01
1.11874029e-01 4.69295643e-02 1.87582150e-01 5.84884644e-01
1.13853049e+00 2.48257577e-01 2.46988878e-01 -4.03886199e-01
2.07393989e-01 -1.10866487e-01 3.83854583e-02 9.12828922e-01
9.46170613e-02 2.92771697e-01 1.04280543e+00 -1.97372168e-01
-7.70177722e-01 -9.72409129e-01 -3.60948592e-02 1.48017919e+00
-1.11882994e-02 -3.32536340e-01 -9.84227002e-01 -2.81426162e-01
1.58421531e-01 1.41489923e+00 -1.08641922e+00 -2.71590799e-01
-8.98751467e-02 -6.05329812e-01 1.08266628e+00 4.45343018e-01
5.50139904e-01 -8.98374200e-01 -1.05405426e+00 -9.81282964e-02
-3.78561288e-01 -1.12003160e+00 2.55122453e-01 2.47556597e-01
-5.06074071e-01 -9.53810573e-01 -2.95810074e-01 -2.86544949e-01
5.62851727e-01 -4.64251935e-02 1.63558710e+00 3.93207401e-01
1.23502329e-01 3.83061409e-01 -9.23363045e-02 -6.38017118e-01
-5.08882880e-01 -7.12248683e-02 3.34236681e-01 -2.08159104e-01
8.87396634e-01 -7.31875241e-01 -3.06351095e-01 -8.91462788e-02
-5.78530312e-01 6.15712583e-01 3.74666661e-01 7.55668879e-01
1.06719099e-01 -1.19656865e-02 3.09256166e-01 -9.39610720e-01
6.66742086e-01 -6.34007096e-01 -2.34489262e-01 3.54852051e-01
-4.89039838e-01 4.78002012e-01 4.33930576e-01 -4.43224132e-01
-1.03186691e+00 -6.29562080e-01 -3.17575485e-02 -2.06833780e-01
-7.13759661e-02 7.08672225e-01 2.87182573e-02 3.35955292e-01
8.28201115e-01 4.45042193e-01 2.87853271e-01 -2.64385432e-01
4.79623079e-01 2.55539775e-01 8.01379085e-01 -1.18742502e+00
5.37961423e-01 3.89616758e-01 -5.50665334e-02 -5.97870290e-01
-1.37781382e+00 1.57923743e-01 -5.04161358e-01 -9.43965651e-03
7.37881601e-01 -1.12468600e+00 -1.39245331e+00 4.21246111e-01
-1.16985250e+00 -5.72607696e-01 -1.18217058e-01 1.68832824e-01
-6.63682520e-01 3.25936407e-01 -5.98748982e-01 -9.95877326e-01
-1.90504268e-01 -9.92240369e-01 7.97510624e-01 1.52359039e-01
-8.61682594e-01 -1.29028404e+00 -2.04609022e-01 6.87990546e-01
4.10348982e-01 -2.66930535e-02 1.33315432e+00 -8.97937655e-01
-3.13523889e-01 2.08540969e-02 -3.86350423e-01 1.56321228e-01
-4.94767576e-01 -2.26379544e-01 -1.12623513e+00 6.81280121e-02
1.56406745e-01 -8.29102278e-01 9.71069515e-01 1.94177672e-01
1.06183386e+00 -2.51970738e-01 -1.16034538e-01 3.81512552e-01
1.06387651e+00 -3.33300889e-01 7.93634892e-01 3.04372638e-01
5.18619776e-01 5.26563644e-01 6.29463866e-02 3.83820623e-01
1.05096769e+00 3.52785915e-01 4.06278461e-01 2.17929110e-01
2.35178024e-02 -4.79589581e-01 5.58732033e-01 3.14581931e-01
-2.28161827e-01 -1.98924601e-01 -1.35856473e+00 3.95871758e-01
-1.82686722e+00 -1.24673986e+00 -1.10827580e-01 1.82445967e+00
1.37014496e+00 5.27044415e-01 -6.81758858e-03 5.95864616e-02
1.87809661e-01 1.12123787e-01 -8.32986295e-01 -4.33713526e-01
-1.24457940e-01 3.35328758e-01 1.16901495e-01 5.76062799e-01
-7.02279985e-01 1.00671422e+00 6.20046568e+00 6.99210227e-01
-8.98121178e-01 1.34701192e-01 7.11971402e-01 -3.23092699e-01
-4.54300284e-01 -1.31100103e-01 -7.04490602e-01 2.67702103e-01
1.22988772e+00 -1.31774381e-01 8.14339161e-01 4.95077699e-01
-2.86337197e-01 -4.75377500e-01 -1.67101276e+00 9.21051800e-01
3.30217898e-01 -1.43592799e+00 -6.70639202e-02 1.97440639e-01
4.02468920e-01 2.33032897e-01 1.33164227e-01 5.00631452e-01
7.90063024e-01 -1.46664834e+00 1.20725727e+00 8.30386817e-01
6.83861017e-01 -4.56508964e-01 5.61575294e-01 8.23849261e-01
-4.86610025e-01 8.88469815e-02 -2.58735508e-01 -7.31859267e-01
-4.44152541e-02 3.18055779e-01 -5.17675281e-01 -9.38126352e-03
3.61930937e-01 6.98170006e-01 -9.15432692e-01 2.32479796e-02
-2.59117097e-01 7.72605419e-01 -5.09452581e-01 -4.19389307e-01
1.74228966e-01 1.40115455e-01 2.71309495e-01 1.14001918e+00
-1.13118716e-01 2.69193500e-01 -3.01427394e-01 1.56224275e+00
-1.76545799e-01 -3.42464626e-01 -3.44220042e-01 -4.27461118e-01
3.02696526e-01 7.39184141e-01 -6.03283763e-01 -6.55840218e-01
-3.80325645e-01 6.40605628e-01 6.59108996e-01 1.40004680e-01
-1.01997542e+00 2.27297544e-01 7.03859329e-01 7.73467794e-02
1.78583801e-01 -2.09304214e-01 -6.86874628e-01 -1.49861789e+00
-2.24712566e-01 -1.30168855e+00 7.77896792e-02 -9.55795705e-01
-1.31143165e+00 -2.91303117e-02 2.62954950e-01 -2.13055432e-01
-4.38673943e-01 -8.68617237e-01 -4.00591075e-01 8.02833378e-01
-1.02008092e+00 -1.29141057e+00 -3.59258950e-02 4.63502586e-01
4.04813617e-01 -2.44692504e-01 1.04443419e+00 -3.02642196e-01
-5.63907623e-01 5.80840945e-01 -5.14340438e-02 3.49423319e-01
2.65392393e-01 -1.35274839e+00 7.78245151e-01 7.12850213e-01
3.23248684e-01 1.19114077e+00 9.51842844e-01 -5.02178490e-01
-1.61064327e+00 -2.65574336e-01 9.24438000e-01 -1.05813301e+00
1.13387358e+00 -3.78490776e-01 -1.01285219e+00 1.01158047e+00
2.81760424e-01 -5.43210566e-01 6.80004954e-01 8.82884681e-01
-8.41061234e-01 3.53764564e-01 -9.10831213e-01 8.65706503e-01
1.05819285e+00 -8.25526297e-01 -1.14136016e+00 2.98872828e-01
5.96489489e-01 -3.57870758e-01 -7.23166764e-01 -1.95976138e-01
7.19731271e-01 -1.22659600e+00 1.10072720e+00 -8.15612614e-01
1.20270967e+00 9.79747400e-02 -2.41248742e-01 -1.04758370e+00
-3.01340699e-01 -3.63690376e-01 -3.66933346e-01 1.01394892e+00
4.63107437e-01 -7.28393316e-01 7.25174665e-01 1.19887257e+00
3.13407063e-01 -7.80095220e-01 -6.58664465e-01 -3.88391346e-01
4.44372535e-01 -7.72046626e-01 5.77355206e-01 9.67711806e-01
5.51059186e-01 6.59838796e-01 -6.01549707e-02 4.45924029e-02
9.69205678e-01 -1.56625032e-01 6.09434009e-01 -1.45331132e+00
-7.76787281e-01 -6.79127038e-01 -3.41583788e-01 -1.02594411e+00
5.73473990e-01 -1.05774176e+00 -1.20062709e-01 -1.61324799e+00
6.27029479e-01 -1.25368908e-01 -2.03199819e-01 9.02293801e-01
-1.19057551e-01 4.00863923e-02 2.79881984e-01 2.54152808e-02
-9.68408108e-01 2.46723294e-01 1.00638032e+00 -1.30922750e-01
5.20078182e-01 -5.38051844e-01 -1.11649120e+00 1.19349766e+00
5.73633313e-01 -2.83668667e-01 -4.99893188e-01 -7.65405595e-01
9.99704301e-01 1.18151046e-02 9.61213946e-01 -1.00990975e+00
5.42013526e-01 -1.28031716e-01 2.23777816e-01 -3.63627434e-01
6.62644804e-01 -3.21446985e-01 -6.89379275e-02 3.58311385e-01
-7.70746589e-01 3.58382314e-02 2.17568785e-01 4.15806592e-01
2.74086982e-01 -2.89667875e-01 5.86025059e-01 -3.55782449e-01
-5.26250184e-01 -3.91117722e-01 -3.32609624e-01 4.83949363e-01
5.60729861e-01 -8.53479579e-02 -6.78046048e-01 -5.71149051e-01
-3.74925703e-01 1.47118166e-01 5.76210797e-01 1.50134295e-01
3.37670743e-01 -7.00388789e-01 -7.17840075e-01 5.29480800e-02
-4.01430428e-02 -1.12840526e-01 1.04157448e-01 1.00243819e+00
-4.05494839e-01 3.91588986e-01 -1.50987566e-01 -3.97690535e-01
-1.03919506e+00 2.74012834e-01 5.27922988e-01 -1.94637030e-01
-5.17040491e-01 1.24375820e+00 2.06424996e-01 -3.79738182e-01
2.30840117e-01 -4.51996356e-01 -4.57651652e-02 2.29936559e-02
6.08018637e-01 1.35173917e-01 -3.30906838e-01 -4.13138032e-01
-4.28226560e-01 1.29639074e-01 -1.86141998e-01 -2.61824667e-01
1.35883152e+00 -3.10239736e-02 -5.33793092e-01 7.84218132e-01
7.10885584e-01 -2.88871378e-01 -1.08025682e+00 -4.57073331e-01
-9.69620720e-02 -1.01205990e-01 1.04427353e-01 -1.18558621e+00
-5.52937090e-01 1.24918270e+00 -6.36521205e-02 2.02924684e-01
5.38501978e-01 1.09456278e-01 4.66661334e-01 8.94998372e-01
5.56301355e-01 -1.00021541e+00 7.18188612e-03 8.36287260e-01
4.85141337e-01 -1.22457981e+00 2.37548932e-01 1.48656532e-01
-8.76746356e-01 8.87571454e-01 4.48616266e-01 -1.29562523e-02
3.30744416e-01 2.08112374e-01 -3.23215812e-01 -4.86956477e-01
-1.32066393e+00 -2.76331939e-02 7.11112767e-02 4.24049586e-01
6.64234638e-01 2.00989366e-01 2.82078594e-01 8.48833263e-01
-6.21553421e-01 1.44934759e-01 4.93098885e-01 5.26568294e-01
-3.86662841e-01 -5.58986962e-01 -2.23769948e-01 5.34263790e-01
-3.95201504e-01 -5.10506749e-01 -4.23759818e-01 6.92490578e-01
1.60592154e-01 9.04930055e-01 1.50778979e-01 -3.03519785e-01
2.55078822e-02 4.56324160e-01 7.42193758e-01 -5.35013855e-01
-5.50475836e-01 -5.32806575e-01 4.42993253e-01 -5.45206606e-01
-2.51160651e-01 -9.20134485e-01 -1.51519322e+00 -9.73395765e-01
1.62335858e-01 -2.82147020e-01 3.70275646e-01 1.46518362e+00
3.45084876e-01 5.72414994e-01 -4.60175455e-01 -6.96885228e-01
-8.09762776e-01 -9.06569541e-01 -3.80381227e-01 5.28679252e-01
1.69061527e-01 -7.57319450e-01 -2.51707047e-01 1.24893762e-01] | [9.543479919433594, 7.32581901550293] |
ad5cec0d-c772-4afc-a1e4-89373047d3d9 | malware-detection-and-prevention-using | 2206.12770 | null | https://arxiv.org/abs/2206.12770v1 | https://arxiv.org/pdf/2206.12770v1.pdf | Malware Detection and Prevention using Artificial Intelligence Techniques | With the rapid technological advancement, security has become a major issue due to the increase in malware activity that poses a serious threat to the security and safety of both computer systems and stakeholders. To maintain stakeholders, particularly, end users security, protecting the data from fraudulent efforts is one of the most pressing concerns. A set of malicious programming code, scripts, active content, or intrusive software that is designed to destroy intended computer systems and programs or mobile and web applications is referred to as malware. According to a study, naive users are unable to distinguish between malicious and benign applications. Thus, computer systems and mobile applications should be designed to detect malicious activities towards protecting the stakeholders. A number of algorithms are available to detect malware activities by utilizing novel concepts including Artificial Intelligence, Machine Learning, and Deep Learning. In this study, we emphasize Artificial Intelligence (AI) based techniques for detecting and preventing malware activity. We present a detailed review of current malware detection technologies, their shortcomings, and ways to improve efficiency. Our study shows that adopting futuristic approaches for the development of malware detection applications shall provide significant advantages. The comprehension of this synthesis shall help researchers for further research on malware detection and prevention using AI. | ['Fan Wu', 'Akond Rahman', 'Dan Lo', 'Alfredo Cuzzocreak', 'Michael Whitman', 'Md Abdullah Khan', 'Shahriar Sobhan', 'Farhat Lamia Barsha', 'Maria Valero', 'Hossain Shahriar', 'Md Jobair Hossain Faruk'] | 2022-06-26 | null | null | null | null | ['novel-concepts'] | ['reasoning'] | [ 3.41549248e-01 -4.49831128e-01 -5.63293099e-01 2.14649960e-01
9.96291935e-02 -8.32429230e-01 6.84412837e-01 6.39962479e-02
-6.24764711e-02 2.45715275e-01 -2.21974403e-01 -6.65601730e-01
2.57041484e-01 -9.52614784e-01 -2.56006271e-01 -5.50867438e-01
9.96342003e-02 -9.58913341e-02 3.39426875e-01 -1.90518320e-01
8.65957141e-01 6.09697223e-01 -1.48821390e+00 4.98315811e-01
9.25416529e-01 9.19256985e-01 -1.02491081e-02 5.41469693e-01
-3.70535672e-01 1.08392072e+00 -7.97552466e-01 -3.50060672e-01
1.60571411e-01 -2.25418702e-01 -6.35400534e-01 -2.04046860e-01
-3.46255571e-01 -6.00702405e-01 -9.25926343e-02 1.60883832e+00
-1.79724768e-01 -5.34502029e-01 5.34792840e-01 -1.60783947e+00
-7.26471722e-01 3.98565859e-01 -6.55471623e-01 2.62098014e-01
3.84824127e-01 3.50649655e-01 4.11409885e-01 -4.62372571e-01
2.14592129e-01 7.73515105e-01 5.34722388e-01 6.88156962e-01
-5.38653255e-01 -7.46881306e-01 9.57984924e-02 1.95835456e-01
-1.03132951e+00 -6.49918094e-02 8.49386334e-01 -8.28975797e-01
9.15684283e-01 5.09058237e-01 7.09606290e-01 1.46156645e+00
6.14011943e-01 6.09537482e-01 7.12228417e-01 -3.32588583e-01
2.76745111e-01 7.29622781e-01 5.05784273e-01 7.70735323e-01
8.23897541e-01 1.76633224e-01 1.72721639e-01 -6.68478549e-01
1.05351113e-01 6.39618039e-01 -1.07903711e-01 -1.44129187e-01
-7.16548383e-01 1.09533334e+00 8.78178924e-02 6.78204238e-01
-3.26031297e-01 -2.18289778e-01 9.03232813e-01 3.75639535e-02
2.43857518e-01 4.73345757e-01 -3.32901001e-01 -2.81765997e-01
-3.71519834e-01 -2.17072010e-01 9.39363539e-01 3.87344837e-01
2.33985499e-01 3.92160296e-01 5.34522474e-01 3.39374334e-01
5.43463290e-01 5.11165917e-01 7.74112940e-01 -4.88180935e-01
3.26649398e-02 1.14577639e+00 -1.84857175e-01 -1.54375947e+00
7.33839199e-02 6.46880567e-02 -7.30347335e-01 2.26279780e-01
-1.22735694e-01 -2.03482620e-02 -6.04136527e-01 1.05972850e+00
1.15843602e-01 -3.34272049e-02 -1.23230502e-01 2.55317748e-01
4.79862899e-01 9.50126112e-01 1.34589285e-01 -3.50682765e-01
1.38673532e+00 -5.63407481e-01 -7.89520860e-01 -3.39526623e-01
5.74266195e-01 -4.51714635e-01 7.95712829e-01 4.98372734e-01
-4.29068536e-01 -1.77678019e-01 -1.15007985e+00 5.86042047e-01
-9.14370596e-01 -3.21046919e-01 6.75013363e-01 1.34893978e+00
-4.91556108e-01 1.91144556e-01 -8.72200310e-01 -3.70402753e-01
5.68133891e-01 3.61665606e-01 5.76333925e-02 3.53240699e-01
-1.07951295e+00 9.21252191e-01 4.33361590e-01 -2.31344745e-01
-1.07608306e+00 -1.88610271e-01 -6.74871683e-01 -4.17043455e-02
3.34490508e-01 4.73913364e-02 1.02675426e+00 -1.24238694e+00
-1.01248157e+00 8.36754084e-01 3.39664012e-01 -5.56545556e-01
2.06040591e-01 -1.20614499e-01 -5.33026874e-01 -8.55652019e-02
-1.09335504e-01 -2.59349078e-01 1.12555063e+00 -1.24258375e+00
-6.86742067e-01 -5.87651849e-01 -6.60243630e-02 -3.61049086e-01
-9.98794317e-01 4.80970532e-01 1.48674771e-01 -3.30568820e-01
-2.52805948e-01 -7.88078547e-01 1.20346196e-01 -4.45472032e-01
-3.76029104e-01 -7.84786120e-02 1.64998984e+00 -9.10921037e-01
1.59301269e+00 -2.08752394e+00 -1.71866551e-01 2.64780402e-01
3.85118634e-01 9.49386239e-01 2.73917377e-01 2.82744288e-01
1.25506490e-01 6.34022295e-01 -2.32137099e-01 6.48046494e-01
-4.09798354e-01 -6.45965114e-02 -5.86765289e-01 4.35696214e-01
2.88170762e-02 6.25021815e-01 -8.21126223e-01 -1.71892107e-01
3.65173489e-01 3.82603467e-01 -2.60573059e-01 2.06585675e-01
-2.14897975e-01 2.18269855e-01 -8.82351339e-01 1.15199602e+00
5.65958261e-01 -1.51144609e-01 3.66516799e-01 2.02121496e-01
-8.25437531e-02 1.34881958e-01 -5.12579918e-01 4.04792607e-01
-3.33757430e-01 6.93414330e-01 1.59969434e-01 -1.18222368e+00
6.12208128e-01 4.13371921e-01 5.40132642e-01 -4.25563127e-01
4.88177299e-01 2.36627921e-01 1.47649899e-01 -1.05742228e+00
9.99315456e-02 4.98838603e-01 -4.14156131e-02 6.37836099e-01
-5.74843109e-01 1.41484499e-01 -1.61914632e-01 -5.76085551e-03
1.47811615e+00 -4.21198577e-01 8.06291580e-01 1.03534259e-01
1.01183188e+00 2.04698250e-01 3.40874314e-01 3.76770347e-01
-7.37703979e-01 -4.31818217e-01 6.24473035e-01 -5.41750848e-01
-8.71019363e-01 -7.82668829e-01 1.54418275e-01 1.17540622e+00
2.33251508e-02 -1.64815515e-01 -1.07758987e+00 -1.14103770e+00
-6.22336380e-02 5.36638379e-01 -3.84197563e-01 -5.00945687e-01
-6.15364313e-01 -7.91952252e-01 6.71147585e-01 2.15168029e-01
6.99558675e-01 -1.32013893e+00 -7.92756379e-01 -5.55648804e-02
-7.49559402e-02 -8.85472059e-01 -9.61966068e-02 -1.06033370e-01
-7.97096372e-01 -1.52085507e+00 -9.63388756e-02 -8.61716568e-01
7.46069491e-01 5.78760564e-01 6.55894041e-01 6.40089154e-01
-3.90181273e-01 3.48696858e-01 -5.29521227e-01 -7.21981406e-01
-1.06236982e+00 -5.65975234e-02 1.56484604e-01 -1.92085400e-01
8.37381065e-01 -2.00108096e-01 -1.10408314e-01 1.21970415e-01
-9.91343796e-01 -4.74688530e-01 5.49110174e-01 4.83914346e-01
-2.69886971e-01 6.17999494e-01 2.96997070e-01 -1.02620959e+00
9.28385973e-01 -8.15821588e-01 -7.03908086e-01 2.07606152e-01
-7.06857622e-01 -2.63732284e-01 8.34614754e-01 -6.54311121e-01
-8.04884553e-01 -1.45818274e-02 2.75996942e-02 -1.35369942e-01
-2.69855052e-01 1.43169001e-01 -4.15898114e-01 -3.18852544e-01
8.23926806e-01 5.70378602e-01 2.62920439e-01 -5.37158586e-02
-9.02192295e-02 1.33048093e+00 -5.36064617e-02 -4.77973372e-03
7.79795587e-01 5.37426531e-01 -4.89055842e-01 -1.02049029e+00
-2.84407288e-01 -3.77716571e-01 -1.01071380e-01 -4.38236713e-01
7.60677457e-01 -1.99668080e-01 -9.14798379e-01 8.02756429e-01
-1.31725693e+00 4.43237215e-01 4.09331322e-01 2.63289623e-02
3.89823988e-02 6.85855865e-01 -7.01505005e-01 -1.18904138e+00
-6.56857669e-01 -1.35162055e+00 5.46960175e-01 4.87024300e-02
-3.62473667e-01 -8.05151105e-01 1.61702543e-01 5.86884201e-01
4.63809878e-01 2.04240039e-01 1.07002687e+00 -1.01627028e+00
-5.13511360e-01 -7.10221171e-01 -4.55584861e-02 6.49868667e-01
4.70026284e-01 6.13512158e-01 -8.09063792e-01 -4.12143432e-02
6.47512197e-01 9.66033116e-02 3.12830180e-01 8.60724971e-02
1.22320426e+00 -9.25444961e-01 -6.82810843e-01 1.91227853e-01
1.19826961e+00 1.20123458e+00 5.76928973e-01 4.10374373e-01
6.98463976e-01 8.01087677e-01 3.91425818e-01 3.34274411e-01
-1.62160426e-01 8.05180445e-02 8.18553686e-01 3.92893374e-01
7.42018282e-01 7.83201307e-02 8.14649284e-01 6.43547952e-01
-1.57727882e-01 -1.11915715e-01 -1.16704631e+00 1.97519362e-01
-1.54158449e+00 -1.31388652e+00 -1.29662588e-01 2.03944087e+00
3.29296619e-01 2.86209285e-01 1.30704328e-01 4.37823653e-01
1.05497336e+00 4.66304161e-02 -5.53480923e-01 -6.60760045e-01
5.41716397e-01 -1.66340649e-01 4.04027402e-01 5.56599386e-02
-1.38180196e+00 7.21017122e-01 6.50916433e+00 6.81457639e-01
-1.31600296e+00 2.98814863e-01 4.71298248e-01 3.30819815e-01
4.65377234e-02 -1.64428979e-01 -4.49288338e-01 9.84532952e-01
1.00398207e+00 -1.52507856e-01 6.15738153e-01 1.48592460e+00
2.31004909e-01 3.71027701e-02 -5.80952942e-01 6.89740002e-01
2.55093157e-01 -1.08655763e+00 5.73922060e-02 3.01569551e-01
3.69105011e-01 -2.72294343e-01 2.27274418e-01 2.80215293e-01
1.13440372e-01 -7.22854435e-01 2.31431261e-01 3.25887874e-02
1.39796168e-01 -8.49348485e-01 9.04771924e-01 6.80314243e-01
-1.00947428e+00 -6.06057465e-01 3.86296958e-02 -1.93701461e-01
-2.87553489e-01 2.96347708e-01 -8.14543784e-01 -2.83172093e-02
7.16269612e-01 4.45079803e-01 -3.98012906e-01 4.09749597e-01
-9.19936001e-02 5.46098351e-01 9.98919681e-02 -6.40046775e-01
4.86444347e-02 -2.06765637e-01 4.80064422e-01 8.89986694e-01
2.25394685e-02 -1.42682284e-01 1.84708431e-01 8.03750455e-01
1.91016376e-01 -4.46101204e-02 -1.38602042e+00 -9.31576848e-01
3.82678092e-01 1.18907118e+00 -8.49444032e-01 -3.04967403e-01
-6.73914254e-01 7.59006441e-01 -6.27695024e-02 -1.33132860e-01
-9.69425559e-01 -4.21646982e-01 7.95686185e-01 3.57642949e-01
-2.99349427e-01 -2.05652907e-01 -6.25086129e-01 -9.26770627e-01
-1.09359525e-01 -1.30148089e+00 3.19713801e-01 -1.76978782e-01
-1.16503143e+00 5.64189672e-01 -2.54007280e-01 -1.19104135e+00
-9.56497267e-02 -9.14548874e-01 -7.82191873e-01 3.14158827e-01
-7.75737762e-01 -9.39799130e-01 -2.81764597e-01 2.74510026e-01
5.95001101e-01 -5.92847764e-01 6.21867239e-01 7.80388489e-02
-7.99835324e-01 3.43474858e-02 -2.10255310e-02 5.42914212e-01
7.50742480e-02 -5.33869147e-01 2.58024931e-01 1.08048487e+00
-2.56713122e-01 8.93186390e-01 3.56954694e-01 -1.15499938e+00
-1.59716642e+00 -8.47529709e-01 5.54484606e-01 -6.90986872e-01
7.60708511e-01 -3.12775731e-01 -9.40663278e-01 7.56012261e-01
1.52613580e-01 -4.39573109e-01 8.08428466e-01 -3.19602460e-01
-4.43561018e-01 1.18639253e-01 -1.46363556e+00 6.44220710e-01
5.21155894e-01 -5.87353885e-01 -4.72848088e-01 3.83606613e-01
5.30750036e-01 1.85090452e-01 -1.77430674e-01 4.80426639e-01
4.55303341e-01 -1.10360146e+00 8.46774757e-01 -6.52200043e-01
3.35271358e-01 -5.72711378e-02 1.12500437e-01 -4.97346550e-01
-1.76182479e-01 -3.42873394e-01 -7.44309008e-01 9.33567405e-01
2.15275474e-02 -8.01398456e-01 8.06596398e-01 6.26196921e-01
3.89359981e-01 -5.33340693e-01 -4.43211883e-01 -6.52588606e-01
-2.49125436e-01 -4.45293158e-01 3.27041775e-01 1.23972571e+00
3.58905882e-01 4.53587808e-02 -3.02171141e-01 1.37956336e-01
4.80372936e-01 -3.35431546e-01 5.26481032e-01 -1.24353814e+00
3.97771820e-02 -7.10447073e-01 -4.72570062e-01 -3.14834118e-01
1.06239900e-01 -5.74741602e-01 -3.43843400e-01 -1.03355563e+00
5.34309626e-01 1.52444974e-01 -2.49854904e-02 4.58023846e-01
5.94497919e-02 1.37909257e-04 -6.77049160e-02 4.87429231e-01
-2.37644345e-01 8.38495195e-02 5.73926687e-01 -3.89695585e-01
-2.43489742e-01 3.04470032e-01 -7.77322233e-01 1.21740139e+00
1.05409002e+00 -4.55889791e-01 -2.89484471e-01 7.49516189e-02
1.85325429e-01 -3.71217251e-01 2.14562908e-01 -8.16500723e-01
1.94036644e-02 -5.69741845e-01 1.74930885e-01 -4.64389354e-01
-8.85781460e-03 -1.17646480e+00 2.00159140e-02 1.38189495e+00
1.27916679e-01 9.45513248e-02 3.93502973e-02 5.35608649e-01
4.83285971e-02 -5.41175604e-01 9.61866140e-01 -4.52800423e-01
-8.20473969e-01 5.06235734e-02 -1.23249412e+00 -3.92540753e-01
1.87542319e+00 -3.16664517e-01 -4.16823596e-01 -2.91514784e-01
-1.20178208e-01 -1.15946673e-01 4.65396672e-01 6.76891446e-01
7.09898233e-01 -8.81812394e-01 -8.69560912e-02 3.35990071e-01
5.59453517e-02 -8.17024589e-01 2.51430757e-02 5.13193846e-01
-9.26672757e-01 6.17659330e-01 -3.76529187e-01 -3.02567035e-01
-1.74666274e+00 1.15292943e+00 1.63189501e-01 -3.02078605e-01
-1.95720822e-01 3.59379619e-01 -2.78657191e-02 -1.94530547e-01
3.91051918e-01 1.19940847e-01 -5.97086310e-01 -1.95385680e-01
9.04099047e-01 7.20770001e-01 -1.42023951e-01 -5.46672285e-01
-6.04699075e-01 1.09669738e-01 -3.71695340e-01 4.35218632e-01
8.47701430e-01 2.54828900e-01 -4.96046126e-01 5.91749214e-02
9.71923053e-01 1.19839706e-01 -4.04936045e-01 3.14047366e-01
3.09613049e-01 -5.97323596e-01 -1.82222590e-01 -6.86862588e-01
-9.88110483e-01 8.55770171e-01 6.25682592e-01 1.06080377e+00
1.17306840e+00 -3.01901549e-01 7.93947577e-01 3.88802677e-01
5.56823850e-01 -8.60221088e-01 4.15139526e-01 4.78267312e-01
5.29370725e-01 -1.35665655e+00 -2.05510199e-01 -5.07423520e-01
-6.20566964e-01 1.08579969e+00 9.43008304e-01 -1.73076037e-02
7.81496048e-01 5.60958564e-01 -2.30343696e-02 -2.10209429e-01
-2.99040735e-01 2.58491248e-01 -8.69662091e-02 7.98093677e-01
3.65583479e-01 5.29611371e-02 -4.31572080e-01 4.69199866e-01
3.65791291e-01 -2.55456358e-01 5.21252632e-01 1.40656698e+00
-8.76048505e-01 -1.12305796e+00 -7.87510931e-01 6.01414382e-01
-8.25684607e-01 -1.03139700e-02 -1.13834321e+00 5.55547178e-01
2.49521822e-01 1.07982147e+00 -1.85935706e-01 -8.89869690e-01
-1.32158920e-01 9.67123508e-02 -8.00664127e-02 -6.27372086e-01
-7.82598555e-01 -4.69522148e-01 -2.47223198e-01 -3.70567709e-01
-9.68905911e-02 -3.65154386e-01 -1.19761539e+00 -7.44618237e-01
-4.98182863e-01 -3.92707027e-02 9.98166680e-01 8.05318117e-01
4.23857868e-01 3.66552293e-01 8.18185687e-01 -3.18334699e-01
-5.72001278e-01 -7.46297181e-01 -1.57833770e-01 3.56539220e-01
4.24587801e-02 -5.05663514e-01 -5.70374072e-01 -4.38193232e-02] | [14.413896560668945, 9.675056457519531] |
5f1ef12b-ee58-47dc-a9b3-7f060429b294 | a-hybrid-neuro-symbolic-approach-for-text | null | null | https://openreview.net/forum?id=NzjyY2Z9zJd | https://openreview.net/pdf?id=NzjyY2Z9zJd | A Hybrid Neuro-Symbolic approach for Text-Based Games using Inductive Logic Programming | Text-based games (TBGs) have emerged as an important test-bed, requiring reinforcement learning (RL) agents to combine natural language understanding with reasoning. A key challenge for agents solving this task is to generalize across multiple games and shows good results on both seen and unseen objects. Currently, pure deep learning-based RL systems can perform well to known entities and states. They, however, perform poorly in novel situations e.g., when handling out-of-vocabulary (OOV) objects. In the perspective of generalization, recent efforts in infusing external commonsense knowledge into an RL agent show better results than pure deep-learning systems. However, the policies learned by these systems are not interpretable or easily transferable. To tackle these issues, we have designed a hybrid neuro-symbolic framework for TBGs that uses symbolic reasoning along with the neural RL model. It employs inductive logic programming (ILP) to learn the symbolic rules (policies) as default theory with exceptions and is represented in the form of an answer-set-program (ASP) that allows performing non-monotonic reasoning in the partially observable game environment. We use WordNet as an external knowledge source to lift the learned rules to their generalized versions. These rules are learned in an online manner and applied with an ASP solver to predict an action for the agent. We show that the agents that incorporate the neuro-symbolic hybrid approach with the generalized rules outperform the baseline agents. | ['Gopal Gupta', 'Mrinmaya Sachan', 'Murray Campbell', 'Tim Klinger', 'Kartik Talamadupula', 'Pavan Kapanipathi', 'Mattia Atzeni', 'Keerthiram Murugesan', 'Kinjal Basu'] | 2021-11-21 | null | null | null | aaai-workshop-clear-2022-2 | ['inductive-logic-programming', 'text-based-games'] | ['methodology', 'playing-games'] | [-9.33727846e-02 5.52825332e-01 -8.10359195e-02 -1.50240004e-01
-1.54785082e-01 -5.70443928e-01 6.88564599e-01 -1.46505788e-01
-5.42991161e-01 1.00301921e+00 -1.13325574e-01 -4.94109333e-01
-2.71278381e-01 -1.57950509e+00 -9.24993813e-01 -2.59522676e-01
-1.25244990e-01 1.05958974e+00 6.64028466e-01 -1.03699350e+00
-1.12602703e-01 2.68581510e-01 -1.65081501e+00 5.56202292e-01
1.04340267e+00 7.14399457e-01 1.31208599e-01 4.56126630e-01
-2.40568593e-01 1.64029121e+00 -5.15493572e-01 -5.47998846e-01
4.73607540e-01 -4.67330515e-01 -1.09764969e+00 -4.04648036e-01
-1.71879232e-02 -5.82749605e-01 -3.25182855e-01 1.18162441e+00
1.96441244e-02 5.08131742e-01 3.19380403e-01 -1.40334773e+00
-7.17228770e-01 1.03690302e+00 3.78244609e-01 -1.03102110e-01
7.14746594e-01 7.13830233e-01 1.20267558e+00 -2.68571407e-01
7.36571193e-01 1.48828852e+00 3.85596275e-01 1.10033011e+00
-1.13823533e+00 -4.28843737e-01 2.71595418e-01 7.08099246e-01
-1.08523858e+00 3.48606729e-03 4.86251414e-01 -1.63180128e-01
1.75885129e+00 1.58543333e-01 9.76904392e-01 1.13213575e+00
1.01455145e-01 8.86351407e-01 1.30947089e+00 -5.88773012e-01
6.17477596e-01 1.94412500e-01 -1.05787583e-01 8.04659784e-01
2.87168801e-01 7.47611642e-01 -4.97980237e-01 2.94498485e-02
7.38985479e-01 -2.09410667e-01 1.23697303e-01 -4.23909277e-01
-8.61333072e-01 1.08304358e+00 6.07117176e-01 2.14906693e-01
-4.55119133e-01 4.58498895e-01 6.25673592e-01 6.55704081e-01
-2.44246535e-02 1.08459377e+00 -6.37599170e-01 -8.82957205e-02
-4.51959610e-01 7.92210698e-01 1.19598782e+00 8.83412242e-01
7.05582798e-01 3.55143815e-01 -4.53768522e-02 4.91101742e-01
3.83443713e-01 5.82535267e-01 6.42486393e-01 -1.19922876e+00
1.58402577e-01 8.60165894e-01 -1.10810094e-01 -8.43049347e-01
-4.41448510e-01 -2.27778539e-01 -2.75416166e-01 5.16619444e-01
2.85424441e-01 -2.04821885e-01 -8.31507981e-01 2.03873563e+00
1.62740827e-01 4.72239792e-01 6.95954084e-01 6.16993904e-01
1.01016295e+00 6.41684473e-01 3.10457408e-01 -3.77744995e-02
1.09833205e+00 -9.14334655e-01 -3.87571603e-01 -6.67789876e-01
8.29211593e-01 5.00737846e-01 1.17520905e+00 5.85528016e-01
-1.09377432e+00 -2.23478869e-01 -1.09318292e+00 1.21106125e-01
-8.52227151e-01 -7.43761420e-01 8.65588129e-01 3.54965359e-01
-1.21223855e+00 5.78927994e-01 -6.42584860e-01 -4.12184268e-01
3.18818629e-01 6.66006327e-01 -3.85442339e-02 -2.14784145e-02
-1.90621471e+00 1.46307850e+00 1.32368875e+00 -2.43757159e-01
-1.17754459e+00 -3.43732566e-01 -1.17639554e+00 1.17125377e-01
1.12315321e+00 -8.73621523e-01 1.51737964e+00 -1.14739406e+00
-2.19492602e+00 7.51067519e-01 5.94172001e-01 -9.38584745e-01
3.54605168e-01 4.58032601e-02 -3.03553313e-01 1.50641147e-02
-2.99493242e-02 7.31981039e-01 5.04859984e-01 -1.11405885e+00
-6.52492523e-01 -2.21523777e-01 9.99505281e-01 4.15790796e-01
2.05505058e-01 -1.31176561e-01 -4.35581841e-02 -2.69152939e-01
-1.58801943e-01 -9.95187342e-01 -2.33518556e-01 -5.71930587e-01
-2.16446042e-01 -4.48218048e-01 3.11006069e-01 -2.53675491e-01
9.06907380e-01 -1.69225287e+00 3.22418809e-01 2.05636874e-01
-2.77088750e-02 3.86752278e-01 -1.72646612e-01 3.07216436e-01
3.15897539e-02 -1.44804548e-02 8.66652653e-02 3.54777962e-01
4.51772511e-01 8.59552622e-01 -5.49317658e-01 -3.78889114e-01
1.68475702e-01 1.31020558e+00 -1.30526590e+00 -2.34020621e-01
2.58872986e-01 -4.02317256e-01 -1.03712177e+00 2.41228744e-01
-1.26722062e+00 1.59194589e-01 -4.55147892e-01 1.90760657e-01
1.02481268e-01 -8.11381415e-02 5.24407923e-01 3.33100647e-01
3.98408473e-01 4.71850455e-01 -1.20873284e+00 1.50757420e+00
-4.77320611e-01 1.01486661e-01 -4.66419876e-01 -1.06499207e+00
7.16100514e-01 1.63923621e-01 -2.33511120e-01 -8.01278949e-01
1.21716872e-01 1.42146617e-01 4.51814204e-01 -6.66476369e-01
2.28835508e-01 -6.88680828e-01 -3.45569789e-01 3.32992405e-01
3.99521023e-01 -6.42172277e-01 3.79520237e-01 2.67868131e-01
1.16769934e+00 5.56505919e-01 7.39484549e-01 9.38014984e-02
6.53616846e-01 6.11179590e-01 6.35645807e-01 1.14317286e+00
1.06016539e-01 -3.21428150e-01 5.92026472e-01 -7.41497755e-01
-5.90948641e-01 -8.83136749e-01 4.76288676e-01 1.61179030e+00
7.45331720e-02 -4.67444688e-01 -7.00606942e-01 -6.50499046e-01
-5.87343946e-02 1.33337963e+00 -4.47011620e-01 -6.44161642e-01
-5.53474486e-01 -3.61280322e-01 9.03154433e-01 5.79153657e-01
7.38224268e-01 -1.89990938e+00 -1.03454149e+00 4.76716906e-01
8.70037153e-02 -1.04936218e+00 4.41195726e-01 4.49853092e-01
-6.39277339e-01 -1.06926906e+00 3.62336844e-01 -5.44266045e-01
1.04264900e-01 -4.84607786e-01 1.24334252e+00 2.57778227e-01
2.34026909e-01 4.49040532e-01 -4.06986147e-01 -6.27495050e-01
-8.42924237e-01 -3.69246267e-02 2.28013068e-01 -5.69393218e-01
6.69277728e-01 -5.09407938e-01 5.00038601e-02 2.84403097e-02
-9.82768178e-01 2.01566786e-01 2.61332601e-01 9.52156961e-01
1.36809558e-01 9.82965007e-02 7.50935912e-01 -9.84350324e-01
9.54739034e-01 -4.55938548e-01 -8.43204021e-01 3.78199130e-01
-4.87518460e-01 3.89446586e-01 1.03356254e+00 -4.34152603e-01
-1.17073441e+00 -3.21491420e-01 8.33943561e-02 -1.60953313e-01
-3.43238682e-01 9.07139778e-01 -3.00634533e-01 6.42633662e-02
1.09155309e+00 4.18876112e-01 -4.94254045e-02 2.05900341e-01
6.57783329e-01 3.23919863e-01 5.72951972e-01 -1.30926740e+00
7.71888971e-01 5.50220385e-02 -7.20131770e-02 -2.59849459e-01
-1.05419540e+00 2.64547374e-02 -3.05054426e-01 -6.19383641e-02
7.13331759e-01 -6.79954171e-01 -1.07501590e+00 3.14217538e-01
-9.00964141e-01 -1.11969447e+00 -6.53781474e-01 2.13192701e-01
-1.13761723e+00 -8.38363767e-02 -6.60028577e-01 -6.77758873e-01
-7.80379251e-02 -1.13820422e+00 4.30826157e-01 2.34466583e-01
-2.63604909e-01 -1.06334698e+00 2.30976313e-01 1.43361896e-01
4.10404652e-01 7.60374814e-02 1.29708087e+00 -1.43832457e+00
-5.19380629e-01 1.57240063e-01 8.74035954e-02 3.64395052e-01
-2.23826915e-01 -3.69799197e-01 -8.58324409e-01 5.32172509e-02
-5.32631688e-02 -9.29822743e-01 4.32115257e-01 1.24735266e-01
9.02297556e-01 -4.96346563e-01 2.06769397e-03 4.17284817e-01
1.29944599e+00 5.33108950e-01 7.42038846e-01 9.12999690e-01
1.95419490e-01 4.14868146e-01 5.73791981e-01 2.97024280e-01
6.63311601e-01 6.83392107e-01 5.89139044e-01 6.50913477e-01
1.00871816e-01 -3.75002921e-01 6.99319363e-01 3.05172026e-01
-4.62720275e-01 1.77376881e-01 -1.18366241e+00 2.34680846e-01
-2.14775324e+00 -1.39527941e+00 4.43882585e-01 1.80547476e+00
1.41339004e+00 4.69989181e-01 1.18089374e-03 -2.57740736e-01
2.18976498e-01 -1.20297618e-01 -8.84502172e-01 -7.89063513e-01
-1.47805110e-01 6.44155681e-01 8.98021534e-02 6.16587460e-01
-7.02017725e-01 1.81745780e+00 6.05925369e+00 7.33006954e-01
-9.41252410e-01 4.55341255e-03 -1.11926021e-02 3.01892441e-02
-2.00196374e-02 7.06549734e-02 -6.17581725e-01 -4.50832732e-02
1.06871212e+00 -2.77388841e-01 1.15815079e+00 1.06534338e+00
-1.53630987e-01 -1.50298834e-01 -1.34695804e+00 6.46666408e-01
5.27021773e-02 -1.34081745e+00 3.20663631e-01 -1.99850440e-01
5.84316552e-01 1.88989893e-01 -8.59801099e-02 1.36138725e+00
1.18066311e+00 -1.18675053e+00 8.44366491e-01 4.42192554e-01
4.37489510e-01 -5.76203287e-01 7.46514261e-01 7.54499078e-01
-5.95029950e-01 -5.07144749e-01 -4.51627254e-01 -6.22447431e-01
-2.68305421e-01 -3.96547735e-01 -1.02804649e+00 4.56096977e-01
4.09327358e-01 5.50454736e-01 -3.89842480e-01 6.08132422e-01
-9.56367671e-01 2.67105192e-01 -3.79270971e-01 -3.00056159e-01
4.30124402e-01 -1.50211379e-01 5.29826939e-01 7.53456354e-01
-7.06053302e-02 6.67525947e-01 3.44101697e-01 1.23352611e+00
8.87514353e-02 -2.43363366e-01 -9.07782912e-01 1.03526548e-01
3.19670796e-01 7.61582375e-01 -3.85456592e-01 -6.98470414e-01
-3.04317594e-01 5.80778480e-01 6.61018312e-01 4.09385353e-01
-7.87815511e-01 -3.44790891e-02 4.74272847e-01 -1.85595304e-01
2.18077064e-01 3.82276513e-02 1.68354977e-02 -1.35850430e+00
-4.65460628e-01 -1.47740257e+00 5.61763883e-01 -9.86037374e-01
-1.21417332e+00 5.16469121e-01 4.62456226e-01 -6.99457228e-01
-8.78409147e-01 -1.12065721e+00 -4.83010530e-01 4.86159176e-01
-1.36782253e+00 -1.08619773e+00 -1.37284279e-01 9.32761669e-01
4.27352130e-01 -5.26599705e-01 1.18701279e+00 -4.86249655e-01
-3.14167142e-01 2.80681491e-01 -4.45929080e-01 1.74210668e-01
1.77391961e-01 -1.38936722e+00 4.48126234e-02 5.56096971e-01
8.00934732e-02 6.96813762e-01 7.09241152e-01 -6.55549526e-01
-1.42048776e+00 -1.06385601e+00 2.23352939e-01 -5.36845744e-01
9.33655381e-01 -2.33671129e-01 -9.09916043e-01 1.24879813e+00
2.21677795e-01 -2.38526352e-02 5.16884089e-01 5.00927344e-02
-5.22882581e-01 6.90388978e-02 -1.15284240e+00 9.47583139e-01
1.22126865e+00 -4.99767482e-01 -1.51559818e+00 2.90366054e-01
8.26885521e-01 -7.02493429e-01 -5.97269773e-01 2.02351034e-01
8.06472823e-02 -8.83644164e-01 8.87273669e-01 -1.43723261e+00
5.12486279e-01 -4.64056224e-01 -3.17289561e-01 -1.70570195e+00
-3.04695696e-01 -4.25883293e-01 -3.05265099e-01 5.09003520e-01
3.97834122e-01 -8.68867338e-01 5.44255197e-01 8.87092352e-01
-4.71043326e-02 -5.80014288e-01 -8.12120020e-01 -1.00149238e+00
3.02679390e-01 -9.14065421e-01 7.50064313e-01 8.16127896e-01
7.35032797e-01 4.97632504e-01 -5.32234367e-03 1.00459620e-01
2.38497883e-01 1.03542261e-01 7.09161401e-01 -1.22381556e+00
-6.56579971e-01 -4.31780636e-01 -4.32755560e-01 -6.96605623e-01
8.29707384e-01 -1.27364278e+00 1.58582032e-01 -1.46019101e+00
5.37578054e-02 -4.04177755e-01 -1.96942359e-01 1.04072511e+00
1.79960519e-01 -2.61657447e-01 3.54072273e-01 -2.12470099e-01
-9.23971355e-01 3.91781300e-01 1.18129623e+00 -2.90515810e-01
-2.94462889e-01 -1.78710938e-01 -6.47270143e-01 1.09694207e+00
9.13384557e-01 -3.26325029e-01 -6.20178640e-01 -2.10674673e-01
1.05533171e+00 1.96124669e-02 5.73293149e-01 -9.74150062e-01
5.57152569e-01 -8.39179158e-01 -1.90280527e-01 1.34605378e-01
2.43080288e-01 -8.17622781e-01 -8.88825879e-02 6.45758271e-01
-4.53062624e-01 -1.50085494e-01 4.84154373e-01 3.36387545e-01
-1.49464801e-01 -5.36457181e-01 5.76622248e-01 -6.73947811e-01
-1.16236305e+00 -5.30715622e-02 -5.07937968e-01 4.94733602e-01
1.30172729e+00 -2.24180758e-01 -4.40905094e-01 -5.31246483e-01
-1.11311722e+00 3.32085758e-01 2.64312088e-01 2.03482524e-01
6.78472400e-01 -1.14314568e+00 -5.02708673e-01 3.87566835e-02
1.65973514e-01 2.76806384e-01 -1.91394329e-01 5.09957016e-01
-6.29636049e-01 4.04766798e-01 -5.07997811e-01 -1.66858509e-01
-6.65241182e-01 7.49874115e-01 9.79938030e-01 -5.71145892e-01
-5.26079178e-01 6.80513740e-01 1.48645937e-01 -9.88444746e-01
1.42288163e-01 -6.99397147e-01 -4.41392094e-01 -3.15955609e-01
5.43823957e-01 -2.23030197e-03 2.21692082e-02 -1.16894484e-01
-3.12935352e-01 3.47535126e-02 -8.61091614e-02 -1.58419177e-01
1.64659464e+00 4.75633472e-01 -2.66441852e-01 2.75439054e-01
2.60831505e-01 -4.29415762e-01 -9.08065498e-01 -4.94618833e-01
2.49302149e-01 4.00215611e-02 -2.68329293e-01 -1.22942293e+00
-6.41913772e-01 4.43887979e-01 -8.56326520e-02 3.87083441e-01
8.25988054e-01 1.22616723e-01 1.93258762e-01 1.10824776e+00
7.72666097e-01 -1.20658481e+00 1.40447289e-01 1.38755810e+00
9.71960545e-01 -1.00401962e+00 -2.09663689e-01 -1.12070538e-01
-8.03336680e-01 1.22771776e+00 1.13439214e+00 -3.88291657e-01
7.78641775e-02 2.65733689e-01 -2.37237737e-01 -3.72116685e-01
-1.19071758e+00 -4.26236182e-01 -5.23267139e-04 9.68422949e-01
-1.94867119e-01 2.13494897e-01 2.31453955e-01 8.26540709e-01
-5.14088690e-01 3.22968453e-01 6.31513476e-01 8.95105898e-01
-4.85908657e-01 -1.05222487e+00 -2.76345730e-01 3.66993189e-01
-1.01945668e-01 -3.32755983e-01 -3.79046381e-01 1.00654566e+00
3.07084233e-01 8.31250429e-01 -3.33981156e-01 -3.81738305e-01
5.10803878e-01 3.44565094e-01 8.19660306e-01 -1.22277129e+00
-8.44657898e-01 -7.05684066e-01 2.85879195e-01 -8.08413982e-01
-3.11678797e-01 -4.01608229e-01 -1.74001396e+00 -2.73787469e-01
-7.94295520e-02 2.16285954e-03 4.56607118e-02 1.36978698e+00
1.14087900e-02 6.61440670e-01 -1.05398446e-01 -4.72042173e-01
-1.02463913e+00 -7.46902347e-01 -3.27027470e-01 5.43541551e-01
-4.41589355e-02 -6.89702332e-01 -1.68980971e-01 -2.33495727e-01] | [3.869879722595215, 1.312391996383667] |
09c2b7e0-d525-4cb8-96dc-08d9ef6b620e | guided-transformer-leveraging-multiple | 2006.07548 | null | https://arxiv.org/abs/2006.07548v1 | https://arxiv.org/pdf/2006.07548v1.pdf | Guided Transformer: Leveraging Multiple External Sources for Representation Learning in Conversational Search | Asking clarifying questions in response to ambiguous or faceted queries has been recognized as a useful technique for various information retrieval systems, especially conversational search systems with limited bandwidth interfaces. Analyzing and generating clarifying questions have been studied recently but the accurate utilization of user responses to clarifying questions has been relatively less explored. In this paper, we enrich the representations learned by Transformer networks using a novel attention mechanism from external information sources that weights each term in the conversation. We evaluate this Guided Transformer model in a conversational search scenario that includes clarifying questions. In our experiments, we use two separate external sources, including the top retrieved documents and a set of different possible clarifying questions for the query. We implement the proposed representation learning model for two downstream tasks in conversational search; document retrieval and next clarifying question selection. Our experiments use a public dataset for search clarification and demonstrate significant improvements compared to competitive baselines. | ['W. Bruce Croft', 'Hamed Zamani', 'Helia Hashemi'] | 2020-06-13 | null | null | null | null | ['conversational-search', 'question-selection'] | ['natural-language-processing', 'natural-language-processing'] | [ 3.72279614e-01 3.55772823e-01 -3.41116130e-01 -4.35365081e-01
-1.27992678e+00 -6.01804316e-01 8.89025390e-01 -5.43977367e-03
-5.04674196e-01 9.30348814e-01 9.79000628e-01 -6.15053892e-01
-4.23986554e-01 -4.09922093e-01 -1.79726765e-01 -1.02108665e-01
6.39544129e-01 9.22333777e-01 2.50022382e-01 -6.71626151e-01
7.25366056e-01 1.59507439e-01 -1.46993446e+00 8.63008380e-01
1.21509695e+00 9.11050558e-01 4.83978689e-01 7.39348233e-01
-8.76941264e-01 7.40407884e-01 -1.02197516e+00 -3.94915193e-01
-2.51028061e-01 -3.47637773e-01 -1.66389918e+00 -1.68125898e-01
4.04916584e-01 -3.70785028e-01 -2.71840572e-01 6.11761928e-01
5.45923412e-01 5.34078419e-01 5.24240375e-01 -9.51356113e-01
-1.08598435e+00 8.54329288e-01 -1.00826561e-01 6.84642076e-01
6.29369438e-01 -2.81104952e-01 1.34941339e+00 -9.45904851e-01
3.85264128e-01 1.55518556e+00 1.37840863e-02 6.08101904e-01
-7.40959764e-01 -7.70633519e-01 4.27962750e-01 6.13176227e-01
-7.89347708e-01 -5.09841800e-01 8.38960528e-01 4.99114208e-02
1.15608203e+00 6.00678742e-01 1.24428459e-01 1.13754249e+00
-1.62768900e-01 1.11760306e+00 8.27124357e-01 -2.80515403e-01
-2.12347820e-01 4.67360139e-01 9.13023591e-01 3.35334808e-01
-4.08433581e-04 -2.86882281e-01 -5.55891156e-01 -5.67643166e-01
2.50401258e-01 -2.07943410e-01 -6.30392253e-01 9.74833816e-02
-8.06449533e-01 8.93364310e-01 4.26232338e-01 3.85919601e-01
-5.29115796e-01 6.52781501e-02 1.98736072e-01 5.12207210e-01
3.80919367e-01 1.04358256e+00 -5.23137093e-01 -1.40795305e-01
-4.59085166e-01 2.61833757e-01 1.09785688e+00 1.09746718e+00
5.34307837e-01 -4.67981547e-01 -7.87248909e-01 1.32952237e+00
2.22265914e-01 2.69329220e-01 8.47610056e-01 -9.68578696e-01
7.58646727e-01 7.89881945e-01 4.72550422e-01 -1.01333332e+00
-3.03550363e-01 -7.09146261e-01 -6.41144931e-01 -8.91110957e-01
5.33213913e-02 -2.14768738e-01 -4.25873309e-01 1.56816673e+00
6.93510799e-03 -1.75663278e-01 2.25308105e-01 9.86005545e-01
1.47105980e+00 7.34233558e-01 -7.44064301e-02 -1.06963716e-01
1.67479217e+00 -1.42433453e+00 -1.06788409e+00 -2.62139738e-01
2.22920284e-01 -1.10101473e+00 1.35092413e+00 1.85622144e-02
-1.23695314e+00 -3.96192878e-01 -5.07337928e-01 -7.19902098e-01
-3.37663531e-01 3.11007619e-01 4.98141080e-01 2.24653006e-01
-1.16752911e+00 -1.90415785e-01 5.98130859e-02 -4.11934853e-01
-3.00521534e-02 2.30864033e-01 1.59587592e-01 1.77796781e-02
-1.74838138e+00 8.59837890e-01 -4.75712210e-01 2.07641557e-01
-5.92843652e-01 -6.19900703e-01 -4.09653962e-01 7.80444026e-01
7.10145712e-01 -1.00236630e+00 1.90330255e+00 -3.80901784e-01
-1.47945786e+00 7.25862086e-01 -5.98346531e-01 -3.80588621e-01
2.24394463e-02 -6.17704153e-01 -3.86946380e-01 1.63113132e-01
4.02960300e-01 5.58308005e-01 7.24674702e-01 -1.27062523e+00
-4.53686446e-01 -2.42809474e-01 5.67361951e-01 5.50487936e-01
-2.03283593e-01 5.30196540e-02 -7.17191458e-01 -4.24934745e-01
3.06394063e-02 -7.01062202e-01 -7.82146156e-02 -4.70781982e-01
-6.07700229e-01 -9.74379897e-01 7.71271884e-01 -6.38840616e-01
1.44611788e+00 -1.44809186e+00 1.01693459e-01 -2.19668206e-02
1.59761071e-01 2.46947750e-01 -4.52184469e-01 8.10757279e-01
1.15140617e-01 2.95126617e-01 2.08220735e-01 -2.67942309e-01
1.54508531e-01 -1.76873118e-01 -9.02999043e-01 -5.56260228e-01
-3.19351666e-02 1.12603629e+00 -7.80369043e-01 -4.30017591e-01
-3.49285483e-01 1.90101728e-01 -3.99032235e-01 7.57789433e-01
-3.02889347e-01 6.91062957e-02 -9.39114809e-01 5.60533822e-01
4.05007601e-01 -6.72428370e-01 -1.40169799e-01 -2.77878761e-01
3.68360490e-01 8.57900023e-01 -6.92104042e-01 1.53789222e+00
-8.42334986e-01 6.03556395e-01 3.48893613e-01 -6.99415386e-01
1.00138605e+00 3.28298509e-01 3.44546326e-02 -8.18499267e-01
4.38693416e-04 -5.12484387e-02 -2.48143837e-01 -8.42270315e-01
1.01088560e+00 2.25405499e-01 -2.28559181e-01 1.19533539e+00
-2.15963751e-01 -3.99142712e-01 6.62069395e-02 7.74218500e-01
1.04496396e+00 -3.74170095e-01 7.73442760e-02 -2.69636095e-01
9.71348763e-01 -9.56374779e-02 -2.44017497e-01 1.23318839e+00
-5.19312434e-02 2.14600965e-01 4.07802850e-01 -8.97163525e-02
-1.50370315e-01 -6.51847661e-01 2.50358373e-01 1.55796099e+00
3.70601565e-01 -4.01846647e-01 -5.14496922e-01 -7.51213610e-01
-9.18704923e-03 7.95094252e-01 -4.52066988e-01 -3.83937687e-01
-5.19546032e-01 -1.06138282e-01 4.34471846e-01 2.60788918e-01
6.46566689e-01 -1.21491110e+00 -4.11139071e-01 1.42005254e-02
-7.94977307e-01 -9.27764237e-01 -1.17041886e+00 -2.45842338e-02
-6.68659508e-01 -1.26928461e+00 -9.28066254e-01 -9.56010818e-01
3.09792072e-01 9.91024196e-01 1.52243912e+00 4.69354153e-01
1.12760618e-01 9.99091804e-01 -2.04051495e-01 -1.60831988e-01
1.19125925e-01 6.40376270e-01 -5.52756786e-01 -1.45015880e-01
5.54624498e-01 -2.64918566e-01 -9.73715365e-01 6.67802930e-01
-8.91373932e-01 -6.16492108e-02 4.78093624e-01 1.08450544e+00
-5.94480261e-02 -7.85102069e-01 1.12672007e+00 -9.38222170e-01
2.20335793e+00 -8.13872397e-01 -2.83663183e-01 7.36162484e-01
-6.78065836e-01 4.59991425e-01 1.78020835e-01 -2.92978078e-01
-1.55591321e+00 -6.79551363e-01 6.87802152e-04 1.18809633e-01
1.15811303e-01 6.83058262e-01 9.56526995e-02 3.29479635e-01
5.62402010e-01 1.92213237e-01 -2.02823326e-01 -6.79420650e-01
5.74938297e-01 1.03576767e+00 2.66208410e-01 -7.93737113e-01
3.79151911e-01 -1.52571619e-01 -7.51925111e-01 -6.94463789e-01
-1.01876903e+00 -1.04497433e+00 1.56631052e-01 -1.21161260e-01
4.47554350e-01 -5.42433858e-01 -1.11966217e+00 -1.48275957e-01
-1.63732088e+00 2.12229177e-01 4.43250127e-03 9.22997221e-02
-2.85594493e-01 4.31329846e-01 -4.06903476e-01 -9.86458600e-01
-8.37582648e-01 -1.38455153e+00 1.45904160e+00 4.45164233e-01
-4.26378727e-01 -6.00302517e-01 2.10476354e-01 1.09512794e+00
1.00799692e+00 -7.85889387e-01 1.36919820e+00 -9.92093623e-01
-7.95452118e-01 1.24383710e-01 -4.49669451e-01 -1.75047100e-01
3.98711920e-01 -7.73377180e-01 -9.97038901e-01 5.99037521e-02
1.01839125e-01 -5.61143517e-01 1.13822496e+00 2.70435642e-02
1.51710796e+00 -6.37184262e-01 -4.20008063e-01 -1.11125678e-01
5.50827324e-01 2.75307298e-01 3.91436517e-01 9.66141671e-02
-6.28175959e-02 1.13934588e+00 4.91164744e-01 2.28264928e-01
5.91125131e-01 6.44811153e-01 1.83837146e-01 2.00225309e-01
-1.55896127e-01 -2.10151419e-01 -2.34327242e-01 6.99816823e-01
4.21110004e-01 -7.17087686e-01 -6.02027178e-01 6.66416705e-01
-1.66844499e+00 -1.03954053e+00 2.64617205e-01 1.75405598e+00
1.08076251e+00 -3.56246203e-01 -3.61341476e-01 -3.74777138e-01
8.09691787e-01 3.31781864e-01 -8.03902626e-01 -5.16639531e-01
3.21419127e-02 3.05086136e-01 -4.88418490e-01 1.05576456e+00
-5.09561658e-01 9.29654241e-01 5.96015167e+00 4.25479054e-01
-6.94978476e-01 -1.66576385e-01 6.93758428e-01 9.97586027e-02
-9.32683527e-01 -9.37422067e-02 -7.63801217e-01 1.93843663e-01
5.68107128e-01 -6.40989959e-01 3.98264855e-01 7.60694206e-01
1.30675569e-01 -3.47634070e-02 -1.15155816e+00 1.04356909e+00
3.86038750e-01 -1.57368457e+00 6.91476822e-01 -4.78222489e-01
3.49426448e-01 -3.03383410e-01 9.33308154e-02 8.02787960e-01
2.66826659e-01 -1.12505066e+00 -2.24438235e-01 6.30137146e-01
2.99748272e-01 -4.11607325e-01 7.98990786e-01 4.30704117e-01
-8.89637530e-01 -7.63890743e-02 -1.27800494e-01 2.66939521e-01
3.01694840e-01 7.07312971e-02 -1.23405814e+00 2.05721781e-01
5.56737065e-01 8.62755552e-02 -5.29697418e-01 7.15413213e-01
-1.22650251e-01 3.60523939e-01 -5.49647547e-02 -6.96517110e-01
4.47781712e-01 -4.51159775e-02 4.33742672e-01 1.06865263e+00
2.85875529e-01 5.25551200e-01 -2.47783765e-01 8.20604146e-01
-6.39805377e-01 3.16738933e-01 -4.96569484e-01 -8.67071226e-02
7.55967617e-01 1.21190131e+00 1.19208775e-01 -4.87129331e-01
-9.96939614e-02 7.68775403e-01 3.84522974e-01 6.96903527e-01
-3.16378206e-01 -6.63747013e-01 4.92949456e-01 -1.33450791e-01
-6.13586009e-02 1.20574124e-01 3.34458262e-01 -1.25777769e+00
1.17682576e-01 -1.29087591e+00 6.78290188e-01 -1.31041276e+00
-1.33925521e+00 1.05729496e+00 1.13013647e-01 -6.11847937e-01
-6.28535151e-01 -1.76619783e-01 -7.46835589e-01 1.28028262e+00
-2.09059882e+00 -6.21562421e-01 -6.14660323e-01 6.36346400e-01
1.24383235e+00 -1.06978275e-01 7.33248889e-01 1.63571596e-01
-1.97839603e-01 4.75046426e-01 -2.70000100e-01 -2.88160980e-01
1.06615090e+00 -1.11156917e+00 9.90910605e-02 2.09933639e-01
7.18153119e-02 1.31352389e+00 4.69756842e-01 -4.74573404e-01
-1.45830142e+00 -4.54607487e-01 1.36094809e+00 -3.80793214e-01
4.68529940e-01 -1.49524003e-01 -1.17164493e+00 5.60750544e-01
1.20677638e+00 -7.78442979e-01 1.01478899e+00 4.10900503e-01
-1.38228774e-01 -4.70467955e-02 -8.43374610e-01 7.90004492e-01
8.68718147e-01 -6.78316712e-01 -1.23668671e+00 6.32819891e-01
1.13699853e+00 -2.75988817e-01 -2.66689271e-01 1.52108997e-01
4.13634628e-01 -6.23470664e-01 9.52000141e-01 -1.20322227e+00
1.26162559e-01 1.38045266e-01 4.80390191e-02 -1.31159735e+00
-1.94601610e-01 -8.75726819e-01 -7.28335530e-02 1.15631974e+00
6.99156225e-01 -7.15257823e-01 7.21277356e-01 9.67020214e-01
-2.21900687e-01 -8.20481598e-01 -8.75275493e-01 4.05071303e-02
-1.70084283e-01 8.43100846e-02 1.11310494e+00 5.68782806e-01
4.77487475e-01 1.22054935e+00 -3.03777546e-01 -1.02047637e-01
-6.29960746e-02 6.59828365e-01 8.15143466e-01 -1.21701622e+00
-5.38384356e-02 -6.11467659e-01 7.60408998e-01 -2.22327805e+00
4.89577591e-01 -8.72777164e-01 2.41571832e-02 -1.83497226e+00
3.69277716e-01 -2.68231869e-01 -1.61560163e-01 -6.99030608e-02
-3.42841923e-01 -6.25673711e-01 -1.21728085e-01 2.34046355e-01
-6.53238356e-01 9.54915285e-01 1.44925106e+00 -5.78598142e-01
-8.91549811e-02 3.06757122e-01 -1.43206489e+00 2.65463829e-01
5.58025897e-01 -1.36080254e-02 -9.38671708e-01 -8.88972700e-01
2.86315918e-01 5.63626945e-01 2.82294098e-02 -1.52889043e-01
6.76405072e-01 -1.61626890e-01 -1.59175962e-01 -9.19368446e-01
6.16104364e-01 -6.64127111e-01 -5.79106629e-01 6.75710589e-02
-1.46589327e+00 6.14461482e-01 7.64675215e-02 7.36838102e-01
-4.40027922e-01 -4.81687486e-01 1.36406645e-02 -2.53642678e-01
-2.80534893e-01 1.23778515e-01 -6.51650727e-01 3.62550855e-01
1.66398630e-01 3.38017970e-01 -8.45531464e-01 -1.29597223e+00
-3.95446360e-01 7.38746345e-01 -4.56919938e-01 9.41952586e-01
9.66209590e-01 -1.15405464e+00 -6.42533183e-01 -2.18074575e-01
2.27229759e-01 -1.79168597e-01 1.00933075e-01 3.15381706e-01
2.29662508e-01 1.32324100e+00 2.63391554e-01 -3.27734977e-01
-1.29557657e+00 3.13807189e-01 2.53202200e-01 -5.68034470e-01
1.89404376e-02 8.56845319e-01 3.63753021e-01 -5.78058958e-01
5.25956929e-01 -7.42866337e-01 -9.48164105e-01 1.41412854e-01
6.04417026e-01 3.30737621e-01 8.77609849e-02 -1.89120963e-01
-1.43369317e-01 3.92804027e-01 -5.20598233e-01 -1.81592703e-01
8.52699399e-01 -6.43784583e-01 -1.52825639e-01 -1.21254073e-02
1.04127491e+00 6.85876086e-02 -1.92083657e-01 -6.51357889e-01
4.27150905e-01 -3.34969193e-01 -1.16572507e-01 -1.06675744e+00
-7.58722663e-01 9.04253960e-01 2.65693188e-01 5.18350840e-01
9.54289436e-01 3.79224032e-01 1.08335841e+00 1.20166516e+00
2.12747091e-03 -9.87005949e-01 5.80601931e-01 7.77879357e-01
1.71390808e+00 -1.37044775e+00 -3.52814585e-01 -3.04466695e-01
-5.71436346e-01 9.83401418e-01 1.00475764e+00 3.46688956e-01
3.41332376e-01 -4.10139263e-01 2.59107471e-01 -5.74338555e-01
-1.23148656e+00 -8.74557719e-02 5.24475753e-01 2.19816998e-01
7.34069645e-01 -5.30299067e-01 -5.22122741e-01 9.26741898e-01
-1.99777827e-01 -1.20928749e-01 3.81109178e-01 5.08702934e-01
-5.44796526e-01 -1.08388543e+00 -1.67870924e-01 6.27400935e-01
-3.52109998e-01 -5.21946907e-01 -9.73978877e-01 3.32880884e-01
-9.29506063e-01 1.54347479e+00 -1.60131276e-01 3.60447392e-02
3.97754580e-01 2.48258948e-01 -2.15632364e-01 -7.33490646e-01
-1.03495181e+00 -2.51404971e-01 6.80644751e-01 -5.45587599e-01
-3.11618179e-01 -1.38336271e-01 -7.65363097e-01 3.00243795e-01
-7.22210109e-01 1.27072608e+00 4.54759747e-01 8.78379881e-01
1.02610314e+00 4.39226210e-01 6.00770295e-01 -1.43247217e-01
-9.51080799e-01 -1.34698367e+00 1.37665316e-01 2.78886288e-01
5.67885876e-01 -6.40150011e-01 -4.71045822e-01 -3.94227862e-01] | [12.149691581726074, 7.789054870605469] |
c0d5440a-fc9c-464b-8048-0ab81e2959ef | exclusive-autoencoder-xae-for-nucleus | 1811.11243 | null | http://arxiv.org/abs/1811.11243v1 | http://arxiv.org/pdf/1811.11243v1.pdf | eXclusive Autoencoder (XAE) for Nucleus Detection and Classification on Hematoxylin and Eosin (H&E) Stained Histopathological Images | In this paper, we introduced a novel feature extraction approach, named
exclusive autoencoder (XAE), which is a supervised version of autoencoder (AE),
able to largely improve the performance of nucleus detection and classification
on hematoxylin and eosin (H&E) histopathological images. The proposed XAE can
be used in any AE-based algorithm, as long as the data labels are also provided
in the feature extraction phase. In the experiments, we evaluated the
performance of an approach which is the combination of an XAE and a fully
connected neural network (FCN) and compared with some AE-based methods. For a
nucleus detection problem (considered as a nucleus/non-nucleus classification
problem) on breast cancer H&E images, the F-score of the proposed XAE+FCN
approach achieved 96.64% while the state-of-the-art was at 84.49%. For nucleus
classification on colorectal cancer H&E images, with the annotations of four
categories of epithelial, inflammatory, fibroblast and miscellaneous nuclei.
The F-score of the proposed method reached 70.4%. We also proposed a lymphocyte
segmentation method. In the step of lymphocyte detection, we have compared with
cutting-edge technology and gained improved performance from 90% to 98.67%. We
also proposed an algorithm for lymphocyte segmentation based on nucleus
detection and classification. The obtained Dice coefficient achieved 88.31%
while the cutting-edge approach was at 74%. | ['Chao-Hui Huang', 'Daniel Racoceanu'] | 2018-11-27 | null | null | null | null | ['miscellaneous'] | ['miscellaneous'] | [ 1.07193477e-02 2.25527778e-01 3.22968394e-01 -1.18033020e-02
-2.83749759e-01 -2.69265443e-01 4.16672826e-01 5.89062274e-01
-9.67690825e-01 9.74522173e-01 -9.24717486e-02 -1.48818985e-01
-1.84541509e-01 -1.00764275e+00 -3.01103562e-01 -1.22543490e+00
9.88479145e-03 5.97308278e-01 4.12312858e-02 1.80080645e-02
1.14261158e-01 8.06827009e-01 -1.64118469e+00 3.61408383e-01
5.55963874e-01 8.37162733e-01 -2.81452574e-02 1.15298450e+00
1.46231269e-02 8.80516529e-01 -5.02501249e-01 -2.18511373e-02
-9.76387486e-02 -2.17085108e-01 -8.47855866e-01 -6.29905015e-02
-3.40818577e-02 -3.27695340e-01 7.33669102e-02 8.76229286e-01
6.16576970e-01 -1.22111805e-01 1.22221053e+00 -1.02056086e+00
-2.19830230e-01 4.02018219e-01 -4.53959256e-01 2.92290986e-01
-9.36116800e-02 -1.69642121e-02 3.66725504e-01 -7.51594484e-01
8.74048591e-01 5.32965124e-01 7.27523863e-01 5.81721783e-01
-8.15358579e-01 -5.04463196e-01 -7.70963490e-01 3.86498362e-01
-1.55297852e+00 9.52061340e-02 4.28532571e-01 -6.47543907e-01
9.39205766e-01 2.79458821e-01 9.63671565e-01 4.36063111e-01
5.40021896e-01 6.53020561e-01 1.54305661e+00 -6.67747259e-01
4.43281710e-01 5.58355570e-01 4.58493769e-01 7.71359265e-01
5.16750872e-01 5.75249409e-03 2.48571217e-01 -9.20923799e-02
6.44128084e-01 1.73394140e-02 -8.66198465e-02 3.35158139e-01
-9.70487773e-01 8.45659137e-01 3.02440226e-01 1.12147570e+00
-7.21993923e-01 -1.29151881e-01 6.10573947e-01 -1.39527554e-02
1.74309716e-01 2.61422783e-01 -1.81755766e-01 2.43107513e-01
-9.98348415e-01 1.10565815e-02 9.54621553e-01 3.18140000e-01
5.60253739e-01 2.16205679e-02 -7.76112601e-02 3.35372061e-01
1.25795782e-01 2.90162295e-01 7.96379268e-01 -4.33392167e-01
-6.40127659e-01 9.01515543e-01 -2.71514952e-01 -9.44809735e-01
-6.23705089e-01 -5.56869686e-01 -1.16481328e+00 4.67587352e-01
4.30408239e-01 -1.90737903e-01 -8.96994412e-01 1.10970831e+00
5.64516425e-01 -1.59627154e-01 4.79485244e-01 7.20758677e-01
1.05186784e+00 7.82734036e-01 2.00402081e-01 -2.39193127e-01
1.52218354e+00 -7.54017770e-01 -1.07996881e+00 5.99048197e-01
7.03439653e-01 -7.70031631e-01 3.02536666e-01 6.32708311e-01
-7.88111210e-01 -5.22637784e-01 -1.21574616e+00 1.72312751e-01
-7.32424140e-01 6.05270088e-01 5.31491220e-01 6.11475766e-01
-1.25364351e+00 4.57033068e-01 -8.55321884e-01 -8.42330217e-01
1.56206205e-01 7.34366596e-01 -8.00881147e-01 2.97507226e-01
-8.37103188e-01 8.73321235e-01 6.97545409e-01 5.02928868e-02
-5.15176356e-01 -2.84583300e-01 -4.51491356e-01 2.48894468e-01
-4.26746532e-02 -6.76366448e-01 9.43826735e-01 -8.95274341e-01
-1.41435194e+00 1.00089347e+00 3.12230915e-01 -5.08337557e-01
2.38114834e-01 3.27244580e-01 -3.83965313e-01 7.93168366e-01
-3.74283195e-01 8.18781734e-01 2.03019887e-01 -1.19008005e+00
-9.63408709e-01 -3.61228257e-01 -1.53398022e-01 -5.18272780e-02
-4.89212751e-01 -2.59009689e-01 3.21412832e-02 -3.50217670e-01
-3.31587136e-01 -8.72469544e-01 4.52029333e-03 -1.07214443e-01
-2.31076539e-01 -4.66143668e-01 1.00167632e+00 -9.74746644e-01
1.16779530e+00 -2.04226947e+00 -2.25485325e-01 5.18460035e-01
4.14041996e-01 4.51910645e-01 4.50640321e-01 3.13624084e-01
-1.74794763e-01 1.09269738e-01 -3.02766204e-01 -6.83428440e-03
-1.59274235e-01 2.19417557e-01 7.21172273e-01 6.56070650e-01
1.32783487e-01 3.18583637e-01 -6.33583009e-01 -9.80643570e-01
3.00999194e-01 9.38310206e-01 -3.16919118e-01 6.40030652e-02
4.52199489e-01 6.43475214e-04 -1.27956882e-01 9.30735290e-01
7.16597557e-01 -8.50571841e-02 5.52587099e-02 -4.85412955e-01
-1.42909467e-01 -9.18928385e-01 -1.19984353e+00 7.29422688e-01
-2.64190018e-01 6.83189392e-01 2.38770872e-01 -8.89866471e-01
1.01445985e+00 6.71035588e-01 5.60966492e-01 -3.02274346e-01
8.00684988e-01 4.90388572e-01 1.84481680e-01 -7.31693625e-01
3.50872338e-01 -2.75528938e-01 4.59301233e-01 1.65832222e-01
4.95866746e-01 2.54000872e-01 5.17142594e-01 -4.58220169e-02
9.76978183e-01 -3.03834230e-01 8.99928153e-01 -5.42862177e-01
1.02067506e+00 1.74457714e-01 1.96896300e-01 3.13354045e-01
-4.08029437e-01 3.28590930e-01 4.40235257e-01 -4.58724976e-01
-1.15930414e+00 -4.60038900e-01 -3.81005108e-01 3.41784805e-01
-3.54082882e-01 1.76518679e-01 -1.22164083e+00 -5.31322539e-01
-2.61936188e-01 3.39832574e-01 -8.48977268e-01 8.80358890e-02
-3.32564384e-01 -9.72322881e-01 6.02249205e-01 3.76550108e-01
6.20398402e-01 -1.09581089e+00 -7.58177400e-01 1.76377252e-01
1.94300205e-01 -7.79288292e-01 2.90181458e-01 4.36494946e-01
-8.05639207e-01 -1.30790687e+00 -9.15138543e-01 -1.08619380e+00
9.12075043e-01 -4.82565492e-01 7.59220541e-01 3.91078621e-01
-4.65477914e-01 3.49567533e-01 -6.18045151e-01 -5.01507282e-01
-7.02677846e-01 -4.85938638e-02 -1.94519490e-01 1.52217731e-01
5.96215069e-01 -3.91582489e-01 -8.80715549e-01 -1.31702676e-01
-1.19018936e+00 -2.82197237e-01 1.05273819e+00 1.03404820e+00
8.38179410e-01 3.16061229e-01 3.75552148e-01 -1.04934907e+00
3.90848428e-01 -3.60247731e-01 -1.59618363e-01 -1.05797015e-01
-8.03447783e-01 -3.31269234e-01 7.57154703e-01 -1.61399975e-01
-7.86907971e-01 1.07009500e-01 -5.69424868e-01 -1.51747748e-01
-5.57213604e-01 5.38077056e-01 3.36872071e-01 -2.14061871e-01
6.10567808e-01 1.88685656e-01 4.22098458e-01 -1.83138937e-01
-4.29661930e-01 1.07666266e+00 6.29845977e-01 4.21533249e-02
1.98096469e-01 5.45937419e-01 2.41357684e-01 -8.86541963e-01
1.05505675e-01 -7.41662562e-01 -3.45940053e-01 -2.61580378e-01
1.20811701e+00 -6.45115376e-01 -8.28579485e-01 7.44452238e-01
-8.82807791e-01 -5.76999858e-02 -2.13993996e-01 8.48924696e-01
-3.54514688e-01 2.94878602e-01 -9.80687320e-01 -8.70551288e-01
-9.60813701e-01 -9.63485062e-01 5.80955744e-01 7.04830587e-01
-1.04346573e-01 -1.09488857e+00 2.19010487e-01 8.16958770e-02
4.76276964e-01 6.88124835e-01 1.06530559e+00 -1.15996492e+00
1.71656996e-01 -6.80177271e-01 -8.25045332e-02 5.36416411e-01
-2.57988811e-01 4.82207537e-01 -1.06653368e+00 -2.16190860e-01
1.00109152e-01 4.52161878e-02 7.59639144e-01 5.00962436e-01
9.55257177e-01 -3.34116668e-01 -4.92152452e-01 3.66749972e-01
2.02865911e+00 6.60344899e-01 7.74799943e-01 5.60469747e-01
1.00268714e-01 4.56949264e-01 5.22104383e-01 4.17427510e-01
-1.45900339e-01 2.18237415e-02 4.98282552e-01 -5.56849539e-01
-2.33655065e-01 3.52836460e-01 -1.57028407e-01 8.12237501e-01
-3.19625586e-01 -6.18861914e-01 -9.22027171e-01 7.93725967e-01
-1.39986074e+00 -8.87568235e-01 -2.33130828e-01 1.57258475e+00
5.66997647e-01 3.14328000e-02 1.55913785e-01 8.43977630e-01
9.12071288e-01 -5.78788936e-01 -3.51049677e-02 -8.05119157e-01
-2.75907461e-02 8.53501439e-01 4.36701000e-01 3.97795826e-01
-1.23472929e+00 2.28380114e-01 5.24148417e+00 1.07752657e+00
-1.28537405e+00 4.91223745e-02 6.73845649e-01 3.88333678e-01
2.75638252e-01 -5.06292045e-01 -6.53631032e-01 4.20249909e-01
9.51990724e-01 1.52672291e-01 -1.68045118e-01 9.04766202e-01
-1.44275844e-01 -5.39401829e-01 -6.83680654e-01 7.69924045e-01
1.99462533e-01 -1.14750206e+00 -1.31193653e-01 2.49180332e-01
5.92933416e-01 -4.13183093e-01 -2.01486096e-01 4.03708220e-01
-4.72423196e-01 -1.00702977e+00 -2.05456801e-02 6.54317915e-01
7.42998958e-01 -1.14096999e+00 1.88333666e+00 2.51209646e-01
-9.70527470e-01 -1.12173799e-02 -2.15879917e-01 -1.90434959e-02
-3.65951449e-01 6.17202520e-01 -1.29550946e+00 4.09347773e-01
4.72320110e-01 1.31365538e-01 -5.28465271e-01 1.10331929e+00
2.09370270e-01 6.41607821e-01 -5.37594259e-01 -6.85003519e-01
5.09518743e-01 -4.60610576e-02 3.50568593e-01 1.72269738e+00
5.21073520e-01 1.73995122e-01 -3.81405979e-01 4.01195377e-01
2.63681114e-01 6.31947041e-01 -3.26996088e-01 -1.27109187e-02
2.65540421e-01 1.86085820e+00 -1.37753892e+00 -6.04082942e-01
-4.53911014e-02 7.29070783e-01 -2.67129928e-01 1.31372781e-02
-6.82443678e-01 -1.01495492e+00 -2.42047250e-01 1.05088331e-01
3.43203992e-01 5.90767026e-01 -2.83206496e-02 -4.78327721e-01
-4.46357042e-01 -6.72131002e-01 5.19941390e-01 -5.78190982e-01
-1.10943437e+00 7.69950032e-01 -1.71137422e-01 -1.12673438e+00
-1.96260452e-01 -1.03781879e+00 -4.95029777e-01 5.12724638e-01
-1.18957376e+00 -1.24538970e+00 -4.40171540e-01 3.64238769e-01
9.47543085e-02 -3.50449502e-01 1.08775985e+00 3.73694777e-01
-4.80607629e-01 5.58529377e-01 3.98594558e-01 3.11694264e-01
4.81670409e-01 -1.64051318e+00 -8.99314940e-01 5.78110814e-01
-5.53127289e-01 4.41204548e-01 6.41242921e-01 -3.58814150e-01
-8.31968904e-01 -9.32147920e-01 1.04309857e+00 3.07197094e-01
2.32537046e-01 2.12768555e-01 -6.00833416e-01 3.52075636e-01
5.84387898e-01 1.61539793e-01 1.22931159e+00 -3.79196316e-01
4.18929458e-01 5.00709713e-02 -1.82569385e+00 2.26717308e-01
-1.00074284e-01 -2.67465889e-01 -4.82895255e-01 1.63994715e-01
4.72189598e-02 -2.53138483e-01 -1.43870735e+00 4.56418186e-01
7.55365193e-01 -1.20784152e+00 5.63648880e-01 3.77890952e-02
5.86814463e-01 -5.12496471e-01 5.14257662e-02 -9.08937275e-01
-4.73263443e-01 1.92154720e-01 -1.24525622e-01 1.04806435e+00
2.11932465e-01 -4.50835019e-01 9.21928406e-01 -1.23024010e-03
-1.67324021e-01 -1.32180512e+00 -7.45264053e-01 -2.10400328e-01
-7.06717893e-02 4.24811929e-01 1.11753508e-01 9.13021564e-01
-2.18829345e-02 1.02229647e-01 1.90892085e-01 -4.24972698e-02
4.10130203e-01 -2.93827623e-01 3.57041001e-01 -1.14950788e+00
-2.83258617e-01 -4.34216887e-01 -9.84359682e-01 1.15172081e-01
-1.21224016e-01 -8.70235205e-01 -4.32360917e-01 -1.56889606e+00
4.77827430e-01 9.53288972e-02 -5.63121915e-01 4.58984584e-01
-3.98937762e-02 5.52495360e-01 -1.82239302e-02 -6.71804473e-02
-2.63451342e-03 -1.31688908e-01 1.09148288e+00 -1.45336941e-01
-8.45779553e-02 -2.56239802e-01 -4.04442400e-01 8.22452784e-01
9.50265110e-01 -4.34795260e-01 5.68652898e-02 4.41834867e-01
1.08547267e-02 -6.11521155e-02 2.43240282e-01 -1.58173919e+00
4.06819552e-01 3.84217858e-01 8.96864057e-01 -9.87996399e-01
-8.80990624e-02 -1.03808892e+00 3.89036030e-01 1.03138924e+00
7.31477467e-03 -8.96341205e-02 1.34384587e-01 3.63619357e-01
-5.84801793e-01 -5.36205351e-01 9.44121122e-01 -1.74262270e-01
-7.57040381e-01 -1.42765015e-01 -8.26494753e-01 -8.50327909e-01
1.47883999e+00 -7.14759350e-01 -3.11371088e-01 4.00473252e-02
-9.08876300e-01 -7.36508369e-02 3.26560229e-01 -7.12156177e-01
4.66042995e-01 -9.39704537e-01 -6.69504881e-01 -2.20433846e-02
6.20452836e-02 -5.96672520e-02 5.49595177e-01 1.18196797e+00
-1.34124231e+00 4.10383910e-01 -7.15973854e-01 -5.13985395e-01
-1.59705198e+00 4.80435550e-01 4.39665049e-01 -7.09438384e-01
-1.53278396e-01 7.55489647e-01 -2.96625018e-01 -2.46451318e-01
7.87586570e-02 -3.41373146e-01 -1.08581746e+00 1.86572120e-01
4.36428964e-01 8.01584482e-01 3.80067199e-01 -5.42569757e-01
-2.20513090e-01 5.54858923e-01 2.78707314e-02 1.03738740e-01
1.29990220e+00 3.44773740e-01 -4.17449147e-01 2.90810615e-01
1.33784187e+00 9.49767828e-02 -4.04342234e-01 4.39439327e-01
-3.53708416e-01 2.81303763e-01 4.66855079e-01 -1.06966996e+00
-9.80705440e-01 6.97133601e-01 1.24408960e+00 1.61533162e-01
1.40690196e+00 -4.69314128e-01 7.43324518e-01 3.60295653e-01
-6.46268576e-02 -1.24776268e+00 -3.01159590e-01 2.75878042e-01
3.59777778e-01 -9.89189148e-01 7.45753795e-02 -2.37543941e-01
-3.02899361e-01 1.81797826e+00 4.31103587e-01 -4.17052358e-01
7.74781466e-01 5.65777600e-01 -4.79354747e-02 -3.59303653e-01
-6.13990903e-01 -2.12653309e-01 3.08760237e-02 5.45795679e-01
5.51011264e-01 1.50977790e-01 -1.02926278e+00 7.38058150e-01
-5.65818921e-02 5.43023646e-01 6.49036944e-01 1.04206574e+00
-6.00914657e-01 -6.01727903e-01 -5.71438253e-01 5.81206679e-01
-8.81755590e-01 2.52622008e-01 -4.04042929e-01 1.42353821e+00
6.79228723e-01 6.31407559e-01 2.52445221e-01 -4.17284131e-01
-6.51335642e-02 1.14801124e-01 4.12964702e-01 -9.73502994e-02
-1.06192708e+00 2.64337569e-01 1.76780317e-02 1.46980667e-02
-7.15851128e-01 -3.06383789e-01 -1.53921306e+00 -4.03892905e-01
-4.74327207e-01 4.69700873e-01 9.21896935e-01 7.91897476e-01
-1.91336975e-01 6.77102804e-01 4.42757696e-01 -3.43839318e-01
-1.39806390e-01 -1.19631374e+00 -1.03234780e+00 9.23136249e-02
4.14655566e-01 -3.81975442e-01 -5.93502462e-01 2.36106694e-01] | [15.325451850891113, -2.856750726699829] |
d5b43177-2ec3-43f1-b05b-cce8147a485f | metaphora-kodikon-tes-matlab-se-programmata-c | null | null | http://ikee.lib.auth.gr/record/271194 | http://ikee.lib.auth.gr/record/271194/files/GRI-2015-15043.pdf | Μεταφορά κωδίκων της MATLAB σε προγράμματα C++ για την επίλυση Κανονικών Διαφορικών Εξισώσεων Porting MATLAB routines to C++ for the numerical integration of Ordinary Differential Equations | Στο πλαίσιο της διπλωματικής αυτής εργασίας επιχειρείται ο συγκερασμός της αξιοπιστίας και των δυνατοτήτων που παρέχει η ώριμη ρουτίνα ολοκλήρωσης προβλημάτων αρχικών συνθηκών ode15s του MATLAB με την ταχύτητα που προσφέρει η C++. Για το σκοπό αυτό δημιουργήθηκε σε γλώσσα C++ ένα αλγοριθμικά πιστό αντίγραφο της ode15s και των λοιπών αναγκαίων ρουτίνων, συμπεριλαμβάνοντας όλες τις παρεχόμενες δυνατότητες. Η επιλογή του ολοκληρωτή ode15s έναντι του δημοφιλέστερου ode45 εδράζει στη δυνατότητα του πρώτου να αντιμετωπίζει άκαμπτα προβλήματα και παράλληλα να επιδεικνύει υψηλή επίδοση ακόμη κι όταν χρησιμοποιείται σε μη άκαμπτα προβλήματα. Ο κώδικας έχει δομηθεί με τρόπο που διευκολύνει την προσθήκη νέων ολοκληρωτών του MATLAB όπως η ode45, ode113 κοκ.
Τhis M.Sc. thesis attempts to unify the reliability and capabilities of ode15s, a mature initial value problem integration routine offered by MATLAB, with the high-speed properties of the C++ programming language. In this direction, ode15s and all relevant routines required were ported in C++ yielding an algorithmically identical C++ version of ode15s, which includes all functionality available in MATLAB. The choice of ode15s over the more popular ode45 is justified by the ability of the former to integrate stiff problems while exhibiting only a slight performance impact when applied to non-stiff problems. The source code has been structured to facilitate the addition of other MATLAB integrators such as ode45, ode113 etc. | ['Θεοδωρακόπουλος Αχιλλέας-Κωνσταντίνος Βασιλείου'] | 2015-10-01 | null | null | null | theodorakopoulos-akhilleas-elektrologos-mekh | ['numerical-integration'] | ['miscellaneous'] | [-0.06707186 0.71851075 -1.3376298 0.36580244 0.89169896 -0.7404142
0.76586497 -0.54263496 -0.7221984 1.4804866 1.058006 -1.7825217
-0.5664324 -0.9489754 0.4386211 -0.4071409 0.71024686 0.9571552
0.10713258 0.10720581 0.14636791 0.41291448 -1.5592191 0.23353395
0.533816 0.06062973 0.32202873 0.77736634 0.02787662 1.6208551
-0.26534554 -0.24233419 -0.18384874 -0.16052748 -1.8152683 0.16762733
-0.433837 -0.35414225 -0.6762433 1.163095 -0.10332826 0.31408554
0.8451705 -0.8137864 0.51757646 0.01592172 -0.62347364 0.75217223
0.28093523 0.7832048 0.23100466 -0.99146736 1.1816871 1.7251629
0.47706506 -0.13984418 -0.70164955 -0.8514646 -0.01393156 0.2231451
-0.2999525 0.42039478 0.56133366 -0.6603219 0.7584399 0.37896147
0.8342711 1.6382256 0.058128 0.11227581 1.99698 -0.2205459
-0.30561712 -0.22867967 -0.17251241 0.66011393 0.46208704 0.7898478
0.32217473 0.01350805 1.9145188 0.70229256 0.53049755 0.6066871
-0.84196866 1.2861309 -0.73439896 0.8106431 0.32999074 0.37249845
1.0188751 0.27032316 -0.29547226 0.09399518 -0.39516875 -0.6796571
0.21980265 1.2747443 0.13662171 0.427123 1.0812201 0.19757265
0.2728489 0.39093238 0.6740859 1.1059663 0.29374823 -2.17659
0.11752658 0.91929454 0.72739166 -0.9449576 0.10320961 0.878737
-0.00878168 0.6424313 0.43624282 -0.10751766 -0.5522282 0.75303674
0.13486329 -0.3209654 0.156481 0.6571739 0.7351179 0.25251547
1.0740356 -0.3498219 2.0268614 -0.16133967 -1.1527724 -0.12474753
0.88525546 -1.3905882 1.1298552 -0.20197216 -1.0260043 -0.23336537
-0.5924186 -0.24051538 0.3872188 -0.03490245 1.2426095 1.043742
0.23320463 0.19487086 -0.5549034 0.01273303 0.43807596 -0.10035525
0.03551148 0.06161111 -1.1270001 0.12105313 1.1418025 -0.5328832
-0.20694453 -1.6484828 -0.4568463 -0.47257513 1.3947842 -0.5541568
1.5030919 -0.1023269 -1.7380848 1.0867791 0.334526 0.55724216
1.458757 0.68207264 -0.162422 0.7269811 -0.336248 0.05766292
0.43511087 -0.91901976 -1.9954443 -0.17381223 -0.0246721 -0.23638462
0.8026416 1.683095 1.3360279 -0.38309723 -0.19513796 -0.76231444
-0.19916925 -0.8379112 -0.06545098 -0.33683813 1.6221956 0.26363617
1.2033646 -1.7092078 -0.5369116 0.52121496 0.86400723 1.6257653
0.9227189 0.7170804 -0.04469453 0.10949872 0.05623358 1.3832085
-0.1378866 0.83795583 -0.5120199 -0.22418314 -0.27515373 0.3686259
-1.6924415 -0.46215072 0.7069664 -0.15623423 0.47952807 -0.47125056
-0.06174001 0.80025804 -1.475348 0.24580112 0.4441331 -0.15944664
1.3527954 0.3249076 -0.8856982 1.031656 -1.4130445 0.57783115
0.2310074 0.5337352 -0.04941647 -0.9364734 1.2086047 0.32166028
0.7497262 -1.1977435 1.0653647 0.7784344 0.24360356 -0.96612483
0.9848357 -1.1464671 -0.04005325 0.7895195 -0.58435977 0.05956581
0.95025164 -0.42873886 0.996443 0.5088483 0.91870576 -0.6928871
0.7619596 -0.27016655 0.58404326 0.4760853 -0.52267814 -0.32529652
0.47988316 -1.3603907 -1.576777 -1.6231618 -0.35082233 1.5016997
0.09149867 0.23656279 -0.01942462 -0.76752806 -0.09333119 -0.5401825
0.37906325 0.23181969 -1.1239269 0.01812769 0.45399836 0.5953707
1.0193261 -1.5525556 -2.0583353 0.76607585 0.09206127 -1.5033337
1.1681013 -0.48845893 -0.8301582 -1.3228326 -0.02388885 -0.2149262
-0.1012148 0.5385976 0.9623234 0.55936646 -1.431857 -0.33218172
-0.71300095 -0.3414445 -0.8061211 0.12840354 0.11060607 -1.3049147
0.38988268 -0.97372323 -0.79053146 0.04328942 -0.3772204 0.5070042
0.15164408 0.21957894 -0.09513457 -0.91451895 0.30711097 -1.6664996
0.09399953 0.07926936 -1.1141417 -0.4415668 -0.37357625 -0.08242022
0.58828735 -0.642426 -1.4901161 -1.669868 -0.14140396 -0.9827162
-0.6195289 -0.01104936 0.08076185 0.8145486 0.20358512 -0.5278225
0.11711797 -0.37988544 -0.19899353 -0.13763481 1.4124546 -1.5627867
0.21045595 1.0747887 0.47317648 -1.285188 -1.0435265 0.01208098
0.31970823 -0.10478207 1.3141778 -0.67214596 -2.3135862 -0.11839626
0.18013018 -0.00943335 -0.62655956 1.5487695 -0.5063996 0.51142085
-1.5204004 -0.9588836 0.72277844 -0.77280325 -0.29781115 1.2147956
-0.7462785 -2.179407 -0.04303091 0.550196 0.04089509 0.75991887
0.9289602 0.26275823 -0.3707587 0.3970475 -0.848959 -0.11152728
0.16830373 0.39155743 -0.9981206 0.3543725 -0.18918538 -0.34003708
0.34393892 -0.17909266 0.84892106 -0.8081212 -0.13184464 0.04483499
2.3636162 0.35533163 2.0162163 1.1733252 0.84377456 0.72996575
0.76371396 0.5984713 0.30910432 0.21068229 -0.485752 0.6680199
-0.03251333 -0.24287581 -0.6430539 0.16545728 -1.0421383 0.60340166
-0.95618653 0.9539437 -2.1265154 -1.2001619 -0.05495199 1.4858353
0.25919825 0.26581252 -0.09860487 0.20198151 0.72277707 0.99253887
0.4341116 -0.7502196 0.21928887 1.0132084 0.508896 0.9907656
-0.5728605 1.1004567 -0.19908455 1.7705038 -1.3256829 -0.402336
-0.2834552 0.06557041 -0.96143836 0.73125875 -0.66351527 0.9009909
1.0625639 -0.22323841 -0.62884927 0.29646173 0.3773725 -1.1871157
-0.26463318 0.72470254 -1.29531 -1.999352 -0.8920982 0.34328318
1.0192541 -0.6952779 0.66600156 0.18300878 1.8019981 -0.973784
0.11058941 0.24296695 0.09907208 -1.1805404 0.16671444 -0.21484248
-0.56818604 -0.68290377 -0.4502234 -0.6182983 -0.37577063 0.7923157
-1.0625005 1.078597 0.5478394 -0.38714877 1.2257146 0.3102895
-1.2529936 0.2647882 0.10420998 0.00862446 0.96785235 -0.89201546
0.33992863 0.9168955 0.08189882 0.57717526 0.64069057 0.36583075
-0.04645855 0.16689053 -0.21463831 0.12697965 1.1220763 0.98209524
-1.1139728 -0.48375103 -0.92390084 -1.2542504 0.11099698 -0.05677069
-0.39537358 0.18144292 1.126431 0.7295648 0.6216222 0.7441261
0.2710456 -0.6726246 -0.8205555 -0.7365174 0.4526217 -0.3966844
-0.34568322 -0.55437577 0.79476315 -0.5965631 -0.63437533 -0.44728512
0.1698307 1.3100266 -1.4818137 -0.69341403 0.45941532 -0.406033
-0.3721457 -0.10535251 0.8309713 -0.37672618 0.56400394 -0.02745585
0.6923117 0.27237892 0.38498992 -0.92365706 -0.25658858 0.04076472
-1.017712 0.40414768 0.96660244 -0.31825674 -0.19296244 -0.1918327
1.0432323 0.4623207 0.80584764 0.44120294 -1.8633881 0.7152642
-0.30210567 0.5664332 0.49963456 -0.27169695 -0.57277465 0.72693515
-0.59625775 0.78005636 0.7811981 -1.2082746 -0.7181301 -0.24425545
-0.0040384 -0.7674477 -1.8586512 0.30204353 1.0832558 -1.2144482
0.5999639 -1.0697781 0.6800792 -0.11014703 0.05802554 0.06545037
0.09183977 -0.50593615 0.5081777 0.6349299 -0.9410945 -0.639619
0.85179734 0.10351054 -0.3872149 -0.9502669 -1.3798884 -1.1712947
0.3497405 -0.02328879 0.2541566 1.9682031 -0.2250165 -0.66442573
0.08910859 -0.551137 0.48686683 0.26582316 0.69935554 -0.8577404
-0.04279643 -0.3022883 -0.356228 0.4341652 -0.17940821 0.31751513
-1.443054 -0.84084535 0.69293046 -1.2388206 0.967565 -0.49834418
0.49174747 0.6387104 0.3824888 0.16422679 -0.45918202 -0.6973147
2.3682795 0.16104586 -0.0761798 -0.91918176 -0.54783964 1.2939065
0.36374572 -0.60752356 -1.4457886 0.33577693 0.9051247 0.6777051
0.61328715 -0.09218384 -0.58587635 -1.0555947 -0.17056099 -0.2857556
-0.33675706 0.2613768 0.7797938 0.70612305 0.27932948 -0.24952662
0.04731495 -0.4371096 0.5534184 -0.8856175 0.81218696 -1.0726197
-1.2049015 -0.23942485 -1.6457645 1.05075 0.6847609 -0.668239
-0.8952955 0.47347668 -1.3448858 -0.15908857 0.6688733 0.59612626
0.1943668 -1.0920813 -0.38917765 -0.857465 -0.03164187 0.59727407
1.2387569 0.6481336 -0.65549695 0.33006173 -1.4186577 -0.43057927
-1.6113921 0.20804986 0.44383445 -1.0398754 -1.257338 0.6309025
0.49110502 -0.67196405 -0.8721169 -0.4076965 0.11816019 -0.16396287
0.71026343 1.5090117 -0.44752717 -0.25855824 -0.6746185 0.02038855
-1.0275581 0.17380953 0.5724344 -0.33882606 -0.3237717 0.25691044
1.8014767 0.06429183 -0.8142657 -0.42483953 0.42019448 -1.2240682
-1.4245768 -0.04909031 0.14354298 0.7509492 0.53835905 -0.21560666
-0.21898119 -0.10425958 0.95998347 -0.13740769 0.58072424 -2.308665
0.6382208 0.613661 -0.48777717 -0.73570335 1.0833671 -0.75771415
-0.7137669 1.8301356 -0.16716929 -1.0543059 0.11356928 0.796723
0.7063714 -0.79464805 -0.32879466 -0.15011732 -0.3688305 0.6355537
0.67543113 0.53859234 -0.6969471 -0.07809854 -1.1269021 -0.02581412
1.2027471 1.3770497 -0.6647671 -1.6919475 -1.0719967 0.14730039
-0.99295217 1.1342418 0.5822998 1.9687994 0.54450446 0.202578
0.923497 0.11582796 -0.23909327 -0.68099296 1.4005983 -0.01890725
-0.4331435 1.1511455 0.42977917 -0.43134892 -0.8518846 -0.7584988
-2.274465 -1.6141372 0.9330014 -0.00648175 -0.14459851 1.4624585
-0.4190393 0.93373024 0.46221393 0.02628031 -0.2598876 -0.42743552
-0.41839367 0.32710624 0.54172754 -1.262835 0.07516776 -0.31580696] | [9.881379127502441, 9.199470520019531] |
fe9895ee-63f7-44dd-860c-45d688f4ab0f | demonstrating-the-risk-of-imbalanced-datasets | 2201.03559 | null | https://arxiv.org/abs/2201.03559v1 | https://arxiv.org/pdf/2201.03559v1.pdf | Demonstrating The Risk of Imbalanced Datasets in Chest X-ray Image-based Diagnostics by Prototypical Relevance Propagation | The recent trend of integrating multi-source Chest X-Ray datasets to improve automated diagnostics raises concerns that models learn to exploit source-specific correlations to improve performance by recognizing the source domain of an image rather than the medical pathology. We hypothesize that this effect is enforced by and leverages label-imbalance across the source domains, i.e, prevalence of a disease corresponding to a source. Therefore, in this work, we perform a thorough study of the effect of label-imbalance in multi-source training for the task of pneumonia detection on the widely used ChestX-ray14 and CheXpert datasets. The results highlight and stress the importance of using more faithful and transparent self-explaining models for automated diagnosis, thus enabling the inherent detection of spurious learning. They further illustrate that this undesirable effect of learning spurious correlations can be reduced considerably when ensuring label-balanced source domain datasets. | ['Michael Kampffmeyer', 'Robert Jenssen', 'Stine Hansen', 'Marina M. -C. Höhne', 'Srishti Gautam'] | 2022-01-10 | null | null | null | null | ['pneumonia-detection'] | ['medical'] | [ 6.28853798e-01 2.18488216e-01 -6.14644051e-01 -4.90743130e-01
-1.23484266e+00 -3.93795907e-01 3.64021331e-01 2.56135643e-01
-1.34570286e-01 7.25589037e-01 4.20993924e-01 -5.20359635e-01
-4.97922450e-01 -5.59152663e-01 -7.74914801e-01 -7.14952111e-01
1.36057153e-01 5.34806609e-01 -1.38838282e-02 3.94973934e-01
-1.41147435e-01 2.91317195e-01 -1.21520054e+00 7.55352914e-01
5.80427349e-01 5.91149688e-01 3.41687091e-02 7.88147569e-01
1.20976545e-01 1.48578393e+00 -6.75429761e-01 -3.04760188e-01
8.07358995e-02 -8.56530666e-01 -6.81908369e-01 2.44750246e-01
5.75130045e-01 -3.30579966e-01 -1.18068568e-01 7.10159540e-01
4.66338933e-01 -4.62261081e-01 1.14754391e+00 -1.18977189e+00
-4.87837106e-01 4.76817697e-01 -7.31132925e-01 7.16374218e-01
-1.14391886e-01 3.16188306e-01 1.17384505e+00 -3.16594809e-01
8.47505748e-01 7.60387897e-01 1.08523762e+00 5.01241207e-01
-9.93655860e-01 -7.00719357e-01 -2.57744342e-01 -3.11648287e-02
-1.17418218e+00 -1.72634616e-01 4.38411117e-01 -6.93707228e-01
6.86249316e-01 4.96878296e-01 2.02378750e-01 1.32168126e+00
3.71439129e-01 4.16636348e-01 1.32261074e+00 -3.74344558e-01
-8.49822760e-02 3.38867694e-01 1.45783514e-01 9.56201613e-01
4.88738060e-01 1.90221205e-01 -5.36932290e-01 -6.74987018e-01
6.06328785e-01 1.37198865e-01 -2.48325393e-01 -2.95232892e-01
-1.28509891e+00 8.06366503e-01 4.06280011e-01 3.66162539e-01
-3.96209717e-01 -1.45682888e-02 3.96590501e-01 -2.78715715e-02
5.27961850e-01 5.68451643e-01 -7.01247990e-01 1.61568940e-01
-1.14475155e+00 1.16128020e-01 4.32469070e-01 4.63439941e-01
3.29972863e-01 -2.45191902e-01 -2.85646349e-01 8.04773510e-01
1.71798304e-01 5.62328756e-01 5.31536937e-01 -9.57862914e-01
3.58917058e-01 7.36732244e-01 -2.92440951e-01 -6.04566097e-01
-7.16306210e-01 -9.33460951e-01 -6.82233095e-01 6.77921996e-02
5.91694653e-01 -2.43335739e-01 -1.03697264e+00 1.69692278e+00
3.18895161e-01 1.98014960e-01 -6.05534390e-03 6.71273112e-01
5.96984446e-01 -6.49818778e-02 4.83305156e-01 -7.66860023e-02
1.56442952e+00 -6.36731505e-01 -5.43052137e-01 -2.40780026e-01
9.66444969e-01 -6.27839565e-01 9.61734712e-01 1.68661669e-01
-8.01372588e-01 -3.11461240e-01 -8.45852256e-01 1.98547184e-01
-6.77142069e-02 2.17523232e-01 5.57382047e-01 6.63265824e-01
-3.62408400e-01 6.50176346e-01 -7.93867052e-01 -3.63699973e-01
6.91765189e-01 8.84893239e-02 -1.61412701e-01 -1.30915895e-01
-8.76045465e-01 9.49997008e-01 2.08365962e-01 -6.76417828e-01
-6.75614119e-01 -1.53055739e+00 -3.68991286e-01 -4.20322418e-02
3.33469808e-01 -8.37355137e-01 1.23883379e+00 -9.50471342e-01
-4.72933471e-01 1.29857135e+00 2.96275527e-03 -4.28380996e-01
6.46566570e-01 7.20094368e-02 -4.33185607e-01 4.83021080e-01
3.73479486e-01 4.24377441e-01 8.74136865e-01 -1.34662402e+00
-8.40986729e-01 -5.01729786e-01 -4.11468089e-01 -6.64032772e-02
-2.56031156e-01 -3.50354984e-02 1.17276460e-01 -8.38607252e-01
-9.67572033e-02 -1.05361247e+00 -1.37614638e-01 -8.92350376e-02
-3.68167579e-01 7.32954312e-03 6.22273803e-01 -7.25073457e-01
1.01375401e+00 -2.16918802e+00 -3.30311924e-01 8.78266245e-02
5.69722533e-01 -1.21255405e-01 1.13730974e-01 3.29722278e-02
-4.99536484e-01 2.42057279e-01 -4.36537981e-01 -2.05935404e-01
-4.87627208e-01 2.58129150e-01 -3.45617026e-01 5.52497745e-01
6.39271498e-01 8.76102805e-01 -8.08827937e-01 -7.75692284e-01
6.41102716e-02 5.08298934e-01 -5.42553246e-01 7.35041499e-02
-1.01904534e-01 5.88755727e-01 -3.20731133e-01 7.05255151e-01
2.99644291e-01 -9.88616526e-01 2.76198775e-01 -2.01356605e-01
3.14389110e-01 4.67068613e-01 -7.40235090e-01 1.21410978e+00
-5.27600944e-01 4.57074851e-01 -1.44629121e-01 -5.90419531e-01
4.14034009e-01 4.82551873e-01 1.02320290e+00 -5.36172330e-01
2.62119845e-02 3.96176279e-01 2.37712264e-01 -4.69449937e-01
-3.34473401e-02 -7.14931667e-01 3.46382558e-01 7.41967380e-01
-7.81263337e-02 -3.54349762e-02 -3.47722620e-01 7.85914660e-02
1.39807546e+00 -2.80014545e-01 2.23548979e-01 -2.19897106e-01
5.94177097e-02 4.50020015e-01 3.24771583e-01 9.16307628e-01
-1.32862419e-01 1.00629532e+00 4.19534743e-01 9.37604811e-03
-1.02012575e+00 -1.11308134e+00 -6.90303743e-01 8.66694570e-01
-5.53716004e-01 -1.05209716e-01 -3.86140466e-01 -1.07302105e+00
2.28929535e-01 6.40786827e-01 -8.15563858e-01 -3.19638908e-01
-5.47090530e-01 -1.39093530e+00 8.51704478e-01 6.54011071e-01
-1.83195192e-02 -6.82342887e-01 -1.08290875e+00 -4.32074554e-02
-2.20329374e-01 -9.96120155e-01 -1.88451074e-02 5.11414409e-01
-1.02865934e+00 -1.54687548e+00 -1.02017748e+00 -3.66362423e-01
6.49653256e-01 9.56911594e-02 1.51091349e+00 3.50533932e-01
-6.23692274e-01 6.88614249e-01 -2.47262508e-01 -4.99211729e-01
-9.15853977e-01 2.61389732e-01 -4.72758830e-01 -3.39631319e-01
4.14492399e-01 -2.13564888e-01 -4.90876824e-01 1.49283394e-01
-8.96025240e-01 -7.35185072e-02 7.08512723e-01 8.14680457e-01
6.63706183e-01 5.78430332e-02 7.70171106e-01 -1.53100240e+00
2.73014754e-01 -8.34994972e-01 -6.14142902e-02 3.89371634e-01
-9.27975416e-01 8.89644492e-03 2.59023547e-01 -2.02514648e-01
-1.08115888e+00 -2.16245744e-02 8.43200609e-02 -3.55183303e-01
-2.56151378e-01 3.00281912e-01 5.38502872e-01 1.07032970e-01
9.84550476e-01 -1.72418863e-01 -1.19986698e-01 -3.52863431e-01
-3.19093578e-02 5.64717531e-01 4.87901419e-01 -2.85323709e-01
5.62922359e-01 7.02160478e-01 3.64455074e-01 -4.45527226e-01
-1.39285874e+00 -7.02665746e-01 -4.27382350e-01 9.21943318e-03
8.84456754e-01 -1.05085862e+00 8.12700391e-03 1.09999411e-01
-6.52121782e-01 -2.96045274e-01 -5.85296273e-01 5.44881046e-01
-4.38207656e-01 7.59886876e-02 -4.69005793e-01 -5.12031615e-01
-2.38823276e-02 -1.02676153e+00 1.14405012e+00 -1.85909465e-01
-5.03785372e-01 -1.18227863e+00 2.84624487e-01 6.21008098e-01
2.25527674e-01 4.30078655e-01 1.33046556e+00 -8.10690880e-01
-5.30541241e-01 -3.49853002e-02 -6.05212808e-01 1.39298007e-01
5.60653389e-01 -1.02169579e-02 -1.28278458e+00 -5.84161952e-02
3.07044387e-01 -3.18879396e-01 1.03667605e+00 5.75122714e-01
1.08272481e+00 -1.71554647e-02 -3.82210642e-01 5.25614798e-01
1.54838991e+00 -1.59288391e-01 9.06012282e-02 2.55958468e-01
7.66971231e-01 8.03795636e-01 3.15618873e-01 3.42153519e-01
3.28818083e-01 2.55947948e-01 1.13955744e-01 -5.91336727e-01
-7.17025518e-01 -1.51408598e-01 -3.36935848e-01 2.81725764e-01
2.12712213e-01 -9.01726410e-02 -1.30057538e+00 7.23747253e-01
-1.32112360e+00 -6.37120724e-01 -5.01674592e-01 1.81337643e+00
1.10214376e+00 8.57580900e-02 2.41140410e-01 1.15000412e-01
5.09106994e-01 9.59056243e-02 -6.00912988e-01 2.17316911e-01
-2.98269123e-01 2.21595719e-01 7.92401612e-01 1.86220571e-01
-9.13724303e-01 9.15223658e-02 7.66435337e+00 3.90315205e-01
-1.18421948e+00 5.43071091e-01 9.82931614e-01 -2.82765567e-01
-3.42322797e-01 -3.42714161e-01 -5.22145331e-01 3.92139733e-01
1.17162311e+00 1.40471518e-01 -1.13056518e-01 6.93179250e-01
-1.26271378e-02 -1.60821378e-01 -1.09679270e+00 5.56160212e-01
1.74495265e-01 -1.38372111e+00 -2.09040821e-01 1.04535609e-01
8.99578810e-01 3.36203575e-01 2.50769734e-01 -1.25752240e-01
2.01229006e-01 -1.00240898e+00 4.04730886e-01 2.28284389e-01
9.17930961e-01 -4.65184331e-01 7.15621352e-01 1.25704020e-01
-5.59798777e-01 -1.41474560e-01 -1.99794360e-02 3.93407762e-01
-4.91148792e-02 8.69493484e-01 -1.34997785e+00 4.62838978e-01
5.65411329e-01 6.74574435e-01 -8.68672907e-01 7.79853761e-01
2.72346642e-02 1.01959026e+00 -3.76064479e-02 6.51553631e-01
-4.24082913e-02 3.26520175e-01 2.61851072e-01 1.45343733e+00
1.24301694e-01 5.59610464e-02 -1.52296633e-01 8.23106945e-01
-8.49112421e-02 1.76945589e-02 -5.60749054e-01 4.29466181e-02
1.05379492e-01 9.59628522e-01 -7.92593300e-01 -5.14741898e-01
-5.69700658e-01 5.05560219e-01 -2.29883362e-02 2.04637665e-02
-1.02215576e+00 3.42077762e-01 2.85538197e-01 5.14962256e-01
1.39443919e-01 3.97944331e-01 -1.02852738e+00 -1.00531948e+00
-2.82603949e-01 -1.00676107e+00 8.21102262e-01 -9.12623227e-01
-1.42654800e+00 3.09036821e-01 -3.76919843e-02 -1.05144680e+00
-1.98054254e-01 -4.91534770e-01 -2.97836661e-01 7.27401137e-01
-1.97643888e+00 -1.04848266e+00 -2.72137105e-01 4.71628755e-01
3.28170657e-01 -7.65939206e-02 6.93930745e-01 4.08922255e-01
-3.08164328e-01 5.01899779e-01 5.32564744e-02 -8.71516168e-02
1.08692825e+00 -1.41272998e+00 -1.25481546e-01 5.73159575e-01
1.78277746e-01 3.92875493e-01 5.26587725e-01 -8.71480465e-01
-6.45237446e-01 -1.30781293e+00 8.78489554e-01 -1.19194341e+00
5.91115534e-01 2.70258307e-01 -1.12043750e+00 6.38699710e-01
-3.82872373e-02 -1.76938638e-01 1.06012964e+00 1.11226015e-01
-6.48168743e-01 1.34563506e-01 -1.33623564e+00 1.29234180e-01
8.99371386e-01 -7.39273250e-01 -6.07087433e-01 7.65186250e-01
5.20852745e-01 -3.30159724e-01 -1.03374159e+00 4.98993337e-01
3.69911104e-01 -1.07317531e+00 1.09464097e+00 -7.18792439e-01
9.47175443e-01 5.59340268e-02 -5.72971292e-02 -1.00502026e+00
-1.53749943e-01 2.00950474e-01 1.03332624e-01 9.76222754e-01
5.32475829e-01 -4.56077069e-01 9.64164495e-01 5.95960915e-01
3.51565219e-02 -5.95690608e-01 -8.82116079e-01 -4.85333085e-01
2.76609749e-01 -4.09586489e-01 4.30008054e-01 1.34551537e+00
-3.91483098e-01 6.39040470e-02 -1.50871590e-01 3.67268085e-01
6.53766096e-01 2.91320551e-02 2.28967085e-01 -1.18978226e+00
-5.45943081e-01 -1.54443070e-01 1.23021290e-01 -1.29743144e-01
1.17487617e-01 -9.10859346e-01 -1.22695431e-01 -1.25997448e+00
5.88163495e-01 -7.75333881e-01 -6.56657338e-01 4.36358899e-01
-5.59051275e-01 5.21937549e-01 1.29111424e-01 6.38668716e-01
-2.55220592e-01 -2.37174407e-01 1.28681839e+00 -6.19694740e-02
2.12621242e-01 1.60191819e-01 -8.66548002e-01 8.03725183e-01
7.60994554e-01 -1.16644573e+00 -5.05503953e-01 -4.08949018e-01
2.38744989e-01 -6.14629174e-03 8.56383443e-01 -9.08831894e-01
-1.65576696e-01 -1.34513497e-01 4.95299250e-01 -5.48386574e-01
-5.45037072e-03 -9.77557123e-01 1.06690042e-01 7.67189562e-01
-6.35373473e-01 1.74812496e-01 1.65544078e-01 7.06256926e-01
-1.28812268e-02 -2.92075366e-01 8.63287687e-01 -4.93083358e-01
-1.96670040e-01 -2.38375947e-01 -1.74398333e-01 4.96442229e-01
8.72034311e-01 3.82713564e-02 -4.22048241e-01 -6.68756012e-03
-3.92715365e-01 -1.62991658e-01 4.54397559e-01 3.23002815e-01
1.37593329e-01 -8.59633625e-01 -8.30415189e-01 1.44199625e-01
1.63783014e-01 -8.39700252e-02 3.34654450e-01 9.13134098e-01
-2.47960761e-01 5.86875439e-01 6.72172308e-02 -8.97790372e-01
-1.37529385e+00 3.49235922e-01 5.98160386e-01 -6.18101060e-01
-5.58546543e-01 8.38709116e-01 4.65839297e-01 -3.14718425e-01
7.48555176e-03 -3.50085407e-01 2.25028440e-01 3.10712177e-02
2.93779016e-01 4.45059836e-01 3.80352736e-01 -3.67659330e-01
-5.71014941e-01 4.00951952e-01 -9.75297466e-02 1.56649217e-01
1.32124901e+00 7.43796155e-02 8.26921389e-02 5.10402322e-01
1.09500861e+00 1.29887804e-01 -1.07471943e+00 -1.20566741e-01
8.87204409e-02 -3.14488143e-01 1.08536616e-01 -1.32318318e+00
-1.00008214e+00 6.77025199e-01 8.11853647e-01 8.96228403e-02
9.93473589e-01 2.46819824e-01 4.74708527e-01 2.65946332e-02
-1.30659088e-01 -8.29673827e-01 1.53966159e-01 -9.68881100e-02
4.37436312e-01 -1.40741515e+00 3.00477624e-01 -3.32175136e-01
-8.70051324e-01 8.01563025e-01 4.36987430e-01 7.55901858e-02
5.23434937e-01 2.94605941e-01 3.63327146e-01 -5.33807755e-01
-7.76008189e-01 -5.59318718e-03 3.00118327e-01 7.76255131e-01
5.75405955e-01 1.88261360e-01 1.34992287e-01 4.42563504e-01
-1.91518024e-03 1.59124643e-01 4.91469473e-01 8.79149854e-01
-1.47570878e-01 -1.02168739e+00 -6.82926714e-01 9.43246543e-01
-9.09059346e-01 -2.47475654e-01 -4.90522712e-01 8.93238544e-01
4.98121142e-01 6.69436216e-01 4.66174446e-02 9.27265882e-02
2.34618098e-01 3.24178338e-01 3.43254745e-01 -8.25369656e-01
-8.46861184e-01 -2.66158953e-02 -5.68375438e-02 -2.04965875e-01
-8.41103435e-01 -7.06712961e-01 -1.16398025e+00 9.98568833e-02
-3.87621492e-01 -1.81770846e-01 4.42538738e-01 9.03334379e-01
4.79844093e-01 1.10974932e+00 4.23301905e-01 1.28529176e-01
-7.78954983e-01 -6.38241112e-01 -4.08842117e-01 6.67882919e-01
7.35496581e-01 -6.26470566e-01 -6.03042662e-01 3.09737921e-01] | [15.034908294677734, -2.093337059020996] |
3b1b214b-3515-4a19-8ac2-5ad877660bd1 | rama-a-rapid-multicut-algorithm-on-gpu | 2109.01838 | null | https://arxiv.org/abs/2109.01838v3 | https://arxiv.org/pdf/2109.01838v3.pdf | RAMA: A Rapid Multicut Algorithm on GPU | We propose a highly parallel primal-dual algorithm for the multicut (a.k.a. correlation clustering) problem, a classical graph clustering problem widely used in machine learning and computer vision. Our algorithm consists of three steps executed recursively: (1) Finding conflicted cycles that correspond to violated inequalities of the underlying multicut relaxation, (2) Performing message passing between the edges and cycles to optimize the Lagrange relaxation coming from the found violated cycles producing reduced costs and (3) Contracting edges with high reduced costs through matrix-matrix multiplications. Our algorithm produces primal solutions and lower bounds that estimate the distance to optimum. We implement our algorithm on GPUs and show resulting one to two orders-of-magnitudes improvements in execution speed without sacrificing solution quality compared to traditional sequential algorithms that run on CPUs. We can solve very large scale benchmark problems with up to $\mathcal{O}(10^8)$ variables in a few seconds with small primal-dual gaps. Our code is available at https://github.com/pawelswoboda/RAMA. | ['Paul Swoboda', 'Ahmed Abbas'] | 2021-09-04 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Abbas_RAMA_A_Rapid_Multicut_Algorithm_on_GPU_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Abbas_RAMA_A_Rapid_Multicut_Algorithm_on_GPU_CVPR_2022_paper.pdf | cvpr-2022-1 | ['3d-instance-segmentation-1', 'graph-partitioning'] | ['computer-vision', 'graphs'] | [ 5.66047393e-02 1.66227780e-02 -2.41292551e-01 1.24255061e-01
-9.96803701e-01 -8.58163059e-01 2.22522944e-01 4.86964881e-01
-2.42535219e-01 7.81724870e-01 -1.91254362e-01 -5.74648082e-01
-3.04428250e-01 -7.97884822e-01 -6.78379476e-01 -9.52496827e-01
-5.18568575e-01 1.00258517e+00 1.09382771e-01 7.01606125e-02
3.06923807e-01 2.22592935e-01 -1.15245378e+00 2.21066758e-01
7.23930478e-01 7.77256966e-01 -6.49420768e-02 1.00127208e+00
1.82948038e-01 2.78138846e-01 -3.23772021e-02 -2.11190194e-01
5.38305342e-01 -4.93064433e-01 -1.14977169e+00 3.53584349e-01
-6.64257035e-02 2.33781189e-01 -2.36384153e-01 1.03159463e+00
2.34637365e-01 1.46880269e-01 2.95125961e-01 -1.45851779e+00
-2.43291005e-01 5.66035330e-01 -1.58608925e+00 2.28547126e-01
3.11220944e-01 -5.37718199e-02 1.21909606e+00 -6.53944910e-01
5.45790195e-01 1.11247468e+00 5.74831545e-01 -2.64222622e-02
-1.71563923e+00 -4.62344915e-01 -8.28952622e-03 1.87969819e-01
-1.57447922e+00 -1.28220558e-01 4.97205049e-01 -1.77247390e-01
9.66374815e-01 8.32619190e-01 6.71434402e-01 2.59124428e-01
-1.25400901e-01 4.12338197e-01 1.16347444e+00 -3.95893991e-01
2.23009691e-01 -8.08763877e-02 3.02323550e-01 8.91997874e-01
5.20228088e-01 -2.49841679e-02 -1.05548762e-01 -5.33978045e-01
3.26278538e-01 -2.85541981e-01 -1.24181405e-01 -2.20051378e-01
-1.20451581e+00 1.19606292e+00 3.91055524e-01 -5.02132438e-02
-2.19089761e-01 5.00576377e-01 5.27546942e-01 2.20355749e-01
2.68183649e-01 5.45335300e-02 -4.33557928e-01 -4.77295928e-02
-9.14803565e-01 1.58823982e-01 9.63594437e-01 8.31747591e-01
9.05447960e-01 -3.82784367e-01 1.67082265e-01 7.56023228e-01
-6.83994517e-02 3.49212885e-01 -2.48628721e-01 -1.25589097e+00
3.88824880e-01 4.30239201e-01 -1.41834160e-02 -1.17940044e+00
-6.69662952e-01 -1.79922774e-01 -9.35452819e-01 -3.33916396e-02
5.19977391e-01 -4.07177985e-01 -4.60049778e-01 1.40383625e+00
8.17119241e-01 3.47156763e-01 -2.14961156e-01 9.66801822e-01
2.93120652e-01 9.13134873e-01 -3.15208286e-01 -6.31116927e-01
1.62563872e+00 -1.12600839e+00 -1.77490160e-01 -1.53598115e-01
1.01424301e+00 -1.16146064e+00 6.01762235e-01 4.95613784e-01
-1.27082574e+00 -9.10676867e-02 -6.59230590e-01 -7.62657374e-02
-1.75769106e-01 -9.07053128e-02 1.02414060e+00 6.75732911e-01
-1.25558352e+00 5.83169818e-01 -9.09636140e-01 -9.22870114e-02
1.17759049e-01 8.18578064e-01 -2.32691601e-01 -1.20033078e-01
-5.34532368e-01 2.76348978e-01 5.68247020e-01 3.41457278e-02
-3.50814074e-01 -6.61363780e-01 -6.46551549e-01 -1.00198388e-01
8.25821757e-01 -6.54175103e-01 8.06619048e-01 -7.43126035e-01
-8.68847489e-01 1.01407814e+00 -3.80561113e-01 -3.32419634e-01
4.06093180e-01 1.86226398e-01 1.25293091e-01 4.61000875e-02
1.87908456e-01 4.15438652e-01 2.86254197e-01 -1.37279069e+00
-7.87462533e-01 -5.50552428e-01 2.77322065e-02 1.28627956e-01
2.38340329e-02 3.30024250e-02 -1.01578844e+00 -4.26649600e-01
4.16894495e-01 -1.50760007e+00 -6.54230416e-01 -3.60717773e-01
-5.36183178e-01 -9.35227349e-02 5.75840116e-01 -6.68201983e-01
1.20699728e+00 -1.82683504e+00 3.71807963e-01 7.19008744e-01
4.61041778e-01 -8.82667974e-02 -9.52090546e-02 5.73119819e-01
-9.85453799e-02 -1.37011170e-01 -3.00101340e-01 -9.02734920e-02
-5.42609878e-02 3.69417965e-01 2.04508692e-01 8.58150125e-01
-4.19824600e-01 6.52999938e-01 -1.03063178e+00 -4.86308515e-01
3.19560915e-02 1.46272585e-01 -8.87292027e-01 -1.52346760e-01
-9.03881788e-02 3.00692230e-01 -1.72951698e-01 4.29996818e-01
9.88690078e-01 -5.18740773e-01 8.51768136e-01 -6.21227808e-02
-1.92722693e-01 5.22913337e-02 -1.71467459e+00 1.37369215e+00
-3.35708559e-01 4.09692615e-01 5.35369694e-01 -1.35256851e+00
5.28712213e-01 1.33957388e-02 8.20062101e-01 -4.89053756e-01
2.31997862e-01 2.17786685e-01 -2.60254610e-02 -1.93353653e-01
4.57572043e-01 1.39996484e-01 -2.13012829e-01 7.74925828e-01
-4.43181664e-01 -8.35486222e-03 7.75211632e-01 5.34998477e-01
1.12090087e+00 -2.38639936e-01 1.36187479e-01 -5.62699497e-01
5.23139060e-01 4.37664509e-01 7.66248405e-01 4.27231610e-01
1.50376305e-01 2.04534337e-01 1.11895978e+00 -5.86374938e-01
-1.08768713e+00 -6.55819893e-01 1.30078465e-01 1.04869914e+00
2.69424379e-01 -7.08637059e-01 -8.20815682e-01 -1.47507221e-01
9.77649167e-02 2.48439118e-01 -5.87551951e-01 1.47285149e-01
-6.31621897e-01 -1.26980042e+00 8.33075121e-02 3.85847092e-01
2.11258337e-01 -3.68774831e-01 -4.28762645e-01 1.90357417e-01
-1.50581166e-01 -1.17965865e+00 -6.67624056e-01 3.09461504e-01
-9.18328106e-01 -1.25553000e+00 -2.56776303e-01 -9.27301884e-01
1.06424499e+00 4.82285351e-01 1.05626428e+00 5.09155035e-01
-5.70065618e-01 3.98951881e-02 -1.22174583e-01 2.12648898e-01
-9.07199904e-02 -2.08204403e-03 -2.29090098e-02 -2.54514664e-01
2.39217669e-01 -5.25786877e-01 -6.34066880e-01 2.73528367e-01
-5.01676023e-01 2.37610325e-01 2.64545441e-01 9.31470990e-01
1.02440321e+00 3.99275154e-01 -1.14742123e-01 -1.18682969e+00
5.04039705e-01 -6.38126910e-01 -1.31933129e+00 1.81032583e-01
-6.74876213e-01 1.52413264e-01 5.51054776e-01 -1.46783426e-01
-5.02099991e-01 4.26732004e-01 1.01152770e-01 -2.48819143e-01
2.83300221e-01 5.16887069e-01 4.32464331e-01 -2.46821955e-01
2.95684397e-01 9.62761492e-02 -1.60427645e-01 -1.47381335e-01
4.72371995e-01 2.51583427e-01 5.04029095e-01 -9.82579768e-01
8.01719725e-01 7.69760907e-01 5.71730196e-01 -6.86091959e-01
-6.78411067e-01 -7.14024007e-01 -5.09554684e-01 -9.16355401e-02
7.16716886e-01 -7.36417472e-01 -1.23146546e+00 -6.77945092e-02
-9.45702255e-01 -3.26420009e-01 1.92945093e-01 2.78388828e-01
-3.45570326e-01 6.97640836e-01 -8.11148465e-01 -7.64657319e-01
-3.24409902e-01 -1.08751750e+00 8.28339040e-01 7.03842118e-02
-1.56556949e-01 -9.94476378e-01 2.38059729e-01 7.65115798e-01
-2.67102331e-01 6.88755512e-01 7.73911834e-01 -2.54713390e-02
-6.64014935e-01 1.45518780e-01 -4.72157419e-01 -2.27098435e-01
-1.60673454e-01 1.33672372e-01 -4.06607121e-01 -7.07310557e-01
-2.85398990e-01 -1.98861118e-02 7.53681600e-01 4.39430982e-01
1.17543626e+00 -5.92114806e-01 -6.85684323e-01 9.02622342e-01
1.78582561e+00 1.26171678e-01 2.53733575e-01 2.40807682e-01
6.91131175e-01 3.77982408e-01 6.33536220e-01 6.86219037e-01
4.04287815e-01 5.86400568e-01 3.09946686e-01 -4.15083140e-01
2.48077586e-01 2.89898872e-01 2.76741926e-02 8.73972416e-01
-3.13918829e-01 5.19277975e-02 -1.14540577e+00 8.63657534e-01
-2.20715165e+00 -6.95784509e-01 -1.00658429e+00 2.10453248e+00
6.89978600e-01 1.74409617e-02 4.20791596e-01 3.46938759e-01
8.17453861e-01 -2.18807191e-01 -2.93925911e-01 -1.04752314e+00
2.62265891e-01 2.67670840e-01 9.34203565e-01 8.61466408e-01
-9.29958642e-01 8.94487798e-01 5.37003708e+00 6.77010000e-01
-7.06506193e-01 1.86141089e-01 8.58552039e-01 -4.24401820e-01
-1.25039667e-01 4.34646338e-01 -3.31826538e-01 3.31873953e-01
8.55172694e-01 -3.27551693e-01 9.12836075e-01 7.30735660e-01
2.62626559e-01 -5.75937569e-01 -9.88363326e-01 1.08214414e+00
-3.15651894e-01 -1.46203828e+00 -5.01359522e-01 5.24413168e-01
1.26628935e+00 -9.07524973e-02 -1.33891806e-01 -2.17509180e-01
5.43384492e-01 -7.92921901e-01 3.69500577e-01 -4.23419803e-01
5.27509749e-01 -1.24622536e+00 4.61926669e-01 1.80587798e-01
-1.47796905e+00 1.96713619e-02 -3.38865608e-01 -1.37024134e-01
6.71211854e-02 9.17927802e-01 -5.99228203e-01 6.40279233e-01
8.74469459e-01 6.97939470e-02 -7.32740909e-02 7.53910601e-01
8.01868066e-02 4.27181214e-01 -6.31965637e-01 3.29950526e-02
5.83800197e-01 -9.65025246e-01 4.33388382e-01 1.33086693e+00
1.34173468e-01 5.10856152e-01 3.39892358e-01 5.67504525e-01
-1.76124647e-01 8.06425959e-02 -2.88492590e-01 1.27037331e-01
3.50036949e-01 1.40178502e+00 -1.51952004e+00 -3.60902637e-01
-4.68175024e-01 1.08268249e+00 3.75993282e-01 1.72328755e-01
-1.10306537e+00 -4.03143257e-01 7.28853106e-01 -6.78893179e-02
4.46173728e-01 -5.46927035e-01 -5.99170804e-01 -8.08346868e-01
1.90314220e-03 -7.58485675e-01 8.57929170e-01 -2.13142797e-01
-8.37876201e-01 3.57442826e-01 -2.11391449e-01 -5.78593791e-01
4.72665504e-02 -4.46027607e-01 -5.13936341e-01 6.80059791e-01
-1.19780374e+00 -8.48456502e-01 -2.93413639e-01 7.31436849e-01
1.16439991e-01 4.93877143e-01 5.96848607e-01 4.22360957e-01
-8.37008953e-01 4.01038647e-01 3.67143691e-01 -2.57586148e-02
3.02390099e-01 -1.27993989e+00 2.75238723e-01 9.42582548e-01
2.04736087e-02 3.88862163e-01 8.98218572e-01 -4.81132925e-01
-1.79915154e+00 -7.57096827e-01 9.07091558e-01 2.61011627e-02
7.98250079e-01 -3.35254937e-01 -6.19759858e-01 6.89550936e-01
3.72547090e-01 6.14145994e-02 7.23325431e-01 2.05532551e-01
-1.31132260e-01 -9.15844068e-02 -1.05836320e+00 5.64732909e-01
9.28212345e-01 -5.43030575e-02 9.71894190e-02 7.98801124e-01
2.98964500e-01 -6.84628606e-01 -9.50335503e-01 -3.18718962e-02
2.57868946e-01 -8.72104228e-01 1.10370362e+00 -4.20338213e-01
2.43157312e-01 -4.50328559e-01 -4.23415098e-03 -1.00542426e+00
-4.77692336e-01 -9.93398488e-01 -1.16538536e-02 8.76964688e-01
5.12238562e-01 -5.35665691e-01 8.59364867e-01 5.45447290e-01
2.21154496e-01 -9.83953774e-01 -8.40380371e-01 -7.11484075e-01
6.76183030e-02 -3.18928927e-01 3.24899495e-01 1.21132004e+00
2.42532209e-01 3.76065314e-01 -2.58092850e-01 3.73400509e-01
1.03018749e+00 7.46496618e-01 7.13974535e-01 -9.77487564e-01
-5.43513060e-01 -3.62167776e-01 6.64017648e-02 -6.43846810e-01
-4.67166603e-02 -1.18976462e+00 -3.32460195e-01 -1.49685264e+00
5.54010153e-01 -7.10863113e-01 8.55060965e-02 5.81622779e-01
-1.45002410e-01 5.93110204e-01 -4.34742495e-03 -1.38808951e-01
-6.77362144e-01 -5.73130846e-02 9.25166428e-01 1.36873145e-02
-2.75710046e-01 -2.43682474e-01 -7.47742832e-01 5.87322235e-01
9.56743121e-01 -6.53566539e-01 -1.52899727e-01 -5.11983693e-01
4.38318878e-01 4.73066956e-01 1.46302372e-01 -6.88101709e-01
2.98501581e-01 -3.68073136e-01 1.05199791e-01 -4.99834210e-01
1.93394735e-01 -6.51282787e-01 4.97958094e-01 1.06620598e+00
8.47360492e-02 4.67773587e-01 4.32623863e-01 4.26907212e-01
5.64851575e-02 -8.46590921e-02 1.06302810e+00 -9.02482718e-02
-4.38505977e-01 1.83324888e-01 -2.42160574e-01 7.95161203e-02
1.49290645e+00 -7.30174631e-02 -2.89962590e-01 -1.57173499e-01
-7.51045048e-01 8.09323668e-01 4.21076953e-01 -1.12948837e-02
1.93075061e-01 -1.12186217e+00 -7.78801441e-01 -1.95521146e-01
-3.44083369e-01 -3.28074656e-02 2.60699362e-01 1.17774367e+00
-1.08451581e+00 4.82150018e-01 -1.83484741e-02 -7.06871867e-01
-1.72427309e+00 9.06618416e-01 1.74201280e-03 -4.76235718e-01
-4.98001724e-01 1.02671158e+00 -1.76228121e-01 -1.61174163e-01
-1.53535485e-01 9.05433223e-02 5.68693399e-01 8.88485461e-02
1.72074050e-01 9.68314171e-01 -1.10796308e-02 -5.59125125e-01
-6.16888821e-01 6.25251889e-01 1.42181562e-02 -1.46810219e-01
1.61222517e+00 -5.72002269e-02 -7.88602889e-01 -3.08919281e-01
1.38627982e+00 6.64198250e-02 -8.85333478e-01 -6.98340759e-02
1.05718791e-01 -5.58022022e-01 1.69895187e-01 -2.14915246e-01
-1.37708163e+00 4.07461554e-01 2.66452461e-01 3.21437508e-01
1.36586034e+00 2.01795697e-01 9.50008690e-01 1.48015976e-01
3.72378320e-01 -1.43916774e+00 -2.96421736e-01 2.17012092e-01
4.80169713e-01 -1.09125483e+00 4.61163521e-01 -8.28141212e-01
-4.64133114e-01 1.04639137e+00 3.56831580e-01 -3.37674081e-01
4.85693544e-01 5.40638924e-01 -3.39356393e-01 -2.48199314e-01
-8.95263791e-01 -2.00703636e-01 -3.38656493e-02 1.63654909e-01
2.67342478e-01 3.78287375e-01 -9.41589952e-01 6.61254749e-02
-2.05759630e-01 -3.78629595e-01 6.59069359e-01 7.58379340e-01
-1.74001545e-01 -1.38545120e+00 -6.77609861e-01 4.62011367e-01
-4.64520663e-01 -6.04128428e-02 -2.83015341e-01 5.99623561e-01
1.20472774e-01 9.97277677e-01 1.63163975e-01 -1.57153472e-01
-6.55047596e-02 -3.63062799e-01 5.24899006e-01 -3.55107456e-01
-5.72890699e-01 3.50216955e-01 2.13599801e-01 -7.00229347e-01
-1.51262894e-01 -9.13429260e-01 -1.50836205e+00 -1.01831305e+00
-1.99512511e-01 4.78157550e-01 5.90875447e-01 5.17509162e-01
2.77943939e-01 3.70409071e-01 7.53391385e-01 -7.90165365e-01
-1.89628173e-02 -1.84850141e-01 -4.56774652e-01 1.57649964e-01
5.06317690e-02 -2.29324937e-01 -2.47494206e-01 -9.76102874e-02] | [7.0462260246276855, 5.172204971313477] |
aba3a5d1-e7c9-4f89-a810-ccf8d0ae956f | differentially-private-adversarial-auto | 2307.02135 | null | https://arxiv.org/abs/2307.02135v1 | https://arxiv.org/pdf/2307.02135v1.pdf | Differentially Private Adversarial Auto-Encoder to Protect Gender in Voice Biometrics | Over the last decade, the use of Automatic Speaker Verification (ASV) systems has become increasingly widespread in response to the growing need for secure and efficient identity verification methods. The voice data encompasses a wealth of personal information, which includes but is not limited to gender, age, health condition, stress levels, and geographical and socio-cultural origins. These attributes, known as soft biometrics, are private and the user may wish to keep them confidential. However, with the advancement of machine learning algorithms, soft biometrics can be inferred automatically, creating the potential for unauthorized use. As such, it is crucial to ensure the protection of these personal data that are inherent within the voice while retaining the utility of identity recognition. In this paper, we present an adversarial Auto-Encoder--based approach to hide gender-related information in speaker embeddings, while preserving their effectiveness for speaker verification. We use an adversarial procedure against a gender classifier and incorporate a layer based on the Laplace mechanism into the Auto-Encoder architecture. This layer adds Laplace noise for more robust gender concealment and ensures differential privacy guarantees during inference for the output speaker embeddings. Experiments conducted on the VoxCeleb dataset demonstrate that speaker verification tasks can be effectively carried out while concealing speaker gender and ensuring differential privacy guarantees; moreover, the intensity of the Laplace noise can be tuned to select the desired trade-off between privacy and utility. | ['Melek Önen', 'Massimiliano Todisco', 'Imen Chihaoui', 'Ismet Kerenciler', 'Oualid Zari', 'Michele Panariello', 'Oubaïda Chouchane'] | 2023-07-05 | null | null | null | null | ['speaker-verification'] | ['speech'] | [ 1.62283748e-01 2.05291957e-01 -1.31678194e-01 -5.83379447e-01
-7.04738021e-01 -8.47369492e-01 4.57037300e-01 2.79390693e-01
-4.21611071e-01 4.52224731e-01 2.34485164e-01 -2.73814410e-01
4.60622050e-02 -6.38988078e-01 -3.20562899e-01 -8.66316855e-01
-6.40465915e-02 -1.43825442e-01 -2.41797313e-01 4.88494411e-02
1.07329935e-01 6.62494540e-01 -1.54547787e+00 -1.93771757e-02
6.62756383e-01 1.23028111e+00 -5.71259856e-01 5.70315003e-01
7.89972320e-02 1.67467386e-01 -8.08420777e-01 -9.81775582e-01
2.50277817e-01 -2.46397331e-01 -4.43645120e-01 -4.73150253e-01
2.85357565e-01 -3.67499232e-01 -3.24933261e-01 1.00404489e+00
7.46017337e-01 -2.02169895e-01 5.31909645e-01 -1.27154136e+00
-6.22840822e-01 3.78929675e-01 -2.36582443e-01 -1.45636097e-01
5.03661990e-01 9.71672013e-02 9.26682413e-01 -7.14922607e-01
2.46710271e-01 1.06379890e+00 7.36240685e-01 7.78918862e-01
-1.30726230e+00 -9.56101358e-01 -2.26470381e-01 2.19239239e-02
-1.87607920e+00 -8.61700892e-01 8.43437552e-01 -4.07942653e-01
2.15429440e-01 6.56939983e-01 4.60619777e-01 1.31569552e+00
9.29980054e-02 3.35845232e-01 8.77630651e-01 -3.17630917e-01
2.79190600e-01 8.13199878e-01 -3.14057380e-01 6.08603776e-01
1.43822819e-01 2.78070122e-01 -7.49207377e-01 -5.63094795e-01
2.73753077e-01 -1.46021262e-01 -4.02331591e-01 -3.37805331e-01
-7.72851408e-01 7.60430157e-01 1.10326491e-01 1.66411400e-01
-1.46162882e-01 -1.42737538e-01 5.67692280e-01 1.55479684e-01
4.13777351e-01 2.44868934e-01 -8.38316679e-02 -3.79941426e-02
-9.94242787e-01 2.44288728e-01 8.81556928e-01 5.66506386e-01
5.28908253e-01 1.56344861e-01 -1.17537491e-01 7.75336862e-01
4.68960494e-01 6.19293392e-01 5.42401910e-01 -6.46385133e-01
2.18216076e-01 5.06492138e-01 -9.83841810e-03 -1.22489524e+00
1.38287410e-01 -3.27051252e-01 -9.11911905e-01 1.86730072e-01
3.51939291e-01 -1.07866071e-01 -5.55713654e-01 2.09027147e+00
4.79062140e-01 3.40053067e-02 4.04279009e-02 6.55888557e-01
5.92044473e-01 2.90518463e-01 2.43594989e-01 4.60701175e-02
1.61318970e+00 7.26224855e-02 -7.87254810e-01 -1.41132653e-01
1.51936442e-01 -6.27633035e-01 8.04521143e-01 -5.52184023e-02
-7.62465596e-01 -2.04238489e-01 -1.13656294e+00 -6.66507930e-02
-5.84042907e-01 -3.06493100e-02 4.77442324e-01 1.60135424e+00
-9.03125286e-01 3.24144572e-01 -6.79043472e-01 -4.91255037e-02
3.74589801e-01 8.10554445e-01 -5.76757193e-01 2.86035448e-01
-1.54065168e+00 5.48476934e-01 -3.32450196e-02 3.58631819e-01
-4.09686595e-01 -7.62407362e-01 -1.12853241e+00 1.82231233e-01
-2.05487102e-01 -3.62710983e-01 7.89659560e-01 -7.89986551e-01
-1.48323298e+00 1.03861940e+00 -2.31058344e-01 -3.95426899e-01
5.57735026e-01 3.20004970e-01 -7.55890906e-01 3.05762701e-02
-1.69330224e-01 3.46954405e-01 1.33014917e+00 -9.35447514e-01
-3.87482941e-01 -7.92774260e-01 -2.76489735e-01 -1.89555719e-01
-7.31041789e-01 2.11414590e-01 -3.19073573e-02 -6.05784655e-01
-7.26997182e-02 -9.28628504e-01 2.75578320e-01 1.60522684e-01
-3.12977403e-01 1.49358228e-01 1.00801969e+00 -1.13018799e+00
1.22488749e+00 -2.58315921e+00 -5.99761978e-02 5.16118050e-01
4.93187942e-02 3.24131519e-01 2.04704210e-01 1.39535218e-01
3.45019288e-02 2.38905758e-01 -4.13676739e-01 -6.42110527e-01
1.79356188e-01 1.89859811e-02 -4.11125511e-01 7.59720623e-01
2.43617713e-01 6.38844371e-01 -4.50945497e-01 -4.87604439e-01
5.37185669e-02 9.50045526e-01 -5.29657960e-01 2.14154392e-01
2.69858867e-01 3.01753521e-01 -2.73481876e-01 8.28253865e-01
8.64399374e-01 4.70748693e-01 1.46889687e-01 -1.74441691e-02
6.64665699e-02 1.91182524e-01 -1.10250270e+00 9.98675168e-01
-3.99811774e-01 4.94273961e-01 7.55482554e-01 -5.91390491e-01
9.99774158e-01 6.11160338e-01 3.42566431e-01 -3.43010485e-01
4.44951057e-01 1.63814381e-01 -1.16409138e-01 -2.33913064e-01
6.49830043e-01 -2.40089476e-01 -2.65524060e-01 3.62812787e-01
-1.82703331e-01 2.04935074e-01 -5.85478127e-01 -4.37369719e-02
5.38268209e-01 -3.54796499e-01 1.01700403e-01 -1.41958967e-01
8.56172800e-01 -6.94655955e-01 5.98011434e-01 1.02967449e-01
-5.67068160e-01 4.60327804e-01 3.91132206e-01 8.54664966e-02
-7.58163035e-01 -9.74650145e-01 -4.67395008e-01 8.58340502e-01
-2.49363706e-01 -1.41249374e-01 -8.27557921e-01 -6.76513016e-01
3.96936357e-01 5.89341700e-01 -7.17076302e-01 -4.98827845e-01
-3.65655273e-01 -3.84120047e-01 1.15872014e+00 3.79119664e-01
2.88563997e-01 -6.16008997e-01 -2.69097120e-01 -5.35018407e-02
-1.07558511e-01 -9.62182462e-01 -7.65626729e-01 -2.15070788e-02
-3.91373992e-01 -7.02306747e-01 -6.02320135e-01 -6.24199867e-01
5.95523953e-01 -2.13780105e-01 3.58776778e-01 -9.25850943e-02
-2.58622080e-01 4.98429060e-01 -1.26894226e-03 -5.44452548e-01
-7.86449015e-01 2.65669376e-01 2.05685541e-01 7.70249784e-01
4.78844434e-01 -6.20397925e-01 -5.16685605e-01 2.61163265e-01
-8.18531454e-01 -5.90935171e-01 1.21644206e-01 7.75774539e-01
1.12029091e-01 1.16640717e-01 8.27909291e-01 -6.33268833e-01
6.09379947e-01 -4.03898835e-01 -4.87304628e-01 2.27814317e-01
-6.10720932e-01 1.95095723e-04 5.87603807e-01 -4.64799583e-01
-9.88207459e-01 4.08604592e-02 -3.45200479e-01 -2.26710439e-01
-8.09099823e-02 1.66855276e-01 -7.42094457e-01 -4.87711489e-01
2.95077860e-01 4.31006402e-01 4.94173527e-01 -2.42375940e-01
2.23691568e-01 1.38901830e+00 6.56553686e-01 -3.28521729e-01
8.80478084e-01 3.89228553e-01 -1.05882823e-01 -1.01067281e+00
-2.70720899e-01 -3.78609359e-01 -4.53810841e-01 -1.35777622e-01
6.91720068e-01 -7.16834307e-01 -1.16004908e+00 6.79832399e-01
-7.78321266e-01 3.51644903e-01 3.33777480e-02 2.31342748e-01
-2.70211846e-01 4.96837169e-01 -4.24183041e-01 -1.17531681e+00
-5.72461069e-01 -1.02037239e+00 9.98415887e-01 9.54780634e-03
-5.71711123e-01 -9.64209378e-01 -2.78260410e-01 5.20447493e-01
7.38987744e-01 2.06431508e-01 9.22722995e-01 -7.67910659e-01
-3.61613214e-01 -7.54836202e-01 6.66056648e-02 5.79422414e-01
4.24802840e-01 -2.11459816e-01 -1.42304969e+00 -3.40596884e-01
2.43028685e-01 -6.76520765e-02 5.27867615e-01 -4.44244780e-02
1.12246561e+00 -8.06485951e-01 -4.38803919e-02 7.55422533e-01
9.52253759e-01 7.89618641e-02 5.55482686e-01 -1.21485673e-01
5.90243816e-01 8.28176498e-01 2.12700088e-02 5.29505670e-01
4.14513350e-01 5.85737705e-01 3.50038469e-01 9.83713716e-02
2.17263728e-01 -4.46524531e-01 3.03175598e-01 4.85793412e-01
1.85763955e-01 8.55173096e-02 -6.57347560e-01 4.44413215e-01
-1.25174880e+00 -1.10100043e+00 4.67536122e-01 2.63702178e+00
1.08159280e+00 -1.48382306e-01 1.95344821e-01 5.21478534e-01
7.00748324e-01 3.42310935e-01 -5.19430578e-01 -6.34159088e-01
-4.26544584e-02 9.20462459e-02 6.43192828e-01 5.46516240e-01
-1.12825537e+00 5.78015149e-01 5.75803614e+00 4.56872284e-01
-1.41480052e+00 -1.37449190e-01 5.96876025e-01 -2.20907405e-02
-4.71386105e-01 -4.31353211e-01 -7.65390933e-01 5.48003614e-01
1.06228161e+00 -3.80573720e-01 5.96909881e-01 7.18442321e-01
9.00470763e-02 3.33878219e-01 -1.11206055e+00 8.36248875e-01
2.14950547e-01 -8.36464822e-01 -3.39949161e-01 3.34466368e-01
6.27440512e-02 -6.92599952e-01 5.39143860e-01 9.04228613e-02
-2.55644500e-01 -1.01524973e+00 7.74927378e-01 2.00136870e-01
1.10040200e+00 -1.10245669e+00 6.56153619e-01 1.42193437e-01
-1.00463581e+00 -2.83107311e-01 3.90194878e-02 4.27218258e-01
-5.11748120e-02 3.46011490e-01 -1.02483320e+00 2.90850759e-01
4.28930908e-01 1.48047209e-01 -3.31784338e-01 4.92312074e-01
-1.73882782e-01 6.54808760e-01 -3.42134356e-01 -7.51611870e-03
-2.76192725e-01 -5.32280989e-02 5.39720237e-01 1.08679283e+00
3.19773316e-01 -3.27349976e-02 -3.89489800e-01 8.99244189e-01
-3.02570492e-01 1.89562932e-01 -6.67563975e-01 -3.23651910e-01
7.63368428e-01 1.00158215e+00 -1.54007837e-01 1.67049244e-01
-1.21952422e-01 1.03215408e+00 -2.94601601e-02 2.12417006e-01
-5.83294690e-01 -4.74348664e-01 1.19934654e+00 3.17835987e-01
1.53546199e-01 -9.95125845e-02 -5.79393923e-01 -8.60313416e-01
9.03962329e-02 -1.12149060e+00 4.21711266e-01 -6.07302273e-03
-1.06744242e+00 3.58146608e-01 -4.42323446e-01 -8.20236504e-01
-3.73529971e-01 -2.45920792e-01 -3.41027230e-01 1.24679160e+00
-1.26834106e+00 -1.28143501e+00 8.34255740e-02 5.37462354e-01
-1.81550294e-01 -2.98409313e-01 1.22168314e+00 4.02690530e-01
-5.40575325e-01 1.33557367e+00 -1.77053437e-02 4.08313692e-01
7.25184798e-01 -8.93007994e-01 2.78985679e-01 6.18400455e-01
-2.18757600e-01 1.10466218e+00 6.16502345e-01 -4.17109698e-01
-1.49958968e+00 -9.34057772e-01 1.23492730e+00 -3.63925725e-01
4.07006711e-01 -9.04738188e-01 -1.07467675e+00 5.61673999e-01
-1.09128855e-01 -1.76129252e-01 1.13059783e+00 1.27321452e-01
-7.92796493e-01 -4.16719854e-01 -1.73946369e+00 4.44807708e-01
4.49190557e-01 -1.17365062e+00 -3.53246659e-01 -2.15183824e-01
5.11869967e-01 -1.08075224e-01 -9.35609519e-01 1.76480949e-01
9.23921108e-01 -7.27777898e-01 1.07080805e+00 -3.57693076e-01
-4.83736284e-02 -2.01131150e-01 -2.68421263e-01 -8.85809302e-01
2.99362633e-02 -6.84801340e-01 -1.56768128e-01 1.87955904e+00
3.60724002e-01 -9.90611196e-01 9.88334835e-01 1.34679139e+00
5.30781865e-01 -4.39075530e-01 -1.29447079e+00 -6.24251902e-01
-1.09123155e-01 -3.08838874e-01 1.01081300e+00 9.64164555e-01
6.93461671e-02 -1.04249753e-01 -5.74527502e-01 5.21169245e-01
9.00425494e-01 -1.96685553e-01 5.88534296e-01 -1.15085316e+00
-4.25426215e-02 -3.67774755e-01 -7.23827839e-01 -3.27516377e-01
4.96454448e-01 -8.74143124e-01 -1.75198689e-01 -6.98550522e-01
-8.27885419e-02 -3.54622066e-01 -3.28016877e-01 4.95021254e-01
-8.97818059e-02 1.90033585e-01 -1.00843117e-01 -1.94204152e-01
3.47468883e-01 6.15412056e-01 5.49513876e-01 -2.11725608e-01
-1.66712493e-01 4.25397098e-01 -8.60332549e-01 3.85423422e-01
7.79079556e-01 -4.53523129e-01 -3.20649385e-01 8.05900544e-02
-1.07897848e-01 5.68540581e-02 3.88422996e-01 -6.92182600e-01
3.40264559e-01 9.92335379e-02 3.25896293e-01 -1.52268767e-01
5.96160531e-01 -1.03045571e+00 -2.20083445e-02 4.44415987e-01
-3.31700861e-01 -2.02515662e-01 2.37329140e-01 4.44173813e-01
-2.96811879e-01 6.59721419e-02 8.57368767e-01 4.21858191e-01
-8.99341181e-02 2.89070725e-01 -2.80169278e-01 -1.65685624e-01
9.10143912e-01 -2.72796482e-01 8.76046717e-02 -4.77794468e-01
-4.85576779e-01 3.68166193e-02 5.34269631e-01 4.69211757e-01
6.19580090e-01 -1.11209369e+00 -6.87250972e-01 8.70792985e-01
1.82123870e-01 -6.39331162e-01 3.30371380e-01 3.96977216e-01
-1.21948510e-01 2.36659735e-01 1.01692649e-02 -2.34843045e-01
-1.72826552e+00 5.36710143e-01 3.54471534e-01 4.00850326e-01
-2.30300844e-01 7.59903073e-01 -1.75740749e-01 -4.58020747e-01
5.23619592e-01 1.32731885e-01 4.06708196e-03 2.56237924e-01
7.42456198e-01 3.20998490e-01 1.94365382e-01 -8.73436511e-01
-6.83126330e-01 2.44250327e-01 1.57071576e-01 -1.54441863e-01
1.09454787e+00 -2.60852844e-01 -1.46370575e-01 8.88529196e-02
1.44472361e+00 5.70984185e-01 -9.01764452e-01 -7.97300711e-02
-5.92603125e-02 -7.08494067e-01 3.41319852e-02 -6.89242959e-01
-1.04243183e+00 8.84449482e-01 7.01870322e-01 2.98726648e-01
1.03154433e+00 -2.30586708e-01 8.58040512e-01 -2.67365426e-01
2.24791974e-01 -8.52553666e-01 -6.83346391e-01 8.77954215e-02
6.39440775e-01 -1.06045139e+00 -2.45659903e-01 -4.10921395e-01
-5.06739497e-01 7.79526353e-01 9.75052193e-02 5.24861276e-01
7.52748907e-01 3.80918235e-01 3.07005584e-01 2.14567259e-01
-2.35036105e-01 4.36159939e-01 3.22718740e-01 7.85186470e-01
2.33587116e-01 3.10254753e-01 -1.38468042e-01 9.47987676e-01
-6.76359415e-01 -3.67551029e-01 1.09056316e-01 7.37104356e-01
-8.24414939e-02 -1.26863647e+00 -6.39615774e-01 1.30344555e-01
-8.78380120e-01 7.01853409e-02 -5.57741463e-01 2.82370567e-01
7.18254298e-02 9.07217383e-01 -1.34931341e-01 -4.03578997e-01
2.01146051e-01 4.66115922e-01 1.29231870e-01 -1.36380792e-01
-7.89738774e-01 -4.88358080e-01 2.03949273e-01 -2.83291280e-01
3.33064795e-02 -8.16194713e-01 -9.57073867e-01 -6.85116887e-01
-1.96430564e-01 1.90924361e-01 1.16581070e+00 6.26364529e-01
5.20064354e-01 -8.26581474e-03 9.84070480e-01 -3.79643232e-01
-1.01283669e+00 -5.45720756e-01 -7.76334047e-01 3.77216727e-01
6.81526661e-01 -3.08921069e-01 -6.43842757e-01 -5.28391637e-02] | [14.000777244567871, 5.860822677612305] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.