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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
9086cbf8-9ee7-4808-a9c3-270a0fdece10 | key-frame-extraction-with-attention-based | 2306.13176 | null | https://arxiv.org/abs/2306.13176v1 | https://arxiv.org/pdf/2306.13176v1.pdf | Key Frame Extraction with Attention Based Deep Neural Networks | Automatic keyframe detection from videos is an exercise in selecting scenes that can best summarize the content for long videos. Providing a summary of the video is an important task to facilitate quick browsing and content summarization. The resulting photos are used for automated works (e.g. summarizing security footage, detecting different scenes used in music clips) in different industries. In addition, processing high-volume videos in advanced machine learning methods also creates resource costs. Keyframes obtained; It can be used as an input feature to the methods and models to be used. In this study; We propose a deep learning-based approach for keyframe detection using a deep auto-encoder model with an attention layer. The proposed method first extracts the features from the video frames using the encoder part of the autoencoder and applies segmentation using the k-means clustering algorithm to group these features and similar frames together. Then, keyframes are selected from each cluster by selecting the frames closest to the center of the clusters. The method was evaluated on the TVSUM video dataset and achieved a classification accuracy of 0.77, indicating a higher success rate than many existing methods. The proposed method offers a promising solution for key frame extraction in video analysis and can be applied to various applications such as video summarization and video retrieval. | ['Senem Tanberk', 'Samed Arslan'] | 2023-06-21 | null | null | null | null | ['video-retrieval'] | ['computer-vision'] | [ 2.29031086e-01 -5.13357460e-01 -3.01830441e-01 8.52539092e-02
-8.71460557e-01 -2.65765190e-01 3.07112873e-01 4.14281934e-01
-3.09274107e-01 3.76813620e-01 3.84999335e-01 4.42671776e-01
-1.71147659e-01 -4.53305870e-01 -6.09275043e-01 -9.92171764e-01
-1.19007573e-01 -5.00365309e-02 2.88011104e-01 1.36771366e-01
7.03910828e-01 5.69415748e-01 -1.93974662e+00 6.35132313e-01
5.31016648e-01 1.35604692e+00 7.41189480e-01 7.80162454e-01
-1.81039304e-01 1.16716850e+00 -9.15572405e-01 -6.09795675e-02
1.05072141e-01 -6.46195829e-01 -6.11839890e-01 4.49552923e-01
5.28054416e-01 -6.71756327e-01 -4.31731910e-01 1.00157702e+00
4.93375808e-01 6.40295446e-01 4.71575201e-01 -1.04059350e+00
3.09389010e-02 6.24848604e-01 -6.53292060e-01 5.87428868e-01
5.03050923e-01 -4.10831451e-01 8.79756391e-01 -7.55042434e-01
7.09313989e-01 9.35988724e-01 3.75332177e-01 2.25002781e-01
-6.73218906e-01 -5.29790521e-01 -2.36510441e-01 8.07305515e-01
-1.30081713e+00 -5.75208902e-01 1.02045429e+00 -4.86095369e-01
8.83652925e-01 4.10226971e-01 1.05556738e+00 4.85852361e-01
3.36713374e-01 1.08228683e+00 5.11888206e-01 -3.52703810e-01
2.66911983e-01 -1.03003658e-01 1.73956126e-01 5.68334341e-01
-1.24591567e-01 -7.89708436e-01 -7.44990408e-01 2.87291426e-02
5.82685053e-01 3.66290927e-01 -3.74927580e-01 2.24745378e-01
-1.23810482e+00 7.41810918e-01 3.52395087e-01 4.85330939e-01
-8.86852920e-01 5.98105490e-02 6.31489694e-01 1.49199307e-01
4.26796615e-01 3.65823120e-01 -4.43968037e-03 -1.08372614e-01
-1.52375031e+00 3.73162180e-01 3.59810948e-01 6.25123739e-01
5.88502645e-01 7.17870146e-02 -2.84328341e-01 8.89970422e-01
-3.07938363e-02 1.96083672e-02 7.98907161e-01 -1.28053117e+00
2.80069828e-01 6.39499843e-01 -1.74052522e-01 -1.35017979e+00
-2.47757509e-01 -6.68272227e-02 -7.36382008e-01 -1.23265721e-01
-6.35541081e-02 1.18049771e-01 -6.66208029e-01 9.55332220e-01
3.22396368e-01 2.65214890e-01 -1.32687673e-01 9.19595897e-01
1.16687107e+00 1.38871062e+00 -1.43493474e-01 -7.20739365e-01
1.50918627e+00 -8.44159782e-01 -8.87304604e-01 2.68130720e-01
1.64976850e-01 -9.44483101e-01 5.95404506e-01 7.52582967e-01
-1.13850284e+00 -8.04972172e-01 -1.02342153e+00 -4.94333878e-02
3.10166907e-02 5.50208151e-01 2.55532324e-01 -9.39016566e-02
-9.84836996e-01 6.38629913e-01 -6.12232089e-01 -5.31895161e-01
5.39943874e-01 3.59068155e-01 -3.28233153e-01 1.16854772e-01
-7.55500555e-01 3.32977951e-01 7.70679653e-01 -8.90850052e-02
-7.55793571e-01 -3.84950489e-01 -6.92620158e-01 3.86508673e-01
5.01574278e-01 -4.85650390e-01 1.04677355e+00 -1.34482145e+00
-1.29733884e+00 5.87653995e-01 -1.54324457e-01 -6.73184991e-01
9.96667445e-02 -4.70647603e-01 -1.51183411e-01 8.11441183e-01
1.41571984e-01 7.06967652e-01 1.29136610e+00 -7.02013671e-01
-1.01931083e+00 -2.16256708e-01 -5.79930171e-02 4.32432741e-01
-5.77405632e-01 2.96776533e-01 -7.71734655e-01 -9.53648329e-01
1.03906795e-01 -7.78846145e-01 2.02323034e-01 -4.71408814e-01
-3.32977086e-01 -4.66831475e-01 1.06360471e+00 -1.18048704e+00
1.55886257e+00 -2.17330837e+00 3.44569504e-01 6.98254257e-02
1.27477437e-01 3.77167255e-01 2.75173783e-01 4.44369853e-01
6.15587384e-02 -2.85583675e-01 1.82009175e-01 -5.14300652e-02
-4.56147879e-01 -4.50849444e-01 -1.03251681e-01 2.90040165e-01
-8.95128548e-02 3.08040977e-01 -5.86699963e-01 -9.01305854e-01
5.07448852e-01 3.27045500e-01 -4.74675536e-01 3.74251902e-01
-5.75579070e-02 1.47804365e-01 -3.33998919e-01 4.52076733e-01
2.13784173e-01 3.00170574e-02 -6.24211244e-02 -6.84038162e-01
-2.33565524e-01 -4.03137412e-03 -1.19939566e+00 1.58174980e+00
-1.33460328e-01 1.16053534e+00 -7.90787041e-02 -1.29239154e+00
8.07210386e-01 3.12920928e-01 9.97057736e-01 -3.52416039e-01
4.85209763e-01 -1.41838208e-01 -1.52966186e-01 -8.91782224e-01
8.88137937e-01 2.84652084e-01 1.12849057e-01 2.09110841e-01
2.75530875e-01 8.56566206e-02 7.17358291e-01 3.73024821e-01
7.84190714e-01 -1.21168859e-01 4.34375584e-01 -6.95253722e-03
7.32405901e-01 6.92414194e-02 5.26073337e-01 4.78215665e-01
-2.87227128e-02 6.33466005e-01 3.17121059e-01 -5.84322512e-01
-1.10965204e+00 -3.60998720e-01 3.06581318e-01 9.45461810e-01
1.16008386e-01 -7.90539682e-01 -1.07146859e+00 -3.31843674e-01
-3.73649597e-01 4.12521333e-01 -1.83523789e-01 -1.90227687e-01
-6.07624233e-01 -3.47517639e-01 1.17919274e-01 3.18020970e-01
6.13568008e-01 -1.28206038e+00 -8.02168727e-01 2.51143605e-01
-5.32547176e-01 -1.08527684e+00 -3.59848797e-01 -2.13780120e-01
-7.77396560e-01 -1.05552459e+00 -9.36563373e-01 -1.07325554e+00
4.40424412e-01 5.83060205e-01 5.42469680e-01 7.41369426e-02
-2.65613377e-01 3.21359009e-01 -6.75208569e-01 -2.35928893e-01
-3.60571414e-01 9.49355364e-02 2.97248736e-02 3.00676793e-01
2.97495484e-01 -2.38944829e-01 -6.75593555e-01 -5.65983243e-02
-9.13494289e-01 1.68972716e-01 4.17475194e-01 5.82554042e-01
7.34315991e-01 5.71628273e-01 4.56078142e-01 -4.89959836e-01
7.09701359e-01 -3.05436760e-01 -4.43453401e-01 6.71103364e-04
2.01786477e-02 -3.84144694e-01 7.89277613e-01 -2.53056020e-01
-8.52116287e-01 1.83799490e-01 -7.58561641e-02 -6.61652207e-01
-1.41829923e-01 4.57808971e-01 -1.31193981e-01 2.58110166e-01
1.63019463e-01 4.28236187e-01 -1.47105351e-01 -4.85102266e-01
-6.02850551e-03 9.88290668e-01 4.56059784e-01 7.64223486e-02
1.72226369e-01 2.97356993e-01 -1.14893667e-01 -1.48528862e+00
-5.20049870e-01 -6.42536640e-01 -6.00854039e-01 -7.94139326e-01
1.08530557e+00 -9.77812052e-01 -5.66140413e-01 5.28536022e-01
-1.14112973e+00 2.75877714e-01 -9.67902094e-02 7.46201098e-01
-5.49002588e-01 6.66368484e-01 -6.59633815e-01 -6.55086339e-01
-7.21315384e-01 -1.22664022e+00 1.03617418e+00 6.56078100e-01
-4.40007120e-01 -4.22317654e-01 -2.78287202e-01 5.82819343e-01
-7.35859526e-03 1.77942395e-01 6.61040783e-01 -7.38435924e-01
-5.14269650e-01 -4.05844808e-01 2.06500456e-01 5.38557410e-01
7.15997517e-02 3.07785779e-01 -6.69909298e-01 -1.89804509e-01
9.42173153e-02 -1.59668118e-01 1.13703656e+00 8.91153395e-01
1.54099250e+00 -5.21899402e-01 -2.14533642e-01 3.85318041e-01
1.18152058e+00 5.08130729e-01 6.00913346e-01 4.74451274e-01
8.36768866e-01 4.74124700e-01 8.05901289e-01 6.63068593e-01
-4.47504688e-03 6.61166072e-01 1.60196051e-01 1.91178918e-01
-9.41636264e-02 8.15692320e-02 6.14346862e-01 1.15863883e+00
-2.52919018e-01 -2.47088507e-01 -5.62520802e-01 3.98713440e-01
-1.87754774e+00 -1.48734248e+00 -6.58185482e-02 2.06363225e+00
5.13205707e-01 3.73678356e-02 3.63350868e-01 5.86019397e-01
1.10526550e+00 2.16254041e-01 -3.98082554e-01 -3.42202812e-01
1.13074981e-01 7.30804726e-02 1.06727630e-01 3.04866931e-03
-1.51285362e+00 8.28239620e-01 4.92743397e+00 1.18394136e+00
-1.12916100e+00 -4.57895733e-02 6.66438580e-01 -3.89048427e-01
4.52578753e-01 -3.12400997e-01 -6.11681163e-01 5.83366871e-01
7.34946489e-01 -6.54809177e-02 3.18888456e-01 9.26142693e-01
5.78033864e-01 -5.48015773e-01 -8.31858218e-01 1.48027134e+00
5.12304246e-01 -1.59493339e+00 2.79438674e-01 -2.17349961e-01
7.70449221e-01 -3.73332381e-01 -2.33325288e-01 -1.62262231e-01
-5.39556861e-01 -3.97673607e-01 8.08958411e-01 5.24597287e-01
4.37750459e-01 -1.18136799e+00 8.86575878e-01 2.84880430e-01
-1.18568754e+00 -3.53880793e-01 -4.25824791e-01 -1.62472744e-02
9.46232155e-02 5.94038010e-01 -7.12650836e-01 3.93566966e-01
8.59736681e-01 9.91111398e-01 -3.05604607e-01 1.37328494e+00
1.95226222e-01 6.25951350e-01 -1.26387089e-01 -9.28609073e-02
7.29462579e-02 -2.37888992e-01 7.59497225e-01 1.32455719e+00
6.92109227e-01 1.05205208e-01 2.52695620e-01 1.99549779e-01
-1.78506002e-01 4.63867784e-01 -4.79607761e-01 -3.40817153e-01
4.03805763e-01 1.39739704e+00 -1.43585861e+00 -6.73961937e-01
-2.24304199e-01 1.05639935e+00 -2.18359515e-01 7.01859817e-02
-7.41999984e-01 -6.93026483e-01 3.65605682e-01 2.80348629e-01
4.90994900e-01 -7.25161508e-02 3.66156191e-01 -1.04984760e+00
-9.28319693e-02 -9.94566679e-01 4.69574839e-01 -9.57433224e-01
-5.83307803e-01 6.77937567e-01 1.95256937e-02 -1.54277933e+00
-3.93393308e-01 -2.39518508e-01 -5.63255191e-01 2.07742244e-01
-7.50462472e-01 -7.50936568e-01 -5.04758894e-01 5.79696476e-01
1.28318906e+00 -5.32100916e-01 3.68753821e-01 4.51049745e-01
-7.78736234e-01 1.75937209e-02 3.05427969e-01 2.37626061e-01
5.45375884e-01 -9.69290555e-01 -1.24753237e-01 1.05729210e+00
3.90439540e-01 3.99509728e-01 5.94871283e-01 -5.89224279e-01
-1.50512290e+00 -9.47342098e-01 5.63544631e-01 2.68365413e-01
2.41824880e-01 5.43329082e-02 -6.44599795e-01 2.54774898e-01
5.29109359e-01 -5.74356854e-01 8.25942159e-01 -6.28928661e-01
5.83810329e-01 -4.57989037e-01 -6.74703538e-01 3.12166303e-01
3.51513833e-01 -2.86372334e-01 -5.37497878e-01 2.59784102e-01
4.22469378e-01 -2.77211785e-01 -6.87488258e-01 4.04425897e-02
5.95102489e-01 -1.06887209e+00 8.02434742e-01 -2.19105512e-01
7.13198781e-01 -3.13278317e-01 -9.78370979e-02 -1.20633948e+00
-3.64698917e-01 -6.02987051e-01 -3.37920368e-01 1.46857667e+00
-2.96595037e-01 3.09171885e-01 6.20939195e-01 5.60173541e-02
-1.57133684e-01 -5.56598425e-01 -6.50654733e-01 -7.76851177e-02
-9.56086755e-01 -2.12973759e-01 2.05059111e-01 7.23392308e-01
-7.87469745e-02 4.10338074e-01 -5.16046524e-01 -1.89275473e-01
4.68161702e-01 1.92397654e-01 8.32126439e-01 -1.24375677e+00
5.11832461e-02 -5.16464233e-01 -7.27295160e-01 -9.04143214e-01
1.87056839e-01 -7.49301612e-01 -1.00819848e-01 -1.69909227e+00
4.38859493e-01 2.67008454e-01 -2.91992784e-01 3.38030085e-02
-5.60031347e-02 3.67545813e-01 5.19734919e-01 5.86015224e-01
-7.79320955e-01 3.23349684e-01 7.47897983e-01 -2.84119815e-01
-4.31427181e-01 1.90433353e-01 -4.30759490e-01 9.22704458e-01
7.94827163e-01 -3.18180621e-01 -3.28821629e-01 -5.78767918e-02
9.95299816e-02 1.71106666e-01 8.79898816e-02 -1.48210955e+00
3.29157919e-01 8.46382007e-02 7.30223179e-01 -1.03442228e+00
3.18269759e-01 -8.48277926e-01 1.18657470e-01 4.31065321e-01
-4.31863129e-01 8.77375081e-02 6.57379106e-02 3.79730046e-01
-5.45812905e-01 -4.98232841e-01 6.31031692e-01 -2.83102900e-01
-1.15608859e+00 2.93002278e-02 -7.58250952e-01 -4.47345406e-01
1.09430909e+00 -3.48066181e-01 9.91654396e-02 -6.73803926e-01
-6.29049122e-01 -9.68491063e-02 2.04053998e-01 3.62757117e-01
9.53145623e-01 -1.28385615e+00 -5.96611142e-01 7.08532706e-02
-1.81526884e-01 -7.25472197e-02 4.03433800e-01 7.11426973e-01
-1.05209744e+00 2.45323524e-01 -4.59520847e-01 -7.20016718e-01
-1.77696943e+00 4.36554372e-01 -1.23898961e-01 2.09269434e-01
-7.45418012e-01 6.53480887e-01 -3.66924368e-02 8.17718446e-01
4.37509626e-01 -3.66360843e-01 -9.58485663e-01 6.91357613e-01
9.96402681e-01 7.75458455e-01 7.84438923e-02 -9.93869483e-01
-1.96199834e-01 5.52507281e-01 -1.33159965e-01 1.66906878e-01
1.57749450e+00 -1.81575730e-01 -2.50127345e-01 5.07227838e-01
1.39254260e+00 -1.99394256e-01 -1.00508773e+00 3.02817468e-02
-1.91884056e-01 -4.12279159e-01 4.14189994e-01 -5.34112342e-02
-1.20204294e+00 7.88533151e-01 6.32176161e-01 4.19347197e-01
1.48910642e+00 -6.91253785e-03 1.10189831e+00 4.28162426e-01
-4.80746180e-02 -1.50490677e+00 2.25839555e-01 4.15004417e-02
7.82206655e-01 -9.57793951e-01 3.76175314e-01 -4.81170602e-03
-7.52271593e-01 1.44056749e+00 3.21538836e-01 -2.68442869e-01
3.43220323e-01 -1.59100182e-02 -2.37260625e-01 -8.67792889e-02
-4.21495825e-01 -4.57332917e-02 4.85446781e-01 1.17773704e-01
3.89780492e-01 -1.88908979e-01 -5.38374543e-01 5.85971177e-01
-9.84813720e-02 1.52129726e-02 6.55215383e-01 7.79267251e-01
-9.40129220e-01 -5.44098258e-01 -5.59975743e-01 7.85944641e-01
-8.61467421e-01 1.74465135e-01 -4.98873770e-01 3.25567633e-01
1.33693427e-01 9.10642922e-01 2.50618428e-01 -5.48457563e-01
5.40802553e-02 1.25864938e-01 3.80622029e-01 -3.84340733e-01
-5.71455657e-01 6.58450842e-01 -2.35540330e-01 -4.51505154e-01
-8.79117489e-01 -6.58202648e-01 -1.17437470e+00 -1.21475302e-01
-3.16868097e-01 3.04373533e-01 6.66193008e-01 7.80139446e-01
3.67627084e-01 6.83878481e-01 7.59201884e-01 -1.37713313e+00
3.23769934e-02 -9.37782943e-01 -4.79341835e-01 5.34632981e-01
2.51598924e-01 -4.02452141e-01 2.63525024e-02 6.71616852e-01] | [10.227903366088867, 0.49886563420295715] |
b2d47e49-9e16-4714-a7a9-5fbe71a5e829 | the-hci-aspects-of-public-deployment-of | 2306.04765 | null | https://arxiv.org/abs/2306.04765v1 | https://arxiv.org/pdf/2306.04765v1.pdf | The HCI Aspects of Public Deployment of Research Chatbots: A User Study, Design Recommendations, and Open Challenges | Publicly deploying research chatbots is a nuanced topic involving necessary risk-benefit analyses. While there have recently been frequent discussions on whether it is responsible to deploy such models, there has been far less focus on the interaction paradigms and design approaches that the resulting interfaces should adopt, in order to achieve their goals more effectively. We aim to pose, ground, and attempt to answer HCI questions involved in this scope, by reporting on a mixed-methods user study conducted on a recent research chatbot. We find that abstract anthropomorphic representation for the agent has a significant effect on user's perception, that offering AI explainability may have an impact on feedback rates, and that two (diegetic and extradiegetic) levels of the chat experience should be intentionally designed. We offer design recommendations and areas of further focus for the research community. | ['Jason Weston', 'Y-Lan Boureau', 'Melanie Kambadur', 'Moya Chen', 'Mojtaba Komeili', 'Kurt Shuster', 'Benjamin Babcock', 'Giuliano Morse', 'Joshua Lane', 'William Ngan', 'Morteza Behrooz'] | 2023-06-07 | null | null | null | null | ['chatbot', 'chatbot'] | ['methodology', 'natural-language-processing'] | [-1.07702911e-01 7.45200753e-01 7.03137666e-02 -2.70199269e-01
-1.88538596e-01 -5.27066588e-01 5.03201067e-01 1.27661362e-01
-3.05787563e-01 4.86550927e-01 7.59993792e-01 -7.18704402e-01
-2.72032231e-01 -2.34502599e-01 -1.84025958e-01 -9.53884199e-02
3.38754267e-01 1.20818086e-01 -1.43937826e-01 -2.57060885e-01
7.78636336e-01 3.19912612e-01 -1.52387607e+00 4.59454924e-01
6.03308320e-01 2.88054526e-01 3.56801808e-01 5.53363800e-01
-1.37064740e-01 1.04674101e+00 -1.05657494e+00 -5.29494166e-01
4.85686138e-02 -6.24672949e-01 -1.17487645e+00 -4.87864837e-02
1.23964893e-02 -7.14677572e-01 1.45581380e-01 5.32420397e-01
6.05100930e-01 1.30083680e-01 6.31899774e-01 -1.48166931e+00
-8.53123486e-01 7.09285915e-01 1.22864936e-02 3.38845104e-02
8.81233811e-01 4.89131957e-01 9.52876449e-01 -2.66862035e-01
7.36436009e-01 1.32770967e+00 6.94109440e-01 5.66562891e-01
-1.13474214e+00 -6.29368305e-01 6.64715096e-02 -2.56120801e-01
-8.87936771e-01 -4.41044420e-01 6.12637222e-01 -6.65729702e-01
1.10006618e+00 6.15904093e-01 8.49519432e-01 1.39811277e+00
2.25447193e-01 2.49997675e-01 1.24778068e+00 -7.59647310e-01
2.44028240e-01 1.01620257e+00 2.82414734e-01 2.21934423e-01
4.94296402e-01 -3.23495299e-01 -2.30198398e-01 -5.43021798e-01
9.45803285e-01 -2.35762745e-01 -9.31144878e-02 -1.53519094e-01
-5.56919336e-01 7.25137651e-01 1.50305286e-01 7.71120310e-01
-5.07678568e-01 2.65758038e-01 4.44324762e-01 4.53023940e-01
5.10102749e-01 9.83585179e-01 -2.52353370e-01 -1.01164055e+00
1.74918637e-01 4.90637213e-01 1.23676980e+00 9.82721031e-01
2.50971586e-01 -5.79355180e-01 -1.31000623e-01 7.18732417e-01
6.17430508e-01 -3.50070447e-01 -1.12241454e-01 -1.23740757e+00
1.89269289e-01 8.94654036e-01 5.64118028e-01 -9.94711339e-01
-4.81805831e-01 3.83243740e-01 2.88192153e-01 6.21758223e-01
5.99923670e-01 -5.48681855e-01 -1.85710430e-01 1.29718697e+00
8.05854052e-02 -7.36446738e-01 -3.13024044e-01 1.05791736e+00
7.86542118e-01 2.17673764e-01 5.42892396e-01 -2.82084830e-02
1.43032265e+00 -2.92930812e-01 -9.99326289e-01 -2.39044771e-01
8.50213349e-01 -9.16520655e-01 1.65710700e+00 1.96066350e-01
-1.09772038e+00 -8.93204510e-02 -1.01692355e+00 -5.98059520e-02
-6.11681230e-02 -1.55437455e-01 7.79782295e-01 1.57049918e+00
-7.93326378e-01 5.04341662e-01 -5.90304792e-01 -1.14023650e+00
-8.60695168e-02 -1.17115444e-02 -2.52528694e-02 2.69571424e-01
-7.09359109e-01 1.52250028e+00 -1.90548703e-01 9.94991735e-02
-2.46419683e-02 -6.50637686e-01 -2.88638383e-01 -2.42891908e-01
1.19449131e-01 -5.67810297e-01 1.48472106e+00 -8.80042851e-01
-1.61593699e+00 5.70640624e-01 3.97268862e-01 1.23764083e-01
6.69409811e-01 -1.07429840e-01 1.38289735e-01 -7.05837011e-02
1.40908301e-01 4.68620390e-01 2.47411430e-02 -1.60358322e+00
-4.76688921e-01 -3.15773666e-01 7.38586307e-01 3.99896830e-01
-4.38751221e-01 7.26302147e-01 3.67002726e-01 -2.95274764e-01
-4.00801331e-01 -1.00338507e+00 -1.69703513e-01 1.91216886e-01
1.43516511e-01 -3.75756383e-01 7.13294923e-01 -7.53572226e-01
1.10667539e+00 -1.94316185e+00 -4.85159397e-01 -1.74296468e-01
1.85223639e-01 4.47488427e-02 3.00659359e-01 1.18804610e+00
5.24260879e-01 7.27186441e-01 5.38880885e-01 -1.37841642e-01
3.60712886e-01 -1.31087795e-01 2.44038135e-01 4.18014437e-01
-3.26422863e-02 2.79574186e-01 -7.05420852e-01 -4.29967254e-01
2.32291549e-01 5.76575279e-01 -5.69839776e-01 3.90858412e-01
-4.85511832e-02 2.65706211e-01 -6.02689505e-01 3.55321079e-01
4.07161802e-01 8.22146982e-02 1.99234948e-01 1.85738459e-01
-6.41895652e-01 8.66833329e-01 -9.15564597e-01 1.13914204e+00
-6.43725634e-01 7.20652282e-01 4.47454154e-01 -2.00281903e-01
8.39333355e-01 6.52159572e-01 4.45043027e-01 -3.86458516e-01
5.21573007e-01 4.82780337e-02 5.57177544e-01 -1.06839907e+00
5.11839330e-01 -2.09774062e-01 1.60753444e-01 8.62718284e-01
-4.66380894e-01 -3.73696610e-02 -2.78501123e-01 -4.76642661e-02
1.23132837e+00 3.19179803e-01 8.37387294e-02 -2.98925072e-01
-4.36911643e-01 2.94453591e-01 4.36003208e-02 7.59398341e-01
-4.73396122e-01 4.32904571e-01 6.45187795e-01 -1.70334876e-01
-8.40391994e-01 -3.72593999e-01 3.48626450e-02 1.05375516e+00
2.54392654e-01 -6.12060666e-01 -1.02068329e+00 -5.18051505e-01
-3.36449325e-01 1.30905831e+00 -6.10294402e-01 -8.33362341e-02
2.22749487e-02 -3.19496810e-01 6.23678029e-01 2.76623696e-01
3.84059101e-01 -1.36223555e+00 -1.59127927e+00 3.81298512e-01
-2.37228215e-01 -6.06292963e-01 -2.18592927e-01 7.30226040e-02
-7.96154916e-01 -9.03605700e-01 -3.57255340e-01 -5.38210869e-01
2.75209874e-01 5.81070065e-01 7.26984680e-01 5.29653609e-01
-2.55075842e-01 7.97012091e-01 -9.61651742e-01 -7.96556532e-01
-6.28832161e-01 -1.93944797e-01 -4.98539507e-01 -5.72644055e-01
4.67322677e-01 -7.48763621e-01 -6.70403898e-01 5.39018512e-01
-4.55769658e-01 1.98567033e-01 4.80502188e-01 2.99281448e-01
-7.73030818e-01 -3.56071860e-01 7.28301525e-01 -7.81104147e-01
1.43629432e+00 -5.12503684e-01 2.67568618e-01 -9.03936569e-03
-5.97494364e-01 -4.37247694e-01 1.16502382e-01 -5.19453943e-01
-1.50404394e+00 -3.45912188e-01 -1.46932751e-01 3.26145589e-01
-4.90907103e-01 4.68151003e-01 -2.43232578e-01 -2.44241402e-01
1.06114423e+00 -8.74019802e-01 4.37107533e-01 -3.96817327e-01
1.16934009e-01 1.16475070e+00 -4.38806146e-01 -8.46832275e-01
1.97429165e-01 1.40698880e-01 -8.72046888e-01 -1.02150583e+00
2.72069216e-01 -2.47281358e-01 -3.15647036e-01 -6.31212294e-01
9.55958903e-01 -3.67060810e-01 -1.10400546e+00 -2.08924890e-01
-1.23945391e+00 -7.86612868e-01 2.24702254e-01 4.61344868e-01
-5.17592967e-01 -2.54598558e-02 -5.77455282e-01 -1.60361600e+00
-9.95524228e-02 -1.03761137e+00 4.65601534e-01 2.48141751e-01
-1.25168896e+00 -8.26185167e-01 -4.40316088e-02 9.40716922e-01
5.95133305e-01 1.11035734e-01 9.95164216e-01 -5.90071023e-01
-3.41921657e-01 -2.49983847e-01 -9.47094411e-02 -1.89677253e-01
3.71045113e-01 8.39706883e-02 -1.06171441e+00 1.33892417e-01
1.32439926e-01 -3.60307246e-01 -3.25613499e-01 2.60564089e-02
2.94390023e-01 -7.82024562e-01 -3.46312404e-01 -4.88204330e-01
9.94232357e-01 7.85482347e-01 8.36647928e-01 5.88588238e-01
3.99623394e-01 1.02743924e+00 9.61014032e-01 5.08617342e-01
2.87894934e-01 7.83733785e-01 2.87193328e-01 1.81227535e-01
1.13276772e-01 -2.17659935e-01 4.46603388e-01 1.03123762e-01
-3.58395606e-01 -1.44755915e-01 -7.39359081e-01 4.33157951e-01
-1.72090912e+00 -9.15045381e-01 -2.37874761e-01 1.85084796e+00
3.46246928e-01 1.66015804e-01 6.94012761e-01 -6.28987551e-02
5.48296392e-01 -1.07006051e-01 -5.44249676e-02 -9.64994013e-01
7.03569293e-01 -2.39750579e-01 2.30549037e-01 3.04135203e-01
-3.05769980e-01 3.57920349e-01 6.66227818e+00 -1.09659106e-01
-7.33259797e-01 1.87302172e-01 6.74312651e-01 2.94150203e-01
-4.86719429e-01 3.29035223e-01 -7.93657079e-02 1.81001127e-01
7.43912101e-01 -1.86948359e-01 3.53854567e-01 9.02384996e-01
8.24514270e-01 -4.76596028e-01 -1.04572880e+00 3.12850654e-01
-2.86563486e-01 -9.79847372e-01 -7.39563048e-01 2.55210876e-01
9.10527632e-03 -5.88014126e-01 -2.49675840e-01 3.17824334e-01
1.63610503e-01 -9.16115165e-01 8.21217000e-01 8.27303156e-03
3.06580484e-01 -4.55332547e-01 6.61855459e-01 4.08134282e-01
-5.91416359e-01 -9.52028483e-02 -5.38897626e-02 -1.09732592e+00
1.49591416e-01 -4.74307925e-01 -1.34299278e+00 -1.61113963e-02
9.48881567e-01 -2.41337180e-01 -3.23706418e-01 1.00098360e+00
2.82735854e-01 6.95138156e-01 4.32195142e-02 -8.89051735e-01
-6.25410005e-02 -2.92085558e-01 4.34856951e-01 1.11508703e+00
6.56019896e-02 4.64998543e-01 -2.19818786e-01 1.15978885e+00
5.55961430e-01 9.75567475e-02 -9.70364571e-01 -3.21241856e-01
8.95415366e-01 1.14756024e+00 -1.15326726e+00 3.41669261e-01
-4.97107476e-01 8.00344229e-01 1.51808746e-02 2.69322813e-01
-6.54317617e-01 -4.58981276e-01 1.01173592e+00 4.50184882e-01
-3.43792617e-01 3.77355777e-02 -9.99018490e-01 -3.84954363e-01
1.44888610e-01 -1.02122378e+00 -7.47720376e-02 -9.30786550e-01
-1.15506232e+00 2.22236082e-01 1.71896085e-01 -8.48493338e-01
-1.21208221e-01 -2.85002500e-01 -8.45878839e-01 9.86936867e-01
-9.26993936e-02 -1.14293587e+00 -3.46066087e-01 -3.57999325e-01
4.04684812e-01 4.81994808e-01 9.26321864e-01 4.95045073e-02
-2.29362354e-01 4.61514235e-01 -4.04381245e-01 -2.65215904e-01
8.52288723e-01 -9.47728157e-01 2.25749120e-01 1.55525103e-01
-4.70145375e-01 1.00751066e+00 1.32189870e+00 -7.70806968e-01
-1.60969186e+00 -2.61292934e-01 7.66596794e-01 -8.90060782e-01
4.50158417e-01 -4.45001304e-01 -6.49549901e-01 7.80021489e-01
6.26934767e-01 -1.07341981e+00 6.77271485e-01 4.54508483e-01
5.17342379e-03 4.14655864e-01 -1.73577809e+00 1.09962273e+00
1.20379305e+00 -4.22935575e-01 -5.20583570e-01 2.73995195e-03
5.70861876e-01 -1.53317302e-01 -9.03236210e-01 -2.47249022e-01
1.01462054e+00 -1.24763989e+00 4.00502056e-01 -3.53549689e-01
4.51599568e-01 2.22305655e-01 2.85017967e-01 -1.21983385e+00
-3.35516363e-01 -9.76332188e-01 7.50810564e-01 1.68513262e+00
6.01096570e-01 -6.81614101e-01 8.73259246e-01 1.79280865e+00
-4.30612534e-01 -6.05725408e-01 -5.06508648e-01 -3.45489293e-01
4.01807874e-02 -4.36214715e-01 2.99705535e-01 9.04514134e-01
9.63602662e-01 4.81188715e-01 -2.36665547e-01 -1.13510810e-01
-7.84690753e-02 -6.60433531e-01 1.06163621e+00 -1.15863609e+00
-8.49496797e-02 -6.23984218e-01 -2.19371036e-01 -5.68666160e-01
-4.59307641e-01 -1.93595156e-01 2.24533066e-01 -2.05539393e+00
2.01976284e-01 -6.15324080e-01 6.02419674e-01 2.69594252e-01
2.36802157e-02 -2.18223035e-01 4.90101308e-01 -4.53560241e-02
-2.30525210e-02 2.47116670e-01 9.73709106e-01 4.21198249e-01
-6.60263121e-01 5.61996996e-02 -1.40701199e+00 4.86064464e-01
9.14557338e-01 -3.07626426e-01 -7.75361001e-01 -1.70087725e-01
2.13586167e-01 2.47998655e-01 3.58513951e-01 -6.80691719e-01
-6.36748523e-02 -5.86373329e-01 -6.31466657e-02 1.78995460e-01
3.60989779e-01 -1.09457552e+00 5.35474837e-01 3.79760742e-01
-6.41863167e-01 1.51048571e-01 2.48827174e-01 1.52282387e-01
6.16055608e-01 -6.52455807e-01 1.52152002e-01 -4.84304391e-02
1.41263336e-01 -6.28012300e-01 -1.29729033e+00 -2.12112784e-01
9.48474824e-01 -7.67259717e-01 -2.58415967e-01 -9.69777942e-01
-2.62877584e-01 -8.24061334e-02 8.13244998e-01 6.35986149e-01
2.01141894e-01 -8.82574141e-01 -1.22315310e-01 -5.33987105e-01
1.17760129e-01 -5.82764626e-01 1.37879774e-01 7.39195466e-01
-6.93533599e-01 1.43847480e-01 -5.41795790e-01 6.24617934e-02
-1.38320005e+00 2.23194614e-01 1.73169732e-01 4.53614712e-01
-9.78222847e-01 5.17904401e-01 -1.34160340e-01 -2.48680130e-01
5.55393994e-01 -1.53990522e-01 -2.74762094e-01 8.12907442e-02
3.57127607e-01 8.18715453e-01 -1.74898226e-02 -1.30726233e-01
-1.95043251e-01 -3.88689697e-01 -2.59376951e-02 -6.41734302e-01
1.08563530e+00 -2.24910364e-01 1.10517725e-01 9.04439509e-01
7.23246336e-01 -3.22041772e-02 -1.19383597e+00 7.44002342e-01
2.85620570e-01 -8.06069493e-01 -4.80937153e-01 -1.06199026e+00
-4.86450307e-02 5.25734484e-01 5.12790740e-01 8.84136021e-01
5.85985601e-01 1.36619329e-01 2.41818413e-01 3.74880135e-01
5.33210695e-01 -1.22730160e+00 2.33481258e-01 4.29887474e-02
1.24884331e+00 -8.55709314e-01 -1.98610529e-01 -6.93504930e-01
-1.01633275e+00 8.47157478e-01 9.75820839e-01 1.12671405e-01
4.09638524e-01 3.50621521e-01 3.38580191e-01 -4.08540726e-01
-8.22426438e-01 1.62789315e-01 -4.33244497e-01 9.88614976e-01
1.32970893e+00 4.43628505e-02 -1.09609663e+00 8.46010685e-01
-3.67747903e-01 1.64287001e-01 8.21920693e-01 1.27239299e+00
-4.26588714e-01 -1.13233173e+00 -6.04966164e-01 5.77618420e-01
-4.99945164e-01 4.62281168e-01 -1.15522873e+00 1.25237954e+00
-2.95043647e-01 1.68319559e+00 -2.12099209e-01 -5.94140708e-01
5.91268957e-01 2.64444016e-02 4.41428751e-01 -7.97035933e-01
-1.30093944e+00 -2.12395757e-01 1.04429305e+00 -1.20033465e-01
-1.40005648e-01 -6.75999999e-01 -9.75987077e-01 -3.95083815e-01
-5.45928895e-01 1.77420154e-01 7.01173306e-01 8.95084739e-01
4.57591653e-01 3.32721174e-01 2.45229900e-01 -9.93474305e-01
-4.73515630e-01 -1.47580123e+00 -2.66161233e-01 3.35234255e-01
-1.78706929e-01 -7.60928690e-01 -2.38276288e-01 -3.53220254e-01] | [12.272294044494629, 7.794983863830566] |
144144b3-5c19-4a6b-8e4b-2b2c8526a090 | modeling-text-with-graph-convolutional | 1802.00985 | null | http://arxiv.org/abs/1802.00985v2 | http://arxiv.org/pdf/1802.00985v2.pdf | Modeling Text with Graph Convolutional Network for Cross-Modal Information Retrieval | Cross-modal information retrieval aims to find heterogeneous data of various
modalities from a given query of one modality. The main challenge is to map
different modalities into a common semantic space, in which distance between
concepts in different modalities can be well modeled. For cross-modal
information retrieval between images and texts, existing work mostly uses
off-the-shelf Convolutional Neural Network (CNN) for image feature extraction.
For texts, word-level features such as bag-of-words or word2vec are employed to
build deep learning models to represent texts. Besides word-level semantics,
the semantic relations between words are also informative but less explored. In
this paper, we model texts by graphs using similarity measure based on
word2vec. A dual-path neural network model is proposed for couple feature
learning in cross-modal information retrieval. One path utilizes Graph
Convolutional Network (GCN) for text modeling based on graph representations.
The other path uses a neural network with layers of nonlinearities for image
modeling based on off-the-shelf features. The model is trained by a pairwise
similarity loss function to maximize the similarity of relevant text-image
pairs and minimize the similarity of irrelevant pairs. Experimental results
show that the proposed model outperforms the state-of-the-art methods
significantly, with 17% improvement on accuracy for the best case. | ['Yuhang Lu', 'Jing Yu', 'Li Guo', 'Zengchang Qin', 'Yanbing Liu', 'Weifeng Zhang', 'Jianlong Tan'] | 2018-02-03 | null | null | null | null | ['cross-modal-information-retrieval'] | ['miscellaneous'] | [ 1.17475711e-01 -1.94383606e-01 -1.86752498e-01 -1.95870683e-01
-1.00664127e+00 -2.87288994e-01 8.31780732e-01 4.44263667e-01
-4.34371412e-01 9.82543081e-02 1.88879520e-01 6.01962656e-02
-3.05643529e-01 -8.41247201e-01 -4.38462436e-01 -5.55710375e-01
2.80728519e-01 3.51735383e-01 7.08388761e-02 -3.15580428e-01
1.28660336e-01 1.20304838e-01 -1.26589870e+00 5.61582506e-01
5.71756899e-01 1.44789302e+00 2.32881218e-01 4.07445580e-01
-6.27487659e-01 7.49900401e-01 1.23767331e-02 -5.38650870e-01
1.99590445e-01 -4.82302457e-01 -8.76722753e-01 8.81331321e-03
7.98379421e-01 -3.87489796e-02 -1.00999939e+00 1.17978203e+00
7.03912020e-01 2.58841634e-01 8.14237416e-01 -1.28601360e+00
-1.30482328e+00 2.91031867e-01 -6.27172589e-01 -1.17604591e-01
3.59853595e-01 -1.79838687e-01 1.16584301e+00 -9.83935118e-01
6.02154434e-01 1.49763703e+00 4.42163676e-01 2.79537171e-01
-9.09984946e-01 -4.98154879e-01 -7.42826611e-03 5.62693775e-01
-1.78157163e+00 6.37804857e-03 9.83189881e-01 -3.12653393e-01
9.61093724e-01 1.71070531e-01 5.43711960e-01 8.55144501e-01
3.24544340e-01 1.03151321e+00 6.75070286e-01 -4.39988703e-01
-4.19611454e-01 1.11759387e-01 -5.49565256e-02 1.02315927e+00
-8.24611112e-02 -2.75340050e-01 -5.25001526e-01 -1.05896845e-01
4.33702379e-01 5.26160896e-01 -3.91680717e-01 -6.69209361e-01
-1.22088122e+00 1.07428312e+00 9.69951868e-01 7.17172861e-01
-1.92024171e-01 1.43287003e-01 6.75569415e-01 3.71985823e-01
4.25187528e-01 2.72133440e-01 -6.17237948e-02 5.03315032e-01
-7.80426502e-01 -1.16359867e-01 6.24467552e-01 1.06232536e+00
8.14243317e-01 -2.61828780e-01 -4.60328847e-01 1.17604053e+00
7.28553593e-01 5.59846163e-01 8.20320010e-01 -3.46646696e-01
6.23525023e-01 1.02443147e+00 -5.49930930e-01 -1.55739129e+00
-2.94299752e-01 -1.55354694e-01 -1.06693649e+00 -5.66078126e-01
-5.92594733e-04 3.03361356e-01 -9.79928851e-01 1.40986705e+00
9.75365657e-03 1.35068834e-01 4.22401130e-02 1.02784908e+00
1.52519679e+00 6.86359406e-01 4.46261913e-02 1.10928006e-02
1.44791734e+00 -1.36597252e+00 -6.30423009e-01 -1.75190702e-01
7.70571589e-01 -9.44673240e-01 9.37006295e-01 -3.21203828e-01
-9.65224862e-01 -4.79683399e-01 -8.44143987e-01 -4.65814292e-01
-1.09738719e+00 -1.26128569e-01 3.45091581e-01 2.98664898e-01
-1.26410699e+00 2.40822434e-01 -3.11602384e-01 -7.79919386e-01
3.45338702e-01 8.20076168e-02 -6.26795053e-01 -6.17231011e-01
-1.57765210e+00 8.65138233e-01 6.24240041e-01 -8.25897828e-02
-4.88811821e-01 -4.95898306e-01 -1.27343225e+00 2.80681580e-01
2.30101764e-01 -1.13711107e+00 5.39474785e-01 -1.13920116e+00
-1.04051578e+00 1.20178390e+00 5.62014319e-02 -4.42963578e-02
3.35759640e-01 3.22168112e-01 -6.36852384e-01 6.20317280e-01
1.69621259e-01 8.49853516e-01 1.01306534e+00 -1.35453284e+00
-2.97114104e-01 -5.40680528e-01 -6.77452073e-04 6.11490846e-01
-7.84423411e-01 -7.25481939e-03 -1.19088900e+00 -5.17575264e-01
1.29718125e-01 -9.27841961e-01 2.11823896e-01 6.66947290e-02
-4.05052304e-01 -5.40784776e-01 9.14615870e-01 -6.70347929e-01
1.13550436e+00 -1.89892662e+00 1.88120633e-01 2.31217042e-01
3.15211117e-01 3.28349918e-02 -7.31860697e-01 7.08336532e-01
2.57402249e-02 3.91425081e-02 1.03584431e-01 -4.85422522e-01
1.17366888e-01 -5.47548123e-02 1.86977595e-01 4.18861032e-01
9.77918208e-02 1.43565750e+00 -8.81138623e-01 -9.61324573e-01
2.76366830e-01 7.75365055e-01 -6.92575201e-02 1.48407221e-01
3.19455154e-02 -1.60489921e-02 -5.01449227e-01 7.78057218e-01
6.08066499e-01 -6.36953413e-01 1.57445401e-01 -6.47075593e-01
5.63488245e-01 -2.13949502e-01 -7.97493398e-01 2.16189504e+00
-6.60706460e-01 5.80564559e-01 -2.76277065e-01 -1.32428813e+00
7.09782064e-01 2.86506504e-01 5.78759432e-01 -1.09717739e+00
5.18510520e-01 4.01847661e-02 -4.43853676e-01 -7.32521355e-01
6.85988903e-01 -4.27603684e-02 -9.27959532e-02 2.03417033e-01
3.79544765e-01 -1.95387855e-01 2.16465980e-01 4.06634480e-01
7.42351532e-01 -2.95931488e-01 6.02896139e-02 1.91888062e-03
7.58337677e-01 -1.59647599e-01 -2.00775057e-01 6.92315102e-01
-2.98778135e-02 7.79572666e-01 1.49979576e-01 -1.87433973e-01
-6.43982291e-01 -7.26358175e-01 6.26844317e-02 1.22908950e+00
5.47272861e-01 -2.45825395e-01 -3.92713666e-01 -7.43391097e-01
5.76500408e-02 2.41374195e-01 -8.98321390e-01 -4.09304798e-01
-1.18316509e-01 -3.08479190e-01 2.94936895e-01 3.45995754e-01
6.07959569e-01 -1.02309942e+00 2.04018980e-01 -2.21100792e-01
-2.83930361e-01 -1.16639173e+00 -9.91202474e-01 -2.52729267e-01
-6.78368270e-01 -1.19233918e+00 -1.06783855e+00 -1.28045952e+00
6.78357482e-01 8.55792403e-01 1.24736857e+00 3.10451120e-01
-4.38475251e-01 1.16551971e+00 -5.12302399e-01 8.89222175e-02
2.65930686e-02 9.98806432e-02 -4.66464162e-01 3.73959243e-01
6.06858313e-01 -4.97336760e-02 -8.38764548e-01 5.29871471e-02
-1.42182028e+00 -8.44205171e-02 4.74021167e-01 1.08700311e+00
6.80208921e-01 -1.87733844e-01 2.46811435e-01 -5.04290819e-01
7.56561995e-01 -7.58656383e-01 -7.97958300e-02 8.02300990e-01
-5.77058733e-01 -4.79677394e-02 5.07930815e-01 -5.42854607e-01
-7.56520331e-01 -3.05474371e-01 4.21544820e-01 -8.90823126e-01
5.03764488e-02 1.06766844e+00 -4.93455380e-02 -3.24820548e-01
2.74770409e-01 5.30680597e-01 -2.11103912e-02 -1.09910302e-01
7.83567131e-01 6.06908262e-01 6.27717525e-02 -2.24124029e-01
5.78376234e-01 4.42856252e-01 7.24045634e-02 -8.62493634e-01
-8.56423914e-01 -8.83972704e-01 -4.86363918e-01 -2.73480594e-01
1.03281260e+00 -1.11226928e+00 -5.95143020e-01 3.00269157e-01
-9.07244384e-01 7.37433434e-02 9.73381177e-02 5.33312917e-01
-3.56702238e-01 6.01140559e-01 -5.47063053e-01 -2.78138697e-01
-5.22197843e-01 -1.03653598e+00 1.34355617e+00 3.85082245e-01
3.19410980e-01 -1.56099379e+00 -8.44997093e-02 7.77556419e-01
3.71135443e-01 -1.82192162e-01 9.81666863e-01 -9.57774818e-01
-4.88010377e-01 -5.94950080e-01 -6.95504010e-01 2.06448302e-01
2.15814531e-01 -3.26064318e-01 -6.65996850e-01 -5.49675763e-01
-4.15928721e-01 -6.25008464e-01 1.15082276e+00 4.55897361e-01
1.19618213e+00 -1.17378332e-01 -4.39710051e-01 5.19216657e-01
1.69485474e+00 -1.06041007e-01 6.26395285e-01 2.98401803e-01
1.20042109e+00 6.49156868e-01 2.92188317e-01 -7.94764906e-02
7.58486152e-01 5.12634873e-01 4.72861141e-01 -1.71602070e-01
-2.71688282e-01 -3.29932392e-01 -4.69594784e-02 1.06221735e+00
4.65523660e-01 -5.14631093e-01 -9.67681170e-01 8.13043356e-01
-2.00956130e+00 -8.75789940e-01 4.54730839e-02 1.87773323e+00
5.54746270e-01 -4.22830075e-01 -3.09807211e-01 -4.49594527e-01
8.73776674e-01 3.14124495e-01 -4.02162343e-01 -5.52246124e-02
-2.97810763e-01 -1.42715335e-01 2.82794803e-01 4.12770540e-01
-1.22265208e+00 9.27144051e-01 4.90139103e+00 1.19926000e+00
-1.19057572e+00 6.56500831e-02 5.30813396e-01 -8.27035084e-02
-4.13562387e-01 -3.31284255e-01 -4.25560176e-01 3.03847343e-01
4.35089797e-01 -2.34496415e-01 3.63557518e-01 5.68168342e-01
-2.30329186e-01 2.55938441e-01 -9.19949114e-01 1.57060921e+00
8.98314297e-01 -1.36035252e+00 4.95341361e-01 -2.38302842e-01
9.46736991e-01 1.86167404e-01 2.27534264e-01 4.78664726e-01
3.45734693e-02 -8.94507527e-01 2.84372091e-01 8.85578215e-01
6.83129430e-01 -7.44445205e-01 9.42994893e-01 -5.45748957e-02
-1.55886471e+00 1.74802244e-01 -3.04395914e-01 6.00228548e-01
-9.88919437e-02 1.19143859e-01 -5.33831775e-01 9.57863629e-01
5.81165671e-01 1.08538175e+00 -9.50932562e-01 8.20819914e-01
3.45668137e-01 -9.59155560e-02 -7.20738396e-02 -1.09204374e-01
5.65451562e-01 -2.47183606e-01 2.14273408e-01 1.28910160e+00
4.62172240e-01 -2.22311050e-01 4.08225179e-01 6.08952463e-01
-5.88653088e-01 6.50460124e-01 -9.52194154e-01 -3.40453714e-01
1.59553722e-01 1.48346913e+00 -5.26994586e-01 -4.02768940e-01
-9.66464162e-01 1.21340740e+00 3.31816286e-01 5.79147339e-01
-4.49745625e-01 -5.10457575e-01 2.00864062e-01 -1.69845939e-01
2.59884655e-01 2.18780592e-01 1.23900905e-01 -1.30244458e+00
1.85297225e-02 -6.16979003e-01 7.79859841e-01 -1.11592507e+00
-2.05904484e+00 6.90901995e-01 -1.81430995e-01 -1.38075960e+00
-1.10735325e-02 -6.15442038e-01 -2.28356615e-01 1.06351268e+00
-1.83712375e+00 -1.80121791e+00 -4.92931902e-01 1.22012901e+00
4.89670992e-01 -5.09946823e-01 7.54953742e-01 3.18931192e-01
-1.59275070e-01 7.07420409e-01 4.03070271e-01 4.28833276e-01
8.74197125e-01 -9.34340417e-01 -2.49107301e-01 3.12259465e-01
1.95548147e-01 5.18163264e-01 3.22435386e-02 -5.71107328e-01
-1.66747618e+00 -1.26483774e+00 1.04899919e+00 -2.62817480e-02
1.03328919e+00 1.55407684e-02 -9.72245753e-01 3.86858195e-01
6.70027971e-01 3.88802737e-01 8.64576697e-01 -8.01892281e-02
-7.06112504e-01 -1.68700218e-02 -8.51996422e-01 7.15163946e-01
6.63431406e-01 -1.13041711e+00 -4.18582112e-01 5.11204600e-01
8.59534621e-01 -1.70445859e-01 -1.14257932e+00 9.74219516e-02
5.13782561e-01 -6.07499182e-01 1.09623075e+00 -8.11139286e-01
5.63154757e-01 -1.22678829e-02 -4.96498287e-01 -1.35509384e+00
-2.51529902e-01 -7.22429976e-02 -1.38244471e-02 1.06579494e+00
2.67324895e-01 -4.12206501e-01 4.22212124e-01 4.21609610e-01
8.46572369e-02 -6.74799860e-01 -7.57761240e-01 -6.49592817e-01
2.82129288e-01 -2.72522956e-01 3.17557275e-01 1.40695310e+00
1.62277311e-01 7.55859733e-01 -1.67880222e-01 -6.88019097e-02
4.46476221e-01 3.38230729e-01 3.15627396e-01 -1.02134871e+00
1.74993619e-01 -7.21418977e-01 -6.81771636e-01 -8.52308810e-01
6.13004088e-01 -1.37692189e+00 -1.39546260e-01 -1.79128492e+00
7.37118483e-01 7.62355775e-02 -5.97549319e-01 2.88516760e-01
-3.98635179e-01 4.45521533e-01 5.22310257e-01 6.56611472e-02
-1.08025229e+00 8.42306137e-01 1.49303329e+00 -9.29134965e-01
6.46898597e-02 -5.98461926e-01 -5.36683261e-01 4.22530860e-01
5.93904972e-01 -3.53248447e-01 -5.29108465e-01 -7.50913262e-01
4.91452903e-01 3.41769338e-01 4.58090752e-01 -5.64493239e-01
5.98110437e-01 -2.66067963e-03 2.37682551e-01 -6.47421122e-01
3.96170974e-01 -1.25570285e+00 -2.31587261e-01 1.21628009e-01
-6.35016620e-01 -5.86345382e-02 5.18942475e-02 8.91099513e-01
-7.77446508e-01 -1.98443606e-01 5.25187910e-01 -2.56421864e-01
-6.54570699e-01 8.23843002e-01 -2.54331082e-02 1.61996514e-01
7.43737578e-01 -4.61876020e-02 -3.62403810e-01 -6.72954738e-01
-7.67486930e-01 4.59592789e-01 2.82332003e-01 9.18584704e-01
9.35545206e-01 -1.96512616e+00 -6.48009717e-01 -1.20713860e-01
8.53205383e-01 -5.15640736e-01 5.68710446e-01 8.67300868e-01
-2.38435164e-01 8.17508221e-01 4.29684632e-02 -7.26177275e-01
-1.34333241e+00 7.83428431e-01 4.41637069e-01 -5.47471702e-01
-8.85272771e-02 8.18195820e-01 2.83261746e-01 -6.35721743e-01
1.65461093e-01 1.84469283e-01 -3.98483276e-01 4.53970343e-01
4.17199671e-01 -3.16019654e-02 6.15589730e-02 -1.08832169e+00
-2.94733763e-01 8.61416996e-01 -1.06916502e-01 1.73566133e-01
1.07923651e+00 -4.80068296e-01 -3.47688854e-01 4.21975344e-01
2.11605215e+00 -7.06647933e-01 -6.20489240e-01 -9.07507181e-01
-3.24769825e-01 -4.07862306e-01 4.99696195e-01 -6.02899909e-01
-1.49645579e+00 1.03423440e+00 9.61076438e-01 3.85101348e-01
1.15926480e+00 1.37287498e-01 9.26616311e-01 4.63248044e-01
5.15022874e-03 -1.15268397e+00 5.29297590e-01 6.40888095e-01
9.51640666e-01 -1.75632608e+00 -2.19200507e-01 -6.99673146e-02
-7.02924013e-01 1.24145055e+00 5.01673162e-01 8.90562534e-02
9.11486745e-01 -6.50128007e-01 -1.77393779e-02 -5.52577674e-01
-4.47398752e-01 -6.35096371e-01 1.03571928e+00 5.24232805e-01
4.93670195e-01 -1.27275199e-01 -1.31153882e-01 2.61172980e-01
4.33447450e-01 -2.78383791e-01 -2.18880162e-01 7.45169282e-01
-1.45492792e-01 -9.06275332e-01 -2.13283867e-01 5.48884571e-01
-2.73401499e-01 -5.10760367e-01 -4.02238786e-01 7.28361011e-01
-1.85431510e-01 1.05028772e+00 2.51358122e-01 -3.47826093e-01
2.31337264e-01 1.83041960e-01 5.27260125e-01 -4.29910213e-01
-5.17488301e-01 1.63284242e-01 -2.70527512e-01 -5.00774741e-01
-7.06154346e-01 -3.27560067e-01 -7.79553771e-01 -2.97582150e-01
-4.13502127e-01 -1.41860247e-01 6.47619009e-01 9.80052531e-01
4.14887518e-01 5.28430223e-01 6.53345168e-01 -6.05686069e-01
-2.14342922e-02 -1.03182828e+00 -7.53494620e-01 8.08718741e-01
1.20762013e-01 -4.82500464e-01 -1.65969357e-01 1.05596110e-01] | [10.815652847290039, 1.2604072093963623] |
9a55522b-d7f0-4120-a682-5039339bd707 | a-comparison-of-three-heart-rate-detection | 2101.09144 | null | https://arxiv.org/abs/2101.09144v4 | https://arxiv.org/pdf/2101.09144v4.pdf | A comparison of three heart rate detection algorithms over ballistocardiogram signals | Heart rate (HR) detection from ballistocardiogram (BCG) signals is challenging because the signal morphology can vary between and within-subjects. Also, it differs from one sensor to another. Hence, it is essential to evaluate HR detection algorithms across several datasets and under different experimental setups. In this paper, we studied the potential of three HR detection algorithms across four independent BCG datasets. The three algorithms are as follows: the multiresolution analysis of the maximal overlap discrete wavelet transform (MODWT-MRA), continuous wavelet transform (CWT), and template matching (TM). The four datasets were obtained using a microbend fiber optic sensor, a fiber Bragg grating sensor, electromechanical films, and load cells, respectively. The datasets were gathered from: a) 10 patients during a polysomnography study, b) 50 subjects in a sitting position, c) 10 subjects in a sleeping position, and d) 40 subjects in a sleeping position. Overall, CWT with derivative of Gaussian provided superior results compared with the MODWT-MRA, CWT (frequency B-spline), and CWT (Shannon). That said, a BCG template was constructed from DataSet1. Then, it was used for HR detection in the other datasets. The TM method achieved satisfactory results for DataSet2 and DataSet3, but it did not detect the HR of two subjects in DataSet4. The proposed methods were implemented on a Raspberry Pi. As a result, the average time required to analyze a 30-second BCG signal was less than one second for all methods. Yet, the MODWT-MRA had the highest performance with an average time of 0.04 seconds. | ['Bessam Abdulrazak', 'Ibrahim Sadek'] | 2021-01-22 | null | null | null | null | ['template-matching'] | ['computer-vision'] | [ 3.55320990e-01 -3.03526133e-01 3.14597607e-01 1.68068763e-02
-4.76211727e-01 -1.99621141e-01 -2.02611297e-01 4.10964526e-03
-4.38328624e-01 8.42601776e-01 -1.23528138e-01 -9.40643251e-02
-1.40635625e-01 -4.83834952e-01 1.63954303e-01 -9.29114759e-01
-3.56715590e-01 -2.02293888e-01 2.06834167e-01 8.80321786e-02
1.25926465e-01 3.31079036e-01 -1.13736010e+00 3.44562791e-02
8.87751639e-01 1.18178487e+00 2.28443906e-01 7.85960376e-01
5.33376157e-01 1.85373291e-01 -8.70113015e-01 2.20821798e-01
1.87545419e-01 -9.92568195e-01 -2.08104312e-01 -2.52703696e-01
-1.54037178e-01 -2.04514146e-01 -4.00878675e-02 5.87262571e-01
1.10707593e+00 7.84646794e-02 2.24075258e-01 -7.98064291e-01
-1.15661733e-01 3.89072120e-01 -4.89941269e-01 5.68060458e-01
4.88343358e-01 6.20566234e-02 3.89880598e-01 -5.11300802e-01
2.25419462e-01 7.47037292e-01 1.18507290e+00 2.61888236e-01
-1.53739285e+00 -5.75581193e-01 -5.73267758e-01 1.02862336e-01
-1.59339142e+00 -3.06110650e-01 1.02162123e+00 -5.09233236e-01
5.98488927e-01 7.26366639e-01 1.14289236e+00 7.47393787e-01
5.26626706e-01 -2.04467699e-02 1.80042446e+00 -4.18065101e-01
1.70207337e-01 1.81813389e-01 2.91065633e-01 2.01307654e-01
2.55502313e-01 1.68742061e-01 -3.40488493e-01 -4.58563119e-01
7.89340198e-01 1.09285407e-01 -6.83336794e-01 4.65067387e-01
-1.21847057e+00 5.41013122e-01 -1.15944356e-01 7.80945897e-01
-3.96730125e-01 -2.24010572e-01 4.07483310e-01 3.75726998e-01
1.72387674e-01 4.29150343e-01 -4.43920493e-01 -2.95476705e-01
-9.71002817e-01 1.63045838e-01 8.19705963e-01 4.10280973e-01
1.49712950e-01 8.74228552e-02 -3.14025432e-01 7.06007481e-01
1.67088330e-01 5.21698773e-01 9.47602987e-01 -6.90069079e-01
6.72254562e-02 1.06718227e-01 2.84812480e-01 -1.23703682e+00
-8.59576404e-01 -4.44052398e-01 -8.65703583e-01 -2.64058143e-01
5.19239366e-01 -2.13880002e-01 -3.39338779e-01 1.33393633e+00
2.31761351e-01 3.09460163e-01 9.50261019e-03 1.18273675e+00
8.73264253e-01 4.30262357e-01 -1.62885338e-01 -9.29697692e-01
1.66916609e+00 -2.37528607e-01 -1.10628986e+00 7.79543296e-02
1.69976726e-01 -7.26947725e-01 9.38575685e-01 6.55179560e-01
-1.02428365e+00 -1.12393403e+00 -1.03106153e+00 4.18815345e-01
1.63569450e-01 1.23101413e-01 -7.02299997e-02 1.04833627e+00
-6.82060361e-01 1.06501126e+00 -9.02936161e-01 -3.08189631e-01
-1.48377597e-01 -1.00561911e-02 8.19186866e-02 3.65292400e-01
-1.43259037e+00 6.96775854e-01 1.27111033e-01 3.93574864e-01
-2.14895144e-01 -4.15095627e-01 -5.56741655e-01 -1.45194426e-01
-2.27623612e-01 -3.60529870e-01 6.83381438e-01 -4.25929576e-01
-1.67791140e+00 6.14278674e-01 -1.79075390e-01 -3.17697406e-01
5.46891868e-01 -2.42054850e-01 -9.52406228e-01 3.14224064e-01
-2.26181611e-01 -2.99646050e-01 9.37168896e-01 -6.52690291e-01
-6.76288903e-02 -5.04797041e-01 -4.66805607e-01 -8.04411918e-02
6.87350705e-02 6.13007434e-02 2.62335628e-01 -6.76463902e-01
7.78367758e-01 -9.31420982e-01 1.43517122e-01 -4.73430365e-01
-3.72774601e-01 3.27232838e-01 4.41561103e-01 -1.07869017e+00
1.64584172e+00 -2.54661560e+00 -7.94708952e-02 3.51986051e-01
9.87005904e-02 2.13992208e-01 5.70889652e-01 4.31789577e-01
-1.95936356e-02 2.11067513e-01 -1.94800049e-01 1.11183599e-01
-4.44822013e-01 -1.18105315e-01 5.99819981e-03 8.84284973e-01
-3.18327934e-01 4.00329143e-01 -5.47385812e-01 -6.52167141e-01
3.78373295e-01 4.57932264e-01 6.66547790e-02 3.81193191e-01
6.21338069e-01 9.10454154e-01 -1.67988285e-01 5.96840620e-01
6.67396486e-01 1.25167564e-01 2.41559610e-01 -7.16672182e-01
-3.76330078e-01 8.47012699e-02 -1.37415934e+00 1.32332623e+00
-1.22436613e-01 5.69928169e-01 -1.83320031e-01 -9.05055106e-01
1.61892343e+00 9.06360388e-01 8.10747445e-01 -7.52174020e-01
2.73033306e-02 3.61072242e-01 3.71923417e-01 -1.12113941e+00
-6.43684566e-02 -6.16853178e-01 1.50612876e-01 3.26631874e-01
-4.54255223e-01 -1.78943247e-01 -6.52897656e-02 -5.03070772e-01
8.13727200e-01 -1.19782627e-01 5.85062444e-01 -4.74425286e-01
6.00456715e-01 -3.99156868e-01 7.84264863e-01 5.76778769e-01
-5.82178831e-01 7.06472576e-01 2.06605598e-01 -6.21121943e-01
-4.93148834e-01 -9.39399362e-01 -6.22647583e-01 3.49719256e-01
2.05344662e-01 -1.96387932e-01 -4.60996330e-01 3.07489663e-01
3.14794928e-02 4.24033493e-01 -3.10613722e-01 -1.30328283e-01
-6.19908869e-01 -8.52495253e-01 8.28435957e-01 3.03906232e-01
7.12512255e-01 -1.19521976e+00 -1.17460811e+00 5.52829504e-01
-4.45547670e-01 -9.80686188e-01 -4.07516181e-01 -1.14141263e-01
-1.22511625e+00 -1.03731453e+00 -7.49260128e-01 -3.84101152e-01
5.89748658e-03 -6.28748611e-02 7.40869105e-01 -9.83633325e-02
-5.94692230e-01 1.83023289e-01 -3.84569734e-01 -4.86254841e-01
-1.17692158e-01 -4.57744777e-01 2.69639045e-01 1.29519045e-01
2.68497646e-01 -8.64911616e-01 -1.00796211e+00 5.82700312e-01
-2.90591121e-01 -2.63696074e-01 6.92804232e-02 5.73705733e-01
5.70731282e-01 6.31284639e-02 8.44105661e-01 -3.25071305e-01
1.04726553e+00 -2.60933787e-01 -3.43647838e-01 -1.94446340e-01
-7.76232362e-01 -6.26137853e-01 6.09098554e-01 -6.17088914e-01
-5.66261709e-01 -1.20157622e-01 -9.68823023e-03 -1.53313130e-01
-2.01004948e-02 4.30972427e-01 3.15501362e-01 1.74844980e-01
9.24802363e-01 2.39057586e-01 4.14882660e-01 -4.65004057e-01
-3.08609366e-01 7.97407627e-01 6.54973030e-01 -4.48799402e-01
3.96773487e-01 1.56487226e-01 -5.82439788e-02 -1.12543881e+00
-1.72535360e-01 -3.93253773e-01 -5.34605801e-01 -5.28285444e-01
1.10550869e+00 -8.54157925e-01 -8.71606052e-01 7.57366955e-01
-9.26868021e-01 -1.03958130e-01 -1.70092002e-01 1.07904732e+00
-4.78586912e-01 4.57669914e-01 -7.75695205e-01 -1.18343997e+00
-8.69006217e-01 -7.14696050e-01 4.57676321e-01 4.26382184e-01
-4.91274774e-01 -8.34956169e-01 5.09251989e-02 4.35055435e-01
6.29519105e-01 1.09698129e+00 3.68355364e-01 -2.32446939e-01
3.35314065e-01 -3.69807452e-01 3.39478046e-01 3.70563060e-01
4.51475888e-01 -1.82527035e-01 -1.19643164e+00 -5.78516483e-01
8.15823138e-01 1.00175561e-02 1.92531735e-01 6.95327401e-01
1.06581950e+00 -1.12358913e-01 -1.67846859e-01 4.68459368e-01
1.43072426e+00 7.43639767e-01 9.73884940e-01 5.42739928e-02
2.31383175e-01 4.90363061e-01 5.53728521e-01 5.06074905e-01
-9.51280147e-02 6.25991940e-01 -2.89121345e-02 -2.88712621e-01
3.15881252e-01 2.28325605e-01 4.43906337e-01 1.01952529e+00
-7.98431158e-01 2.25327492e-01 -5.38103521e-01 1.24331422e-01
-1.39848733e+00 -9.84858871e-01 -6.41356349e-01 2.52906775e+00
7.99173832e-01 1.53573945e-01 5.52806079e-01 7.57440448e-01
9.41575706e-01 -5.18882237e-02 -6.28925562e-01 -4.75764304e-01
1.33394878e-02 4.36831623e-01 1.83951259e-01 1.70835778e-01
-7.75003135e-01 -3.38938326e-01 6.21631765e+00 1.05285577e-01
-1.43391716e+00 1.32375166e-01 4.25869167e-01 6.98651969e-02
2.11919278e-01 -1.19745113e-01 -3.84316742e-01 9.16474640e-01
1.13741231e+00 -5.14469028e-01 3.34742486e-01 5.74042499e-01
6.95970118e-01 -2.88030416e-01 -8.01496387e-01 1.19979942e+00
-1.60498023e-01 -5.95317662e-01 -1.09832466e+00 -1.82761431e-01
1.27357259e-01 -3.19107264e-01 -5.05925357e-01 3.69914770e-02
-9.91709530e-01 -8.72894824e-01 3.18882823e-01 8.34797502e-01
9.60719287e-01 -2.38182560e-01 8.30900133e-01 4.12606627e-01
-1.34290123e+00 -6.65907608e-03 -2.57476896e-01 -2.57435739e-01
2.66777247e-01 1.02000356e+00 -4.21511710e-01 6.45138443e-01
6.13508523e-01 4.46453065e-01 -1.19283721e-01 1.02550149e+00
5.26396632e-02 8.55133414e-01 -4.22814965e-01 4.88259830e-02
-5.93663275e-01 -3.64307880e-01 6.39786899e-01 9.55688119e-01
5.62748909e-01 4.79271024e-01 2.18738601e-01 9.55516458e-01
6.91471219e-01 1.58902183e-01 -3.47968400e-01 2.67764688e-01
6.50381982e-01 1.09938562e+00 -5.72319090e-01 -1.93730354e-01
-4.30966228e-01 3.80761981e-01 -6.01911843e-01 3.29217434e-01
-1.00354815e+00 -7.71866024e-01 5.15172863e-03 4.11928058e-01
-2.56903470e-01 -7.18919709e-02 -4.61068392e-01 -7.63006687e-01
1.88912332e-01 -7.65464425e-01 4.65433419e-01 -6.97677374e-01
-1.10580671e+00 6.42633915e-01 1.30628064e-01 -1.54355633e+00
2.42644474e-02 -1.07896782e-01 -7.96393216e-01 1.39915752e+00
-9.19133842e-01 8.31748396e-02 -8.06932867e-01 6.60582542e-01
2.43215024e-01 2.88625240e-01 9.97565150e-01 6.77720249e-01
-5.33462584e-01 3.41799349e-01 -2.55106091e-01 -2.39787206e-01
5.28815508e-01 -1.04208505e+00 -1.15630597e-01 5.53517282e-01
-5.44795036e-01 7.25306392e-01 8.62317085e-01 -4.61015671e-01
-1.23766625e+00 -6.54766798e-01 7.89160788e-01 2.04707593e-01
1.69842735e-01 5.99312335e-02 -1.06430602e+00 1.79454923e-01
-1.61071807e-01 1.87187001e-01 7.03512311e-01 -3.31954032e-01
4.37944680e-01 -5.49024522e-01 -1.54172337e+00 -3.41146486e-04
4.51475799e-01 -3.48577291e-01 -7.75299609e-01 1.16286762e-01
3.72576326e-01 -6.14365041e-01 -1.80278409e+00 4.88302052e-01
1.12292671e+00 -1.03943908e+00 7.59560406e-01 1.99350379e-02
-6.47373721e-02 -5.33476770e-01 6.42248839e-02 -1.11189532e+00
-3.70104283e-01 -1.00489974e+00 -7.87895918e-02 1.00294161e+00
3.02249063e-02 -1.16399431e+00 9.55565721e-02 6.73147857e-01
-1.29352972e-01 -8.07973981e-01 -8.47415686e-01 -1.03893185e+00
-3.88262868e-01 -1.08600236e-01 2.32921332e-01 9.04595196e-01
4.08648789e-01 1.38922334e-01 -5.29336512e-01 -1.31551921e-03
5.77165663e-01 2.00358197e-01 4.10692245e-01 -1.27811766e+00
-4.31660950e-01 1.10817730e-01 -4.58268106e-01 -5.09144187e-01
-8.24251831e-01 -4.43581462e-01 -3.28930020e-02 -1.25346661e+00
-1.09090805e-01 -4.27795321e-01 -5.75298071e-01 7.90010914e-02
-4.50105548e-01 2.13914722e-01 2.35336442e-02 4.00430650e-01
3.88098568e-01 -1.09271391e-03 1.21000671e+00 2.36673534e-01
-8.10919464e-01 4.80781019e-01 -1.39309078e-01 4.67051566e-01
9.91266906e-01 -5.32609224e-01 -4.40189809e-01 3.76028717e-01
-2.08728746e-01 7.99345136e-01 3.51951152e-01 -1.29235315e+00
-1.22778982e-01 6.58612326e-02 4.51745242e-01 -6.10764086e-01
3.74339253e-01 -6.47851348e-01 1.00192142e+00 9.94117916e-01
6.51059896e-02 1.38041809e-01 8.18201527e-02 1.21407799e-01
-1.64992660e-01 -2.51938775e-02 1.04490948e+00 -2.62527913e-01
3.58151551e-03 -1.26334608e-01 -3.37885171e-01 1.17445372e-01
8.94488931e-01 -7.01457560e-01 -4.26061630e-01 -3.05131897e-02
-1.06474519e+00 -1.85009152e-01 -1.01009384e-02 -8.73833075e-02
7.72452593e-01 -1.33885586e+00 -5.91716707e-01 3.62307280e-01
-6.42740801e-02 -3.03444415e-01 4.83025223e-01 1.61438966e+00
-4.67636913e-01 1.50139570e-01 -4.18128163e-01 -8.80740166e-01
-1.27318144e+00 2.84571826e-01 6.57967746e-01 -8.03927779e-02
-9.39974964e-01 2.10280970e-01 -5.05946279e-01 3.96230221e-01
-7.61709083e-03 -5.00895023e-01 -4.29542273e-01 1.47904068e-01
4.64150548e-01 8.13081324e-01 6.25396818e-02 -2.01690316e-01
-3.62009734e-01 9.41924155e-01 5.52998364e-01 -9.67442915e-02
8.37214887e-01 -2.42130816e-01 -3.85147445e-02 9.47004557e-01
9.91596818e-01 1.30625233e-01 -6.24728501e-01 5.75796552e-02
-1.31818965e-01 -4.89962190e-01 -2.67803669e-01 -5.14278352e-01
-9.23029900e-01 6.85583174e-01 1.01289284e+00 9.05770481e-01
1.51349819e+00 -5.33475637e-01 8.17939341e-01 -1.97538823e-01
3.82069796e-01 -1.03403687e+00 -1.15836412e-01 -3.49933028e-01
8.75439584e-01 -6.32773638e-01 2.21038356e-01 -2.90017694e-01
-5.19408584e-01 1.42766404e+00 2.85182983e-01 -6.07514232e-02
7.68727958e-01 -6.30961433e-02 1.66297376e-01 -1.09402321e-01
-3.09020966e-01 1.50527328e-01 1.98092699e-01 4.61748242e-01
8.17809343e-01 2.17911214e-01 -1.09856403e+00 6.15202725e-01
-2.14001060e-01 4.52538252e-01 4.65969354e-01 8.95798266e-01
-4.85716671e-01 -5.10811210e-01 -5.40356994e-01 6.52255774e-01
-7.09386349e-01 1.53542414e-01 1.87864572e-01 7.17259169e-01
1.68666869e-01 1.47871852e+00 -1.60371676e-01 -5.60273826e-01
7.09952712e-01 3.40777844e-01 3.06492686e-01 -2.79473960e-01
-6.83647931e-01 3.90675217e-01 8.64637792e-02 -5.40290534e-01
-6.31519377e-01 -7.76387691e-01 -1.19655383e+00 -1.41929850e-01
-4.46488708e-01 2.00237319e-01 6.26700401e-01 6.35133147e-01
1.70758516e-01 5.50231218e-01 9.19435740e-01 -3.90746891e-01
-4.37670290e-01 -1.39623141e+00 -1.17160022e+00 3.62717211e-01
3.67953330e-01 -4.37558115e-01 -7.74496913e-01 1.87864035e-01] | [13.961009979248047, 3.0244035720825195] |
6e5b9ef8-fd26-4693-8c1a-64e5ebca671f | cvefixes-automated-collection-of | 2107.08760 | null | https://arxiv.org/abs/2107.08760v1 | https://arxiv.org/pdf/2107.08760v1.pdf | CVEfixes: Automated Collection of Vulnerabilities and Their Fixes from Open-Source Software | Data-driven research on the automated discovery and repair of security vulnerabilities in source code requires comprehensive datasets of real-life vulnerable code and their fixes. To assist in such research, we propose a method to automatically collect and curate a comprehensive vulnerability dataset from Common Vulnerabilities and Exposures (CVE) records in the public National Vulnerability Database (NVD). We implement our approach in a fully automated dataset collection tool and share an initial release of the resulting vulnerability dataset named CVEfixes. The CVEfixes collection tool automatically fetches all available CVE records from the NVD, gathers the vulnerable code and corresponding fixes from associated open-source repositories, and organizes the collected information in a relational database. Moreover, the dataset is enriched with meta-data such as programming language, and detailed code and security metrics at five levels of abstraction. The collection can easily be repeated to keep up-to-date with newly discovered or patched vulnerabilities. The initial release of CVEfixes spans all published CVEs up to 9 June 2021, covering 5365 CVE records for 1754 open-source projects that were addressed in a total of 5495 vulnerability fixing commits. CVEfixes supports various types of data-driven software security research, such as vulnerability prediction, vulnerability classification, vulnerability severity prediction, analysis of vulnerability-related code changes, and automated vulnerability repair. | ['Leon Moonen', 'Amara Naseer', 'Guru Prasad Bhandari'] | 2021-07-19 | null | null | null | null | ['severity-prediction'] | ['computer-vision'] | [-5.02380610e-01 -2.25359574e-01 -2.19961450e-01 -1.58593491e-01
-8.83374393e-01 -1.32181430e+00 -1.12404600e-01 1.05081546e+00
1.22149229e-01 7.06724375e-02 3.78720701e-01 -8.72586489e-01
-4.43284452e-01 -1.15588701e+00 -5.77113688e-01 2.88891252e-02
-2.10013688e-01 -3.29569578e-01 3.74975801e-01 -3.73993725e-01
1.06222081e+00 3.53541434e-01 -1.57351339e+00 3.33865196e-01
8.18860114e-01 5.77075362e-01 3.89236845e-02 5.60558558e-01
-3.68917763e-01 4.77253139e-01 -6.21707261e-01 -7.43153155e-01
2.89566815e-01 4.54478055e-01 -1.16328740e+00 -1.07871866e+00
1.91116840e-01 -1.27092794e-01 1.44936427e-01 1.46561623e+00
1.98482767e-01 -6.34079933e-01 1.77693740e-02 -1.44135988e+00
-7.93427408e-01 1.01700485e+00 -6.35401964e-01 5.14887691e-01
5.87115824e-01 1.17452592e-01 9.18421745e-01 -8.01142871e-01
9.87445831e-01 7.87815571e-01 1.05101907e+00 4.22064722e-01
-9.98247743e-01 -4.88047779e-01 -3.07885468e-01 -1.57540366e-01
-1.24541175e+00 -5.15595339e-02 6.50084376e-01 -1.23192728e+00
1.39860225e+00 4.02350336e-01 2.35892713e-01 9.27086592e-01
3.77725363e-01 -3.22324276e-01 5.59194565e-01 -5.58863580e-01
2.80575901e-01 3.26562315e-01 1.11899090e+00 4.26744998e-01
8.62910688e-01 3.97010408e-02 1.39165908e-01 -1.15701783e+00
-1.42173087e-02 2.29550332e-01 -1.13137484e-01 5.56808636e-02
-7.63053477e-01 6.67704582e-01 3.72085162e-02 5.15378237e-01
-2.54271656e-01 -1.68144286e-01 1.02253389e+00 7.25518525e-01
3.48487377e-01 6.85315371e-01 -8.45516324e-01 -3.36208850e-01
-5.91724515e-01 3.66295666e-01 7.55712748e-01 9.77153361e-01
1.16285360e+00 -3.01593632e-01 3.17574084e-01 5.79073727e-01
4.26187634e-01 3.27761859e-01 1.15110576e-01 -6.66689634e-01
1.16627920e+00 1.32126272e+00 -3.78159583e-02 -1.42844677e+00
-4.20803815e-01 1.19814999e-01 -6.13540486e-02 4.03293192e-01
-1.50814140e-02 -1.35830164e-01 -2.13790059e-01 1.47552204e+00
2.79914916e-01 -6.08077705e-01 3.82296108e-02 -2.72570141e-02
7.63522089e-01 2.13224038e-01 7.06343353e-02 -6.14471585e-02
1.12421596e+00 7.01480284e-02 -3.22910100e-01 2.66166836e-01
1.07277524e+00 -5.57427168e-01 1.08806741e+00 5.18381596e-01
-5.36871731e-01 -2.10963815e-01 -9.90824163e-01 2.51569182e-01
-7.83459365e-01 -4.67485517e-01 4.93331164e-01 1.25282252e+00
-1.04378736e+00 4.08770293e-01 -6.85127318e-01 -1.12546623e-01
4.91485864e-01 -1.58679307e-01 -5.54444969e-01 1.02260046e-01
-1.08860338e+00 6.78494215e-01 5.45807958e-01 -5.39226890e-01
-1.03298771e+00 -1.50604963e+00 -7.52522707e-01 2.04351082e-01
6.23981416e-01 8.87335278e-03 7.26416409e-01 -2.48395666e-01
-2.91171789e-01 5.94366968e-01 2.07572743e-01 6.18161261e-02
-2.95027673e-01 -1.06715329e-01 -7.40910053e-01 -1.19645789e-01
1.31534815e-01 -5.75739264e-01 2.68864453e-01 -1.04734397e+00
-3.97621542e-01 -3.99844289e-01 3.76432061e-01 -1.01108074e+00
-1.05614591e+00 8.94274116e-01 6.23245500e-02 -6.12584770e-01
-4.43723112e-01 -5.65117180e-01 -2.22068623e-01 -6.10338032e-01
-5.55249810e-01 -7.58799687e-02 5.18042266e-01 -1.04673171e+00
2.21669555e+00 -2.19367194e+00 -1.09414512e-03 4.30124015e-01
5.61616480e-01 3.63379925e-01 -4.04034615e-01 9.11152184e-01
-6.43676877e-01 9.21649456e-01 -5.01017034e-01 1.42046735e-01
-3.84930633e-02 -4.24829185e-01 -8.24109256e-01 1.51422024e-01
5.42243496e-02 6.37047350e-01 -9.01405573e-01 -2.31270209e-01
-2.34786928e-01 8.32628459e-02 -8.62136960e-01 8.87470916e-02
-1.52806178e-01 -1.31876826e-01 -7.37827241e-01 1.06838179e+00
8.91164839e-01 4.37763870e-01 -1.31513938e-01 2.96461970e-01
-7.92253196e-01 4.19170141e-01 -9.26024497e-01 1.50690460e+00
-4.57182527e-01 4.69996840e-01 -2.04050869e-01 -3.59377295e-01
1.23680055e+00 3.74605209e-01 3.89894247e-01 -3.26972276e-01
-1.65403739e-01 2.69787699e-01 -4.23289061e-01 -8.00437629e-01
5.90565145e-01 7.62301803e-01 -7.56148398e-01 9.40265656e-01
9.93000567e-02 2.40284622e-01 2.53578037e-01 4.79710132e-01
1.80716181e+00 -2.67705917e-01 4.05698478e-01 -6.19036019e-01
7.33307719e-01 1.64794520e-01 7.48728991e-01 2.80781835e-01
7.58685693e-02 7.39900693e-02 1.21689487e+00 -6.32133663e-01
-1.20229506e+00 -7.42023349e-01 -3.40557963e-01 7.51239061e-01
-5.16959667e-01 -1.17333853e+00 -1.21856022e+00 -9.77662325e-01
-1.10209577e-01 5.36384106e-01 -8.27232838e-01 -2.91743219e-01
-4.42080617e-01 -6.58041418e-01 8.53544414e-01 4.48780060e-01
1.24570549e-01 -1.13436449e+00 -6.94865346e-01 1.05841972e-01
-1.34973437e-01 -2.67436534e-01 -2.02525526e-01 -2.94674933e-01
-3.88211042e-01 -1.55120790e+00 2.33592719e-01 -1.19800568e-01
4.63493615e-01 -1.71586610e-02 1.18891931e+00 6.85597718e-01
-8.17191899e-01 2.40005761e-01 -6.86040103e-01 -3.77441823e-01
-8.67812872e-01 9.26450565e-02 -1.12946041e-01 -5.78428686e-01
4.78983343e-01 -5.88357270e-01 6.02601320e-02 1.23872921e-01
-1.07148898e+00 -1.09710002e+00 -1.74238726e-01 1.54736459e-01
3.63202780e-01 2.60223001e-01 7.16778934e-01 -8.97008657e-01
7.16356397e-01 -1.33948278e+00 -1.26560926e+00 6.29622102e-01
-7.36994088e-01 -2.60822505e-01 4.52047795e-01 2.66211070e-02
-1.09605026e+00 -2.26754293e-01 -4.09335136e-01 -4.90760840e-02
-7.61406571e-02 1.19415009e+00 -3.38170528e-01 -2.09381148e-01
9.78473425e-01 -3.90552670e-01 -4.66166943e-01 -8.61854255e-01
1.67709216e-01 8.36631060e-01 4.74317133e-01 -8.63560498e-01
1.18369472e+00 -3.22073817e-01 -4.14671630e-01 -3.08664948e-01
-2.29677051e-01 -1.48136064e-01 -7.36662865e-01 2.79707834e-02
5.25090873e-01 -5.07710397e-01 -5.81646621e-01 5.05879939e-01
-1.14493704e+00 -4.91015166e-02 -2.57943142e-02 -1.47735372e-01
8.68434981e-02 6.04369462e-01 -3.19342583e-01 -7.03347921e-01
-6.21186495e-01 -1.21415436e+00 3.30075145e-01 -1.58720464e-01
-3.77748162e-01 -8.23537707e-01 9.86434042e-01 -1.91902295e-01
6.25128329e-01 7.86451280e-01 1.62618256e+00 -5.58885813e-01
-3.93355876e-01 -4.08958316e-01 -6.85521141e-02 2.82853395e-01
7.02746725e-03 1.02331305e+00 -6.35737538e-01 -3.19847822e-01
-2.50221658e-02 -3.08767445e-02 4.70394373e-01 -2.37632945e-01
1.08735430e+00 -2.79927194e-01 -3.64290059e-01 5.38215041e-01
1.99633789e+00 2.37694591e-01 8.26226294e-01 8.06535125e-01
7.02932000e-01 1.06748033e+00 6.75444365e-01 5.64681351e-01
3.58423680e-01 4.49930012e-01 9.74006951e-01 5.92617512e-01
4.63951111e-01 1.47057533e-01 3.70836765e-01 6.29740477e-01
-1.29188418e-01 9.57268104e-02 -1.81053340e+00 9.72466767e-01
-1.25564516e+00 -1.14266515e+00 -7.92403340e-01 2.26934004e+00
8.79814982e-01 -6.40801117e-02 1.95214048e-01 3.59375000e-01
7.81599343e-01 -1.95509568e-01 -1.33444503e-01 -6.77892804e-01
3.59821737e-01 4.34525348e-02 2.70238906e-01 2.97098339e-01
-7.81661749e-01 5.44482589e-01 6.07042646e+00 3.21202010e-01
-6.44933283e-01 3.47124338e-01 -3.36513370e-02 1.45104423e-01
-9.00599241e-01 4.77113664e-01 -8.49027395e-01 4.76910949e-01
1.56332779e+00 -5.75274825e-01 2.52046287e-01 1.25871694e+00
-3.02448541e-01 6.36689141e-02 -8.19517553e-01 2.11989403e-01
-2.86119759e-01 -1.75366211e+00 -4.68862265e-01 2.40034059e-01
6.87183678e-01 9.62141678e-02 4.03313413e-02 1.95724115e-01
3.99591535e-01 -5.11074066e-01 8.57432723e-01 7.51798570e-01
9.61698294e-01 -1.11952090e+00 7.40793109e-01 1.41404405e-01
-1.31278300e+00 -8.01898301e-01 -4.11959708e-01 2.91597158e-01
-2.80658633e-01 9.76979792e-01 -2.72402495e-01 8.14208269e-01
1.25574815e+00 5.84458172e-01 -1.20077038e+00 9.57112730e-01
1.02132605e-02 5.99391401e-01 2.82262772e-01 4.68045235e-01
-3.06956649e-01 2.43937910e-01 5.66204548e-01 1.25774312e+00
2.57235199e-01 -1.34039521e-01 -3.05144072e-01 1.25758994e+00
2.06963643e-01 -6.13798294e-03 -5.35538614e-01 -2.43667230e-01
1.10517001e+00 1.44592750e+00 -5.19103169e-01 -3.52968462e-02
-5.76815665e-01 -2.82589972e-01 3.45569104e-01 3.16628851e-02
-5.60562015e-01 -1.17119253e+00 1.12270820e+00 1.50245383e-01
1.58375457e-01 -1.22761250e-01 -2.47918501e-01 -1.13639843e+00
3.94743443e-01 -9.73291159e-01 4.74043608e-01 -2.80906051e-01
-1.22817004e+00 1.16554940e+00 4.67716396e-01 -1.16316521e+00
-8.06874707e-02 -2.87718832e-01 -1.05891407e+00 1.17424607e+00
-9.35517848e-01 -6.95086122e-01 -3.20576370e-01 5.96604288e-01
-1.37880981e-01 -5.39724708e-01 9.38120186e-01 5.68143129e-01
-9.84585464e-01 6.43465638e-01 -4.60214354e-02 3.76268864e-01
4.81389374e-01 -1.07159245e+00 1.21576774e+00 1.29171944e+00
-4.58853960e-01 1.55749559e+00 2.05473855e-01 -1.32018745e+00
-1.31006491e+00 -1.29407728e+00 9.05118823e-01 -1.27657437e+00
1.17176414e+00 -6.21770263e-01 -1.52666891e+00 6.73427045e-01
-1.67151898e-01 1.62501916e-01 7.99237490e-01 1.01653539e-01
-1.02905381e+00 1.90711871e-01 -1.46541703e+00 -1.62723009e-02
8.41247141e-01 -9.23476398e-01 -7.11595893e-01 -6.60057440e-02
1.03874350e+00 2.22080071e-02 -1.51534474e+00 -4.37328890e-02
1.73858106e-01 -8.88896883e-01 9.35984313e-01 -5.43106973e-01
6.31733656e-01 -2.51797557e-01 -3.45333934e-01 -8.75341773e-01
-2.63422906e-01 -6.62576258e-01 3.56923696e-03 2.07173657e+00
4.80268866e-01 -8.18190813e-01 2.79481765e-02 8.15563083e-01
-3.15079868e-01 -4.21847939e-01 -7.82262087e-01 -6.18676484e-01
3.63497108e-01 -7.18932509e-01 1.43582606e+00 1.33520079e+00
4.40156877e-01 -7.68339276e-01 2.62858093e-01 5.79427183e-01
8.00133586e-01 -3.99471857e-02 5.49885988e-01 -1.63824463e+00
-1.01442106e-01 -2.86433846e-01 -4.20519501e-01 6.65567815e-01
2.87545323e-01 -9.45161879e-01 -4.42545444e-01 -1.11063194e+00
3.57794851e-01 -7.02346683e-01 -2.03542128e-01 1.06458104e+00
-2.45716020e-01 -2.72432208e-01 -1.38438821e-01 2.49665990e-01
4.01019119e-02 -2.30820239e-01 -1.24064803e-01 -1.87085371e-03
-1.99112892e-01 -1.90727398e-01 -9.57145512e-01 7.42495179e-01
6.86472654e-01 -1.13597286e+00 -2.05942467e-01 -3.97902727e-01
8.79775882e-01 2.05996826e-01 4.88049656e-01 -7.84743607e-01
-3.63906175e-02 -3.56973559e-01 -3.20829630e-01 -5.84708691e-01
-8.72484326e-01 -6.23859227e-01 6.57550752e-01 4.41239178e-01
-2.76389956e-01 4.11592841e-01 4.80552733e-01 2.27897272e-01
1.58970617e-02 -1.02235174e+00 2.95980632e-01 7.01371580e-02
-6.57322347e-01 4.30416286e-01 -4.41050917e-01 8.75192359e-02
1.20535064e+00 2.45715342e-02 -1.07284641e+00 8.90332699e-01
-3.50625336e-01 -5.47634885e-02 9.40889060e-01 9.06708062e-01
4.61298615e-01 -1.15872180e+00 -6.32074475e-01 2.62104362e-01
5.39925456e-01 -4.93906975e-01 5.19139826e-01 2.77471781e-01
-3.64218891e-01 2.21764416e-01 -5.96970558e-01 9.75249335e-02
-1.27557993e+00 1.35438073e+00 -9.07136127e-02 -1.19164787e-01
-7.17443228e-01 4.52860653e-01 -3.48812610e-01 -5.85096121e-01
-2.01338619e-01 -1.57613873e-01 -5.73357522e-01 1.92204639e-01
1.19390166e+00 7.63762236e-01 4.57329512e-01 -4.34470326e-01
-7.39319086e-01 6.69238210e-01 -1.53157234e-01 1.80138826e-01
1.80870688e+00 1.19068280e-01 -9.67156827e-01 1.11870013e-01
1.40442109e+00 4.63469595e-01 -8.08129370e-01 1.14466429e-01
6.37721300e-01 -8.09500873e-01 -1.81242242e-01 -7.81364501e-01
-1.26985383e+00 7.32786596e-01 2.13045686e-01 4.64633495e-01
1.07131720e+00 1.45381363e-02 3.21289808e-01 2.88726509e-01
9.77271974e-01 -3.96163940e-01 -2.72954732e-01 4.43967015e-01
9.61064339e-01 -7.64625490e-01 -1.21076517e-02 -3.89326334e-01
-1.24493457e-01 1.24811912e+00 5.93974710e-01 3.58520038e-02
9.55288291e-01 6.85400009e-01 -1.78748086e-01 -4.67262805e-01
-7.97114074e-01 5.65801084e-01 -5.04894182e-02 9.51401949e-01
2.21267626e-01 -2.19607592e-01 -1.95259690e-01 1.15856051e+00
-1.08406700e-01 -3.10035825e-01 1.17026663e+00 1.22569311e+00
-5.20476103e-01 -1.49813724e+00 -5.61844945e-01 2.05753118e-01
-6.40531719e-01 -4.90297198e-01 -3.90888631e-01 4.62773204e-01
2.51606613e-01 1.02969563e+00 -2.86210835e-01 -6.51885509e-01
6.28545761e-01 -2.04571337e-01 -2.15197921e-01 -9.02591527e-01
-1.44788206e+00 -8.70120347e-01 -2.02753637e-02 -7.66615510e-01
3.74490410e-01 -8.94139707e-01 -1.08127177e+00 -4.21297014e-01
-2.37299651e-01 2.81605452e-01 9.30478811e-01 5.04362166e-01
7.49049842e-01 5.40517688e-01 8.38412762e-01 -4.63811398e-01
-3.82619172e-01 -4.82058972e-01 -2.63193071e-01 1.06090337e-01
4.33328748e-01 -6.35514498e-01 -4.27689791e-01 1.08191304e-01] | [7.085632801055908, 7.779716968536377] |
d5baf002-923c-42cd-a29f-e9dd494e9513 | a-slot-is-not-built-in-one-utterance-spoken | null | null | https://openreview.net/forum?id=AjLizwlktx | https://openreview.net/pdf?id=AjLizwlktx | A Slot Is Not Built in One Utterance: Spoken Language Dialogs with Sub-Slots | A slot value might be provided segment by segment over multiple-turn interactions in a dialog, especially for some important information such as phone numbers and names. It is a common phenomenon in daily life, but little attention has been paid to it in previous work. To fill the gap, this paper defines a new task named Sub-Slot based Task-Oriented Dialog (SSTOD) and builds a Chinese dialog dataset SSD for boosting research on SSTOD. The dataset includes total 40K dialogs and 500K utterances from four different domains: Chinese names, phone numbers, ID numbers and license plate numbers. The data is well annotated with sub-slot values, slot values, dialog states and actions. We find some new linguistic phenomena and interactive manners in SSTOD which raise some new challenges of building agents for the task. We test three state-of-the-art models on SSTOD and find they cannot handle the new task well on any of the four domains. We also investigate an improved model by involving slot knowledge in a plug-in manner. More work should be done to meet the new challenges raised from SSTOD which widely exists in real-life applications. | ['Anonymous'] | 2021-11-16 | null | null | null | acl-arr-november-2021-11 | ['sstod'] | ['natural-language-processing'] | [-3.89403671e-01 3.13779175e-01 -2.42515817e-01 -8.88433814e-01
-4.04568672e-01 -8.75598907e-01 8.63971055e-01 -3.66111398e-01
-3.91234368e-01 1.30318630e+00 7.08488703e-01 -5.14759481e-01
2.16575101e-01 -3.73205662e-01 1.52063802e-01 -3.02590162e-01
2.80116290e-01 1.21634936e+00 7.49388218e-01 -1.01111686e+00
1.31817460e-01 1.77266687e-01 -8.81849289e-01 4.96605545e-01
6.93033159e-01 6.60215914e-01 3.01722556e-01 3.74688238e-01
-1.06824017e+00 1.05053782e+00 -9.90543008e-01 -5.62727690e-01
-3.31906080e-02 -5.48455894e-01 -1.50469875e+00 2.10916847e-01
-1.75172880e-01 -4.15951192e-01 -6.36489689e-02 7.30937302e-01
5.15053928e-01 3.76276225e-01 6.06930077e-01 -1.59117579e+00
-4.19441968e-01 1.15284908e+00 1.16192950e-02 5.21421880e-02
3.83685201e-01 -6.37513176e-02 9.56551552e-01 -5.61440825e-01
6.93660319e-01 1.67542601e+00 4.96267349e-01 1.02179170e+00
-7.48748541e-01 -4.41338986e-01 3.06391805e-01 -3.38344388e-02
-7.45018542e-01 -3.94904226e-01 5.68606496e-01 -3.02581191e-01
1.13839638e+00 3.86749834e-01 1.08129308e-01 1.25686216e+00
-2.27147594e-01 8.75780642e-01 1.19806409e+00 -3.62950563e-01
2.35708669e-01 8.27950299e-01 5.47995329e-01 1.71904624e-01
-5.11016667e-01 -2.89415300e-01 -4.67046410e-01 -2.06230223e-01
6.70272112e-01 -2.27009848e-01 1.99711636e-01 -1.36543298e-02
-1.45722055e+00 1.04903483e+00 1.08921379e-01 6.61928773e-01
2.65084326e-01 -5.38790226e-01 7.05039561e-01 5.37451863e-01
4.68233734e-01 6.12110138e-01 -1.24425209e+00 -6.54910743e-01
-4.49539758e-02 5.70389688e-01 1.43793941e+00 1.45661855e+00
6.42646670e-01 -3.15876812e-01 -1.44168511e-01 1.33986092e+00
1.63240641e-01 3.01549733e-01 6.37862265e-01 -8.77044141e-01
7.30961204e-01 8.40004742e-01 4.72872674e-01 -3.26146334e-01
-8.13904822e-01 4.46575671e-01 -5.23334980e-01 -2.74199218e-01
8.09556901e-01 -6.98681891e-01 -5.60929596e-01 1.72413623e+00
5.83727479e-01 -2.97558844e-01 4.14627194e-01 8.49649489e-01
1.44492650e+00 6.60804808e-01 2.68101066e-01 -1.69676244e-01
1.99918520e+00 -1.28678036e+00 -1.21475840e+00 -3.85832489e-01
7.66841173e-01 -9.00922954e-01 1.25727296e+00 -2.76664476e-04
-4.85780895e-01 -4.68764752e-01 -4.48697418e-01 -2.60116965e-01
-8.89109194e-01 2.31030453e-02 8.07333410e-01 6.96415961e-01
-8.86159718e-01 -6.58173934e-02 -2.31798187e-01 -7.14329541e-01
-4.54959303e-01 4.80576783e-01 -1.99166551e-01 5.42881906e-01
-1.67557371e+00 1.14283442e+00 4.91178155e-01 -1.35258019e-01
-4.44093496e-01 -1.78188488e-01 -8.46090376e-01 -2.13661984e-01
9.09510255e-01 -1.30232841e-01 1.93535185e+00 -5.31542599e-01
-1.97166097e+00 7.44236112e-01 -1.69940799e-01 -2.72011310e-01
3.09720874e-01 -1.82822064e-01 -6.29029632e-01 -5.15804052e-01
2.11171404e-01 8.28927279e-01 2.61192560e-01 -9.85352576e-01
-8.19358230e-01 -2.66967475e-01 5.51043153e-01 4.86790836e-01
-2.48117894e-01 4.24288750e-01 -2.57644743e-01 -4.69587535e-01
-2.36613870e-01 -1.19023287e+00 -3.22858185e-01 -5.50925195e-01
-6.67954087e-01 -8.98226082e-01 9.53521490e-01 -4.66826171e-01
1.18521976e+00 -2.02918649e+00 -5.88612445e-03 -5.20474553e-01
2.67303139e-02 3.34138334e-01 2.56447583e-01 8.85659575e-01
4.43312049e-01 -9.99463126e-02 -1.28980324e-01 -4.13062632e-01
3.92278373e-01 4.73996699e-01 -4.31093723e-01 -3.38058025e-01
-4.90447730e-02 9.11144495e-01 -5.51142395e-01 -6.00170791e-01
2.82166660e-01 -1.95103168e-01 -2.13238493e-01 5.63526630e-01
-8.83223474e-01 7.18476176e-01 -7.37619400e-01 5.87572873e-01
7.10740507e-01 -1.35397106e-01 3.15732837e-01 1.43174544e-01
-2.78946310e-01 5.01079798e-01 -9.54436600e-01 1.67320323e+00
-3.29533368e-01 3.74293536e-01 1.80304021e-01 -8.16441655e-01
1.19716859e+00 4.59304094e-01 1.04209103e-01 -4.24063951e-01
3.17057133e-01 2.97984898e-01 1.20795801e-01 -5.93261302e-01
9.55652773e-01 -2.09327176e-01 -7.51920521e-01 2.87049502e-01
9.88945886e-02 -3.52350235e-01 1.20890113e-02 2.36859605e-01
7.22388744e-01 -2.43764251e-01 4.75837231e-01 -3.84406328e-01
8.41888905e-01 4.13797200e-01 3.20545614e-01 5.91538966e-01
-5.04134715e-01 1.14648260e-01 8.31832945e-01 -4.25854266e-01
-7.77223170e-01 -5.59523940e-01 -2.66368419e-01 1.80080128e+00
3.37294668e-01 -2.76034772e-01 -7.46294975e-01 -1.13036728e+00
-1.39857098e-01 5.89674175e-01 -4.17993039e-01 4.86476362e-01
-5.18038869e-01 -4.88060325e-01 6.33790195e-01 3.71697992e-01
1.17464006e+00 -1.55445075e+00 -7.34248683e-02 4.67072904e-01
-3.43610883e-01 -1.61992764e+00 -5.97774088e-01 3.83224130e-01
-5.04518449e-01 -7.66113818e-01 -8.63642275e-01 -1.29136157e+00
1.38963386e-01 1.06171072e-01 1.13185596e+00 -3.56197268e-01
4.10542309e-01 1.24618672e-01 -8.52801621e-01 -5.43801010e-01
-5.82459688e-01 3.61307055e-01 -5.91633804e-02 -3.61421883e-01
1.00085306e+00 2.13598292e-02 -5.73651341e-04 1.08010805e+00
-4.16707069e-01 6.17861934e-02 1.03437543e-01 9.46455240e-01
-3.35029900e-01 -1.25150710e-01 9.15358186e-01 -1.25758862e+00
1.08828318e+00 -5.65446615e-01 -4.90696967e-01 1.27555236e-01
8.92904028e-02 7.14551061e-02 5.66792965e-01 -5.30825257e-01
-1.60407269e+00 2.61052884e-02 -4.34996873e-01 5.24958193e-01
-6.37969434e-01 3.21675658e-01 -4.16759551e-01 3.64051461e-01
4.74048316e-01 1.43417880e-01 9.29422751e-02 -9.15477097e-01
3.86972249e-01 1.43468404e+00 2.68091738e-01 -7.13263392e-01
3.53157818e-01 3.06501379e-03 -7.14648187e-01 -9.95406866e-01
-7.38902509e-01 -9.02654111e-01 -6.11561358e-01 4.24969979e-02
1.03438604e+00 -7.94675410e-01 -1.03282428e+00 7.97611237e-01
-1.28721070e+00 -7.41154313e-01 7.71197528e-02 2.84210503e-01
-3.28736335e-01 2.46177018e-01 -8.74947667e-01 -9.13390398e-01
-1.70501158e-01 -1.29589081e+00 1.05908227e+00 3.66944730e-01
-2.83581555e-01 -1.18691325e+00 8.41667950e-02 5.54043233e-01
6.14693820e-01 -4.23761427e-01 9.69778419e-01 -1.47884214e+00
-2.06720471e-01 3.74486417e-01 -1.42783105e-01 1.86481446e-01
4.36836541e-01 -6.98207438e-01 -8.55007946e-01 1.79575533e-01
8.25139806e-02 -7.86407232e-01 2.77227342e-01 9.24359169e-03
5.61064243e-01 -3.00417781e-01 -3.39366108e-01 -3.66078317e-01
7.18729377e-01 9.19209898e-01 3.40066195e-01 4.20046866e-01
3.88750196e-01 1.02850258e+00 1.15752554e+00 5.06862164e-01
7.15089321e-01 1.22859848e+00 2.78701726e-02 -5.80534451e-02
1.76150024e-01 -2.38181222e-02 2.82042533e-01 7.36981153e-01
4.59872246e-01 -3.37036252e-01 -8.53512108e-01 5.68876982e-01
-2.08358717e+00 -6.12054229e-01 -1.82323888e-01 1.52864838e+00
1.39662123e+00 8.92549530e-02 5.84911525e-01 -2.97477156e-01
7.67489493e-01 3.19766551e-01 -5.78899026e-01 -5.77436090e-01
-4.25351821e-02 -3.27553272e-01 -9.31609869e-02 6.99448526e-01
-1.32598305e+00 1.68592298e+00 5.87301970e+00 8.85596335e-01
-8.47788095e-01 3.72998565e-02 7.44095325e-01 8.81498158e-01
5.89220747e-02 1.17492743e-01 -1.38148475e+00 5.05267620e-01
8.04451942e-01 9.53104161e-03 3.06774229e-01 1.24103642e+00
-3.90579775e-02 -2.79639393e-01 -9.75948632e-01 7.38112569e-01
-1.73697829e-01 -1.04303825e+00 -3.27678651e-01 3.44551867e-03
3.54711622e-01 -2.61718959e-01 -2.04782590e-01 8.18745732e-01
7.91321039e-01 -8.59870791e-01 2.22292632e-01 -1.28621534e-01
5.46382248e-01 -2.08357319e-01 1.07953274e+00 5.91133416e-01
-1.09839797e+00 5.53275794e-02 -5.44855952e-01 -2.48362273e-01
3.08226407e-01 -2.05911770e-02 -1.48028231e+00 2.26284623e-01
5.21969140e-01 5.20438671e-01 -2.17659101e-01 4.80662763e-01
6.67641163e-02 3.21958959e-01 -2.19913498e-01 -6.70995355e-01
5.46843708e-01 -1.59560069e-01 2.38262534e-01 1.20533967e+00
-5.46762086e-02 2.48309791e-01 4.24462944e-01 4.80248153e-01
1.70177460e-01 2.31727898e-01 -6.09238625e-01 9.94796231e-02
7.91534662e-01 1.20107114e+00 -8.43196630e-01 -4.62809324e-01
-6.17278636e-01 8.46882164e-01 3.49606201e-02 1.61343887e-01
-7.63081908e-01 -3.71715605e-01 6.49153829e-01 -2.82061636e-01
1.16373457e-01 -3.20622146e-01 -1.89197525e-01 -1.09157491e+00
-2.09518835e-01 -1.15036356e+00 3.94114196e-01 -5.29972315e-01
-1.59480941e+00 1.08642626e+00 3.81923735e-01 -1.25656450e+00
-5.73706567e-01 -7.90092528e-01 -3.58684719e-01 7.10035741e-01
-1.18431270e+00 -1.11662030e+00 -2.32225299e-01 8.48501742e-01
1.36189115e+00 -7.08930790e-01 1.03477609e+00 1.61752477e-01
-2.53181785e-01 3.35315734e-01 1.06118321e-02 4.77366120e-01
1.16303086e+00 -1.47275400e+00 7.12495983e-01 1.13068648e-01
-1.38029337e-01 6.57932222e-01 7.47896910e-01 -6.06877685e-01
-8.28929901e-01 -6.09061003e-01 1.25407076e+00 -8.05467725e-01
6.96381569e-01 -8.09880316e-01 -7.17195034e-01 8.93415272e-01
6.75441802e-01 -5.64899325e-01 6.48900449e-01 4.09383237e-01
1.17791677e-02 1.85317218e-01 -1.26219118e+00 5.67346156e-01
9.61401284e-01 -2.60629237e-01 -1.07477546e+00 6.13303185e-01
1.22080863e+00 -7.95258701e-01 -5.52692831e-01 9.48505327e-02
6.18532710e-02 -9.31538463e-01 7.40794718e-01 -7.32723713e-01
4.95052189e-02 6.34052278e-03 -1.24447964e-01 -1.24863589e+00
2.58785903e-01 -7.21521854e-01 3.87286693e-01 1.83245659e+00
6.46378756e-01 -7.70150602e-01 9.32751596e-01 1.01964152e+00
-2.77626604e-01 -2.68585116e-01 -1.08435702e+00 -7.02712834e-01
1.86726704e-01 -1.73638538e-01 1.02266240e+00 9.64773417e-01
6.91320181e-01 8.84885252e-01 -8.09990168e-01 -3.64902467e-01
-2.19409615e-01 -2.99232844e-02 1.08348060e+00 -1.25325251e+00
-2.37202700e-02 3.77615578e-02 1.44738043e-02 -1.71312690e+00
2.87364364e-01 -2.59394467e-01 1.44420460e-01 -1.26473761e+00
-9.84705836e-02 -8.97591591e-01 4.78862822e-01 7.31922388e-01
-6.49951547e-02 -3.35564345e-01 3.63051057e-01 2.12444112e-01
-8.02581906e-01 7.03329206e-01 1.19672561e+00 -8.93700346e-02
-6.37983561e-01 3.97735536e-01 -7.24355459e-01 5.46305418e-01
8.74703526e-01 -2.10869387e-01 -6.64341688e-01 -1.36691496e-01
-3.33350718e-01 6.29352093e-01 -1.52944967e-01 -6.26023829e-01
3.25562358e-01 -4.63020712e-01 -3.28729779e-01 -5.77526391e-01
7.47713387e-01 -1.02799952e+00 -3.04603815e-01 6.02076985e-02
-6.20741308e-01 -7.04462454e-02 2.07520470e-01 2.36260667e-01
-3.99101049e-01 -4.12286252e-01 4.60661858e-01 -5.11102200e-01
-1.19239032e+00 -1.89947903e-01 -6.73936248e-01 2.84179568e-01
1.05044127e+00 5.70976138e-02 -7.43444979e-01 -7.22084582e-01
-8.96001220e-01 5.37803471e-01 2.32258830e-02 8.23901236e-01
9.99764949e-02 -1.15251327e+00 -3.24118465e-01 8.04280490e-03
1.81504682e-01 -1.42492637e-01 3.31584394e-01 3.24854165e-01
-3.58763248e-01 9.04963970e-01 -3.58738840e-01 -3.74892205e-01
-1.39005375e+00 3.01458955e-01 2.57695079e-01 -5.26175678e-01
-2.67394662e-01 8.80374968e-01 1.95541635e-01 -1.25666916e+00
5.41409075e-01 -3.39697331e-01 -7.67808974e-01 3.43465447e-01
4.44543600e-01 -2.46503092e-02 -2.32199505e-01 -7.17101634e-01
-4.48478490e-01 2.50403851e-01 -3.76592249e-01 -4.27940756e-01
9.40897822e-01 -4.66114163e-01 -3.18994485e-02 6.61655366e-01
8.30096424e-01 -3.44971418e-02 -1.01788425e+00 -6.00048542e-01
3.96860391e-01 -2.42689282e-01 -8.40584278e-01 -1.21687198e+00
-5.55155456e-01 6.14660501e-01 3.14503044e-01 7.63045788e-01
6.05634570e-01 4.12553430e-01 9.95062411e-01 7.62560010e-01
6.55487835e-01 -1.29036582e+00 2.87229985e-01 1.26879156e+00
9.07146037e-01 -1.60455966e+00 -5.90050161e-01 -6.48094535e-01
-1.63310909e+00 9.39793646e-01 1.16819358e+00 5.07455409e-01
7.03626394e-01 3.55057508e-01 8.03358376e-01 -1.80713192e-01
-6.71268225e-01 -3.09732437e-01 -2.28574321e-01 8.45033467e-01
6.05133235e-01 9.57633927e-02 -5.62144578e-01 8.64257991e-01
-4.65131909e-01 -2.71177143e-01 5.58466196e-01 1.03527331e+00
-5.90457976e-01 -1.66747582e+00 -3.09090495e-01 2.71287441e-01
-2.24255800e-01 -2.22552419e-01 -9.04274523e-01 1.00536013e+00
-1.69568449e-01 1.45548952e+00 -1.59577131e-01 -4.96197701e-01
3.33231181e-01 4.02445436e-01 -4.12684321e-01 -8.10070395e-01
-8.95915568e-01 -1.01936618e-02 7.60106623e-01 -1.29240975e-01
-5.93280435e-01 -3.58110934e-01 -1.28227437e+00 -3.65141273e-01
-4.00374889e-01 7.00408399e-01 3.99395287e-01 9.51098800e-01
1.58112466e-01 1.33416593e-01 4.69201654e-01 -6.19764626e-01
-6.49606049e-01 -1.39139366e+00 -8.41172278e-01 4.26148802e-01
1.31717160e-01 -8.35038662e-01 -2.40719959e-01 -2.59222806e-01] | [12.757564544677734, 7.8600382804870605] |
2608763c-6269-4e7d-b650-07edbaa6201e | learning-discriminative-representations-for | 2011.02120 | null | https://arxiv.org/abs/2011.02120v1 | https://arxiv.org/pdf/2011.02120v1.pdf | Learning Discriminative Representations for Fine-Grained Diabetic Retinopathy Grading | Diabetic retinopathy (DR) is one of the leading causes of blindness. However, no specific symptoms of early DR lead to a delayed diagnosis, which results in disease progression in patients. To determine the disease severity levels, ophthalmologists need to focus on the discriminative parts of the fundus images. In recent years, deep learning has achieved great success in medical image analysis. However, most works directly employ algorithms based on convolutional neural networks (CNNs), which ignore the fact that the difference among classes is subtle and gradual. Hence, we consider automatic image grading of DR as a fine-grained classification task, and construct a bilinear model to identify the pathologically discriminative areas. In order to leverage the ordinal information among classes, we use an ordinal regression method to obtain the soft labels. In addition, other than only using a categorical loss to train our network, we also introduce the metric loss to learn a more discriminative feature space. Experimental results demonstrate the superior performance of the proposed method on two public IDRiD and DeepDR datasets. | ['Yupeng Xu', 'Shaorong Xie', 'Zhijie Wen', 'Liyan Ma', 'Li Tian'] | 2020-11-04 | null | null | null | null | ['diabetic-retinopathy-grading'] | ['medical'] | [-1.84386298e-02 -3.52966249e-01 -2.00293288e-01 -7.23386168e-01
-5.29755712e-01 -2.91622460e-01 1.99625790e-01 -7.25594312e-02
-3.44353557e-01 6.63210928e-01 3.84321839e-01 -2.29908392e-01
-2.12293997e-01 -6.38037741e-01 -2.14748591e-01 -9.36797976e-01
3.12616050e-01 -1.21797631e-02 9.56894681e-02 2.21332163e-01
1.94896325e-01 4.19696063e-01 -1.40669119e+00 2.18086988e-01
1.38804984e+00 1.24205983e+00 -3.86723410e-03 1.06010608e-01
-6.09331504e-02 9.74390090e-01 -2.83879161e-01 -1.92756623e-01
3.24222714e-01 -5.50581515e-01 -4.50436234e-01 1.79453775e-01
5.06852686e-01 -7.24373162e-01 -4.96848106e-01 1.33247328e+00
6.76640451e-01 -1.47430673e-01 6.97861135e-01 -6.44393504e-01
-8.28377724e-01 6.94807321e-02 -8.78684938e-01 3.06098819e-01
-5.46113513e-02 3.68338436e-01 9.25836027e-01 -6.44180715e-01
1.51943833e-01 1.13008249e+00 4.33441222e-01 3.57853681e-01
-1.06735766e+00 -5.96006095e-01 1.81079097e-02 5.36893904e-01
-1.29624569e+00 -3.04545492e-01 7.64313936e-01 -7.68728256e-01
-3.47409165e-03 1.19737223e-01 5.06681502e-01 6.27488673e-01
-3.04625127e-02 7.00352728e-01 1.61322951e+00 2.74641886e-02
4.38579321e-02 -1.21209033e-01 1.51745170e-01 7.44749963e-01
2.87644029e-01 9.57010686e-02 1.80913538e-01 7.28919357e-02
7.30242074e-01 4.44997787e-01 -4.40984368e-01 -2.91723192e-01
-9.53945398e-01 7.65108705e-01 7.87387490e-01 -8.12872946e-02
-3.15079182e-01 -1.10883988e-01 3.83453697e-01 1.07448930e-02
2.32469872e-01 6.18263446e-02 -1.70632660e-01 5.13294563e-02
-7.22004533e-01 -1.85261928e-02 1.24034755e-01 3.09653133e-01
7.51647949e-01 -3.82464141e-01 -7.34623015e-01 1.09308195e+00
3.31586063e-01 9.85916406e-02 4.37503666e-01 -9.93739784e-01
3.12560439e-01 1.02433097e+00 1.42066684e-02 -9.45232272e-01
-2.73922026e-01 -7.38848686e-01 -1.21835160e+00 3.80940467e-01
4.87193137e-01 -5.49531318e-02 -1.20585060e+00 1.32348013e+00
2.07117856e-01 3.16128619e-02 -2.57318527e-01 1.34930694e+00
6.37157977e-01 2.66889870e-01 -5.40607758e-02 -1.52744085e-01
1.27191627e+00 -7.27101862e-01 -5.08062422e-01 5.40458933e-02
5.70037961e-01 -5.37282884e-01 1.17819810e+00 4.30306941e-01
-7.98115730e-01 -4.59604859e-01 -8.20642471e-01 -4.77829188e-01
-1.07890088e-02 6.58240318e-01 7.66984940e-01 2.41983339e-01
-8.20491493e-01 2.65329361e-01 -7.66973913e-01 -1.78616062e-01
8.96982789e-01 1.46905243e-01 -1.63270950e-01 -5.39570272e-01
-9.66772020e-01 6.86416924e-01 1.26534477e-01 4.81928378e-01
-5.43983161e-01 -5.10420978e-01 -5.87936103e-01 -1.46199092e-01
5.13950363e-02 -8.19508433e-01 8.39384675e-01 -9.66568589e-01
-1.27518272e+00 9.98042941e-01 -2.17647731e-01 -2.72530854e-01
7.57421434e-01 -9.49438885e-02 -2.92789251e-01 1.84499413e-01
2.73527224e-02 4.31577861e-01 4.88736629e-01 -9.33504999e-01
-1.05171275e+00 -5.55575788e-01 2.76064962e-01 -2.55059320e-02
-2.20427692e-01 8.27896968e-02 -3.27819288e-01 -4.76960659e-01
1.77317277e-01 -6.51622176e-01 -2.10994944e-01 4.17170495e-01
-5.90860605e-01 -4.90054876e-01 5.34838736e-01 -8.48073065e-01
1.19033599e+00 -2.30415320e+00 4.55513448e-02 8.86890367e-02
7.11178422e-01 3.46218318e-01 1.41175374e-01 -3.16977501e-01
1.52881160e-01 8.23614076e-02 -2.29598492e-01 -7.46060461e-02
-2.35293895e-01 1.04821175e-01 -6.43642023e-02 5.06996572e-01
2.71533221e-01 6.78784013e-01 -7.43386805e-01 -5.55816531e-01
6.15514554e-02 4.50810611e-01 -4.75462258e-01 2.54172027e-01
-1.46538178e-02 5.79282939e-01 -7.59536803e-01 7.88611650e-01
7.23249376e-01 -5.00645399e-01 -2.98874676e-01 -5.94929695e-01
-2.75420398e-01 1.83694214e-01 -7.41624415e-01 1.32925475e+00
-7.77595863e-02 4.98490334e-01 -3.02521586e-01 -1.19265890e+00
8.03501725e-01 -5.65126166e-02 3.72321337e-01 -7.85709262e-01
1.85593948e-01 4.09759462e-01 3.03133607e-01 -7.45654166e-01
-2.51214862e-01 2.50514727e-02 3.79837483e-01 -3.54947113e-02
-5.30247092e-01 5.55844903e-01 2.14843273e-01 -2.82604724e-01
9.49074090e-01 4.61257957e-02 2.38839626e-01 1.59386903e-01
6.14339113e-01 -1.33920431e-01 1.04861212e+00 2.50434548e-01
-4.17269439e-01 9.68558013e-01 7.88497210e-01 -6.87483728e-01
-7.73890734e-01 -9.05121207e-01 -5.87482810e-01 3.30620825e-01
4.17669445e-01 2.36820877e-01 -3.32052201e-01 -7.19980896e-01
1.51919693e-01 2.44104385e-01 -7.04375863e-01 -2.06289738e-01
-4.99775529e-01 -1.09555793e+00 2.76656330e-01 5.23091853e-01
8.23583305e-01 -5.39367974e-01 -1.89348519e-01 -3.16299945e-02
-1.03512101e-01 -7.62732089e-01 -5.47705233e-01 -3.90940517e-01
-9.45355415e-01 -1.15473437e+00 -1.13379884e+00 -9.10574079e-01
7.71556973e-01 4.62138861e-01 7.19399095e-01 1.27811730e-01
-4.94851559e-01 -3.03019345e-01 -1.77706972e-01 3.47109139e-02
1.47138998e-01 -1.15451746e-01 -2.95476556e-01 5.02260566e-01
6.61888003e-01 -5.83545327e-01 -1.25052011e+00 2.73324072e-01
-6.78032041e-01 2.89467089e-02 1.25521326e+00 8.05423021e-01
8.70500147e-01 7.96156824e-02 4.72858727e-01 -6.81855321e-01
3.47527206e-01 -3.53372514e-01 -5.60201705e-01 2.72775143e-01
-7.53971696e-01 3.32645476e-02 4.66637462e-01 -2.08526284e-01
-9.12964463e-01 -3.07015907e-02 -2.55825501e-02 -3.31542760e-01
-2.05396950e-01 7.62079954e-01 -3.12290132e-01 -1.14836305e-01
3.28977793e-01 2.38438860e-01 -2.68693939e-02 -7.32422590e-01
5.35704643e-02 9.46420372e-01 4.19480622e-01 -2.09923118e-01
5.96204698e-01 4.46403950e-01 7.18802586e-02 -4.39255238e-01
-1.31789720e+00 -4.32239473e-01 -2.30218709e-01 8.75321627e-02
1.09636676e+00 -1.01446474e+00 -6.33000910e-01 4.90930557e-01
-9.12280738e-01 -2.20159162e-02 6.47230595e-02 7.14249551e-01
-1.23329177e-01 4.89845634e-01 -4.77541864e-01 -4.41005915e-01
-8.85943025e-02 -1.13991630e+00 8.96129847e-01 6.39383256e-01
3.78098845e-01 -6.46545291e-01 1.37787402e-01 5.72940290e-01
3.00881684e-01 4.07479495e-01 1.23458183e+00 -1.86932698e-01
-7.93756247e-01 -2.56868005e-01 -1.07100213e+00 7.14918017e-01
5.47866106e-01 1.36324063e-01 -7.72596776e-01 -2.00374544e-01
-3.33258420e-01 -3.62703741e-01 1.32450080e+00 5.45586050e-01
1.67572474e+00 -2.27215543e-01 -2.06851155e-01 1.05343938e+00
1.44128692e+00 2.58807302e-01 8.71008813e-01 3.18607926e-01
8.97155464e-01 5.35760581e-01 4.59848553e-01 2.33330101e-01
8.23555291e-01 5.15770853e-01 4.22570378e-01 -4.93684262e-01
-3.85765672e-01 -5.76461405e-02 4.14911658e-02 5.68961143e-01
-4.71975535e-01 1.48193046e-01 -7.84503639e-01 5.68906128e-01
-1.68351638e+00 -7.70153940e-01 -1.71414196e-01 2.30850601e+00
1.14698386e+00 2.76790950e-02 -2.43250597e-02 -2.60816425e-01
8.60811532e-01 -5.62918447e-02 -7.68563628e-01 1.58847973e-01
-2.49396399e-01 -6.10186234e-02 4.08997893e-01 3.53563279e-02
-1.14585030e+00 4.45119172e-01 5.08283901e+00 5.39971292e-01
-1.53015757e+00 -2.29595408e-01 9.13553298e-01 -9.74840224e-02
-1.15564860e-01 -1.83231104e-02 -6.81465745e-01 9.22133505e-01
1.65403619e-01 1.00343246e-02 2.68857837e-01 4.96231139e-01
6.78692877e-01 1.01637512e-01 -9.67113078e-01 9.88785505e-01
-1.56811893e-01 -8.91844690e-01 1.77072704e-01 2.03368738e-01
8.19277823e-01 -3.15939374e-02 2.71625817e-01 2.16159791e-01
1.49289578e-01 -1.22149575e+00 1.94674043e-03 9.97417867e-01
8.29097688e-01 -7.62651026e-01 9.90713120e-01 3.20149809e-02
-8.51537049e-01 -2.37868458e-01 -6.41939759e-01 1.03674963e-01
-1.36515766e-01 8.77297282e-01 -4.43861216e-01 3.14698219e-01
7.37462699e-01 1.09358919e+00 -7.72121549e-01 1.78737175e+00
-2.33720630e-01 6.26985013e-01 1.04368247e-01 4.19041187e-01
2.02646740e-02 -5.11703730e-01 2.68874347e-01 7.38399386e-01
4.49676514e-01 1.97265297e-01 2.61194795e-01 8.11594009e-01
-2.39011869e-01 9.14685726e-02 -2.04547882e-01 3.47535461e-02
9.90655273e-02 1.20576596e+00 -1.03767358e-01 5.89217693e-02
-6.94323182e-01 7.44782507e-01 2.35656649e-01 5.59350610e-01
-4.37586963e-01 -5.81860662e-01 7.60988176e-01 6.22444488e-02
2.06092462e-01 -6.09945133e-02 -3.91920388e-01 -1.34914792e+00
2.59957850e-01 -8.11141849e-01 3.14448148e-01 -6.03390634e-01
-1.57714891e+00 3.63754630e-01 -6.24581635e-01 -1.62677908e+00
1.20286137e-01 -4.66474324e-01 -6.94998801e-01 1.12676072e+00
-2.05984259e+00 -1.10137415e+00 -7.93156385e-01 4.72300678e-01
2.25833096e-02 -5.36044873e-02 3.31715584e-01 6.32436037e-01
-9.75910127e-01 5.72422266e-01 2.47522503e-01 4.42160815e-01
1.08607030e+00 -1.41490543e+00 -4.30067420e-01 6.99988425e-01
-6.58207953e-01 5.40305674e-01 4.27518368e-01 -3.31137955e-01
-8.23266387e-01 -1.39207315e+00 7.31100380e-01 -9.62309688e-02
5.52400351e-01 1.68842852e-01 -1.09946311e+00 3.65501225e-01
-1.50513783e-01 4.86951888e-01 5.67388594e-01 -5.85063435e-02
-3.34866524e-01 -7.10728586e-01 -1.03129566e+00 4.99733984e-01
9.45283294e-01 -4.88669038e-01 -4.37063634e-01 2.33658209e-01
5.52858174e-01 -1.94036707e-01 -8.22152197e-01 6.31023526e-01
5.97275913e-01 -9.75145102e-01 7.58991539e-01 -6.08373761e-01
7.75835991e-01 -5.86754382e-01 8.16344842e-02 -1.04952514e+00
-3.28006566e-01 -6.64041564e-02 4.14108299e-02 1.17399216e+00
7.28641674e-02 -9.00827110e-01 5.97395480e-01 4.11101639e-01
-1.73343837e-01 -9.22215104e-01 -6.25487387e-01 -4.38535631e-01
9.54948217e-02 2.93550611e-01 3.81956369e-01 8.61121893e-01
-6.61835015e-01 1.70682251e-01 -2.37212792e-01 4.06070203e-01
8.05494010e-01 4.98428553e-01 5.13198316e-01 -1.51933527e+00
-1.64641961e-01 -6.41997695e-01 -6.96067691e-01 -1.04833555e+00
-1.52232856e-01 -8.43862534e-01 -2.96075977e-02 -1.86641800e+00
5.72594643e-01 -5.44613898e-01 -8.66076767e-01 4.53179955e-01
-4.09762263e-01 3.46418679e-01 -4.85511184e-01 5.19897759e-01
-4.89830554e-01 6.94035172e-01 1.57456481e+00 -3.26571345e-01
-1.56145513e-01 1.88835472e-01 -1.07762194e+00 7.21922934e-01
7.08741426e-01 -2.20631629e-01 -2.93411523e-01 -6.49129868e-01
2.10472509e-01 -2.45287865e-01 8.13092649e-01 -8.33665729e-01
1.81089669e-01 -3.19057405e-01 4.67433870e-01 -4.97535199e-01
-6.48623332e-02 -5.18376052e-01 -4.41838056e-01 2.10881323e-01
-3.81649137e-01 -5.31936646e-01 -2.72906810e-01 5.28926611e-01
-5.93824625e-01 2.55056441e-01 1.01248288e+00 2.24268287e-01
-6.12094820e-01 8.01082730e-01 1.31327555e-01 5.33894673e-02
7.71173358e-01 -1.01827629e-01 -5.27842104e-01 2.05476005e-02
-5.06132901e-01 4.36193913e-01 5.26758730e-01 2.04758197e-01
5.67758024e-01 -1.30168033e+00 -9.33294415e-01 -1.12292834e-01
4.04643387e-01 2.21566603e-01 4.41989928e-01 1.33674300e+00
-5.24752975e-01 3.02454978e-01 -2.44435251e-01 -6.62367284e-01
-1.05911124e+00 2.72876501e-01 6.86062574e-01 2.16750409e-02
-7.31692612e-01 5.41748345e-01 5.65776706e-01 8.87940302e-02
2.87550509e-01 -4.79774743e-01 -7.11022019e-01 5.44492528e-02
6.69559360e-01 3.16093147e-01 -5.35194688e-02 -2.15959653e-01
-2.22016826e-01 8.57010007e-01 -3.98392260e-01 5.72275817e-01
1.25032556e+00 -2.88421810e-01 -3.26320678e-01 1.68145031e-01
1.19448161e+00 1.24510070e-02 -1.39264178e+00 -5.29219747e-01
-5.41270971e-02 -6.36316001e-01 5.70512891e-01 -9.18334007e-01
-1.40913689e+00 1.19213974e+00 1.02754390e+00 2.92615797e-02
1.39365685e+00 -1.00805350e-01 9.27481115e-01 1.98065609e-01
1.15587398e-01 -7.00904608e-01 -2.38610238e-01 1.36046559e-01
6.82712853e-01 -1.44237530e+00 -9.44541171e-02 -2.17137352e-01
-4.02955562e-01 1.14297748e+00 5.63863099e-01 -2.29025353e-02
3.80334407e-01 -4.82666761e-01 3.97182465e-01 5.13012148e-02
-3.70868891e-01 -5.19086421e-01 5.20415664e-01 5.01085937e-01
4.32792306e-01 3.45616862e-02 -6.60471380e-01 6.67351842e-01
1.44649431e-01 4.49060231e-01 5.13679385e-01 3.95188838e-01
-5.16387463e-01 -1.08696759e+00 1.86390400e-01 9.15778697e-01
-6.28032982e-01 -3.88843007e-02 -2.06046283e-01 4.35929924e-01
4.06351268e-01 6.93483710e-01 7.04024211e-02 -2.35493496e-01
2.91152149e-01 -3.94560814e-01 2.51513720e-01 -5.30363679e-01
4.76953946e-02 -1.76456440e-02 -1.46063149e-01 -4.67953682e-01
-4.27495062e-01 -6.29440010e-01 -1.07372046e+00 -7.94251040e-02
1.22524157e-01 -2.57204652e-01 2.90580809e-01 9.88712370e-01
3.83222371e-01 4.47023392e-01 1.02816856e+00 -1.69822961e-01
-7.11398542e-01 -8.05490494e-01 -6.72057509e-01 2.31577694e-01
7.49395907e-01 -7.46670365e-01 -4.83518809e-01 -1.66111905e-02] | [15.810710906982422, -3.982178211212158] |
d291a2d6-df28-4b83-a7e0-54d11e8d4c84 | few-shot-scene-classification-of-optical | null | null | https://www.webofscience.com/wos/woscc/full-record/WOS:000818847600008 | https://ieeexplore.ieee.org/document/9799778 | Few-Shot Scene Classification of Optical Remote Sensing Images Leveraging Calibrated Pretext Tasks | Small data hold big artificial intelligence (AI) potential. As one of the promising small data AI approaches, few-shot learning has the goal to learn a model efficiently that can recognize novel classes with extremely limited training samples. Therefore, it is critical to accumulate useful prior knowledge obtained from large-scale base class dataset. To realize few-shot scene classification of optical remote sensing images, we start from a baseline model that trains all base classes using a standard cross-entropy loss leveraging two auxiliary objectives to capture intrinsical characteristics across the semantic classes. Specifically, rotation prediction learns to recognize the 2-D rotation of an input to guide the learning of class-transferable knowledge, and contrastive learning aims to pull together the positive pairs while pushing apart the negative pairs to promote intraclass consistency and interclass inconsistency. We jointly optimize two such pretext tasks and semantic class prediction task in an end-to-end manner. To further overcome the overfitting issue, we introduce a regularization technique, adversarial model perturbation, to calibrate the pretext tasks so as to enhance the generalization ability. Extensive experiments on public remote sensing benchmarks including Northwestern Polytechnical University (NWPU)-RESISC45, aerial image dataset (AID), and Wuhan University (WHU)-remote sensing (RS)-19 demonstrate that our method works effectively and achieves best performance that significantly outperforms many state-of-the-art approaches. | ['Tiancan)', 'TC (Mei', 'Can) [1] ; Mei', 'C (Li', 'Yu) [1] ; Li', 'Y (Wan', 'Yongjun) [1] ; Wan', 'YJ (Zhang', 'Zhi) [1] ; Zhang', 'Z (Gao', 'Hong) [1] ; Gao', 'H (Ji', 'Ji'] | 2022-07-06 | null | null | null | journal-2022-7 | ['scene-classification'] | ['computer-vision'] | [ 5.74737191e-01 -4.84076925e-02 -2.31910974e-01 -5.15778661e-01
-9.77105737e-01 -2.59301513e-01 3.25107276e-01 -2.29575559e-01
-2.10545152e-01 6.72000706e-01 -1.41614318e-01 -1.54310599e-01
-3.53106141e-01 -9.79691446e-01 -8.07148635e-01 -8.81070256e-01
9.22511145e-02 1.53962806e-01 1.24022178e-02 -1.33097991e-01
1.33470688e-02 3.33591878e-01 -1.55277014e+00 2.52726942e-01
1.11708426e+00 1.22269678e+00 2.39842057e-01 3.67460966e-01
1.44941032e-01 1.06652093e+00 -1.59331933e-01 -6.62182719e-02
6.70382261e-01 -3.37868184e-01 -7.40567625e-01 1.44472912e-01
5.50210834e-01 -3.25794429e-01 -2.71375477e-01 1.29202271e+00
6.62949860e-01 5.95765412e-01 6.13373101e-01 -1.12549555e+00
-1.02495694e+00 3.77638698e-01 -5.95055997e-01 1.51924476e-01
-2.67534316e-01 4.06356931e-01 1.03941548e+00 -1.11378372e+00
6.57180130e-01 9.92594123e-01 5.96677959e-01 6.39080644e-01
-1.06189847e+00 -7.55874634e-01 3.17119569e-01 4.00959253e-01
-1.49226904e+00 -1.99384660e-01 9.13013399e-01 -4.40559119e-01
7.03381181e-01 2.16235340e-01 4.68850642e-01 8.70741069e-01
-2.14464799e-01 7.29129791e-01 8.20022643e-01 -1.13645494e-01
6.29011929e-01 3.46444882e-02 2.48891428e-01 6.08703852e-01
-6.66054934e-02 2.47925855e-02 -2.26620771e-02 6.82234913e-02
4.61532533e-01 6.60843492e-01 -4.27460045e-01 -6.51977420e-01
-9.88508463e-01 8.43769789e-01 9.44369972e-01 4.28060368e-02
-1.89833209e-01 -1.37960777e-01 2.21902117e-01 4.54578668e-01
6.27731860e-01 4.35388923e-01 -6.71862781e-01 2.91844934e-01
-4.64468807e-01 2.22557243e-02 4.04191256e-01 9.05428708e-01
1.19444156e+00 2.06462920e-01 -1.97858781e-01 1.13955855e+00
-2.14702766e-02 6.83225334e-01 5.34447312e-01 -6.50434196e-01
5.40615857e-01 6.13372564e-01 -1.63711712e-01 -9.21912968e-01
-2.55243033e-01 -5.01186013e-01 -1.20268261e+00 1.52649969e-01
2.45573223e-02 -2.13687614e-01 -1.27512884e+00 1.51655030e+00
3.88920486e-01 5.17115116e-01 3.13999295e-01 1.00246966e+00
9.03170586e-01 9.50458467e-01 1.07725374e-01 -1.71969458e-01
8.10535848e-01 -1.04203355e+00 -2.08632007e-01 -5.84338605e-01
4.30178821e-01 -2.14568630e-01 1.39214158e+00 1.86347533e-02
-4.96303439e-01 -6.59572363e-01 -1.21523929e+00 1.04379192e-01
-5.01565754e-01 -2.73590884e-03 5.52096546e-01 1.93880394e-01
-1.95481911e-01 6.84280992e-01 -7.12160766e-01 -1.68611571e-01
1.01702559e+00 -1.76068582e-02 -2.65613317e-01 -4.41440433e-01
-1.04566896e+00 5.61888218e-01 5.74357390e-01 1.80238262e-01
-1.05677640e+00 -1.23510158e+00 -1.05685139e+00 1.75475031e-01
6.87343955e-01 -4.19156611e-01 8.60035121e-01 -1.11415446e+00
-1.41416597e+00 7.43671715e-01 3.97207767e-01 -2.30576172e-01
3.80677193e-01 -3.77938375e-02 -2.93828934e-01 1.17251776e-01
-2.67718881e-02 6.28262877e-01 9.82768953e-01 -1.26216769e+00
-4.72737223e-01 -5.00291348e-01 -9.36948359e-02 1.29060209e-01
-5.36193550e-01 -3.22548151e-01 7.60513172e-02 -9.24140751e-01
2.47913480e-01 -9.63039160e-01 -2.99263448e-01 3.41574728e-01
-2.50993252e-01 6.03645816e-02 1.01631069e+00 -3.46332490e-01
8.12960982e-01 -2.26234984e+00 1.27296627e-01 8.65787640e-02
2.52262317e-03 5.27814388e-01 -4.85213816e-01 1.13187439e-03
-2.34136462e-01 -9.54874307e-02 -6.58284485e-01 1.75158203e-01
-2.69736856e-01 2.83478320e-01 -7.24132001e-01 4.88665015e-01
5.01885891e-01 1.00003827e+00 -1.14219749e+00 -1.19823990e-02
3.76878649e-01 4.08021808e-01 -5.14772594e-01 2.35638097e-01
-3.10904056e-01 2.88927525e-01 -5.89386463e-01 9.32528377e-01
7.95124233e-01 -5.02336979e-01 -1.91082418e-01 -1.72798499e-01
3.51943493e-01 -4.00443405e-01 -1.22838569e+00 1.68632066e+00
-3.72035980e-01 3.19362462e-01 -2.99507827e-01 -1.51019943e+00
1.14799213e+00 -1.25366807e-01 4.66038287e-01 -4.32593375e-01
-4.11747163e-03 -3.21940333e-02 -1.93590656e-01 -6.15478516e-01
1.98228955e-01 -3.46923262e-01 3.73623222e-02 3.65347601e-02
9.06970426e-02 -4.37329203e-01 -3.53501111e-01 -9.21314806e-02
8.46037567e-01 2.11882256e-02 4.61481035e-01 -5.36749624e-02
2.52791584e-01 -4.51816358e-02 9.47949767e-01 6.80822849e-01
-3.81251544e-01 8.26896131e-01 1.22788288e-02 -7.91249990e-01
-9.50894535e-01 -1.12533283e+00 -1.97022542e-01 1.14631236e+00
3.10330451e-01 9.42489207e-02 -2.53524721e-01 -7.69592702e-01
1.00618035e-01 5.99491537e-01 -8.98179352e-01 -6.28485501e-01
-1.82851076e-01 -8.94576967e-01 1.13001965e-01 6.17018938e-01
7.90592253e-01 -9.96564090e-01 -6.02604270e-01 1.29504040e-01
-8.39999095e-02 -9.98391807e-01 -1.44892544e-01 2.09798217e-01
-8.00941586e-01 -1.00695372e+00 -7.79951751e-01 -7.91785717e-01
6.38330877e-01 5.97486854e-01 8.03062439e-01 -1.44004315e-01
-6.83541954e-01 -5.30978013e-03 -5.35299778e-01 -4.43712294e-01
5.62297925e-02 -1.16713420e-02 -1.24166399e-01 2.38000751e-01
4.60492074e-01 -5.79082310e-01 -5.36904871e-01 2.15333045e-01
-1.04674017e+00 3.44518684e-02 6.54679060e-01 1.27419412e+00
6.39875054e-01 2.20529493e-02 7.25524247e-01 -9.18635249e-01
-1.49538353e-01 -6.59044027e-01 -4.28202182e-01 5.60113907e-01
-5.53921342e-01 -1.95160806e-01 9.86170113e-01 -6.03894413e-01
-1.02930498e+00 1.44305050e-01 2.38206759e-01 -8.06282818e-01
-1.24500379e-01 6.63912654e-01 -3.39895368e-01 -8.80509317e-02
9.25484359e-01 3.07032466e-01 -1.38744101e-01 -2.30893075e-01
3.99855942e-01 8.28687489e-01 5.12903094e-01 -3.30980420e-01
9.58574116e-01 6.17707551e-01 -5.19084781e-02 -1.10689294e+00
-1.40947056e+00 -5.92478454e-01 -6.33417010e-01 1.79128852e-02
7.34668672e-01 -1.21774614e+00 -1.83528274e-01 6.80087090e-01
-6.42721474e-01 -4.62195486e-01 -6.11667633e-01 4.37752753e-01
-4.89965558e-01 1.29835442e-01 -1.95343643e-01 -6.24764502e-01
-5.50457835e-01 -7.43613005e-01 9.60964918e-01 3.99772823e-01
5.02047777e-01 -5.97350955e-01 1.50557263e-02 3.24790835e-01
3.47607464e-01 3.81249011e-01 9.47720528e-01 -6.17169201e-01
-5.19712865e-01 -1.77762374e-01 -4.12734240e-01 7.36412704e-01
1.88347876e-01 -4.88418806e-03 -1.05380344e+00 -3.79557252e-01
-9.78763178e-02 -8.86239469e-01 1.21053576e+00 1.87782332e-01
1.47701192e+00 -2.90849954e-01 -1.65520161e-01 9.32273090e-01
1.65954673e+00 7.03945011e-02 8.50653350e-01 1.53931201e-01
1.00770319e+00 2.99068987e-01 7.02317953e-01 6.40501320e-01
2.89596081e-01 2.36177325e-01 3.87059510e-01 1.17340647e-01
1.41678289e-01 -2.70286858e-01 9.83881280e-02 3.68495077e-01
-1.67685691e-02 1.24742754e-01 -9.27135527e-01 4.13648546e-01
-1.90068197e+00 -1.34474409e+00 5.64642668e-01 2.24815702e+00
8.87868583e-01 -8.83523077e-02 -4.30699676e-01 3.98268551e-02
6.17516756e-01 4.79809850e-01 -1.03950250e+00 4.29344118e-01
-1.74183622e-01 2.44871825e-01 3.62831622e-01 2.06139877e-01
-1.38328242e+00 1.10205030e+00 4.67755413e+00 8.61191571e-01
-1.36761618e+00 -7.28018358e-02 7.92943716e-01 -1.90575749e-01
-1.89572051e-01 6.08668365e-02 -6.71231270e-01 1.94940284e-01
4.37619418e-01 -2.91658659e-02 5.10645568e-01 1.08237052e+00
-1.36346340e-01 2.34127715e-01 -8.34465027e-01 9.76597905e-01
1.98630795e-01 -1.40116882e+00 1.92794561e-01 -3.42041641e-01
1.16180897e+00 3.78639549e-01 9.56495628e-02 7.25804210e-01
2.08524868e-01 -8.25439513e-01 3.62176776e-01 6.93201900e-01
9.35995400e-01 -6.14639759e-01 4.56466675e-01 4.72103417e-01
-1.20347607e+00 -5.91483057e-01 -9.51243997e-01 -1.12968870e-01
-3.49829495e-01 4.42308336e-01 -4.54414725e-01 3.76731515e-01
6.79612339e-01 1.24719620e+00 -5.23772657e-01 9.11369741e-01
-1.99380159e-01 4.78598028e-01 -1.25237331e-01 4.49418910e-02
2.83158839e-01 -3.61685753e-01 5.17401218e-01 5.16080439e-01
1.56019270e-01 6.62995577e-01 4.69058037e-01 8.51849139e-01
-3.31606835e-01 -2.86935456e-02 -6.57033682e-01 -2.47649532e-02
5.70343554e-01 1.19626570e+00 -3.42217803e-01 -4.03451055e-01
-3.90385181e-01 8.24755073e-01 4.80856061e-01 3.94952714e-01
-6.58769011e-01 -6.65579736e-01 7.91552365e-01 -3.04826200e-01
5.30067444e-01 1.23529933e-01 -3.01903546e-01 -1.58229434e+00
9.62788910e-02 -8.43259990e-01 5.01525044e-01 -8.74658167e-01
-1.56264937e+00 4.24022883e-01 -2.36847103e-01 -1.62586379e+00
5.96649237e-02 -5.33992112e-01 -7.94001043e-01 6.20999515e-01
-1.91444528e+00 -1.43074632e+00 -8.69806826e-01 6.38373911e-01
7.00457394e-01 -3.27504426e-01 8.04522574e-01 9.76959020e-02
-6.68835819e-01 6.41956806e-01 4.36200827e-01 3.15834165e-01
6.32463157e-01 -9.98466134e-01 8.80949870e-02 8.29465568e-01
-5.72645478e-02 2.47011006e-01 3.46486807e-01 -5.16168356e-01
-1.36954892e+00 -1.69244087e+00 2.98622489e-01 -1.66017935e-01
7.05081940e-01 -1.35174453e-01 -1.33774316e+00 6.53547943e-01
-4.82040346e-01 8.22327077e-01 7.76564240e-01 -1.74167588e-01
-7.03364015e-01 -5.78498423e-01 -1.16832173e+00 3.21164459e-01
1.03860736e+00 -6.08907282e-01 -6.41989589e-01 5.38937211e-01
7.38380134e-01 -1.64891362e-01 -7.77115941e-01 6.82857215e-01
3.90817076e-01 -5.32148063e-01 1.13941348e+00 -1.11575341e+00
8.22522223e-01 -1.84918374e-01 -5.38866341e-01 -1.47886479e+00
-3.51059705e-01 -3.69353175e-01 2.95226760e-02 9.60782886e-01
2.83689260e-01 -5.14950931e-01 7.13673353e-01 6.69622838e-01
-2.86165297e-01 -7.01715350e-01 -6.71691775e-01 -1.06240225e+00
2.92821437e-01 -2.44052529e-01 4.50972289e-01 1.38374424e+00
-1.71057001e-01 4.81448501e-01 -5.94829798e-01 3.52342308e-01
7.58535445e-01 4.43803787e-01 8.55156124e-01 -1.33800018e+00
-2.01664910e-01 -1.89915732e-01 -5.83813846e-01 -7.09160030e-01
3.49231303e-01 -9.77407634e-01 2.57599503e-01 -1.26563537e+00
3.43674481e-01 -5.65201044e-01 -4.88976032e-01 9.15053189e-01
-5.54642081e-01 1.81705788e-01 2.48824432e-01 3.28857154e-01
-5.70335925e-01 1.09844589e+00 1.18914366e+00 -5.47068119e-01
-2.57303625e-01 5.45826778e-02 -7.19370425e-01 6.22223377e-01
6.19330108e-01 -4.42707062e-01 -5.07333398e-01 -4.36342418e-01
7.98650384e-02 -2.08109364e-01 5.42688429e-01 -1.23500347e+00
1.75324470e-01 -6.21293843e-01 5.39461136e-01 -4.43463027e-01
1.13915555e-01 -9.45862889e-01 -2.55672485e-01 3.53496015e-01
-4.06828105e-01 -7.57774472e-01 -5.37405796e-02 7.81301737e-01
-1.65556058e-01 -9.21013430e-02 1.30210459e+00 -2.03960150e-01
-1.12358046e+00 9.16934669e-01 2.63282865e-01 5.05373657e-01
1.32386553e+00 3.86300720e-02 -6.25459254e-01 4.71557602e-02
-5.10695040e-01 4.06080812e-01 2.59088337e-01 5.23720324e-01
8.81882846e-01 -1.30570590e+00 -9.02308047e-01 4.54572082e-01
7.65226603e-01 3.23608547e-01 6.87134027e-01 3.88833344e-01
-3.95858824e-01 -1.09847948e-01 -3.44299942e-01 -5.44505835e-01
-8.29317689e-01 7.76907980e-01 3.64638925e-01 1.08053014e-01
-8.03169847e-01 9.92656350e-01 7.57128596e-02 -7.18216419e-01
2.52956778e-01 -2.66449600e-01 8.73126462e-03 3.93243618e-02
9.18243706e-01 3.20215851e-01 -3.64617705e-02 -2.88853586e-01
-3.55497986e-01 5.94484985e-01 -8.20012987e-02 5.03128648e-01
1.84089243e+00 -9.53751132e-02 1.86657637e-01 6.17696166e-01
1.52071261e+00 -5.26931047e-01 -1.66392899e+00 -7.59554744e-01
-3.28822434e-01 -6.27762556e-01 1.83968663e-01 -8.80490065e-01
-1.09878099e+00 1.04916465e+00 8.69689524e-01 -1.32651970e-01
1.20326114e+00 -7.34416991e-02 7.06197500e-01 1.07069254e+00
2.08203435e-01 -1.26474619e+00 1.22417651e-01 7.63513029e-01
7.16297209e-01 -1.95330107e+00 -4.54463735e-02 -1.11585371e-01
-8.25396478e-01 1.10751295e+00 7.50095248e-01 -1.81887701e-01
9.57976997e-01 -1.66239753e-01 9.85114947e-02 -7.27242678e-02
-6.38755322e-01 -2.86693811e-01 4.88729060e-01 6.75907969e-01
-1.38651446e-01 1.65959373e-01 3.65172237e-01 6.72759771e-01
2.66826391e-01 2.39732355e-01 2.80363917e-01 1.02749789e+00
-7.41991878e-01 -3.51014197e-01 5.35893440e-03 5.67039490e-01
-6.18486032e-02 -2.68219352e-01 -1.84770554e-01 4.87116963e-01
-5.97153902e-02 5.84113598e-01 1.20602623e-01 -5.51606774e-01
3.17604214e-01 -7.31105208e-02 1.03812315e-01 -8.44204485e-01
-1.64691091e-01 -3.97970200e-01 -4.68455076e-01 -4.00304168e-01
-3.01735431e-01 -4.12822396e-01 -1.11775494e+00 5.44996075e-02
-3.12545836e-01 -1.17461374e-02 3.19745362e-01 1.08127499e+00
3.77028406e-01 2.83258706e-01 1.18725228e+00 -7.87171364e-01
-1.21003413e+00 -1.03003705e+00 -8.04823577e-01 5.91833651e-01
4.52873528e-01 -5.02666831e-01 -5.11924028e-01 6.96240216e-02] | [9.92615032196045, 2.6768438816070557] |
f6ad9a21-7be2-482d-a610-c5d414c04f2d | levin-tree-search-with-context-models | 2305.16945 | null | https://arxiv.org/abs/2305.16945v2 | https://arxiv.org/pdf/2305.16945v2.pdf | Levin Tree Search with Context Models | Levin Tree Search (LTS) is a search algorithm that makes use of a policy (a probability distribution over actions) and comes with a theoretical guarantee on the number of expansions before reaching a goal node, depending on the quality of the policy. This guarantee can be used as a loss function, which we call the LTS loss, to optimize neural networks representing the policy (LTS+NN). In this work we show that the neural network can be substituted with parameterized context models originating from the online compression literature (LTS+CM). We show that the LTS loss is convex under this new model, which allows for using standard convex optimization tools, and obtain convergence guarantees to the optimal parameters in an online setting for a given set of solution trajectories -- guarantees that cannot be provided for neural networks. The new LTS+CM algorithm compares favorably against LTS+NN on several benchmarks: Sokoban (Boxoban), The Witness, and the 24-Sliding Tile puzzle (STP). The difference is particularly large on STP, where LTS+NN fails to solve most of the test instances while LTS+CM solves each test instance in a fraction of a second. Furthermore, we show that LTS+CM is able to learn a policy that solves the Rubik's cube in only a few hundred expansions, which considerably improves upon previous machine learning techniques. | ['Levi H. S. Lelis', 'Marcus Hutter', 'Laurent Orseau'] | 2023-05-26 | null | null | null | null | ['rubik-s-cube'] | ['graphs'] | [ 1.66659907e-01 3.99805784e-01 -6.25979364e-01 1.25750184e-01
-1.13448071e+00 -7.47020662e-01 2.02027470e-01 3.09775949e-01
-6.04156911e-01 8.67816925e-01 8.37421566e-02 -6.80205584e-01
-5.24592459e-01 -7.77809620e-01 -1.17024875e+00 -9.93649006e-01
-4.66233075e-01 8.46868813e-01 2.17117146e-01 -3.91919538e-02
4.30479735e-01 4.19197887e-01 -1.17517865e+00 2.96710759e-01
5.77996969e-01 1.27551711e+00 3.33329678e-01 8.67027342e-01
-5.52965552e-02 7.72308052e-01 -3.80654722e-01 -3.70693535e-01
7.13034868e-01 -2.85965532e-01 -1.13308680e+00 -3.64081919e-01
2.07877636e-01 -2.05950454e-01 -3.73066097e-01 9.33742702e-01
4.19733346e-01 1.94280744e-01 3.14866841e-01 -1.13107991e+00
-1.54972494e-01 1.15983486e+00 -3.27484101e-01 1.74490467e-01
1.22780010e-01 1.77048251e-01 1.21853387e+00 -1.16555922e-01
8.81638527e-01 9.97383356e-01 7.34827459e-01 3.10236692e-01
-1.37440073e+00 -3.10167760e-01 1.81053445e-01 4.07064795e-01
-1.10684288e+00 -2.07009912e-01 3.41192424e-01 2.07210220e-02
1.21574748e+00 4.09826875e-01 6.53326988e-01 8.74035060e-01
9.32042748e-02 9.77377176e-01 9.08599555e-01 -5.43130457e-01
6.51142716e-01 -2.75808871e-01 -2.94275880e-02 7.59365976e-01
1.20928720e-01 4.71019834e-01 -4.02938068e-01 -3.24806750e-01
4.96606559e-01 -4.66209501e-01 -4.05659020e-01 -6.19218946e-01
-9.09868956e-01 9.64541018e-01 4.52097237e-01 2.59246588e-01
-2.62586921e-01 5.89797616e-01 7.11689115e-01 6.14581585e-01
3.28382641e-01 6.84122860e-01 -6.67969942e-01 -5.96799254e-01
-9.95409250e-01 3.66969883e-01 1.26040685e+00 8.70064676e-01
4.51720685e-01 1.34348869e-02 -1.62343860e-01 5.03883302e-01
-3.49802643e-01 3.28052193e-01 3.79578680e-01 -1.44788718e+00
8.03723276e-01 1.58592924e-01 1.82766095e-01 -7.74395108e-01
-4.41616774e-01 -5.79090416e-01 -4.91553783e-01 2.40696520e-01
6.30131304e-01 4.73510772e-02 -5.23223042e-01 1.87822890e+00
2.50678033e-01 1.77205741e-01 -3.12852450e-02 6.78618431e-01
-5.98702431e-02 8.15300763e-01 -5.10114372e-01 -4.42986995e-01
7.28512645e-01 -1.14018750e+00 -2.92524576e-01 -1.67327583e-01
1.09627795e+00 -2.60455608e-01 1.14959490e+00 7.36833096e-01
-1.55034614e+00 1.47914961e-01 -9.24833179e-01 1.78597614e-01
-1.37275890e-01 -3.47690552e-01 7.81757951e-01 6.21808112e-01
-1.18881214e+00 9.70786572e-01 -1.03553820e+00 -1.12174936e-01
3.90258074e-01 5.05761683e-01 -2.20983505e-01 -2.18827128e-01
-5.24786472e-01 9.49403167e-01 7.02138603e-01 -1.67465180e-01
-1.05334556e+00 -7.56035030e-01 -6.46895885e-01 4.22464639e-01
9.09234822e-01 -5.35871089e-01 1.49265254e+00 -7.97219694e-01
-1.63879013e+00 5.10063231e-01 -9.37251896e-02 -1.05847096e+00
7.12865353e-01 -4.31526639e-03 3.13160390e-01 2.19347313e-01
-2.08310187e-01 4.65839684e-01 4.81035888e-01 -1.04827201e+00
-6.50764465e-01 -2.26516977e-01 2.99081475e-01 -4.10967730e-02
-1.81270570e-01 -2.26663187e-01 -4.74899113e-01 -4.32558179e-01
-1.00466311e-02 -9.14787352e-01 -4.93388563e-01 -1.84390858e-01
-3.24586123e-01 -2.99970657e-01 2.40995541e-01 -4.90762889e-01
1.33606184e+00 -1.79871476e+00 5.47420025e-01 4.26572382e-01
-1.55544654e-01 4.49086353e-02 -5.12761116e-01 6.02626085e-01
4.32544760e-02 2.19062269e-01 -3.80119950e-01 -2.75815815e-01
2.77540892e-01 6.47328675e-01 -4.08587039e-01 5.75280905e-01
-4.80573833e-01 8.38979244e-01 -7.66480029e-01 -5.30349612e-02
-1.92052096e-01 -2.98707068e-01 -9.09309864e-01 -1.89471856e-01
-7.32147455e-01 -1.57808751e-01 -1.44603297e-01 5.36143005e-01
3.68655384e-01 -3.41801584e-01 2.74979651e-01 3.49575460e-01
3.56108621e-02 4.05850857e-01 -1.10563707e+00 1.86431372e+00
-6.64069355e-01 5.12663603e-01 1.47273779e-01 -1.32960463e+00
5.18119574e-01 2.00478896e-01 5.03766596e-01 -6.61053240e-01
2.12952673e-01 4.06223893e-01 -1.18774466e-01 -4.84578252e-01
2.41829008e-01 1.64755329e-01 6.84028026e-03 7.43929684e-01
-9.79907289e-02 1.12280712e-01 5.09771645e-01 3.01707499e-02
1.60694337e+00 -8.58678669e-02 1.33255810e-01 -2.80891508e-01
9.22612473e-02 6.20067939e-02 6.69488609e-01 1.27628231e+00
1.96841881e-01 3.18997890e-01 1.03528166e+00 -7.08770394e-01
-1.31214070e+00 -8.25348973e-01 2.17583343e-01 1.15239894e+00
1.98768778e-03 -4.65724647e-01 -7.62015820e-01 -8.48511636e-01
8.33223090e-02 7.25012958e-01 -6.11779749e-01 -1.07752405e-01
-8.43341410e-01 -4.50164020e-01 5.53695381e-01 4.96680588e-01
2.02728078e-01 -1.10562980e+00 -8.29818368e-01 2.69639015e-01
-1.41260684e-01 -8.98835003e-01 -4.86871123e-01 6.43974125e-01
-1.11020589e+00 -1.24008167e+00 -3.70821148e-01 -5.43431282e-01
5.66808999e-01 -1.53303996e-01 1.15911090e+00 1.04269266e-01
-1.32682726e-01 3.44196767e-01 -2.61553913e-01 -9.78476852e-02
-4.41193283e-01 4.52563584e-01 -1.06137916e-01 -5.23283958e-01
-1.62458882e-01 -8.42647791e-01 -3.46281201e-01 2.31900159e-02
-8.51581395e-01 -7.68705904e-02 4.65558350e-01 1.04647279e+00
5.76897383e-01 2.31172964e-01 2.12275177e-01 -8.20379555e-01
5.66837490e-01 -4.91343498e-01 -1.15104830e+00 3.54580015e-01
-7.56691873e-01 4.92585838e-01 8.19542944e-01 -5.45647144e-01
-3.55804771e-01 -3.04641835e-02 -1.68016199e-02 -6.30033493e-01
3.13385487e-01 6.87267959e-01 1.99811563e-01 -2.69346595e-01
6.12120628e-01 2.62997419e-01 -7.07507730e-02 -5.11991143e-01
3.24545056e-01 1.48519352e-01 6.58859313e-01 -1.15326738e+00
5.31210601e-01 2.69203365e-01 2.60169625e-01 -2.40565553e-01
-7.71529496e-01 -1.71980456e-01 -2.58802939e-02 1.68886736e-01
2.90158093e-01 -2.68736362e-01 -1.23814869e+00 -1.00200042e-01
-1.11839509e+00 -9.64127421e-01 -6.50301337e-01 1.93382204e-01
-1.17549777e+00 2.11978346e-01 -7.42780447e-01 -1.03381395e+00
-2.16623440e-01 -1.09438562e+00 6.82072937e-01 -1.00845527e-02
2.12336369e-02 -8.70127380e-01 4.59533669e-02 1.83206216e-01
4.67336327e-01 2.93425888e-01 1.23775434e+00 -6.34770870e-01
-8.06468368e-01 -3.01943570e-01 7.00905845e-02 2.35768989e-01
-6.32847428e-01 -3.17977726e-01 -4.34870869e-01 -6.51223898e-01
9.76046771e-02 -5.00100791e-01 9.31757689e-01 6.13193035e-01
1.68856204e+00 -9.29306209e-01 -1.99216619e-01 9.85755086e-01
1.64696109e+00 2.97821492e-01 5.37306011e-01 7.88316905e-01
2.46203691e-02 2.46473387e-01 4.44988847e-01 5.70311844e-01
4.15852182e-02 7.07371712e-01 7.77607858e-01 5.78220487e-01
3.38306904e-01 -2.54520804e-01 3.84718299e-01 5.55108190e-01
-7.00660422e-02 -2.46687591e-01 -7.98199415e-01 5.79994440e-01
-2.07859468e+00 -9.01037753e-01 4.09790337e-01 2.45802188e+00
9.45329130e-01 3.02310020e-01 3.22864443e-01 2.26590067e-01
3.43343943e-01 1.33917853e-01 -7.15387225e-01 -8.28677595e-01
-3.29059139e-02 6.17090166e-01 1.06467462e+00 6.83408499e-01
-7.39722729e-01 6.90859616e-01 7.04120016e+00 1.17206430e+00
-9.36941922e-01 2.12356672e-01 6.72069430e-01 -4.78812367e-01
-2.11018726e-01 1.63516045e-01 -5.11455595e-01 5.77778161e-01
1.13339472e+00 -2.17559308e-01 1.21779561e+00 1.10216236e+00
-5.66016957e-02 -1.46956429e-01 -1.31767738e+00 8.15758824e-01
-2.75476515e-01 -1.73619854e+00 -4.60576594e-01 3.01132530e-01
8.37324142e-01 1.45070195e-01 1.78048849e-01 4.14318770e-01
7.27623165e-01 -1.14935017e+00 7.03256309e-01 8.97318274e-02
6.00713015e-01 -1.02356112e+00 6.31290376e-01 7.57282495e-01
-7.70000339e-01 -7.31241584e-01 -5.58930635e-01 5.52640334e-02
6.61125407e-02 3.07069689e-01 -7.39028633e-01 5.14975488e-01
6.27114952e-01 4.37346771e-02 -1.35946274e-02 1.33320904e+00
-1.17050834e-01 7.95678675e-01 -7.69997656e-01 -6.25158250e-02
6.27678752e-01 -1.06920958e-01 6.60748601e-01 8.73037994e-01
3.85310441e-01 -4.22021188e-02 1.31007314e-01 7.77037382e-01
-1.62690461e-01 -1.18434504e-01 -3.78920138e-01 -2.83423644e-02
5.94370663e-01 7.74954140e-01 -5.43667018e-01 1.96417589e-02
5.41286804e-02 7.12516189e-01 6.73461974e-01 1.72684729e-01
-7.83412755e-01 -1.36955351e-01 2.47054949e-01 -5.24165891e-02
6.64044499e-01 -1.30015805e-01 -4.22117263e-01 -7.75087118e-01
2.80520469e-01 -1.13134360e+00 6.03205323e-01 -4.44376677e-01
-9.65104401e-01 3.71193916e-01 4.95839007e-02 -8.16986740e-01
-3.95703524e-01 -6.15162790e-01 -5.96281409e-01 4.10310745e-01
-1.22186661e+00 -6.24243140e-01 1.82152569e-01 2.27481693e-01
4.21587735e-01 -1.97112873e-01 7.28962421e-01 -1.19841108e-02
-4.96619493e-01 8.18921149e-01 3.68894607e-01 -3.05309892e-01
5.25787771e-02 -1.40113842e+00 2.08087385e-01 6.76976681e-01
2.84119487e-01 1.98302284e-01 8.80032718e-01 -2.57289916e-01
-1.60588157e+00 -5.94678164e-01 6.74996018e-01 5.45452312e-02
5.80942035e-01 -7.43434131e-02 -6.94933593e-01 7.50908613e-01
1.18113488e-01 1.41779706e-01 1.97366714e-01 2.56498039e-01
-3.97360921e-01 -1.62305906e-01 -1.12762737e+00 4.90677834e-01
1.19075596e+00 -1.84020698e-01 -2.76731521e-01 6.48819387e-01
8.21274340e-01 -7.58487880e-01 -7.38748670e-01 2.88453370e-01
4.90276217e-01 -1.10956264e+00 8.96426022e-01 -9.01025355e-01
4.73450720e-01 1.85068280e-01 -2.53734469e-01 -1.25110102e+00
-1.81111738e-01 -1.08287513e+00 -5.89838982e-01 6.12897873e-01
3.36795926e-01 -7.20042527e-01 1.04286981e+00 5.44440210e-01
-1.70166433e-01 -1.38345420e+00 -1.45426583e+00 -1.26223028e+00
3.07894498e-01 -6.17774665e-01 7.10912764e-01 6.22307777e-01
8.15393478e-02 -2.92005658e-01 -3.76059353e-01 4.66124155e-03
4.61241931e-01 2.23073363e-01 5.80630362e-01 -7.10721850e-01
-1.01282609e+00 -8.77045274e-01 -1.09782442e-01 -1.23843861e+00
2.66834557e-01 -1.10188818e+00 -2.06884563e-01 -1.23172319e+00
7.26728812e-02 -6.91654205e-01 -3.58766228e-01 5.83094239e-01
4.16697443e-01 -3.38778406e-01 3.52317840e-01 -7.88302049e-02
-6.73713565e-01 4.40766871e-01 1.10121334e+00 -5.50448745e-02
-1.89385608e-01 2.08120719e-01 -4.88301903e-01 6.00124061e-01
7.59314537e-01 -7.11474121e-01 -2.60518998e-01 -5.54020882e-01
6.03141010e-01 6.14925563e-01 9.03763548e-02 -9.34161365e-01
5.07033706e-01 -2.81029880e-01 -2.09912255e-01 -4.74601328e-01
1.06529944e-01 -8.36317003e-01 9.17048603e-02 8.13413084e-01
-6.92328453e-01 2.22499698e-01 1.13123663e-01 5.82966208e-01
1.11276388e-01 -6.46161973e-01 5.87693214e-01 -1.45635325e-02
-9.85301584e-02 2.79969603e-01 -6.94575235e-02 2.27755278e-01
9.24262285e-01 -1.06494389e-01 -4.09694761e-01 -4.98216271e-01
-6.75752878e-01 5.47101855e-01 3.74384552e-01 -3.04997470e-02
3.52636665e-01 -1.25298321e+00 -4.93021518e-01 2.11051460e-02
-3.32797796e-01 1.47351131e-01 9.77779701e-02 1.05553424e+00
-7.30518281e-01 5.43324769e-01 1.72304630e-01 -3.14796984e-01
-1.04271173e+00 8.24735641e-01 4.07012224e-01 -9.35116529e-01
-7.50159800e-01 7.40069807e-01 -3.59987408e-01 -3.12096238e-01
6.97673678e-01 -4.68967557e-01 6.06397927e-01 -3.87424350e-01
3.55619550e-01 5.71334600e-01 1.45825080e-03 3.03728610e-01
1.86747890e-02 1.77709073e-01 1.22535318e-01 -2.35199273e-01
1.62467134e+00 3.43328446e-01 -1.05249189e-01 8.24834928e-02
1.12883961e+00 -1.49821207e-01 -1.21630275e+00 -2.46485204e-01
1.73916847e-01 -5.40661514e-01 -6.76207095e-02 -9.14579451e-01
-1.12879384e+00 5.84660828e-01 3.25052619e-01 3.57437253e-01
1.20872557e+00 -1.39099792e-01 1.00221777e+00 8.77785981e-01
7.87746072e-01 -1.24727225e+00 -1.44170269e-01 6.52414799e-01
8.95414591e-01 -7.11580932e-01 -1.48287684e-01 2.29400005e-02
-2.49604598e-01 1.25144207e+00 1.54660732e-01 -3.82309079e-01
2.85820305e-01 5.82136035e-01 -6.68409705e-01 1.67962313e-01
-1.25055337e+00 -3.91604332e-03 -1.96562037e-01 3.14468384e-01
-4.07792181e-01 3.48257460e-02 -3.64894718e-01 7.09762514e-01
-4.24615800e-01 -5.81699200e-02 3.17014039e-01 9.08375382e-01
-5.84411979e-01 -1.27369189e+00 -2.14712694e-01 3.93121243e-01
-4.67037290e-01 -8.30618478e-03 -3.23328041e-02 8.60228062e-01
-7.97935724e-02 5.95144510e-01 -4.94813174e-02 -3.82978082e-01
-2.30539516e-02 8.73373672e-02 8.52176070e-01 -2.25266054e-01
-6.94777906e-01 -8.53269547e-02 2.00477347e-01 -1.04755795e+00
5.70065081e-02 -5.31763077e-01 -9.82083201e-01 -7.19090700e-01
-2.49500006e-01 2.48094678e-01 6.09562576e-01 1.00535154e+00
2.15355664e-01 2.00853467e-01 6.37844145e-01 -6.84718490e-01
-1.22914040e+00 -5.63460052e-01 -4.21584964e-01 -4.55325581e-02
1.29658714e-01 -3.38565499e-01 -5.40087819e-01 -5.02827585e-01] | [5.088741302490234, 3.00252628326416] |
1ecf34b8-06f2-445c-bd33-3515a30be60b | a-tree-transducer-model-for-grammatical-error | null | null | https://aclanthology.org/W13-3606 | https://aclanthology.org/W13-3606.pdf | A Tree Transducer Model for Grammatical Error Correction | null | ['Jan Buys', 'Brink van der Merwe'] | 2013-08-01 | null | null | null | ws-2013-8 | ['grammatical-error-detection'] | ['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] | [-7.478485584259033, 3.7403757572174072] |
b8a86210-f90b-49e2-92d0-4be6ed865ff8 | generative-ai-for-programming-education | 2306.17156 | null | https://arxiv.org/abs/2306.17156v2 | https://arxiv.org/pdf/2306.17156v2.pdf | Generative AI for Programming Education: Benchmarking ChatGPT, GPT-4, and Human Tutors | Generative AI and large language models hold great promise in enhancing computing education by powering next-generation educational technologies for introductory programming. Recent works have studied these models for different scenarios relevant to programming education; however, these works are limited for several reasons, as they typically consider already outdated models or only specific scenario(s). Consequently, there is a lack of a systematic study that benchmarks state-of-the-art models for a comprehensive set of programming education scenarios. In our work, we systematically evaluate two models, ChatGPT (based on GPT-3.5) and GPT-4, and compare their performance with human tutors for a variety of scenarios. We evaluate using five introductory Python programming problems and real-world buggy programs from an online platform, and assess performance using expert-based annotations. Our results show that GPT-4 drastically outperforms ChatGPT (based on GPT-3.5) and comes close to human tutors' performance for several scenarios. These results also highlight settings where GPT-4 still struggles, providing exciting future directions on developing techniques to improve the performance of these models. | ['Gustavo Soares', 'Adish Singla', 'Rupak Majumdar', 'Tobias Kohn', 'Sumit Gulwani', 'José Cambronero', 'Victor-Alexandru Pădurean', 'Tung Phung'] | 2023-06-29 | null | null | null | null | ['benchmarking', 'benchmarking'] | ['miscellaneous', 'robots'] | [-3.69277596e-01 1.66081652e-01 1.55051365e-01 -2.87573010e-01
-6.81404054e-01 -8.17755520e-01 3.78164232e-01 4.48427379e-01
-2.18518674e-01 4.29265797e-01 -2.17556000e-01 -8.28960836e-01
-1.16760932e-01 -1.05105388e+00 -9.31851029e-01 -1.59689620e-01
-2.82935381e-01 6.11750782e-01 7.19097018e-01 -6.04137301e-01
4.50549513e-01 7.20227091e-03 -1.92502403e+00 4.16953266e-01
1.50895011e+00 -1.71763688e-01 1.64885208e-01 8.72541785e-01
-3.53107303e-01 1.01445401e+00 -8.86953354e-01 -5.42739630e-01
-3.53364795e-02 -3.61218542e-01 -9.79149163e-01 -7.97715008e-01
7.86789060e-01 -3.81181277e-02 1.99866489e-01 1.16273093e+00
6.71335697e-01 5.13642617e-02 2.84316279e-02 -1.44008529e+00
-6.66895986e-01 1.09715152e+00 -1.79774851e-01 8.80489722e-02
7.75043786e-01 2.35441789e-01 7.98164845e-01 -4.25090820e-01
6.68449104e-01 1.26175582e+00 1.08021808e+00 8.44443142e-01
-1.25299513e+00 -4.83807594e-01 2.25278571e-01 5.35933189e-02
-1.06591880e+00 1.60780653e-01 4.39644426e-01 -6.90360367e-01
1.10239697e+00 2.52547979e-01 1.27640343e+00 8.71511221e-01
4.59643453e-02 9.67737079e-01 1.24812794e+00 -8.10325146e-01
1.56823471e-01 3.68069738e-01 4.99016076e-01 1.07507956e+00
1.15859672e-01 -1.09148972e-01 -5.10244489e-01 -2.79976249e-01
6.64495766e-01 -6.44651949e-01 -2.76098669e-01 -3.92535955e-01
-1.05636978e+00 7.86428273e-01 -5.12679806e-03 5.09621561e-01
2.07908556e-01 4.16075289e-01 3.45341712e-01 6.34557724e-01
8.91499817e-02 7.94311404e-01 -4.16890264e-01 -8.61458421e-01
-8.66589725e-01 9.20158863e-01 1.46821272e+00 1.43298364e+00
3.78448486e-01 7.67360330e-02 3.91766131e-02 6.71328366e-01
4.57587302e-01 -5.13250567e-02 5.12832046e-01 -9.67996478e-01
4.71401900e-01 8.96008551e-01 -1.71338916e-01 -6.33916438e-01
-1.65033683e-01 -2.61507511e-01 1.06322445e-01 4.53060508e-01
7.56256104e-01 -2.69035250e-01 -6.23496294e-01 1.54848182e+00
1.89173639e-01 3.78307551e-01 2.57176571e-02 6.40579104e-01
1.38645434e+00 7.03866959e-01 2.56042868e-01 5.34921467e-01
1.30048943e+00 -1.32191145e+00 -1.12367719e-01 -1.65953308e-01
1.10982263e+00 -7.31616318e-01 1.41817534e+00 6.99100435e-01
-1.46922398e+00 -4.63086128e-01 -8.03812861e-01 1.13392193e-02
-3.66771042e-01 -2.19723955e-01 7.74291098e-01 1.23397839e+00
-1.42928493e+00 7.12880611e-01 -9.83816803e-01 -4.88169342e-01
3.93402874e-02 2.22017765e-01 1.23874396e-01 -1.91625997e-01
-9.89237785e-01 1.00404775e+00 2.29430079e-01 -5.74190140e-01
-9.23328638e-01 -1.46918797e+00 -7.21512318e-01 3.27392250e-01
2.56204486e-01 -6.07700825e-01 1.79986298e+00 -4.05364692e-01
-1.75441909e+00 9.99751568e-01 3.63737971e-01 -2.23638862e-01
8.32263470e-01 -8.77583623e-02 2.01663151e-01 -1.84902504e-01
1.90367177e-02 6.58434272e-01 -1.00754373e-01 -1.20886409e+00
-4.54728007e-01 1.73440829e-01 7.84785330e-01 1.87459379e-01
-2.56148517e-01 3.98562729e-01 -2.87585735e-01 -4.16563630e-01
-6.20695353e-02 -8.59155059e-01 -3.44553888e-01 -2.90543675e-01
9.85002741e-02 -5.70574284e-01 4.72526282e-01 -2.80686200e-01
1.27424371e+00 -1.74209929e+00 2.10376620e-01 1.69308316e-02
3.28292102e-01 4.03179497e-01 -1.41788647e-01 5.97290814e-01
2.67998129e-02 5.26636541e-01 -1.14831515e-01 -5.40161133e-02
5.68018258e-01 1.70249462e-01 -1.77361906e-01 -2.28283674e-01
-1.22917064e-01 8.10978711e-01 -1.27501190e+00 -6.14453852e-01
2.14790732e-01 3.35176855e-01 -1.11165130e+00 2.80737251e-01
-7.09587574e-01 3.23513478e-01 -4.90890652e-01 6.05681181e-01
2.68681139e-01 6.63094223e-02 8.41966122e-02 8.65379155e-01
-4.80335385e-01 4.47409898e-01 -1.22871017e+00 1.74804223e+00
-6.65436447e-01 6.85926497e-01 -1.19035356e-01 -7.93122232e-01
8.64557326e-01 4.14615989e-01 2.41197906e-02 -3.33848715e-01
-1.81613117e-01 3.06410134e-01 3.94645572e-01 -7.42746711e-01
4.57880646e-01 1.68365791e-01 -1.77926213e-01 7.01892674e-01
2.94096023e-01 -8.27357888e-01 6.95393741e-01 2.86077738e-01
1.39461172e+00 6.01439774e-01 -6.24487773e-02 -6.13373876e-01
3.00054222e-01 1.15253776e-01 3.08751673e-01 1.22223294e+00
-1.22596473e-01 4.82519895e-01 9.62691486e-01 -4.65831280e-01
-9.22387123e-01 -6.18300974e-01 -1.97060201e-02 1.54495001e+00
-2.40051955e-01 -1.00128460e+00 -1.14880896e+00 -6.40400469e-01
-1.76867113e-01 1.06568813e+00 -3.35949421e-01 1.97875813e-01
-8.64957213e-01 -6.87707901e-01 8.65740001e-01 3.77707839e-01
4.47451562e-01 -1.10808361e+00 -9.99640584e-01 3.90318692e-01
4.98154573e-02 -7.93566823e-01 1.76462203e-01 -1.68796163e-02
-8.50221276e-01 -9.58128631e-01 -5.53544641e-01 -1.04507875e+00
5.66885412e-01 2.05691457e-02 1.77231908e+00 9.30616617e-01
-1.97779506e-01 7.91783392e-01 -4.81716514e-01 -6.23408377e-01
-9.33408797e-01 2.55069882e-01 -5.22939980e-01 -1.20819962e+00
3.88944000e-01 -7.79881060e-01 -7.00958371e-02 2.34657768e-02
-5.67191541e-01 2.65426546e-01 1.67007744e-01 7.77903378e-01
-1.67171478e-01 5.23223728e-02 1.99193254e-01 -1.24956834e+00
8.23951721e-01 -4.07281011e-01 -1.04173934e+00 4.65519845e-01
-5.53387344e-01 -1.69197414e-02 6.22472286e-01 -4.15851444e-01
-1.07833731e+00 -4.92109597e-01 -4.33854669e-01 6.67856187e-02
-2.50635445e-01 8.70192289e-01 9.95289907e-02 -6.76813722e-01
8.85669351e-01 8.30919668e-02 -6.43488944e-01 -3.33127022e-01
2.04099175e-02 1.21179156e-01 2.58995831e-01 -1.66727686e+00
6.40195131e-01 -6.22569859e-01 -3.14975172e-01 -6.89727187e-01
-5.17734706e-01 6.53053168e-03 -4.68301177e-01 -2.88770467e-01
3.41916531e-01 -8.73679459e-01 -8.37900519e-01 3.86546314e-01
-1.22170770e+00 -8.47817540e-01 -1.53956622e-01 2.61730641e-01
-4.39810723e-01 2.40539283e-01 -9.69556034e-01 -5.82881749e-01
1.38548273e-03 -1.58571315e+00 7.61773348e-01 5.09165525e-01
-3.41207594e-01 -1.30635202e+00 4.38224941e-01 4.25484687e-01
5.34809351e-01 1.66175738e-01 1.33831942e+00 -7.20059514e-01
-8.31559062e-01 -1.12552285e-01 2.52690792e-01 -8.34863484e-02
-7.97078311e-01 6.23077929e-01 -8.57225001e-01 6.55595735e-02
-2.80610204e-01 -3.14894199e-01 3.01498413e-01 -7.89004788e-02
1.06382263e+00 -1.07883543e-01 -1.73085243e-01 4.57905263e-01
1.45544899e+00 1.33650705e-01 5.76740444e-01 7.72822738e-01
6.58689201e-01 5.32808602e-01 3.11362505e-01 1.14443727e-01
6.85362637e-01 4.68352824e-01 2.52421498e-01 5.52790403e-01
-1.58347632e-03 -3.94964844e-01 4.42301184e-01 1.07805789e+00
-3.97215277e-01 -1.56002149e-01 -1.71695077e+00 7.66962886e-01
-1.85107529e+00 -8.07421803e-01 -5.37060380e-01 1.98337638e+00
1.17479718e+00 2.05352664e-01 8.22093338e-02 -1.35269552e-01
4.83950645e-01 -2.40702868e-01 7.57743493e-02 -6.81062996e-01
3.83987546e-01 8.17365468e-01 -3.94151628e-01 5.00791967e-01
-5.62438309e-01 1.07471526e+00 6.71511364e+00 5.90444505e-01
-8.73074412e-01 1.40408218e-01 5.25704436e-02 3.56639534e-01
-5.54840505e-01 1.96829841e-01 -8.42438281e-01 4.29904789e-01
1.30084729e+00 -5.86754322e-01 3.37473452e-01 1.23040080e+00
-3.62244934e-01 -1.02093503e-01 -1.42098463e+00 3.77744496e-01
-6.09136559e-02 -1.34302902e+00 -2.58465797e-01 -2.48440251e-01
1.17269981e+00 -1.14826813e-01 -6.95651323e-02 1.07207930e+00
9.66615081e-01 -9.60151494e-01 8.44836891e-01 1.54651657e-01
-3.94007005e-02 -6.12852573e-01 7.61250138e-01 4.93041307e-01
-8.68069112e-01 1.21885315e-01 -6.26373589e-02 -4.04525310e-01
-3.98225129e-01 1.71569467e-01 -8.39719713e-01 3.75734568e-01
1.02023542e+00 4.13504452e-01 -8.76386046e-01 1.55409884e+00
-7.72568941e-01 8.77853274e-01 -2.47151107e-01 -4.81140107e-01
1.38810769e-01 6.59625679e-02 5.15021741e-01 1.34476674e+00
4.83263314e-01 3.40254754e-01 3.89182985e-01 1.31339788e+00
1.10212930e-01 6.62396327e-02 -4.93373424e-01 -7.19742328e-02
7.24761546e-01 1.26083553e+00 -7.44645596e-01 -4.45050329e-01
-5.59193790e-01 3.10152233e-01 3.44594151e-01 6.03875779e-02
-8.89672577e-01 -6.07860744e-01 4.14982140e-01 6.91346228e-02
-4.14870158e-02 -3.19309950e-01 -3.35648537e-01 -1.22792506e+00
-1.84299752e-01 -1.31619763e+00 1.58812985e-01 -9.63560820e-01
-1.00360930e+00 5.26988268e-01 4.06972140e-01 -8.77495348e-01
-1.16775572e-01 -5.42399228e-01 -1.07865775e+00 6.79953635e-01
-1.19106829e+00 -9.86556411e-01 -6.89717352e-01 3.17225419e-02
4.42934334e-01 -3.34588699e-02 1.10870874e+00 2.80285895e-01
-5.97954273e-01 6.35083854e-01 -1.84230015e-01 -4.76478674e-02
5.39057016e-01 -1.77954149e+00 7.90063739e-01 8.65472376e-01
-1.72674984e-01 9.35064912e-01 1.03443933e+00 -4.66990620e-01
-1.50912762e+00 -6.71031594e-01 8.94758105e-01 -8.34450901e-01
8.45101655e-01 -3.62244636e-01 -1.22250450e+00 9.61769402e-01
3.82943779e-01 -2.48761952e-01 6.31221831e-01 2.61706978e-01
-2.44395196e-01 5.25300562e-01 -1.12387657e+00 7.33739197e-01
9.58720505e-01 -2.06358761e-01 -9.20498312e-01 4.08538759e-01
6.75171494e-01 -1.24674129e+00 -1.27596033e+00 -6.13305196e-02
4.47064579e-01 -1.22570765e+00 7.91941464e-01 -4.89453316e-01
9.86079276e-01 -8.69681016e-02 2.42445543e-01 -1.45144355e+00
1.19154133e-01 -7.37809598e-01 5.47843874e-02 1.34930885e+00
2.28495836e-01 -6.21336401e-01 9.16947484e-01 1.02535760e+00
-5.75789511e-01 -6.78188622e-01 -3.15930843e-01 -5.70082784e-01
6.52262211e-01 -3.90120417e-01 7.87394881e-01 1.29583895e+00
4.46343243e-01 5.58507964e-02 3.78502429e-01 1.75344840e-01
4.26267803e-01 2.78517306e-01 1.07584977e+00 -1.44683981e+00
-4.25842285e-01 -8.33844423e-01 -3.10542494e-01 -7.81975567e-01
2.54939407e-01 -8.66417110e-01 1.16242595e-01 -1.43165159e+00
1.13037050e-01 -8.32873106e-01 3.37347597e-01 7.14402556e-01
-2.94268310e-01 -4.63230275e-02 3.43573689e-01 -2.56809115e-01
-6.53592169e-01 1.85555518e-01 1.00080085e+00 6.84289187e-02
-2.31815442e-01 2.06628209e-03 -5.37350237e-01 1.08365715e+00
6.75780356e-01 -5.37228525e-01 -2.01691270e-01 -6.61195695e-01
5.59540212e-01 5.40140048e-02 3.78818750e-01 -1.39808297e+00
4.57222819e-01 -2.64922321e-01 -1.10524401e-01 -2.52777394e-02
-1.34168684e-01 -5.35414577e-01 -8.30364525e-02 4.68050927e-01
-3.72202933e-01 2.74276525e-01 6.68396413e-01 -1.66258305e-01
-2.17051521e-01 -1.06515038e+00 4.49724972e-01 -6.17582619e-01
-7.98011959e-01 -2.90851623e-01 -6.99423790e-01 3.27751994e-01
1.03607619e+00 -2.07603976e-01 -6.65549576e-01 -1.59051612e-01
-4.61089432e-01 5.33656180e-01 4.68815327e-01 3.33683431e-01
3.07251811e-01 -9.70436692e-01 -6.44377232e-01 1.55789629e-01
2.54913904e-02 2.54326642e-01 1.22556537e-01 5.56872010e-01
-1.23329079e+00 4.44848865e-01 -5.43897331e-01 -8.24319899e-01
-1.59359932e+00 1.78607598e-01 3.18805277e-01 -4.33200359e-01
-5.33355176e-01 1.13826287e+00 -2.70453572e-01 -1.30293858e+00
3.71937215e-01 -7.24373698e-01 -1.40370056e-01 -3.62835944e-01
4.08380359e-01 3.92245889e-01 1.59162670e-01 1.33188084e-01
5.96859418e-02 4.54512954e-01 4.07157540e-02 5.44002578e-02
1.71924925e+00 5.29044628e-01 -2.33000875e-01 4.02046978e-01
2.95802772e-01 2.65499484e-02 -7.70743012e-01 2.77400874e-02
2.85977751e-01 -4.57212925e-01 -2.71488905e-01 -1.01269317e+00
-6.77191794e-01 1.04350245e+00 1.98024392e-01 3.01061958e-01
8.22757840e-01 -2.13384241e-01 1.12881459e-01 5.20584166e-01
8.46020341e-01 -5.75856805e-01 1.47483855e-01 9.60702598e-01
6.32516861e-01 -9.63055015e-01 5.47969863e-02 -4.63209480e-01
-3.75472486e-01 1.35710418e+00 1.37765074e+00 -1.00357398e-01
2.64805466e-01 3.67493302e-01 -2.02513769e-01 -3.06038022e-01
-1.10516334e+00 2.43079230e-01 -8.73826370e-02 5.12244701e-01
1.10983145e+00 -1.11570410e-01 -1.66832253e-01 7.35324562e-01
-8.48567367e-01 1.67547107e-01 1.05168486e+00 1.56426978e+00
-4.53367174e-01 -1.49852204e+00 -6.02424920e-01 -7.61721879e-02
-4.08553034e-01 -3.05006087e-01 -3.47936094e-01 1.24150264e+00
2.62811124e-01 5.15784502e-01 -3.35652769e-01 -1.15082383e-01
4.88377035e-01 3.14795315e-01 1.14970434e+00 -1.16181672e+00
-1.50008500e+00 -5.08318722e-01 1.55201346e-01 -1.99254125e-01
-1.61032230e-01 -5.42571306e-01 -1.07816684e+00 -8.59265029e-01
-2.08905101e-01 4.54621792e-01 6.02566719e-01 6.40078962e-01
1.31699696e-01 7.66815603e-01 -3.56633306e-01 -7.72682488e-01
-4.72559750e-01 -7.91323125e-01 -3.90000790e-02 -2.43666664e-01
-1.87508643e-01 -3.55323136e-01 -9.77317058e-03 -1.83162931e-02] | [9.511303901672363, 7.398095607757568] |
882c38ef-0c3e-455a-923c-ff339e3aae55 | hyperedge2vec-distributed-representations-for | null | null | https://openreview.net/forum?id=rJ5C67-C- | https://openreview.net/pdf?id=rJ5C67-C- | Hyperedge2vec: Distributed Representations for Hyperedges | Data structured in form of overlapping or non-overlapping sets is found in a variety of domains, sometimes explicitly but often subtly. For example, teams, which are of prime importance in social science studies are \enquote{sets of individuals}; \enquote{item sets} in pattern mining are sets; and for various types of analysis in language studies a sentence can be considered as a \enquote{set or bag of words}. Although building models and inference algorithms for structured data has been an important task in the fields of machine learning and statistics, research on \enquote{set-like} data still remains less explored. Relationships between pairs of elements can be modeled as edges in a graph. However, modeling relationships that involve all members of a set, a hyperedge is a more natural representation for the set. In this work, we focus on the problem of embedding hyperedges in a hypergraph (a network of overlapping sets) to a low dimensional vector space. We propose a probabilistic deep-learning based method as well as a tensor-based algebraic model, both of which capture the hypergraph structure in a principled manner without loosing set-level information. Our central focus is to highlight the connection between hypergraphs (topology), tensors (algebra) and probabilistic models. We present a number of interesting baselines, some of which adapt existing node-level embedding models to the hyperedge-level, as well as sequence based language techniques which are adapted for set structured hypergraph topology. The performance is evaluated with a network of social groups and a network of word phrases. Our experiments show that accuracy wise our methods perform similar to those of baselines which are not designed for hypergraphs. Moreover, our tensor based method is quiet efficient as compared to deep-learning based auto-encoder method. We therefore, argue that we have proposed more general methods which are suited for hypergraphs (and therefore also for graphs) while maintaining accuracy and efficiency. | ['Ankit Sharma', 'Jaideep Srivastava', 'Shafiq Joty', 'Himanshu Kharkwal'] | 2018-01-01 | null | null | null | iclr-2018-1 | ['probabilistic-deep-learning'] | ['computer-vision'] | [-1.30010769e-01 2.28632912e-01 -1.25716865e-01 -2.07804680e-01
1.38496071e-01 -7.32923806e-01 7.76934087e-01 5.59768736e-01
-2.64349520e-01 5.55190802e-01 3.45704108e-01 -3.10674667e-01
-7.17608333e-01 -1.46839893e+00 -7.45253980e-01 -6.16204500e-01
-5.69923222e-01 9.08202291e-01 -3.72029021e-02 -4.13134813e-01
2.68577375e-02 4.38931435e-01 -1.36268651e+00 2.83276081e-01
2.79579312e-01 6.10421717e-01 -8.51778984e-02 3.91464561e-01
-4.86003757e-01 5.47449887e-01 -3.55922073e-01 -1.06632090e+00
2.35158816e-01 -4.69150543e-02 -1.13270760e+00 2.70492733e-01
6.55930340e-01 2.91471720e-01 -6.17284417e-01 1.08047473e+00
4.37812507e-02 -9.18295756e-02 7.83095002e-01 -1.36024010e+00
-7.61236370e-01 1.17720687e+00 -5.59974551e-01 9.71699059e-02
3.27839375e-01 -3.18396240e-01 1.79079521e+00 -7.10694015e-01
5.96999943e-01 1.25442088e+00 7.85075188e-01 5.81476018e-02
-1.59588718e+00 -3.36536527e-01 2.86790273e-05 1.90020129e-01
-1.28689027e+00 2.63461500e-01 6.96810663e-01 -7.21645832e-01
7.74239302e-01 3.95244062e-01 7.86154211e-01 9.82158005e-01
3.07640940e-01 3.93492728e-01 1.02676356e+00 -4.48267192e-01
-1.13552129e-02 2.48658404e-01 6.06395006e-01 1.02335882e+00
7.08811581e-01 4.02774923e-02 -5.09417415e-01 -3.37991357e-01
8.41203690e-01 1.89960808e-01 -2.55332023e-01 -6.84130490e-01
-1.45184612e+00 1.11854959e+00 6.78863347e-01 8.70182157e-01
-4.85834390e-01 3.38683635e-01 4.99230444e-01 3.90258372e-01
2.77834564e-01 4.57767606e-01 -2.63130903e-01 1.25718057e-01
-6.81374013e-01 2.90366173e-01 1.26573527e+00 1.06420767e+00
8.95705640e-01 -2.19219923e-01 -4.40883078e-02 7.02595353e-01
3.04619074e-01 1.14472933e-01 5.50214946e-02 -5.35526514e-01
4.03743356e-01 9.88232911e-01 -4.29568380e-01 -1.57006001e+00
-5.25068998e-01 -6.11986101e-01 -1.27378821e+00 -3.11624944e-01
3.85520816e-01 1.48881182e-01 -4.28382307e-01 1.87503827e+00
1.59456149e-01 1.40055090e-01 -2.76272833e-01 6.62618637e-01
8.32047343e-01 4.09322709e-01 -3.08007777e-01 -7.69410729e-02
1.79234338e+00 -5.75370610e-01 -6.26654565e-01 2.15940014e-01
7.09148586e-01 -3.82259160e-01 6.61873877e-01 2.92300612e-01
-9.18926954e-01 -2.45071247e-01 -7.24670947e-01 -1.98723629e-01
-7.36478209e-01 -2.39984915e-01 1.04395521e+00 7.08644867e-01
-1.40201354e+00 7.16842949e-01 -4.42442566e-01 -5.75631976e-01
3.81829590e-01 4.77897465e-01 -7.81069160e-01 1.58192348e-02
-1.35014987e+00 6.95304811e-01 4.15265441e-01 -9.58960578e-02
-3.92507821e-01 -5.59021592e-01 -1.00087690e+00 3.86622757e-01
5.49345315e-01 -9.99082327e-01 5.27790308e-01 -5.64726889e-01
-8.63123059e-01 9.74744678e-01 1.59789190e-01 -6.91302121e-01
-3.29789706e-02 2.28330851e-01 -3.18843544e-01 -3.04156020e-02
6.18110271e-03 2.69647181e-01 5.46621740e-01 -1.19834220e+00
-1.85473651e-01 -4.07931894e-01 6.52598500e-01 1.46843698e-02
-8.06972504e-01 -1.90703273e-02 -2.37959296e-01 -5.19556820e-01
-4.83026728e-02 -9.98528957e-01 -1.46352857e-01 -2.11019889e-01
-6.61752105e-01 -5.53914309e-01 4.17578638e-01 -3.50904435e-01
1.53874671e+00 -1.67120969e+00 6.32338345e-01 4.52007920e-01
9.06568944e-01 -6.36663940e-03 -6.45760028e-03 1.11422193e+00
-4.20075595e-01 5.59579968e-01 -4.39681709e-01 -3.06271732e-01
3.01361501e-01 4.58857000e-01 -2.27559526e-02 4.29188520e-01
1.01699866e-01 8.54409873e-01 -9.07374620e-01 -5.23949265e-01
6.97917789e-02 3.98942143e-01 -5.55807829e-01 -1.31462261e-01
-1.09918803e-01 -1.43618762e-01 -2.29011759e-01 1.32392555e-01
4.02872860e-01 -5.17397702e-01 3.87938023e-01 -4.17485923e-01
5.02155088e-02 2.58311331e-01 -1.33646190e+00 1.48307943e+00
-4.93894964e-01 4.91713792e-01 -1.35509700e-01 -1.27786922e+00
7.69523561e-01 4.18151617e-01 6.70220315e-01 6.18299805e-02
2.27527041e-02 -2.72563677e-02 4.22656536e-01 -3.88749838e-01
5.63989878e-01 -2.34224826e-01 -1.33810356e-01 6.63396001e-01
2.48618096e-01 -3.58939432e-02 4.52479303e-01 6.88713253e-01
1.43932807e+00 -3.06933820e-01 4.36462432e-01 -4.84350473e-01
4.66615170e-01 -3.08541894e-01 2.01224297e-01 4.60672677e-01
3.09890985e-01 2.79786974e-01 9.81880546e-01 -5.01755834e-01
-1.15114725e+00 -8.56401682e-01 -3.17679584e-01 7.80523360e-01
-2.90308297e-01 -1.00269842e+00 -3.27827990e-01 -6.61395252e-01
1.85224235e-01 5.14840901e-01 -8.24424803e-01 9.11682099e-02
-3.87528956e-01 -7.83804417e-01 4.96472120e-01 3.52023304e-01
5.91207333e-02 -7.61733651e-01 -1.94323696e-02 1.60610899e-01
-2.82475445e-02 -1.21513021e+00 -6.00884676e-01 3.57144065e-02
-7.37541795e-01 -1.26390553e+00 -2.58987874e-01 -7.92092323e-01
4.98998433e-01 9.78819802e-02 1.64017403e+00 1.75881520e-01
-1.37010306e-01 6.52322173e-01 -3.76405716e-01 -3.24603677e-01
-1.19258650e-01 -7.22670555e-02 2.95667350e-01 3.95988733e-01
5.33186972e-01 -9.62675095e-01 -1.45413086e-01 1.74702913e-01
-1.29517663e+00 -1.38685301e-01 5.87645829e-01 9.11476493e-01
3.04998070e-01 3.00879270e-01 1.19563930e-01 -1.29167223e+00
7.11670220e-01 -7.34334946e-01 -4.44335848e-01 1.91033810e-01
-5.23063540e-01 3.39203209e-01 7.35185623e-01 -2.97302067e-01
-2.93409646e-01 -4.59661067e-01 2.52968639e-01 -4.57617283e-01
8.13540816e-02 9.66903389e-01 -6.04131669e-02 1.08719105e-02
4.51501280e-01 2.18947619e-01 5.15871085e-02 -3.28476548e-01
5.26556432e-01 3.68390173e-01 -1.36850506e-01 -6.80543363e-01
1.00597656e+00 4.79673266e-01 6.37397170e-01 -9.31758761e-01
-5.34078896e-01 -6.04424298e-01 -8.90178025e-01 1.16811525e-02
6.78124726e-01 -5.72561264e-01 -1.00444496e+00 -1.18600234e-01
-1.27787042e+00 2.92202264e-01 -2.43532836e-01 4.75238174e-01
-3.25699151e-01 6.21100903e-01 -7.29177475e-01 -5.71816027e-01
-5.59073351e-02 -8.38235438e-01 1.16380966e+00 -4.13601667e-01
-2.16857567e-01 -1.65119958e+00 3.67238581e-01 2.90362775e-01
6.92814291e-02 3.31965387e-01 1.37679279e+00 -9.78704214e-01
-6.60753667e-01 -2.24068269e-01 -2.01483712e-01 2.00049475e-01
4.30768244e-02 -1.89280137e-01 -4.22398895e-01 -1.85164079e-01
-2.22216472e-01 -1.32117972e-01 8.05617929e-01 1.71188012e-01
1.05999184e+00 -5.70062637e-01 -4.41566318e-01 3.42558146e-01
1.70457339e+00 -4.22419965e-01 5.13379216e-01 4.94784266e-02
1.14863801e+00 9.09006596e-01 -2.10479364e-01 3.55453938e-01
6.42721534e-01 8.51697028e-01 5.12220979e-01 1.22968279e-01
1.17760144e-01 -1.56452388e-01 2.51236975e-01 1.30413330e+00
-2.91219771e-01 -3.68924111e-01 -8.05909395e-01 6.80279434e-01
-1.67109919e+00 -1.18906379e+00 -7.50813663e-01 2.23156977e+00
8.44181776e-01 -1.02642916e-01 3.65419835e-01 2.85569668e-01
7.62335479e-01 2.51560628e-01 1.63612440e-01 -6.08796060e-01
-1.96450070e-01 4.48590338e-01 4.55280364e-01 4.60722744e-01
-8.83562207e-01 6.28588498e-01 5.26814413e+00 6.41865730e-01
-4.99330431e-01 -4.04687710e-02 2.07892805e-01 2.06106842e-01
-6.26107693e-01 2.41308689e-01 -6.71915054e-01 3.54339153e-01
9.44021881e-01 -1.59607947e-01 5.09475231e-01 3.98368180e-01
-1.03299342e-01 4.40443814e-01 -1.69446409e+00 1.11109245e+00
1.32612482e-01 -1.35795712e+00 1.98888555e-01 6.65644765e-01
6.16188765e-01 -2.30518460e-01 -5.85560203e-02 3.74632031e-01
5.11128068e-01 -1.26209128e+00 3.31345350e-01 6.05943620e-01
3.22260588e-01 -4.61407244e-01 7.98226893e-01 1.89581171e-01
-1.32150531e+00 2.06301928e-01 -4.31087434e-01 -1.66601852e-01
3.82685028e-02 9.72475708e-01 -7.51497865e-01 1.19128549e+00
5.11012137e-01 8.17872763e-01 -3.70124578e-01 9.52739060e-01
5.30820675e-02 6.63531423e-01 -3.17498237e-01 -3.20502639e-01
3.42450261e-01 -6.00120306e-01 8.01759422e-01 1.22749531e+00
9.20984522e-02 -2.84238886e-02 3.27062726e-01 1.22242081e+00
-3.35992306e-01 1.89884171e-01 -1.17421424e+00 -3.95217597e-01
2.62493759e-01 1.49163437e+00 -4.83368099e-01 -1.35531589e-01
-5.80862880e-01 5.55079579e-01 4.87853676e-01 1.66774198e-01
-6.81705475e-01 -3.02734554e-01 5.57939351e-01 3.48061115e-01
2.54547417e-01 -4.91677970e-01 -7.03244060e-02 -1.17943943e+00
1.27060199e-02 -8.41138542e-01 4.86405581e-01 -2.91616499e-01
-1.83472168e+00 6.31008089e-01 2.78519958e-01 -1.05107343e+00
-1.21100619e-01 -9.17761326e-01 -2.88437277e-01 7.46646464e-01
-1.09861851e+00 -1.15185249e+00 8.38895738e-02 6.13732219e-01
6.27525300e-02 -1.54235274e-01 8.77196789e-01 2.47583702e-01
-4.17033553e-01 4.19072330e-01 -3.00145037e-02 4.22514528e-01
1.57992825e-01 -1.82120287e+00 1.93864673e-01 5.42631686e-01
9.39998090e-01 9.17037308e-01 5.15906870e-01 -4.98747349e-01
-1.72377658e+00 -1.02716827e+00 1.25008011e+00 -7.60413468e-01
1.09103465e+00 -7.44199634e-01 -9.49483216e-01 9.64182854e-01
3.94318759e-01 -4.21625972e-02 8.28847885e-01 7.26081908e-01
-5.33279240e-01 -1.06516411e-03 -7.00907528e-01 6.21998012e-01
1.34459949e+00 -5.33527792e-01 -5.65928042e-01 7.80223310e-01
7.68383026e-01 8.71277452e-02 -1.26061702e+00 8.30566436e-02
2.28754297e-01 -1.06352150e+00 9.23570037e-01 -7.82415867e-01
4.69323546e-01 -1.33364424e-01 -1.33804545e-01 -1.54141939e+00
-7.21162856e-01 -5.40342391e-01 -2.25800022e-01 1.16656709e+00
2.64650643e-01 -6.43144429e-01 7.74622500e-01 2.13150516e-01
1.10353552e-01 -6.88686371e-01 -7.39098489e-01 -8.92839551e-01
8.05513933e-02 -4.48002219e-01 6.64212048e-01 1.24663031e+00
2.79332399e-01 7.83858120e-01 -2.49509975e-01 9.90847498e-02
7.80244768e-01 1.74159959e-01 6.35473728e-01 -1.59002316e+00
-4.97296512e-01 -7.29571760e-01 -9.66496050e-01 -6.01512194e-01
3.12730759e-01 -1.43483794e+00 -5.34471214e-01 -1.69027758e+00
3.27919126e-01 -3.84897232e-01 -6.21219762e-02 4.36731279e-01
2.04249173e-01 -6.51978329e-02 3.17636169e-02 3.10303755e-02
-3.55272293e-01 5.89284539e-01 1.08320761e+00 -2.38742769e-01
1.42832875e-01 -6.17814995e-02 -6.20049298e-01 4.90712017e-01
3.72852832e-01 -4.46937829e-01 -3.39673460e-01 -3.44596773e-01
8.21523190e-01 5.06869256e-02 4.77530926e-01 -5.74206769e-01
3.13946664e-01 1.12047032e-01 -2.44457379e-01 -1.96335554e-01
3.94971937e-01 -9.92098093e-01 2.16129586e-01 2.70361930e-01
-3.13108116e-01 3.30567956e-01 -6.04869500e-02 8.55709195e-01
-2.69456893e-01 -3.13695103e-01 2.83796906e-01 -2.27060750e-01
-1.89226210e-01 6.93995059e-01 -2.94887703e-02 -4.82950918e-02
9.00207460e-01 -1.56017557e-01 -2.00133219e-01 -4.52646136e-01
-6.77965403e-01 1.52007833e-01 2.87910223e-01 3.15303862e-01
5.44814229e-01 -1.36113191e+00 -9.10830498e-01 -2.08378658e-01
1.85694337e-01 -1.18689388e-01 5.07141352e-02 9.91748273e-01
-4.18300539e-01 6.50566578e-01 2.71719396e-02 -4.64843303e-01
-1.27367592e+00 8.43306720e-01 1.50149688e-01 -7.00057089e-01
-5.13120174e-01 8.53250444e-01 3.25168222e-01 -6.65308833e-01
1.51086718e-01 -4.72207367e-01 -5.75686693e-01 1.61431268e-01
1.68079868e-01 2.43978471e-01 -2.02198140e-02 -8.08073878e-01
-3.97703528e-01 4.76711452e-01 -8.44222084e-02 1.26862913e-01
1.46870506e+00 1.01040646e-01 -9.18420315e-01 8.35038364e-01
1.27792072e+00 9.66891572e-02 -2.11041167e-01 -5.91728449e-01
1.59720406e-01 -2.23151863e-01 -5.00896610e-02 -3.65652204e-01
-1.01754487e+00 8.66567612e-01 7.70533830e-02 8.79671872e-01
7.06196368e-01 2.27174073e-01 4.28457528e-01 3.85261506e-01
5.37755966e-01 -4.27703947e-01 1.39130890e-01 5.33673465e-01
1.04901600e+00 -9.39262569e-01 9.17712077e-02 -8.25938821e-01
-4.88511920e-01 1.22840333e+00 1.40012488e-01 -3.23954642e-01
1.03039932e+00 1.05865709e-01 -7.58127630e-01 -7.51166344e-01
-9.37719524e-01 -4.67576057e-01 5.41842759e-01 4.48254496e-01
5.96722186e-01 2.69495100e-01 -5.84899724e-01 4.05929059e-01
-4.14979041e-01 -3.19921046e-01 7.15757370e-01 5.33127189e-01
3.25965881e-03 -1.42942989e+00 -3.25139731e-01 7.91339636e-01
-4.28422719e-01 -4.70993876e-01 -5.44908822e-01 8.90775442e-01
2.40795955e-01 7.24950910e-01 1.74144804e-01 -6.14346206e-01
1.10171363e-01 -8.87195542e-02 6.22906029e-01 -1.05154812e+00
-7.56202042e-01 -3.86570096e-01 3.30472797e-01 -2.44276017e-01
-6.53002560e-01 -7.43618429e-01 -8.74621093e-01 -5.04285991e-01
-4.09047753e-01 2.52200961e-01 4.46471035e-01 9.08418417e-01
2.33781770e-01 4.51782823e-01 5.11850297e-01 -3.88267130e-01
-4.54584152e-01 -1.05218410e+00 -1.16916680e+00 6.41703308e-01
-7.96916708e-02 -8.00898314e-01 -2.73086101e-01 -8.64211023e-02] | [7.329822540283203, 6.499244213104248] |
298f98b8-055c-4f9f-aef6-d684c72f3b62 | dynamical-softassign-and-adaptive-parameter | 2208.08233 | null | https://arxiv.org/abs/2208.08233v2 | https://arxiv.org/pdf/2208.08233v2.pdf | Dynamical softassign and adaptive parameter tuning for graph matching | This paper studies a framework, projected fixed-point method, for graph matching. The framework contains a class of popular graph matching algorithms, including graduated assignment (GA), integer projected fixed-point method (IPFP) and doubly stochastic projected fixed-point method (DSPFP). We propose an adaptive strategy to tune the step size parameter in this framework. Such a strategy improves these algorithms in efficiency and accuracy. Further, it guarantees the convergence of the underlying algorithms. Some preliminary analysis based on distance geometry seems to support that the optimal step size parameter has a high probability of 1 when graphs are fully connected. Secondly, it is observed that a popular projection method, softassign, is sensitive to graphs' cardinality(size). We proposed a dynamical softassign algorithm that is robust to graphs' cardinality. Combining the adaptive step size and the dynamical softassign, we propose a novel graph matching algorithm: the adaptive projected fixed-point method with dynamical softassign. Various experiments demonstrate that the proposed algorithm is significantly faster than several other state-of-art algorithms with no loss of accuracy. | ['Shengxin Zhu', 'Qiang Niu', 'Binrui Shen'] | 2022-08-17 | null | null | null | null | ['graph-matching'] | ['graphs'] | [ 2.08509430e-01 1.52064875e-01 -2.39981145e-01 3.94428708e-02
-6.67548478e-01 -4.96439397e-01 6.05121553e-01 1.52923957e-01
-1.13839246e-01 6.41240478e-01 -3.18788379e-01 -3.00156534e-01
-8.50467324e-01 -1.10806429e+00 -9.11411941e-01 -9.00973797e-01
-1.71055317e-01 8.05916131e-01 6.05661750e-01 -3.39967996e-01
6.63159668e-01 6.34634435e-01 -1.63128197e+00 -5.54346740e-01
1.06123102e+00 4.73138750e-01 1.64866969e-01 8.10324073e-01
-1.20440714e-01 -1.39971122e-01 -2.35363394e-01 -6.48700237e-01
7.94257343e-01 -3.21567088e-01 -7.90424228e-01 -1.31260734e-02
5.38080454e-01 3.10312182e-01 -3.26778263e-01 1.24082077e+00
4.95846003e-01 3.19270074e-01 6.02960706e-01 -1.65816987e+00
-3.35757405e-01 5.92294216e-01 -8.01387429e-01 1.99007273e-01
6.80324435e-01 -2.06500128e-01 9.91583169e-01 -8.15202296e-01
6.49580657e-01 1.21073878e+00 9.49622154e-01 5.32936044e-02
-1.03312647e+00 -6.30287945e-01 -1.13502137e-01 2.06857488e-01
-1.72176874e+00 -2.08085286e-03 9.10719454e-01 -3.00647050e-01
5.34926414e-01 6.15723729e-01 6.98275685e-01 3.76129419e-01
2.76102126e-01 1.67631701e-01 8.56867135e-01 -9.79106903e-01
9.06150714e-02 -7.39146844e-02 2.48496875e-01 8.97339523e-01
5.43032706e-01 2.08986372e-01 -1.94124117e-01 -7.85152316e-01
5.95917583e-01 1.13132916e-01 -2.63984770e-01 -7.84935474e-01
-1.11089361e+00 7.00516582e-01 1.18892128e-02 3.33052337e-01
-9.19945166e-02 1.24284729e-01 -6.33528754e-02 2.95866430e-01
3.44304711e-01 1.16282560e-01 -4.08954807e-02 -1.83971778e-01
-8.33955228e-01 4.20163900e-01 9.92651284e-01 1.16862583e+00
8.71184945e-01 -4.77415234e-01 -4.53001224e-02 7.52727628e-01
4.24214512e-01 6.73688591e-01 6.04411550e-02 -7.12725103e-01
6.19833827e-01 1.01567042e+00 -6.24401271e-02 -1.66897368e+00
-4.07788187e-01 -3.48040313e-01 -7.30406344e-01 4.65766154e-02
5.01059711e-01 2.67636716e-01 -6.16568148e-01 1.69172752e+00
6.83047950e-01 4.36977535e-01 -2.26179883e-01 7.68145859e-01
4.10789549e-01 5.80557525e-01 -3.42168123e-01 -1.87682971e-01
1.02494693e+00 -8.64152610e-01 -3.41420174e-01 2.90731519e-01
6.92390323e-01 -1.02556217e+00 8.17091763e-01 1.89562216e-01
-1.09933186e+00 -3.32930654e-01 -1.24828970e+00 4.63360310e-01
-1.68983221e-01 -7.62883797e-02 4.16743279e-01 1.20971644e+00
-1.20674920e+00 1.07648778e+00 -5.73988438e-01 -6.35447085e-01
-1.81926966e-01 6.27501905e-01 -3.08581591e-01 9.93124247e-02
-9.60344970e-01 4.36246186e-01 2.48575643e-01 -1.34693310e-01
1.33829862e-02 -6.22786939e-01 -6.03601754e-01 1.14677288e-01
3.72253478e-01 -8.14227104e-01 6.41807675e-01 -5.31198084e-01
-1.43396509e+00 1.02743900e+00 -8.97426978e-02 -3.78810406e-01
6.96674824e-01 2.53889412e-01 -3.11684281e-01 4.09741670e-01
7.48106744e-03 1.85505301e-01 8.19073796e-01 -1.01985753e+00
-1.85105592e-01 -5.12713611e-01 -1.76641703e-01 3.25611383e-01
-4.21305418e-01 -1.39443800e-01 -3.58388901e-01 -5.20317435e-01
5.39409697e-01 -1.30056787e+00 -3.53965491e-01 -2.44159289e-02
-3.07022333e-01 -4.19675201e-01 7.19420910e-01 -1.56958550e-01
1.23577893e+00 -1.95600414e+00 3.73472631e-01 8.41775775e-01
1.93421826e-01 1.27670988e-01 9.82237235e-02 7.82743990e-01
-2.38569051e-01 -8.89645889e-02 -4.07474518e-01 -2.84907222e-02
1.37514263e-01 1.32223532e-01 -1.30181029e-01 8.62762272e-01
-4.91320074e-01 5.46923280e-01 -8.61824572e-01 -7.37352192e-01
2.18930572e-01 7.08919168e-02 -6.43359721e-01 -6.65063336e-02
1.99571028e-01 -1.19875655e-01 -2.42040634e-01 6.39959633e-01
1.18431175e+00 -1.77412301e-01 1.71266347e-01 8.96265358e-02
-2.02957228e-01 -3.90794069e-01 -1.74747682e+00 1.44843686e+00
8.52391049e-02 -6.58972049e-03 -9.75923426e-03 -1.11416256e+00
1.52645934e+00 1.78793639e-01 7.41906345e-01 -9.05971378e-02
5.94565384e-02 5.31413615e-01 -2.57235229e-01 -9.15126950e-02
9.56239700e-01 1.51418179e-01 -2.65932381e-02 4.00848329e-01
-1.64355919e-01 -6.38865381e-02 2.38230601e-01 5.54578483e-01
1.15004051e+00 6.90849200e-02 3.22243840e-01 -8.51838171e-01
9.34518397e-01 -7.14649633e-02 4.12333190e-01 7.60672152e-01
-1.97397605e-01 6.67725623e-01 4.68719065e-01 -2.42749676e-01
-1.11592019e+00 -9.05909598e-01 -2.42143080e-01 7.13121295e-01
6.28244817e-01 -5.52632511e-01 -7.64662623e-01 -4.94445324e-01
2.80225426e-01 1.24182239e-01 -2.69961059e-01 -2.86717147e-01
-6.25162542e-01 -8.01080525e-01 3.21572483e-01 4.81091402e-02
4.11842287e-01 -4.67741221e-01 -3.31389606e-02 5.39837219e-02
1.16878964e-01 -8.83234322e-01 -7.39810646e-01 -2.98267215e-01
-8.91594350e-01 -1.29581940e+00 -8.23580861e-01 -9.08144832e-01
1.01726747e+00 5.54814041e-01 8.85569870e-01 4.25363868e-01
-9.63786691e-02 6.78418458e-01 -5.09831011e-01 2.79241055e-01
-2.82119185e-01 2.14837819e-01 1.69812828e-01 4.96991798e-02
2.29026929e-01 -7.33159900e-01 -6.02360606e-01 7.41367996e-01
-6.64236188e-01 -2.69748956e-01 4.17108625e-01 7.53038585e-01
6.49743140e-01 1.76571980e-01 4.24303599e-02 -8.75300884e-01
5.87860525e-01 -5.03699303e-01 -8.52541447e-01 2.84091473e-01
-1.00832963e+00 2.30912879e-01 6.01651311e-01 -3.87421548e-01
-5.00138700e-01 1.43227145e-01 3.62439156e-02 -6.28148735e-01
3.63985777e-01 1.46165922e-01 -5.61824515e-02 -9.52633321e-01
6.26013637e-01 2.89679945e-01 7.65581727e-02 -3.28576237e-01
1.94061071e-01 4.96833503e-01 5.69491327e-01 -9.01143074e-01
1.15921915e+00 4.25869793e-01 6.97785676e-01 -5.81300318e-01
-2.74966329e-01 -7.12428212e-01 -6.40880644e-01 -3.62353712e-01
1.81865916e-01 -3.85681480e-01 -9.27428186e-01 5.86770475e-01
-7.14917302e-01 1.46317765e-01 4.74649109e-02 3.61551434e-01
-6.40788615e-01 9.00134444e-01 -2.24307060e-01 -6.92102492e-01
-4.29376274e-01 -1.02052534e+00 9.74215806e-01 2.03875616e-01
1.65176392e-02 -1.27564609e+00 3.25632870e-01 1.67803064e-01
1.23302899e-01 5.60182512e-01 5.33160329e-01 -6.56523407e-01
-5.84939599e-01 -3.02772790e-01 -9.00880098e-02 -2.81829923e-01
-2.92157710e-01 2.07660839e-01 -1.93259344e-01 -6.13397181e-01
-1.96010455e-01 4.79187697e-01 3.65414560e-01 3.43517333e-01
9.20327485e-01 -3.23019713e-01 -7.57892430e-01 8.30618024e-01
1.65985596e+00 -1.65519714e-02 6.02555811e-01 4.64016616e-01
4.90927696e-01 2.40503386e-01 8.04526448e-01 4.58825111e-01
3.29683721e-01 9.84589756e-01 3.12698036e-01 1.83232933e-01
1.16948888e-03 -3.99834871e-01 1.01814628e-01 1.12450242e+00
-1.94909066e-01 -1.75387248e-01 -8.87376785e-01 4.41912919e-01
-2.03315353e+00 -1.02228272e+00 -5.91692507e-01 2.51204944e+00
3.05763304e-01 2.06556305e-01 3.53530794e-01 6.69549823e-01
1.42550635e+00 -9.98784900e-02 -5.23586646e-02 -5.78467250e-01
-1.40779242e-01 1.82254001e-01 9.66342092e-01 6.45017803e-01
-6.80127084e-01 6.03079796e-01 6.13777447e+00 1.08214486e+00
-5.49980998e-01 -3.06822639e-02 9.23723578e-02 6.27811551e-01
-4.99003649e-01 4.63558763e-01 -1.01187599e+00 7.15212464e-01
6.75360262e-01 -7.73045540e-01 2.28641719e-01 9.63671148e-01
-3.44660252e-01 -4.86663766e-02 -6.81286693e-01 9.21119988e-01
2.14662209e-01 -1.10823035e+00 -2.28286628e-02 2.09867448e-01
7.94419885e-01 -6.12887502e-01 5.91340289e-02 7.78010041e-02
3.73305045e-02 -5.37608027e-01 2.79491454e-01 6.02420509e-01
5.90724111e-01 -1.00530231e+00 6.04139447e-01 2.55907923e-01
-1.57436502e+00 9.81945470e-02 -4.99870330e-01 -5.53490482e-02
-3.48611735e-02 7.02669501e-01 -7.14124322e-01 9.48345006e-01
2.11936846e-01 4.60549980e-01 -4.89020020e-01 1.48751056e+00
1.47515386e-01 4.22170371e-01 -6.76835358e-01 -2.74816215e-01
1.46335870e-01 -7.21022964e-01 1.18324721e+00 7.29516029e-01
6.69691980e-01 1.95386916e-01 3.10785562e-01 6.14122927e-01
3.46325725e-01 4.28257227e-01 -6.02510571e-01 2.21522570e-01
7.37075865e-01 1.20780277e+00 -1.01810801e+00 -8.30044691e-03
-9.73827913e-02 6.51469111e-01 1.24694400e-01 -6.76339418e-02
-7.63817251e-01 -8.59743416e-01 1.23348489e-01 3.50109309e-01
2.26909563e-01 -3.64138275e-01 -1.75090864e-01 -7.49884725e-01
7.69547448e-02 -5.37157655e-01 8.06351840e-01 -5.11689961e-01
-1.04917324e+00 4.52712476e-01 3.04258108e-01 -1.45557427e+00
-8.98403078e-02 -3.83762509e-01 -6.17822349e-01 6.88107789e-01
-1.13501179e+00 -8.82768095e-01 -2.90344447e-01 8.82574022e-01
-7.59012252e-03 -9.95821282e-02 6.07801437e-01 4.87706363e-01
-3.46297443e-01 1.05892634e+00 2.25126192e-01 -3.98400515e-01
5.07480860e-01 -1.11576414e+00 1.26485676e-01 1.12125075e+00
4.21278290e-02 4.65918779e-01 9.50524628e-01 -6.60048366e-01
-1.80339134e+00 -7.36493647e-01 8.58325362e-01 -2.37325907e-01
5.90536237e-01 -1.50347576e-01 -7.85435975e-01 4.30931836e-01
-3.76511544e-01 7.50468001e-02 3.34142417e-01 1.27177924e-01
4.28089537e-02 -2.45878294e-01 -1.55627441e+00 5.48262954e-01
1.33075356e+00 -1.45867942e-02 -5.69370985e-01 7.34088480e-01
7.17410922e-01 -7.41787076e-01 -1.03296018e+00 5.07672429e-01
2.89590895e-01 -9.49232221e-01 1.07279229e+00 5.32357357e-02
-1.44276649e-01 -2.88194656e-01 -1.79771725e-02 -1.04121709e+00
-6.56869769e-01 -1.21904397e+00 -2.58017451e-01 1.15571713e+00
6.96197748e-02 -1.18894553e+00 1.06388581e+00 4.10090238e-01
-9.93350595e-02 -7.36597598e-01 -1.02360117e+00 -1.11351860e+00
-2.02430859e-01 -6.76177442e-04 8.22681606e-01 9.72400606e-01
3.54862213e-01 -1.83682054e-01 -2.75084347e-01 4.40889835e-01
9.92249906e-01 4.83663648e-01 9.52414870e-01 -1.37447333e+00
-4.36483592e-01 -2.89671749e-01 -1.10354817e+00 -8.17636788e-01
9.14855078e-02 -1.10544193e+00 -6.09202683e-01 -1.18812180e+00
2.17756405e-01 -8.70650172e-01 -1.31840393e-01 1.16760068e-01
-2.28617206e-01 1.79435939e-01 5.34232222e-02 2.08731279e-01
-3.77699435e-01 2.83845633e-01 1.21258247e+00 4.07542974e-01
-3.37537080e-01 4.25140530e-01 -3.62111896e-01 2.85064548e-01
6.69087589e-01 -6.71531141e-01 -1.90830961e-01 1.51933536e-01
2.98171997e-01 2.70388454e-01 1.15128092e-01 -1.14571965e+00
5.11452079e-01 -1.55954674e-01 -2.51361609e-01 -7.71793365e-01
2.67363578e-01 -7.67776132e-01 5.48043728e-01 8.53811145e-01
4.58526872e-02 4.37101513e-01 -1.01387843e-01 7.95196652e-01
1.13426253e-01 -5.03836751e-01 6.44903183e-01 1.45985767e-01
-3.37222666e-01 4.98258978e-01 1.72190025e-01 -2.30150279e-02
1.41407049e+00 -7.29597628e-01 -1.52564734e-01 -2.56034106e-01
-5.14576614e-01 2.05591336e-01 6.96828902e-01 2.14805081e-01
4.35796052e-01 -1.62788308e+00 -5.63148975e-01 2.76027471e-01
-2.60131266e-02 -3.42420369e-01 7.61375353e-02 9.58931088e-01
-7.41147459e-01 2.21987367e-01 -1.07205607e-01 -8.96043956e-01
-1.54666328e+00 6.59184396e-01 6.05789237e-02 -1.23378292e-01
-8.10104430e-01 7.75305092e-01 -5.50756335e-01 -7.18250811e-01
9.29534342e-03 1.72231928e-01 1.67765513e-01 -4.57655191e-01
3.28716114e-02 7.68302262e-01 6.10542633e-02 -7.65988052e-01
-3.45599115e-01 1.14784563e+00 1.55944824e-01 7.73156807e-02
1.18616259e+00 -1.35422662e-01 -3.04725885e-01 9.61539373e-02
1.02585387e+00 2.92512476e-01 -7.25067616e-01 -1.60610765e-01
-1.45812497e-01 -1.01661217e+00 -3.12281609e-01 9.74975452e-02
-8.48170578e-01 2.81047285e-01 4.25832003e-01 4.80219036e-01
9.93853688e-01 -4.61893827e-01 9.34071541e-01 1.87780023e-01
6.60133481e-01 -9.44273889e-01 -1.97201580e-01 2.42677346e-01
4.19612259e-01 -7.37795353e-01 3.81519347e-01 -9.58702087e-01
-1.23764440e-01 1.25314331e+00 5.73958397e-01 -5.35559237e-01
8.18391502e-01 1.90762743e-01 -5.67065656e-01 -1.02820911e-01
-5.76329529e-01 1.02881165e-02 3.52780402e-01 2.96523243e-01
-7.56501630e-02 2.49017514e-02 -9.20735657e-01 9.69193205e-02
-5.35264850e-01 -2.55522579e-01 5.23973882e-01 8.36095154e-01
-6.11161113e-01 -1.30887985e+00 -8.14667821e-01 2.14219794e-01
-1.09400690e-01 3.28591436e-01 -3.63770872e-01 8.09727550e-01
-1.76488370e-01 6.60035133e-01 -1.14665575e-01 -3.59603643e-01
1.63696527e-01 -5.26817702e-02 8.35081518e-01 -9.16784853e-02
-4.22715127e-01 -3.77090275e-01 -2.17106000e-01 -3.86732161e-01
-4.48917449e-01 -5.12217522e-01 -1.07022309e+00 -6.80688739e-01
-6.95464909e-01 5.93871415e-01 6.15069687e-01 7.45068848e-01
4.86101687e-01 -5.82383499e-02 9.89534378e-01 -5.55887341e-01
-6.64743841e-01 -3.58648002e-01 -7.50998795e-01 2.89263725e-01
-3.72183383e-01 -9.20729935e-01 -7.41949975e-01 -5.35127640e-01] | [8.071150779724121, -2.0523314476013184] |
53b7d5c8-3967-4612-a5ef-6a6a9d3a38f0 | leveraging-temporal-context-in-low | null | null | http://openaccess.thecvf.com//content/CVPR2023/html/Fosco_Leveraging_Temporal_Context_in_Low_Representational_Power_Regimes_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Fosco_Leveraging_Temporal_Context_in_Low_Representational_Power_Regimes_CVPR_2023_paper.pdf | Leveraging Temporal Context in Low Representational Power Regimes | Computer vision models are excellent at identifying and exploiting regularities in the world. However, it is computationally costly to learn these regularities from scratch. This presents a challenge for low-parameter models, like those running on edge devices (e.g. smartphones). Can the performance of models with low representational power be improved by supplementing training with additional information about these statistical regularities? We explore this in the domains of action recognition and action anticipation, leveraging the fact that actions are typically embedded in stereotypical sequences. We introduce the Event Transition Matrix (ETM), computed from action labels in an untrimmed video dataset, which captures the temporal context of a given action, operationalized as the likelihood that it was preceded or followed by each other action in the set. We show that including information from the ETM during training improves action recognition and anticipation performance on various egocentric video datasets. Through ablation and control studies, we show that the coherent sequence of information captured by our ETM is key to this effect, and we find that the benefit of this explicit representation of temporal context is most pronounced for smaller models. Code, matrices and models are available in our project page: https://camilofosco.com/etm_website. | ['Aude Oliva', 'Emilie Josephs', 'SouYoung Jin', 'Camilo L. Fosco'] | 2023-01-01 | null | null | null | cvpr-2023-1 | ['action-anticipation', 'action-recognition-in-videos'] | ['computer-vision', 'computer-vision'] | [ 5.67940533e-01 -9.64635238e-02 -4.10194993e-01 -3.23785990e-01
-3.76335055e-01 -5.26958704e-01 8.22687566e-01 -1.57373235e-01
-6.23520017e-01 4.18332577e-01 8.62767339e-01 1.30338566e-02
-1.44813016e-01 -1.52723774e-01 -8.19598496e-01 -5.16530752e-01
-3.56611013e-01 -5.62418476e-02 -5.99572808e-02 4.43715639e-02
1.75920889e-01 2.04695091e-01 -1.55101597e+00 5.40751457e-01
2.99798161e-01 6.95582807e-01 3.85674596e-01 6.71413958e-01
6.23738587e-01 8.62468839e-01 -1.28657818e-01 5.78131601e-02
3.84985626e-01 -5.03766775e-01 -5.43321490e-01 2.31820092e-01
5.31223118e-01 -5.05745709e-01 -7.86890328e-01 5.36722183e-01
3.71942490e-01 3.37647736e-01 5.28194845e-01 -1.32876217e+00
-5.24585806e-02 3.94359708e-01 -3.22947085e-01 6.74918294e-01
4.05536026e-01 6.19234741e-01 1.20725918e+00 -5.59865892e-01
8.68825972e-01 1.01503158e+00 6.84663475e-01 5.33242762e-01
-1.53741479e+00 -4.68860179e-01 4.46730763e-01 5.86745858e-01
-1.06229341e+00 -8.76468122e-01 6.41753197e-01 -7.07904041e-01
1.30643022e+00 7.33197331e-02 9.29638028e-01 1.66909659e+00
2.38932922e-01 1.05754054e+00 8.61469507e-01 -1.87653199e-01
1.46457240e-01 -4.03930008e-01 9.25104320e-02 5.65751731e-01
-1.44199375e-03 3.87102515e-01 -1.04338300e+00 3.06916069e-02
7.99176395e-01 1.30624741e-01 -1.43289745e-01 -5.33444583e-01
-1.37262559e+00 5.56369483e-01 8.18350017e-02 1.82877526e-01
-7.86256731e-01 5.78744650e-01 4.91487294e-01 1.14759117e-01
-2.78186090e-02 3.92897099e-01 -4.60322678e-01 -8.36957753e-01
-7.72004724e-01 7.94656351e-02 3.34553599e-01 6.26977861e-01
6.83289289e-01 -2.55026948e-02 -1.27712771e-01 5.66858172e-01
1.31930709e-02 1.18088886e-01 7.47135699e-01 -1.39676058e+00
2.28639260e-01 4.79668766e-01 1.47318169e-01 -8.01419616e-01
-6.14270985e-01 -9.54580307e-02 -2.78209358e-01 -3.15911062e-02
5.31607211e-01 -3.06058317e-01 -8.25524509e-01 2.32107592e+00
1.24670245e-01 7.28911996e-01 -1.36887893e-01 5.84791541e-01
2.41678193e-01 2.78397650e-01 4.04740274e-01 -1.04561538e-01
1.14269650e+00 -5.15091598e-01 -5.35126805e-01 -8.08291852e-01
1.04467118e+00 -2.53719091e-01 9.01068389e-01 1.72434598e-01
-7.03949749e-01 -4.27761763e-01 -7.66737878e-01 -2.43544299e-02
1.07571855e-01 2.28770390e-01 1.05164444e+00 3.34703058e-01
-8.19847107e-01 8.10087919e-01 -1.32949924e+00 -6.10839307e-01
6.72645569e-01 2.86719978e-01 -6.12288892e-01 -6.37392476e-02
-1.00359499e+00 9.27141249e-01 2.08500385e-01 -2.13474259e-02
-1.21732903e+00 -5.92753887e-01 -9.81153727e-01 -8.02858640e-03
4.38712656e-01 -5.87953985e-01 1.33925509e+00 -1.32087004e+00
-1.07688773e+00 8.00947130e-01 -4.20290023e-01 -6.49990261e-01
1.94298491e-01 -4.35645521e-01 -1.17706694e-01 2.29344904e-01
9.00685638e-02 7.86584795e-01 8.95863652e-01 -6.10886753e-01
-4.81495738e-01 -3.80687416e-01 1.74094245e-01 3.69946539e-01
-1.45900995e-01 -1.00369616e-04 -3.07681590e-01 -4.87849921e-01
-1.55349761e-01 -1.31168211e+00 -2.00172037e-01 5.93134835e-02
-3.89906652e-02 -1.34717654e-02 4.73324090e-01 -7.94286907e-01
1.19729877e+00 -2.36128330e+00 8.72200206e-02 -9.69475321e-03
-1.00336790e-01 2.63966262e-01 -2.74199843e-01 6.29941344e-01
-3.90607834e-01 -1.17260396e-01 -4.72222976e-02 -1.48634329e-01
4.29281220e-02 4.30181026e-01 -2.12489471e-01 6.38687551e-01
1.02107041e-01 8.96671712e-01 -9.47190821e-01 -1.88193575e-01
2.93114752e-01 3.65351379e-01 -8.09859931e-01 -1.47951901e-01
-2.65294015e-01 6.70317054e-01 -2.71472424e-01 3.16486239e-01
-9.92736667e-02 -2.75617391e-01 5.03165483e-01 -1.50233641e-01
1.17495693e-01 4.74911332e-01 -1.19401801e+00 1.93602967e+00
-2.94424683e-01 9.13955867e-01 -2.65105873e-01 -1.13070238e+00
1.66597351e-01 4.71474439e-01 7.68981695e-01 -6.53728247e-01
7.25546200e-03 -2.99935341e-01 5.14532804e-01 -7.48185039e-01
1.10360563e-01 -1.19039631e-02 1.55924082e-01 5.81353903e-01
8.60880390e-02 4.13617492e-01 4.28022474e-01 3.49282831e-01
1.50813735e+00 4.72875893e-01 3.16567808e-01 1.21643953e-01
-8.28842372e-02 -1.64816622e-02 9.49277759e-01 8.19800913e-01
-4.26750630e-01 2.57790267e-01 6.58441246e-01 -3.82269979e-01
-9.05331075e-01 -8.06823850e-01 -6.44960478e-02 1.15136087e+00
-3.99925321e-01 -8.31252277e-01 -4.67911005e-01 -3.73351574e-01
-7.65422136e-02 8.04284871e-01 -8.22963834e-01 -6.03966594e-01
-6.24667466e-01 -6.21684849e-01 3.16667646e-01 7.58647501e-01
3.17185968e-01 -1.07456815e+00 -1.11271119e+00 4.87377495e-02
-3.17450494e-01 -1.34495461e+00 -4.43846256e-01 9.66605358e-03
-1.06978703e+00 -1.16415334e+00 -9.96752232e-02 -2.70700753e-01
4.78297085e-01 1.57021806e-01 9.87079859e-01 -5.73230647e-02
-4.82253253e-01 9.77782488e-01 -2.58705497e-01 -3.25123906e-01
-2.95961555e-02 -4.25730824e-01 2.95946896e-01 2.41568401e-01
3.83583039e-01 -8.74545157e-01 -7.85226941e-01 2.41058588e-01
-7.50180840e-01 3.64877403e-01 6.47936285e-01 7.14811087e-01
4.36788142e-01 -1.95844755e-01 2.90245920e-01 -7.52379715e-01
1.50141031e-01 -5.39412618e-01 -1.63785666e-01 -7.36280605e-02
-2.98979908e-01 1.42984092e-02 2.28400916e-01 -6.72094107e-01
-8.71032417e-01 5.08378744e-01 2.53884435e-01 -5.10058641e-01
-2.58040637e-01 5.95496058e-01 4.31016982e-02 2.57213801e-01
6.92917764e-01 3.73203278e-01 4.81085628e-02 -2.93018341e-01
2.63061702e-01 1.65219694e-01 2.55960673e-01 -4.45821464e-01
2.68769592e-01 7.17308044e-01 1.80036485e-01 -1.02073264e+00
-7.35939622e-01 -4.88430411e-01 -5.97890973e-01 -4.17465150e-01
7.02455759e-01 -1.15140700e+00 -7.33787000e-01 3.44597906e-01
-6.68123841e-01 -8.93720031e-01 -3.48174036e-01 9.31622803e-01
-1.04730380e+00 3.81613225e-01 -3.75589967e-01 -6.76541388e-01
3.38852018e-01 -8.68820012e-01 8.30889523e-01 1.80891212e-02
-6.74713612e-01 -9.62926984e-01 3.15118730e-01 3.43657345e-01
6.80661574e-02 1.75178319e-01 7.24655330e-01 -5.40274382e-01
-5.11148095e-01 -3.33791435e-01 2.86808759e-01 2.80361503e-01
1.44915283e-01 -6.06778031e-03 -9.52840328e-01 -1.45055383e-01
-4.17313725e-02 -3.98987144e-01 1.02664256e+00 6.78850591e-01
1.12372053e+00 -1.93690851e-01 -4.44025338e-01 4.45996016e-01
9.18312788e-01 8.40450600e-02 6.50459647e-01 1.83184713e-01
6.19042337e-01 6.27919078e-01 6.29556298e-01 5.81822455e-01
3.36684108e-01 9.38080966e-01 2.95153886e-01 3.44198465e-01
-8.89085159e-02 -3.71102870e-01 7.31713235e-01 1.65623635e-01
-2.74724305e-01 5.48489913e-02 -9.82985914e-01 6.96096480e-01
-2.05313158e+00 -1.41823614e+00 2.19451204e-01 2.26641870e+00
6.86636329e-01 1.25766397e-01 2.30491668e-01 -1.53608307e-01
4.10093576e-01 3.57263744e-01 -7.77058363e-01 -3.52096744e-03
7.98737556e-02 -2.40408629e-02 3.05816114e-01 4.01846081e-01
-1.19727111e+00 8.43278825e-01 6.85979271e+00 4.83110905e-01
-8.97027552e-01 -4.94779199e-02 3.99437457e-01 -6.64434731e-01
1.19194642e-01 2.70256370e-01 -5.56486607e-01 5.20114243e-01
1.34961760e+00 -5.36545441e-02 4.99176383e-01 5.36422729e-01
7.44313002e-01 -4.75709438e-01 -1.63748753e+00 9.96939301e-01
-6.98327925e-03 -1.07732618e+00 -1.30462974e-01 1.28241166e-01
4.82740611e-01 2.70447224e-01 1.25434369e-01 3.79542589e-01
2.88948625e-01 -8.46085787e-01 4.23977464e-01 6.52879179e-01
3.35955173e-01 -2.30698436e-01 -3.52806482e-03 4.24233973e-01
-7.86228657e-01 -3.47979188e-01 -1.36936739e-01 -4.25722748e-01
7.46744722e-02 2.31139228e-01 -8.54902208e-01 1.25904232e-02
4.00006115e-01 1.40330982e+00 -6.62601590e-01 9.76727009e-01
-2.39535362e-01 8.16989243e-01 -3.40518385e-01 3.13424170e-01
1.75123468e-01 -1.44517928e-01 6.32599175e-01 9.92768288e-01
1.27055854e-01 2.71041274e-01 4.58973786e-03 4.08269316e-01
1.49549767e-01 -3.51236999e-01 -8.74339461e-01 -4.82804000e-01
1.89024284e-01 1.00351632e+00 -4.80409771e-01 -3.01530093e-01
-5.16706884e-01 9.65690374e-01 3.27110976e-01 4.80380207e-01
-6.96758986e-01 2.24767223e-01 1.06020749e+00 6.89496752e-03
4.03914511e-01 -4.82183188e-01 2.53961291e-02 -1.25959146e+00
3.83034721e-02 -9.77680206e-01 4.89803463e-01 -8.79920065e-01
-6.81501806e-01 -1.37848720e-01 9.33865607e-02 -1.28370249e+00
-6.71021700e-01 -5.67872465e-01 -6.01719320e-01 4.96286273e-01
-9.46158648e-01 -7.50202239e-01 -5.48974909e-02 6.66298687e-01
5.90107441e-01 4.36610579e-02 7.57849395e-01 2.05420852e-01
-7.63543069e-01 3.24164599e-01 -4.00554202e-02 1.41248077e-01
6.38713896e-01 -8.43115926e-01 3.94047558e-01 9.47160184e-01
7.27178037e-01 7.46568859e-01 8.07138383e-01 -7.68772423e-01
-1.43458140e+00 -7.74514735e-01 6.34551048e-01 -1.00326407e+00
7.07369447e-01 -3.51969481e-01 -5.79744697e-01 1.37050235e+00
-1.46421805e-01 -2.90023349e-02 9.19049978e-01 4.17547196e-01
-3.91425937e-01 1.34232774e-01 -6.43372416e-01 9.42425251e-01
1.47384322e+00 -8.36465061e-01 -6.02546811e-01 5.27991116e-01
2.45413944e-01 -8.74682069e-02 -6.49264634e-01 1.84619948e-01
8.61794055e-01 -9.44120169e-01 9.90280926e-01 -1.15208340e+00
4.69103724e-01 -1.78417563e-02 -2.20965266e-01 -1.32508147e+00
-4.00643826e-01 -6.68947458e-01 -3.32681894e-01 7.19794631e-01
2.48944432e-01 -3.65649521e-01 8.25357497e-01 9.46821630e-01
-1.23184033e-01 -5.93679667e-01 -1.03027380e+00 -8.82469952e-01
-3.92242670e-01 -6.32643700e-01 -5.60148619e-02 8.67289424e-01
2.22948506e-01 3.36299062e-01 -4.91136044e-01 -1.41332559e-02
2.54678369e-01 -4.13974449e-02 6.75487936e-01 -8.97053361e-01
-5.25997221e-01 -2.09444493e-01 -6.65006816e-01 -1.00527334e+00
1.99522987e-01 -8.54384661e-01 -3.81900668e-02 -1.36974669e+00
3.06079656e-01 -1.18087590e-01 -3.67945254e-01 5.92743158e-01
-1.55564830e-01 -5.20769805e-02 2.60806590e-01 2.42967904e-01
-7.12350965e-01 5.77016056e-01 9.91929710e-01 1.97723150e-01
-1.19495928e-01 -9.85026509e-02 -5.34292281e-01 1.04786754e+00
8.17916691e-01 -5.03648818e-01 -6.71936512e-01 -4.99370217e-01
2.05018386e-01 1.02736495e-01 6.47696137e-01 -1.21521151e+00
1.89584002e-01 -3.04558218e-01 4.75429356e-01 -9.53973271e-03
8.53506863e-01 -7.57710218e-01 3.18807423e-01 5.29973209e-01
-5.63381910e-01 6.27008229e-02 3.92176569e-01 9.17868376e-01
1.99599564e-01 -1.25164539e-01 3.68795305e-01 -2.89206654e-01
-1.11647236e+00 2.88661599e-01 -7.95180857e-01 1.53293118e-01
1.00550306e+00 -2.75884628e-01 -1.62191302e-01 -6.58276498e-01
-1.01541615e+00 -6.58870768e-03 4.20186102e-01 3.23577523e-01
4.16015387e-01 -1.21954536e+00 -4.92078036e-01 1.75964117e-01
1.52614549e-01 -6.39022291e-01 5.16239345e-01 1.27951813e+00
8.38850066e-02 4.66404915e-01 -5.28811395e-01 -5.31610727e-01
-1.35681391e+00 4.19871986e-01 3.61233145e-01 -4.91738692e-02
-7.32787192e-01 6.70925200e-01 4.32615161e-01 6.02975637e-02
1.51964605e-01 -1.70938998e-01 -2.74039954e-01 9.97550860e-02
4.09607470e-01 3.07469636e-01 -3.04482102e-01 -6.70961738e-01
-4.82697666e-01 3.19374353e-01 -2.26916343e-01 -3.26248825e-01
1.39381623e+00 -3.79084572e-02 2.78103799e-01 6.48764968e-01
9.59290028e-01 -2.90417820e-01 -1.87416768e+00 -2.80089170e-01
9.68528911e-02 -6.21814847e-01 6.93879426e-02 -6.61427855e-01
-7.99339652e-01 7.09119380e-01 6.77099168e-01 -3.50725502e-01
9.17520285e-01 -1.28147721e-01 4.70865816e-01 6.39489114e-01
2.88826853e-01 -1.18366277e+00 3.38604659e-01 4.64683741e-01
7.36193359e-01 -1.21124482e+00 1.40104353e-01 -1.78861544e-02
-1.04544342e+00 5.42175353e-01 4.46594983e-01 -3.14874709e-01
6.04500294e-01 -2.26895079e-01 -2.96832472e-01 -2.68886447e-01
-1.21483696e+00 -3.13192993e-01 6.78456575e-02 6.09202266e-01
3.16182256e-01 4.00962830e-02 -9.22024250e-02 4.26070124e-01
9.70613733e-02 1.62455052e-01 6.75566316e-01 1.07800794e+00
-2.13435531e-01 -9.03268218e-01 1.14823371e-01 7.85510898e-01
-4.99905735e-01 -1.09364726e-01 -1.97138622e-01 4.88351643e-01
-1.81257818e-02 7.99665034e-01 2.55589247e-01 -5.04600108e-01
1.95872590e-01 3.98834527e-01 6.81583464e-01 -7.18123555e-01
-1.17912412e-01 -1.98505372e-01 4.75050598e-01 -1.27434433e+00
-5.71142733e-01 -1.28482020e+00 -9.79512572e-01 -1.89797878e-02
1.80384427e-01 -3.72525901e-01 4.85965461e-01 9.10532236e-01
7.15243220e-01 4.19874638e-01 2.41624102e-01 -8.99701774e-01
-3.89336944e-01 -8.23076308e-01 -4.58151132e-01 6.17889881e-01
4.27620828e-01 -1.00403535e+00 -3.48599076e-01 4.20949399e-01] | [8.283102989196777, 0.5676915049552917] |
772c4364-bcea-4d6c-a2f3-a08afd52be56 | kernel-estimation-from-salient-structure-for | 1212.1073 | null | http://arxiv.org/abs/1212.1073v2 | http://arxiv.org/pdf/1212.1073v2.pdf | Kernel Estimation from Salient Structure for Robust Motion Deblurring | Blind image deblurring algorithms have been improving steadily in the past
years. Most state-of-the-art algorithms, however, still cannot perform
perfectly in challenging cases, especially in large blur setting. In this
paper, we focus on how to estimate a good kernel estimate from a single blurred
image based on the image structure. We found that image details caused by
blurring could adversely affect the kernel estimation, especially when the blur
kernel is large. One effective way to eliminate these details is to apply image
denoising model based on the Total Variation (TV). First, we developed a novel
method for computing image structures based on TV model, such that the
structures undermining the kernel estimation will be removed. Second, to
mitigate the possible adverse effect of salient edges and improve the
robustness of kernel estimation, we applied a gradient selection method. Third,
we proposed a novel kernel estimation method, which is capable of preserving
the continuity and sparsity of the kernel and reducing the noises. Finally, we
developed an adaptive weighted spatial prior, for the purpose of preserving
sharp edges in latent image restoration. The effectiveness of our method is
demonstrated by experiments on various kinds of challenging examples. | ['Xianfeng GU', 'Zhixun Su', 'Jinshan Pan', 'Risheng Liu'] | 2012-12-05 | null | null | null | null | ['blind-image-deblurring'] | ['computer-vision'] | [ 2.44908899e-01 -6.41494930e-01 4.21192855e-01 -2.07093800e-03
-4.68567520e-01 -3.59109432e-01 2.04360068e-01 -3.68992031e-01
-1.38025671e-01 7.14008093e-01 5.67904174e-01 1.31712809e-01
-4.24070805e-01 -3.27194870e-01 -5.26348054e-01 -9.84138250e-01
1.42328680e-01 -5.74238777e-01 2.85337418e-01 1.79616198e-01
6.08706713e-01 1.91077188e-01 -1.14512444e+00 -1.39708728e-01
1.46507275e+00 7.26528406e-01 5.39170682e-01 3.03851813e-01
2.01438710e-01 9.06682849e-01 -4.22630578e-01 -2.65135467e-02
1.72245786e-01 -3.91988516e-01 -3.95127386e-01 4.35223877e-01
2.78107464e-01 -5.47792375e-01 -5.18347383e-01 1.59664345e+00
3.87064755e-01 2.61302829e-01 6.05750859e-01 -6.79763973e-01
-1.02675712e+00 2.76108176e-01 -9.52475011e-01 2.95725703e-01
-2.93790642e-02 2.66380999e-02 4.18150127e-01 -1.04331613e+00
2.90054470e-01 1.07382560e+00 6.48051739e-01 9.26407427e-02
-1.01037323e+00 -3.75254452e-01 -5.61447069e-02 5.26527047e-01
-1.52740216e+00 -5.34738243e-01 1.14583170e+00 -3.76415521e-01
1.80011883e-01 2.97600120e-01 3.06230873e-01 5.13351262e-01
3.40234101e-01 6.25965416e-01 1.49774516e+00 -4.96223927e-01
1.13817342e-02 -1.75476633e-02 2.47785673e-01 6.73337758e-01
3.40904206e-01 -4.53010760e-02 -2.24537462e-01 -3.79932255e-01
9.28637803e-01 8.75653103e-02 -1.07648218e+00 -3.17238599e-01
-1.29972804e+00 4.70284760e-01 3.60126942e-01 4.54691857e-01
-3.46120954e-01 -1.14970997e-01 1.60331771e-01 -2.82850768e-02
8.19003403e-01 2.54176944e-01 -1.89793095e-01 1.18043378e-01
-1.03944278e+00 1.52001986e-02 3.87423247e-01 4.52913314e-01
6.76290154e-01 -5.30842878e-02 -4.92507070e-01 1.13349962e+00
2.24541292e-01 4.34656650e-01 5.65656066e-01 -8.23931873e-01
1.80123895e-01 2.35433504e-01 4.72201645e-01 -1.23869944e+00
7.99571797e-02 -4.45752561e-01 -1.26652348e+00 2.30836228e-01
5.43858290e-01 -4.12719548e-02 -1.03561354e+00 1.46232951e+00
2.71106482e-01 6.81563854e-01 2.17196010e-02 1.32246971e+00
3.85875583e-01 6.01170003e-01 -2.87410289e-01 -5.08783579e-01
1.28211582e+00 -9.46673512e-01 -1.00527596e+00 -1.54292822e-01
1.07929505e-01 -1.25386560e+00 7.97123432e-01 3.34375173e-01
-1.00243723e+00 -6.84938312e-01 -9.54945266e-01 8.17051008e-02
1.12687998e-01 4.41265196e-01 3.57991695e-01 4.67556655e-01
-9.99908566e-01 6.75642192e-01 -6.36421323e-01 -9.33436379e-02
2.39180461e-01 -3.18978503e-02 -2.84718961e-01 -3.31385493e-01
-1.13457346e+00 1.02398479e+00 2.28869766e-01 5.52264571e-01
-6.25338376e-01 -4.78813648e-01 -7.42305338e-01 1.07914604e-01
3.74438912e-01 -6.08060420e-01 7.50135362e-01 -1.04346883e+00
-1.36476743e+00 3.65391791e-01 -3.40424418e-01 -1.54270440e-01
6.00228190e-01 -5.16963065e-01 -4.30034786e-01 2.36815050e-01
-4.96464828e-03 -2.78223045e-02 1.57721269e+00 -1.31123066e+00
-1.89876363e-01 -3.87086064e-01 -3.70063514e-01 3.17381918e-01
-3.28111231e-01 -7.14927390e-02 -5.39213002e-01 -1.09659910e+00
3.07897002e-01 -7.83463657e-01 -2.79009879e-01 -3.86350416e-02
-3.20809722e-01 2.05293387e-01 8.40723336e-01 -1.22176266e+00
1.24337959e+00 -2.46277785e+00 1.92299828e-01 8.81725177e-02
2.01893821e-01 4.40364987e-01 1.35020450e-01 1.53911382e-01
-5.80950938e-02 -2.30466470e-01 -6.62671804e-01 -4.63447385e-02
-4.01103288e-01 -2.00853065e-01 -4.84979331e-01 9.19480741e-01
-1.17257506e-01 4.58223760e-01 -7.01139271e-01 -3.79960388e-01
4.26832587e-01 6.30474389e-01 -1.51876882e-01 4.07399207e-01
2.61856377e-01 4.81960893e-01 -3.69020551e-01 3.58564734e-01
1.36167777e+00 -1.24038309e-01 -2.95762479e-01 -6.39263153e-01
-3.41821223e-01 -4.87687856e-01 -1.28513587e+00 1.44669318e+00
-3.64140242e-01 5.87957978e-01 4.47100997e-01 -9.01502609e-01
7.68480361e-01 2.17754826e-01 2.84710914e-01 -1.35989577e-01
2.08669696e-02 3.00349623e-01 -1.63096696e-01 -7.50371516e-01
3.64157379e-01 -4.32900079e-02 5.11563718e-01 1.50655761e-01
-4.36405420e-01 9.18040350e-02 -1.87888294e-01 1.54752672e-01
8.11418593e-01 1.71678923e-02 3.19333881e-01 -4.70510453e-01
8.25712323e-01 -2.50011772e-01 4.12259191e-01 6.09088302e-01
-3.25644255e-01 9.22905266e-01 1.92325730e-02 -1.79084584e-01
-8.66185963e-01 -8.14551890e-01 -2.59664953e-01 4.59331006e-01
6.98532224e-01 2.67838929e-02 -9.92470384e-01 -3.30484807e-01
-1.44741103e-01 4.37458485e-01 -3.37384701e-01 -2.46476158e-01
-4.80584651e-01 -1.05349004e+00 2.66926587e-01 1.34598359e-01
1.09343302e+00 -6.78128004e-01 -1.71346441e-01 2.03507632e-01
-3.34891707e-01 -1.07748711e+00 -1.04606140e+00 -3.64673942e-01
-8.57771456e-01 -9.87873912e-01 -1.42846155e+00 -9.54255521e-01
9.32732701e-01 1.05084324e+00 4.65900332e-01 1.56686768e-01
-1.74008295e-01 1.60540104e-01 -2.55090505e-01 1.93215504e-01
-6.32134154e-02 -5.22521079e-01 -5.14000887e-03 5.32404780e-01
-1.34557605e-01 -4.22166109e-01 -8.64704251e-01 4.10440475e-01
-9.98401523e-01 2.14814261e-01 7.88033724e-01 9.88650322e-01
2.00337693e-01 7.17998564e-01 3.69139284e-01 -5.99545121e-01
8.96600187e-01 -2.70078689e-01 -4.87440735e-01 2.22658083e-01
-6.06167793e-01 1.26165763e-01 6.93636179e-01 -5.91883600e-01
-1.66795766e+00 -3.55020091e-02 2.37455130e-01 -5.87827802e-01
-6.42969906e-02 4.91139591e-01 -2.00835876e-02 -3.64225745e-01
4.68326211e-01 5.58345437e-01 6.19562678e-02 -7.49879360e-01
3.44936132e-01 7.96033919e-01 6.29948080e-01 -3.30311984e-01
9.16230381e-01 5.53024352e-01 -1.75397456e-01 -1.02620590e+00
-5.90522349e-01 -6.85209751e-01 -2.71210760e-01 -1.21942297e-01
7.79527724e-01 -9.45373893e-01 -4.20285642e-01 1.05738378e+00
-1.12329817e+00 -9.68762487e-02 3.96233737e-01 6.55074239e-01
-1.36906758e-01 1.14690709e+00 -8.16079676e-01 -8.55170727e-01
-4.29112107e-01 -1.06363750e+00 6.46913290e-01 4.22058284e-01
3.02896827e-01 -1.01783860e+00 -4.50820811e-02 2.75604069e-01
7.72754014e-01 -1.13328911e-01 5.65735400e-01 2.04702362e-01
-6.56565666e-01 4.61858846e-02 -7.10996449e-01 5.76498568e-01
6.03529632e-01 -2.81665325e-01 -8.64354491e-01 -3.54686618e-01
6.27108812e-01 3.56979072e-01 1.10331905e+00 7.21895814e-01
1.13060653e+00 -3.99760514e-01 -3.51094604e-01 7.57429540e-01
1.48011768e+00 -1.54111072e-01 9.29717779e-01 3.12551409e-01
9.09248650e-01 3.44402194e-01 5.77727616e-01 2.04524517e-01
-9.48950276e-02 6.58969104e-01 1.49157763e-01 -2.74968565e-01
-4.10433143e-01 9.24893990e-02 3.00179064e-01 8.97488952e-01
-1.51015744e-01 2.22852957e-02 -4.07433599e-01 7.13342071e-01
-1.99619317e+00 -7.41584718e-01 -3.60870361e-01 2.22288966e+00
8.73462737e-01 -5.11298925e-02 -4.59005237e-01 -1.31587848e-01
1.12442839e+00 2.30840251e-01 -4.01117057e-01 2.97888011e-01
-1.38468117e-01 -1.18542843e-01 6.94061041e-01 6.59577787e-01
-1.08475113e+00 7.88944304e-01 5.66279030e+00 1.16617715e+00
-8.82358074e-01 -1.14579743e-04 5.35019279e-01 5.53521276e-01
-5.21054491e-02 2.05841005e-01 -3.55709344e-01 8.10682058e-01
2.30880812e-01 -1.44181550e-01 6.53252840e-01 5.63466728e-01
5.90961337e-01 -3.55382293e-01 -1.88779905e-01 1.06031609e+00
1.10890009e-01 -8.91662955e-01 -1.39652833e-01 -1.70989349e-01
6.84797704e-01 -4.18158978e-01 8.20853785e-02 -1.78545311e-01
-9.14500281e-02 -7.40352929e-01 4.35530275e-01 8.21207166e-01
4.94093984e-01 -5.33657134e-01 8.15336704e-01 4.25541282e-01
-9.77109611e-01 1.76116928e-01 -5.94662666e-01 -7.01365769e-02
1.38925254e-01 1.19103158e+00 -3.29388738e-01 6.80099547e-01
6.60877705e-01 8.51334214e-01 -4.79260474e-01 1.50151253e+00
-2.59615839e-01 5.72672784e-01 6.02954533e-04 4.50818330e-01
-6.82635903e-02 -6.18424296e-01 7.43231595e-01 1.16000581e+00
5.96028030e-01 3.47392857e-01 5.23180189e-03 8.67859721e-01
1.28339589e-01 5.13730720e-02 -3.75693530e-01 3.87265563e-01
2.48536482e-01 1.36533201e+00 -6.65346324e-01 -3.11168879e-01
-4.10029799e-01 1.41439712e+00 1.35698244e-01 7.74404168e-01
-8.17169845e-01 -4.83364016e-01 4.44780320e-01 -8.28610361e-02
3.99418235e-01 -3.76688957e-01 -2.53051698e-01 -1.63668776e+00
1.59804627e-01 -7.85248101e-01 3.86533849e-02 -8.95627439e-01
-1.38323331e+00 3.14601719e-01 -1.63061604e-01 -1.32531095e+00
4.17853594e-01 -2.81386435e-01 -7.30486393e-01 1.06571496e+00
-1.57838368e+00 -9.27922487e-01 -5.40464938e-01 6.05517507e-01
6.24184668e-01 2.38019824e-01 2.30367318e-01 2.49036178e-01
-6.01827264e-01 1.34311646e-01 4.11860913e-01 -5.75909391e-02
1.10830414e+00 -9.66500521e-01 1.05850806e-04 1.31680393e+00
-2.28974670e-01 8.82327139e-01 8.46888304e-01 -8.11747611e-01
-1.31128240e+00 -9.79736507e-01 3.47780257e-01 7.78994486e-02
5.38989663e-01 1.68467879e-01 -1.29690862e+00 3.33282292e-01
3.39180529e-01 9.25106257e-02 3.53851467e-02 -3.81469339e-01
-6.89031109e-02 -1.32937714e-01 -1.11791503e+00 5.14963508e-01
6.40850365e-01 -3.98246735e-01 -6.16521776e-01 2.23746225e-01
4.37042058e-01 -4.95332003e-01 -6.00789785e-01 5.30369282e-01
4.13258880e-01 -9.47771370e-01 1.13765502e+00 1.72799736e-01
1.86967015e-01 -8.25768054e-01 1.72711074e-01 -1.55352962e+00
-7.94854164e-01 -5.87619960e-01 -3.29521269e-01 1.35437989e+00
-2.56248534e-01 -6.80790365e-01 2.52850443e-01 3.54297549e-01
-1.44966260e-01 -3.20818901e-01 -5.23702383e-01 -5.91968238e-01
-5.38168252e-01 1.28387377e-01 1.37064397e-01 9.86314714e-01
-1.75671667e-01 5.51216900e-02 -1.00693774e+00 5.23093641e-01
1.03976107e+00 1.29944757e-01 4.40809876e-01 -7.04958320e-01
-2.78819531e-01 -1.65077850e-01 -2.11391181e-01 -1.27134168e+00
-7.33723864e-02 -3.22834224e-01 3.13997954e-01 -1.43641281e+00
4.87440079e-01 -7.56336749e-02 -4.07475054e-01 -3.86706553e-02
-9.00886297e-01 3.00897360e-01 -6.07848503e-02 7.10581660e-01
-1.96054608e-01 5.65728366e-01 1.58999598e+00 -9.59084034e-02
-1.07419223e-01 -1.69011224e-02 -5.91611624e-01 8.12238455e-01
6.93815470e-01 -1.80088043e-01 -3.46167058e-01 -4.44938540e-01
-3.08634043e-01 -6.11658581e-02 5.11427283e-01 -9.75551307e-01
3.08482856e-01 -1.90674514e-01 6.01876974e-01 -3.10275018e-01
1.57770798e-01 -8.20848346e-01 1.75098062e-01 2.66077667e-01
-6.14610910e-02 -5.00929117e-01 1.03560090e-01 8.43385100e-01
-3.81428838e-01 -3.35086942e-01 1.01637328e+00 -1.42374888e-01
-8.90263498e-01 5.72438389e-02 -3.44046861e-01 -2.77562141e-01
8.84564757e-01 -3.76355238e-02 -4.11691129e-01 -4.33967292e-01
-4.30381745e-01 -9.71414745e-02 6.23429000e-01 1.10879451e-01
7.79620886e-01 -1.08381391e+00 -7.33353972e-01 2.00378224e-01
-3.04289192e-01 -3.41592908e-01 6.53420508e-01 1.13705075e+00
-6.22843564e-01 6.91453889e-02 -3.83645520e-02 -2.61409074e-01
-1.30810416e+00 7.98854768e-01 2.46713474e-01 -1.75036881e-02
-7.44594514e-01 6.11418366e-01 4.49138343e-01 3.33364934e-01
2.84818280e-02 -3.40622574e-01 -3.36627185e-01 -3.66261363e-01
7.76085019e-01 5.52883804e-01 -7.76187703e-02 -6.78015769e-01
-1.74532890e-01 9.67315972e-01 -1.40898228e-01 -4.21118475e-02
1.07544875e+00 -5.80003381e-01 -4.95303571e-01 -6.02788627e-02
1.08686173e+00 4.19201761e-01 -1.39918697e+00 -2.22258002e-01
-1.36111498e-01 -9.59700465e-01 5.40030062e-01 -3.85283977e-01
-1.03088975e+00 5.80145836e-01 7.32398331e-01 3.42478395e-01
1.44112480e+00 -4.49703932e-01 9.80102479e-01 -2.46591270e-01
1.79606184e-01 -7.56307840e-01 -4.08995859e-02 1.67993888e-01
7.48323202e-01 -1.07091963e+00 2.79810846e-01 -7.21691608e-01
-4.73112106e-01 1.23854113e+00 2.50508338e-01 -2.95388639e-01
5.65825760e-01 -9.77363884e-02 1.54957809e-02 3.32358219e-02
-1.34507596e-01 -1.94147721e-01 5.44602811e-01 2.79237777e-01
2.62512058e-01 -1.40946329e-01 -5.06832898e-01 2.34064803e-01
4.60148275e-01 1.56511098e-01 6.22967541e-01 6.40010417e-01
-7.41329849e-01 -6.87282085e-01 -9.12549257e-01 1.34059981e-01
-6.92231238e-01 -3.02318662e-01 3.15364785e-02 2.02974588e-01
8.39771628e-02 1.08728826e+00 -4.45340991e-01 -1.42748237e-01
8.73487368e-02 -3.40828419e-01 4.44188386e-01 -1.28144845e-01
5.56353405e-02 4.15028274e-01 -2.90012687e-01 -1.74966261e-01
-4.90249485e-01 -4.26741004e-01 -7.80075550e-01 -1.38970345e-01
-7.37360179e-01 2.49266192e-01 3.97467256e-01 8.13269138e-01
1.97125942e-01 2.81959623e-01 6.10455096e-01 -8.97612393e-01
-5.96066475e-01 -1.16161394e+00 -8.67175460e-01 5.43989718e-01
6.43471420e-01 -6.17080152e-01 -9.12685633e-01 3.37285101e-01] | [11.558659553527832, -2.6910159587860107] |
2352e875-eb7a-487f-950f-829023df687c | bayesian-phylogenetic-cognate-prediction | null | null | https://aclanthology.org/2022.sigtyp-1.8 | https://aclanthology.org/2022.sigtyp-1.8.pdf | Bayesian Phylogenetic Cognate Prediction | In Jäger (2019) a computational framework was defined to start from parallel word lists of related languages and infer the corresponding vocabulary of the shared proto-language. The SIGTYP 2022 Shared Task is closely related. The main difference is that what is to be reconstructed is not the proto-form but an unknown word from an extant language. The system described here is a re-implementation of the tools used in the mentioned paper, adapted to the current task. | ['Gerhard Jäger'] | null | null | null | null | naacl-sigtyp-2022-7 | ['cognate-prediction'] | ['natural-language-processing'] | [-7.92762041e-02 6.87200278e-02 -1.45959169e-01 -1.89681813e-01
-4.19336736e-01 -7.16744006e-01 1.11052835e+00 9.03310999e-02
-7.98586071e-01 1.10883498e+00 5.46716452e-01 -6.32383585e-01
-3.12959015e-01 -5.99197090e-01 -2.35978886e-01 -2.27602407e-01
1.30339518e-01 7.98802078e-01 5.66679180e-01 -7.48316705e-01
2.37991065e-01 1.26377851e-01 -1.70164573e+00 4.19415355e-01
5.61481833e-01 1.90036237e-01 6.84972763e-01 6.01525843e-01
-7.02632725e-01 6.17295682e-01 -5.60315847e-01 -6.16182268e-01
7.68459737e-02 -6.28199100e-01 -1.17086971e+00 -2.83348918e-01
3.10808182e-01 1.20476373e-01 -2.95997877e-02 1.03265417e+00
7.03390315e-02 -3.30830216e-02 6.55764341e-01 -7.84330845e-01
-1.07986569e-01 1.28413773e+00 3.42922419e-01 2.42055550e-01
6.90506160e-01 -3.81021678e-01 1.03683960e+00 -8.93191040e-01
1.12240827e+00 1.27321517e+00 6.69218540e-01 2.73069918e-01
-1.03383231e+00 -5.31986952e-01 -1.19548917e-01 2.82275200e-01
-1.83706117e+00 -5.26424587e-01 2.87604898e-01 -5.58011770e-01
1.32683802e+00 4.67459500e-01 9.00209069e-01 1.01133597e+00
2.58063644e-01 1.91455200e-01 1.30969381e+00 -8.68955910e-01
1.57596186e-01 5.11886179e-01 3.39593142e-01 4.30588335e-01
7.96511829e-01 4.85624373e-02 -4.72652555e-01 -2.47330755e-01
4.60317671e-01 -5.95713913e-01 -4.85519692e-02 -1.51489303e-01
-1.22136927e+00 5.83204031e-01 -2.52106786e-01 1.31774020e+00
-2.25724712e-01 -2.18811363e-01 5.55411100e-01 5.52295268e-01
4.78582829e-01 3.77343297e-01 -7.19960988e-01 -1.62729789e-02
-1.11254013e+00 5.43549955e-01 1.51558745e+00 8.37590933e-01
7.96417713e-01 5.28001599e-02 5.06965101e-01 3.17738473e-01
7.85457790e-01 3.08578134e-01 8.83196354e-01 -3.61441106e-01
1.16770677e-01 4.68805820e-01 -1.15975209e-01 -7.18475342e-01
-1.09442577e-01 -5.07604241e-01 -1.71844259e-01 -2.06996351e-01
5.21351516e-01 1.08455636e-01 -4.89268422e-01 1.43044949e+00
1.93564102e-01 2.05827653e-01 2.71727443e-01 2.88301378e-01
1.03222120e+00 4.53146368e-01 8.77694860e-02 -4.53291804e-01
1.32858407e+00 -5.43830156e-01 -9.59313333e-01 -4.39207792e-01
6.50437713e-01 -1.02997291e+00 5.43384612e-01 4.60799843e-01
-7.36488581e-01 -5.28576076e-01 -1.36046255e+00 -1.06078640e-01
-9.24357176e-01 -4.44114715e-01 5.88223577e-01 7.38855243e-01
-1.28132629e+00 3.16733867e-01 -2.23556191e-01 -9.09554899e-01
-6.04582608e-01 -1.93422083e-02 -3.66388172e-01 3.75738651e-01
-1.65091085e+00 1.30319774e+00 1.38514447e+00 -3.55249137e-01
-5.12631536e-01 -3.68540674e-01 -6.95602000e-01 -5.47146499e-01
6.98651433e-01 -7.22314477e-01 1.28062475e+00 -1.17103398e+00
-1.59349692e+00 1.53539455e+00 8.92225057e-02 -5.90066731e-01
4.30195451e-01 -5.24195060e-02 -9.84445810e-01 -5.39837241e-01
1.08043469e-01 -1.16872407e-01 6.11548066e-01 -1.16795874e+00
-5.11628568e-01 -3.14523309e-01 -8.16900209e-02 1.13356370e-03
7.57551417e-02 6.66090131e-01 -2.87000984e-01 -7.78392851e-01
-1.68814614e-01 -5.62823176e-01 -1.22100078e-02 -5.71003616e-01
-3.06378920e-02 -3.20926219e-01 1.08356588e-02 -8.37376177e-01
1.68708909e+00 -1.64658284e+00 3.68820399e-01 1.88106775e-01
1.59908101e-01 4.55605924e-01 -8.34080204e-02 1.06135774e+00
-3.89422923e-01 1.49524957e-01 -3.37079585e-01 5.57827242e-02
-3.66353983e-04 6.01254344e-01 -6.28129423e-01 4.94383246e-01
-5.05824745e-01 6.69394314e-01 -1.13890946e+00 -5.10541856e-01
1.93155915e-01 1.01391591e-01 -1.67016625e-01 -1.47458583e-01
-3.97066385e-01 -3.50094512e-02 -2.55368084e-01 2.81327128e-01
2.51584053e-01 3.67384464e-01 8.33932281e-01 6.78383708e-02
-6.66471779e-01 7.65837431e-01 -1.50967085e+00 1.97288370e+00
-3.69832009e-01 3.10951471e-01 -7.57566607e-03 -8.39086592e-01
1.19419432e+00 6.68639481e-01 2.18970269e-01 -3.02843094e-01
3.87233853e-01 8.03861916e-01 3.46457720e-01 -4.11827445e-01
6.85355186e-01 -5.88566363e-01 -1.71864435e-01 3.13362449e-01
5.30130327e-01 -4.58288968e-01 4.31223243e-01 -5.09510711e-02
7.50969648e-01 5.26360750e-01 1.66814542e+00 -9.89884555e-01
1.11315179e+00 3.20465475e-01 4.54167575e-01 3.84286106e-01
2.65844434e-01 -4.84235622e-02 3.04240827e-02 -6.96976244e-01
-1.05280924e+00 -9.10817683e-01 -3.99303466e-01 1.07238317e+00
-2.36706182e-01 -1.17128396e+00 -3.04774582e-01 -2.83643067e-01
-1.67989254e-01 1.30080450e+00 -3.93773705e-01 5.22615433e-01
-7.95381308e-01 -1.81203876e-02 8.43777001e-01 -1.97416812e-01
-1.69079274e-01 -1.28315234e+00 -5.25426030e-01 4.54220980e-01
-6.54846281e-02 -8.31261754e-01 -8.12022686e-02 -9.35607925e-02
-5.50898969e-01 -9.90650594e-01 -4.43994612e-01 -8.88061821e-01
2.40309238e-02 -1.65918484e-01 1.46319914e+00 1.49377406e-01
-1.54180944e-01 2.52445549e-01 -2.83913016e-01 -7.95753241e-01
-9.93767321e-01 -2.26309546e-03 1.75685555e-01 -1.49062708e-01
5.62478900e-01 -5.17987967e-01 5.23707628e-01 -1.84198245e-01
-1.19423497e+00 3.67136560e-02 -3.01699135e-02 3.28895241e-01
3.93586963e-01 -2.83179462e-01 3.98620129e-01 -1.19876885e+00
4.32140112e-01 -5.92410386e-01 -5.55677295e-01 3.82167131e-01
-6.69232965e-01 1.57827049e-01 5.10573566e-01 -2.74051756e-01
-8.32137287e-01 -1.16036385e-01 -4.79405701e-01 3.13923359e-01
-8.31456259e-02 9.10985589e-01 -3.35930109e-01 3.92170310e-01
5.43326557e-01 3.78035188e-01 -9.73324031e-02 -8.85023594e-01
6.80428743e-01 6.67481959e-01 5.27745843e-01 -5.74709177e-01
6.48970425e-01 1.79874583e-03 -3.18704367e-01 -1.22973537e+00
-4.80486006e-01 -5.43843746e-01 -9.69868302e-01 -1.07494824e-01
4.68228787e-01 -9.25477684e-01 -1.86588243e-02 1.71166509e-01
-1.69571304e+00 -6.26319349e-02 -6.15309596e-01 4.39234912e-01
-5.39421558e-01 3.62439513e-01 6.49861172e-02 -8.57888639e-01
-3.98491234e-01 -5.33528566e-01 6.49555504e-01 -2.66428232e-01
-6.98324025e-01 -1.32431304e+00 9.32362795e-01 -2.58389592e-01
2.75801241e-01 -1.90033466e-01 1.05325890e+00 -1.32913208e+00
1.40902102e-01 -8.56800675e-02 3.30457658e-01 3.44875306e-01
1.52787998e-01 1.72109649e-01 -4.71937507e-01 -2.48354778e-01
3.60121965e-01 -1.19680194e-02 4.63808656e-01 -2.11788625e-01
2.17042804e-01 -3.20683062e-01 -2.43627369e-01 4.54961002e-01
1.89092886e+00 7.20982626e-02 6.38239622e-01 5.44554114e-01
2.66576022e-01 9.25608933e-01 2.22228497e-01 6.71516955e-02
4.58035111e-01 6.78393245e-01 -1.71511963e-01 5.54795265e-01
-3.09080869e-01 -3.14746827e-01 6.73429489e-01 1.63799548e+00
1.95202287e-02 -2.77852356e-01 -1.47790682e+00 6.63583219e-01
-1.83104718e+00 -8.33635807e-01 -5.42316556e-01 2.31203318e+00
1.00981867e+00 1.56944469e-01 1.21228211e-01 -1.25983551e-01
5.88373780e-01 2.74818450e-01 1.76234469e-01 -5.49896300e-01
-4.25033957e-01 5.60509920e-01 2.63751924e-01 1.06969571e+00
-5.61683595e-01 1.19003844e+00 7.82093716e+00 9.46685791e-01
-7.13554859e-01 2.63358951e-01 -4.62742925e-01 4.14495051e-01
-7.33580172e-01 5.56551218e-01 -8.91147256e-01 2.28560999e-01
1.49711978e+00 -1.13802719e+00 4.90902960e-01 4.46646839e-01
-2.97260076e-01 8.47764835e-02 -1.06920242e+00 7.48107612e-01
6.67562008e-01 -1.21320260e+00 4.09532100e-01 1.05489470e-01
3.20895880e-01 2.88353711e-01 -6.62955701e-01 2.41699576e-01
4.95661080e-01 -8.67974758e-01 1.06607080e+00 1.00178003e+00
8.71323764e-01 -4.44410503e-01 8.44962895e-01 5.18435061e-01
-1.28186882e+00 5.76479197e-01 -3.11156750e-01 -3.49783897e-01
3.15065444e-01 3.11225027e-01 -6.05988741e-01 1.01776862e+00
1.19807675e-01 6.47208273e-01 -6.53646827e-01 8.43492687e-01
-3.60083401e-01 6.10696316e-01 -2.45774403e-01 -2.61353999e-01
-7.82153308e-02 -3.58028531e-01 1.21982896e+00 1.52893770e+00
3.08216423e-01 -3.49550188e-01 3.12374592e-01 6.19827747e-01
5.16738951e-01 9.12244916e-01 -1.14812398e+00 -2.01917917e-01
3.08001548e-01 1.07800341e+00 -5.69484651e-01 -7.13319302e-01
-5.19723356e-01 7.38122106e-01 5.63341901e-02 -5.95806018e-02
-3.66132766e-01 -3.83069724e-01 5.67555130e-01 2.74609149e-01
2.20807239e-01 -4.46376890e-01 1.80854816e-02 -1.11778831e+00
-2.87469327e-01 -9.97551501e-01 3.65928054e-01 -5.08660555e-01
-1.39619851e+00 1.07291162e+00 4.01787609e-01 -8.99455488e-01
-9.13127959e-01 -9.09560263e-01 -1.54322490e-01 1.09275126e+00
-8.10567617e-01 -1.24052000e+00 3.85242730e-01 4.69287723e-01
4.29460466e-01 -5.68879783e-01 1.30548036e+00 -2.20946260e-02
-7.10273311e-02 -3.94069739e-02 1.88213110e-01 -1.69166893e-01
7.25568891e-01 -1.11590731e+00 4.47349280e-01 8.97090077e-01
4.16367501e-01 9.60562944e-01 1.06812060e+00 -1.13766623e+00
-1.40463567e+00 -5.77251375e-01 1.96595240e+00 -5.32328486e-01
1.72085893e+00 -4.92646307e-01 -8.61943066e-01 8.34492266e-01
7.16717064e-01 -3.18201393e-01 6.26474380e-01 3.07094604e-02
-4.46408480e-01 -6.52164295e-02 -6.61054730e-01 6.68461919e-01
1.23392558e+00 -8.20223093e-01 -1.51758611e+00 1.99743852e-01
8.06055486e-01 -9.65053961e-02 -7.77515054e-01 9.52729583e-02
6.90788805e-01 -7.48294413e-01 7.22801268e-01 -6.88545227e-01
-1.07466847e-01 -4.14237261e-01 -5.40850282e-01 -1.15165234e+00
-1.86147884e-01 -8.74239087e-01 2.33928159e-01 1.41683924e+00
2.42862284e-01 -9.92052019e-01 5.89328669e-02 -6.93494827e-02
-4.83511165e-02 9.97284651e-02 -1.29261661e+00 -1.07004535e+00
3.84123176e-01 -8.11022282e-01 6.16390765e-01 8.86251509e-01
3.05611640e-02 8.73109519e-01 -1.24593414e-01 -4.36978668e-01
3.08527201e-01 -4.56132330e-02 7.25483835e-01 -1.61340284e+00
-3.33826423e-01 -4.78479296e-01 -4.35229123e-01 -5.33462882e-01
5.67685783e-01 -1.48848760e+00 1.01631463e-01 -1.23395717e+00
-1.14757568e-01 -7.16090426e-02 -3.75492096e-01 2.41904095e-01
2.28585452e-01 9.25843120e-02 2.20036536e-01 2.40415066e-01
-2.72145510e-01 3.51702273e-01 4.75955278e-01 3.63596454e-02
-1.26503929e-01 -3.55886400e-01 -7.29926527e-01 1.00676572e+00
4.80993807e-01 -7.99791336e-01 -2.09403172e-01 -2.78993756e-01
8.76433671e-01 -1.77435786e-01 1.76036716e-01 -9.57430899e-01
3.53234112e-01 -2.82100737e-01 -3.98801148e-01 -5.57296336e-01
-1.87564686e-01 -9.47562993e-01 8.66314769e-01 7.13844836e-01
-5.03244065e-02 6.81017578e-01 4.00382429e-01 2.94108748e-01
-1.78435519e-01 -8.78393412e-01 2.77146935e-01 -3.78612667e-01
-1.11267579e+00 -2.41467327e-01 -8.26755226e-01 2.22068399e-01
8.41612220e-01 -2.86061645e-01 9.15192738e-02 1.02495372e-01
-3.91030610e-01 -4.55778629e-01 5.87188959e-01 5.27008235e-01
3.12141746e-01 -1.13234544e+00 -1.22458434e+00 -6.29771203e-02
3.80859941e-01 -7.89099991e-01 -2.53014445e-01 3.58798772e-01
-7.29289830e-01 7.81419218e-01 -3.26501399e-01 5.67659475e-02
-1.05018353e+00 7.03511596e-01 1.63646266e-01 -3.00640970e-01
-6.57582521e-01 3.39992434e-01 -1.64225578e-01 -5.80436647e-01
-2.06171438e-01 -1.33799359e-01 -5.50640166e-01 2.62140125e-01
7.94047654e-01 2.53484309e-01 1.83184937e-01 -1.14847147e+00
-8.40898395e-01 3.67966413e-01 1.61263570e-01 -6.70721769e-01
1.41766262e+00 -2.50767320e-01 -8.77819180e-01 1.53576231e+00
9.21977639e-01 6.32433534e-01 -2.54251086e-03 -7.17495382e-01
5.52953243e-01 -5.88553175e-02 2.45460719e-02 -8.63466442e-01
-1.69195056e-01 9.70189273e-02 2.26922527e-01 2.03385919e-01
5.96096396e-01 2.06524938e-01 3.53861690e-01 2.45933458e-01
8.15616250e-01 -8.56142879e-01 -7.94091880e-01 1.04277313e+00
1.26561105e+00 -4.84999210e-01 4.79583651e-01 -3.15208435e-01
-3.67461443e-01 1.36610782e+00 8.03699270e-02 -2.86000550e-01
9.23318565e-01 3.84212524e-01 7.87189305e-02 -3.36063147e-01
-7.68705368e-01 -5.03500104e-01 6.71880007e-01 4.84898299e-01
8.42940331e-01 2.48662189e-01 -1.39613104e+00 7.41288781e-01
-5.90732157e-01 -8.60032812e-03 4.22380894e-01 7.66023517e-01
-3.39347482e-01 -1.68779218e+00 -3.24593276e-01 1.67200983e-01
-2.61483014e-01 -5.02881825e-01 -7.83774257e-01 1.27745271e+00
7.87201226e-01 7.40875185e-01 6.66497275e-03 -4.81588453e-01
1.52830422e-01 5.84197998e-01 6.55139029e-01 -1.14048874e+00
-9.16186750e-01 -6.38456568e-02 4.85966921e-01 -3.69425535e-01
-7.96294034e-01 -7.83912420e-01 -8.72481763e-01 -3.17796677e-01
-8.18595216e-02 2.47830242e-01 8.59622121e-01 1.33556998e+00
-3.86553526e-01 3.37818921e-01 -4.26960178e-02 -5.35231948e-01
-3.01634401e-01 -9.18312371e-01 -9.14701581e-01 -6.24836981e-02
-3.96254584e-02 -3.55592668e-01 -3.18426341e-01 -6.28134841e-03] | [10.40386962890625, 10.06480598449707] |
9e02206c-7f8c-4bcd-85f6-0ae045455e7c | leveraging-pre-trained-acoustic-feature | null | null | https://ieeexplore.ieee.org/document/9980083 | http://www.apsipa.org/proceedings/2022/APSIPA%202022/ThAM1-3/1570839332.pdf | Leveraging Pre-Trained Acoustic Feature Extractor For Affective Vocal Bursts Tasks | Understanding humans’ emotions is a challenge for computers. Nowadays, research on speech emotion recognition has been conducted progressively. Instead of a speech, affective information may lay on short vocal bursts (i.e., cry when sad). In this study, we evaluated a recent self-supervised learning model to extract acoustic embedding for affective vocal bursts tasks. There are four tasks investigated on both regression and classification problems. Using similar architectures, we found the effectiveness of using a pre-trained model over the baseline methods. The study is further expanded to evaluate the different number of seeds, patiences, and batch sizes on the performance of the four tasks. | ['Akira Sasou', 'Bagus Tris Atmaja'] | 2022-12-21 | null | null | null | apsipa-2022-12 | ['speech-emotion-recognition'] | ['speech'] | [ 5.19722849e-02 8.19417387e-02 1.45720318e-01 -7.59194076e-01
-4.25340921e-01 -2.83681959e-01 4.77090359e-01 1.35957133e-02
-6.37742639e-01 4.39863563e-01 2.32264102e-01 5.09997867e-02
4.58914697e-01 -2.46011421e-01 -3.33057046e-01 -5.77167690e-01
-1.53884873e-01 -4.09086682e-02 -3.36543135e-02 -2.04686701e-01
1.09580874e-01 2.59627521e-01 -1.67048490e+00 5.03896356e-01
4.10166174e-01 1.17893565e+00 1.21088535e-01 9.60309148e-01
-1.19176120e-01 7.76530027e-01 -1.06896484e+00 -5.78508079e-01
-2.72130996e-01 -5.17984569e-01 -6.33557141e-01 8.79492611e-02
1.51802283e-02 -6.95479214e-02 -1.46234542e-01 7.94279575e-01
7.39564538e-01 3.59962523e-01 6.20286942e-01 -1.22774482e+00
-5.70694566e-01 7.80705750e-01 -1.01309575e-01 4.54962790e-01
3.51636142e-01 9.09736380e-03 1.19071794e+00 -8.30043972e-01
1.71125174e-01 1.12543249e+00 6.12101734e-01 7.25061178e-01
-9.55695033e-01 -8.18174422e-01 1.26686007e-01 3.62921089e-01
-9.33642089e-01 -7.38032401e-01 1.05083919e+00 -3.02497894e-01
1.36080015e+00 3.30118656e-01 6.99680507e-01 1.65514290e+00
-2.24152133e-02 7.95957088e-01 1.17782378e+00 -5.67749143e-01
3.01474571e-01 6.68166637e-01 4.24987972e-01 3.02964658e-01
-4.92963344e-01 2.99274325e-01 -8.52975249e-01 -2.53544152e-01
5.12342378e-02 -4.93009061e-01 -5.63123971e-02 3.44974458e-01
-7.00096369e-01 1.07873142e+00 -1.56435445e-02 6.01384759e-01
-4.83480603e-01 5.56742363e-02 7.41579890e-01 4.05344337e-01
8.74515474e-01 5.23363650e-01 -5.40021718e-01 -6.12223804e-01
-8.90534699e-01 -3.10373336e-01 9.95963216e-01 4.68390554e-01
3.81543666e-01 4.88735616e-01 -1.48302376e-01 1.29552519e+00
1.80690780e-01 1.15818024e-01 7.40319610e-01 -7.47733474e-01
-6.61996612e-03 5.79233505e-02 -1.57396212e-01 -8.82887661e-01
-5.68748593e-01 -2.03230843e-01 -5.66546738e-01 -1.57224789e-01
-2.41791643e-03 -4.58234370e-01 -6.80016875e-01 1.80849874e+00
1.10845012e-03 2.29205132e-01 1.53780490e-01 8.06839764e-01
1.03406394e+00 9.88316000e-01 2.57516950e-01 -4.77813870e-01
1.24271905e+00 -1.22061443e+00 -1.20491779e+00 -5.48719704e-01
4.30870265e-01 -9.50188279e-01 1.38010788e+00 6.68908060e-01
-8.55413020e-01 -6.05804622e-01 -9.88939464e-01 1.63339749e-01
-4.92911011e-01 2.80898631e-01 7.22666919e-01 1.16374588e+00
-9.52472985e-01 5.38926363e-01 -6.61873221e-01 -3.19976985e-01
8.74079615e-02 2.13830620e-01 -6.15597367e-02 6.82791233e-01
-1.38673818e+00 9.58216488e-01 2.56820414e-02 9.96038914e-02
-5.09966552e-01 -4.69969749e-01 -7.83824503e-01 6.26302883e-02
9.61057469e-02 -5.37737831e-02 1.50703526e+00 -1.32021129e+00
-2.07545924e+00 9.09223557e-01 -3.42527956e-01 -4.37191308e-01
-7.79808536e-02 -4.03900981e-01 -7.08340049e-01 1.74677014e-01
-4.18879747e-01 6.79653883e-01 1.11478078e+00 -1.04090464e+00
-2.17556730e-01 -6.23305850e-02 -2.58272022e-01 -4.43184003e-03
-8.17256868e-01 5.96813500e-01 -4.95113768e-02 -6.75991714e-01
-4.06686336e-01 -1.10645151e+00 -4.01670299e-02 -6.50268316e-01
-1.61989644e-01 -5.57073832e-01 5.82201004e-01 -6.71535075e-01
1.41058099e+00 -2.63312650e+00 -7.33565092e-02 -5.05473539e-02
-8.58587995e-02 1.82274282e-01 -1.43802091e-01 4.94723916e-01
-3.44117314e-01 2.65446603e-01 -1.45650193e-01 -6.78059399e-01
2.57621467e-01 2.14406207e-01 -4.54732776e-01 9.98336673e-02
4.18146938e-01 5.96425772e-01 -5.67324877e-01 -3.73467654e-01
7.83125386e-02 5.15500546e-01 -4.42642868e-01 5.20353913e-01
-6.44247308e-02 3.37937295e-01 -1.72447175e-01 4.10102010e-01
2.81217366e-01 3.36531624e-02 3.44080105e-03 -1.23140775e-02
-1.05072640e-01 5.11425555e-01 -7.47884750e-01 1.47731495e+00
-8.19305658e-01 9.65084732e-01 7.20521361e-02 -8.90433550e-01
9.87542987e-01 7.18136191e-01 3.61881614e-01 -4.46268022e-01
5.40654600e-01 2.74033807e-02 2.38743842e-01 -6.68296754e-01
5.22756934e-01 -4.93082911e-01 -2.80302137e-01 5.22644818e-01
2.15198100e-01 -3.91933858e-01 -1.08101405e-02 -5.18472679e-02
9.57790434e-01 -3.39685768e-01 2.21555784e-01 9.62143168e-02
3.89840335e-01 -3.25526774e-01 2.36729607e-01 4.76957798e-01
-4.95334864e-01 4.17373419e-01 5.02993107e-01 -2.38386318e-01
-5.97058654e-01 -8.40624154e-01 -8.41220543e-02 1.51510072e+00
-2.07758740e-01 -6.00870550e-01 -6.41431987e-01 -5.28397858e-01
-2.96498358e-01 1.10667157e+00 -4.78312612e-01 -5.10797203e-01
-3.49059612e-01 -7.57017255e-01 8.09470177e-01 4.59869444e-01
-1.93309978e-01 -1.67770851e+00 -6.36081874e-01 1.54649705e-01
-2.91923553e-01 -1.27933419e+00 -4.49133545e-01 6.52276218e-01
-5.34932673e-01 -3.04533660e-01 -2.59324580e-01 -7.51363158e-01
1.25175957e-02 -1.34984821e-01 1.14671826e+00 -1.25977531e-01
-2.37716496e-01 6.09026253e-01 -7.25343347e-01 -8.60314608e-01
-5.17901540e-01 1.74470316e-03 1.53910339e-01 1.65383011e-01
6.21364534e-01 -5.78863978e-01 -2.75198102e-01 7.51464739e-02
-7.71157622e-01 -3.93978655e-01 4.51237887e-01 7.10948467e-01
1.48721278e-01 -7.54305050e-02 8.76568019e-01 -6.61656082e-01
1.26525354e+00 -4.66035724e-01 1.78787142e-01 -1.13098823e-01
-2.40502670e-01 -1.00740358e-01 6.42380357e-01 -9.37330246e-01
-1.00896001e+00 1.12612816e-02 -5.49505591e-01 -4.55852807e-01
-3.85504693e-01 4.98128027e-01 1.78802162e-01 2.18262777e-01
4.14792091e-01 -1.33065030e-01 -3.74376923e-02 -2.30732366e-01
3.71847987e-01 1.17187881e+00 4.98281308e-02 -3.46874535e-01
1.46424055e-01 -6.77690143e-03 -7.45347142e-01 -1.26687002e+00
-7.65149415e-01 -4.29065168e-01 -2.96715260e-01 -5.50515413e-01
9.33232129e-01 -7.20938444e-01 -6.20685875e-01 4.83136952e-01
-1.20673811e+00 -4.31945235e-01 -1.47959933e-01 7.89797246e-01
-4.78333920e-01 2.74442583e-01 -9.25633252e-01 -1.18131852e+00
-3.22803080e-01 -9.21166301e-01 9.17868257e-01 1.18989475e-01
-8.44558954e-01 -8.21626306e-01 3.43503058e-01 1.92614779e-01
4.53468591e-01 -2.17805237e-01 6.90091670e-01 -9.79023218e-01
2.89556354e-01 -8.87869075e-02 3.38660836e-01 8.13122988e-01
1.49112821e-01 1.09972008e-01 -1.40371776e+00 4.72076121e-04
4.88982677e-01 -9.02435660e-01 8.01780999e-01 2.90974915e-01
1.20960200e+00 -2.98142821e-01 7.94880390e-02 1.97441608e-01
6.25274777e-01 4.87388670e-01 4.34600830e-01 -2.75643263e-02
1.90441608e-02 9.35577273e-01 5.74550033e-01 5.07899880e-01
2.87274197e-02 4.80505049e-01 6.09548986e-02 -1.61834598e-01
4.47079390e-02 1.78599879e-01 6.57272577e-01 1.43941665e+00
1.18294232e-01 -3.98919672e-01 -6.75232768e-01 3.23654830e-01
-1.38414502e+00 -9.46663976e-01 2.04420999e-01 1.69752920e+00
9.74669337e-01 1.79353163e-01 1.59714445e-01 2.92791367e-01
5.91735840e-01 4.24733311e-01 -3.86782795e-01 -1.18616140e+00
1.09337387e-03 6.24938488e-01 -2.51623988e-01 4.01985466e-01
-1.09565842e+00 1.03945827e+00 6.65974617e+00 6.24391735e-01
-1.55847633e+00 1.13271415e-01 5.57687879e-01 -3.18453014e-01
-2.60860147e-03 -4.64986950e-01 -4.45285261e-01 4.42069650e-01
1.45492697e+00 -2.91047692e-02 4.36667144e-01 8.35823059e-01
1.99487686e-01 -5.77272773e-02 -1.12947738e+00 1.02585280e+00
4.23034251e-01 -3.81808788e-01 -3.80954236e-01 -3.30070853e-01
2.61149138e-01 -1.25416890e-01 2.73765981e-01 7.79913664e-01
-1.30643258e-02 -1.10900760e+00 6.73329771e-01 2.08627582e-01
4.01297897e-01 -7.45699704e-01 6.72871292e-01 3.41796130e-01
-7.34699249e-01 -1.01921178e-01 -1.66443974e-01 -2.32026517e-01
3.22209626e-01 5.09722471e-01 -9.50845718e-01 1.38270572e-01
8.60241592e-01 4.54898655e-01 -3.99819493e-01 4.40144539e-01
-2.11742833e-01 1.24513805e+00 -3.15172464e-01 -5.48687994e-01
1.86458305e-01 -2.36530021e-01 4.29182976e-01 1.64642966e+00
1.56575054e-01 2.27535009e-01 -4.36741970e-02 5.13069749e-01
1.29784986e-01 3.98176432e-01 -4.55587596e-01 -6.42347634e-01
2.65387058e-01 1.31663537e+00 -6.58188581e-01 -1.26472309e-01
-5.18705368e-01 1.05017149e+00 3.41443568e-01 2.77005702e-01
-1.09684122e+00 -4.11770165e-01 6.55768275e-01 -2.08194673e-01
3.43885452e-01 -3.28791529e-01 -1.68760538e-01 -7.92865336e-01
-1.11358367e-01 -1.02662146e+00 2.23089948e-01 -8.34103525e-01
-1.44863355e+00 1.04426897e+00 -1.48584515e-01 -7.30624855e-01
-4.03177440e-01 -5.12820542e-01 -7.71078587e-01 5.85240722e-01
-1.18952739e+00 -6.39781833e-01 -4.77711372e-02 2.99154550e-01
7.83228338e-01 -1.70159996e-01 1.07455873e+00 2.87850469e-01
-6.59027636e-01 6.41569972e-01 -3.77023578e-01 -9.00782347e-02
1.03466141e+00 -1.01341796e+00 3.12613472e-02 4.77834910e-01
4.73045051e-01 4.20615971e-01 8.23761046e-01 -2.50622243e-01
-9.05143499e-01 -6.50820553e-01 1.05965090e+00 -1.98307887e-01
8.12551677e-01 -5.90588391e-01 -9.44398880e-01 4.83829439e-01
7.81454325e-01 -4.42125916e-01 1.19164383e+00 4.76407796e-01
-2.66437113e-01 -9.27617699e-02 -7.42090046e-01 4.79125351e-01
5.71849346e-01 -7.05945253e-01 -7.13404357e-01 -1.61166191e-02
9.34341073e-01 -3.00022848e-02 -5.65269351e-01 4.92331982e-01
5.87744474e-01 -1.04009211e+00 7.75710821e-01 -6.65270090e-01
4.55622733e-01 3.28109205e-01 -2.15768769e-01 -1.65028679e+00
-4.86513153e-02 -6.32019520e-01 -6.75194263e-02 1.32035995e+00
4.77195203e-01 -5.30309439e-01 4.73456353e-01 3.40852171e-01
-2.76766092e-01 -8.27123046e-01 -8.81055355e-01 -6.69318318e-01
-1.08643211e-01 -7.63688982e-01 9.86644700e-02 7.79510200e-01
2.75489807e-01 9.27968204e-01 -5.37302077e-01 1.23428300e-01
-2.02412173e-01 -4.38187346e-02 4.60150480e-01 -9.27609503e-01
-2.83909708e-01 -5.73169529e-01 -3.04616421e-01 -6.77776456e-01
7.61892617e-01 -7.89706528e-01 3.31206501e-01 -9.17212009e-01
-1.17240041e-01 -2.76865717e-02 -4.11247909e-01 3.82701337e-01
-3.11845750e-01 3.30506526e-02 1.50586501e-01 -2.83142984e-01
-3.35274726e-01 8.46831083e-01 7.12546945e-01 -1.61798038e-02
-4.09025639e-01 2.27499306e-01 -4.84173417e-01 7.40909934e-01
1.28950393e+00 -6.47109628e-01 -4.77716416e-01 -1.23502910e-02
1.62608951e-01 6.43937141e-02 6.06740490e-02 -7.64957011e-01
1.52848974e-01 1.55365229e-01 -3.40962894e-02 -4.15996224e-01
9.85757232e-01 -5.06752372e-01 -1.62277311e-01 1.32636726e-01
-5.63165724e-01 5.57703748e-02 4.45773065e-01 4.26389128e-01
-5.20542204e-01 -4.60420877e-01 7.67190576e-01 4.78323996e-02
-4.90440220e-01 -1.78668424e-01 -8.74876738e-01 -6.73235133e-02
8.65529299e-01 2.75007216e-03 1.90580159e-01 -6.82138503e-01
-9.57317114e-01 -1.12130329e-01 -2.89806604e-01 6.35780573e-01
7.68481731e-01 -1.00111663e+00 -5.92994690e-01 1.64056525e-01
-4.67164144e-02 -7.07624853e-01 1.58961475e-01 8.09911966e-01
2.63473727e-02 2.80301303e-01 -4.05714400e-02 -4.32022214e-01
-1.72456658e+00 4.10499334e-01 1.72755554e-01 -1.36685655e-01
-9.00865868e-02 1.01117671e+00 6.08100295e-02 -5.14353931e-01
6.81266308e-01 -2.92561620e-01 -4.22696948e-01 4.33957666e-01
3.31043243e-01 1.97149456e-01 1.56898826e-01 -4.32181209e-01
-2.93802410e-01 1.62132442e-01 3.25527303e-02 -4.25625056e-01
1.38952518e+00 1.43499225e-02 -5.51058836e-02 1.19363153e+00
1.17231023e+00 3.29838336e-01 -6.54411197e-01 2.93710753e-02
1.45264715e-01 -6.67248070e-02 2.77995076e-02 -8.16700280e-01
-7.15690613e-01 1.00061381e+00 5.33490598e-01 5.98412275e-01
1.21070647e+00 1.43045247e-01 7.99844682e-01 3.68498594e-01
-3.14670131e-02 -1.30825639e+00 6.07950747e-01 7.12615073e-01
1.05541575e+00 -1.31743634e+00 -3.82888913e-01 -3.81256521e-01
-1.21118760e+00 1.09428191e+00 6.01406574e-01 7.03373104e-02
8.97004843e-01 4.61234152e-01 4.48491812e-01 -1.90163568e-01
-1.11191404e+00 -2.34460250e-01 2.33625203e-01 3.49600822e-01
8.22799265e-01 1.98733166e-01 -4.00801510e-01 1.22602892e+00
-7.17293978e-01 -3.16999257e-01 5.03735483e-01 7.50488281e-01
-3.27675641e-01 -9.89474237e-01 -1.73073933e-01 4.40826476e-01
-7.13314652e-01 -2.56259173e-01 -9.15539205e-01 5.04363537e-01
-4.33632694e-02 1.46137667e+00 1.10678434e-01 -6.31738544e-01
4.19080734e-01 5.48608422e-01 3.28185081e-01 -7.67544746e-01
-9.81182873e-01 1.22176416e-01 4.32350308e-01 -3.03211391e-01
-6.70294702e-01 -6.77837253e-01 -1.12720895e+00 1.83121368e-01
-5.49570978e-01 3.42881233e-01 8.66044581e-01 8.18012536e-01
8.32543299e-02 6.07554615e-01 8.74638855e-01 -6.41007066e-01
-5.99483967e-01 -1.43847096e+00 -5.36716163e-01 3.20701838e-01
2.15159371e-01 -5.03228188e-01 -7.98835158e-01 1.66519403e-01] | [13.574618339538574, 5.796138286590576] |
5060d52a-38aa-4773-b867-f3f549469bec | towards-automated-melanoma-screening-proper | 1604.04024 | null | http://arxiv.org/abs/1604.04024v3 | http://arxiv.org/pdf/1604.04024v3.pdf | Towards Automated Melanoma Screening: Proper Computer Vision & Reliable Results | In this paper we survey, analyze and criticize current art on automated
melanoma screening, reimplementing a baseline technique, and proposing two
novel ones. Melanoma, although highly curable when detected early, ends as one
of the most dangerous types of cancer, due to delayed diagnosis and treatment.
Its incidence is soaring, much faster than the number of trained professionals
able to diagnose it. Automated screening appears as an alternative to make the
most of those professionals, focusing their time on the patients at risk while
safely discharging the other patients. However, the potential of automated
melanoma diagnosis is currently unfulfilled, due to the emphasis of current
literature on outdated computer vision models. Even more problematic is the
irreproducibility of current art. We show how streamlined pipelines based upon
current Computer Vision outperform conventional models - a model based on an
advanced bags of words reaches an AUC of 84.6%, and a model based on deep
neural networks reaches 89.3%, while the baseline (a classical bag of words)
stays at 81.2%. We also initiate a dialog to improve reproducibility in our
community | ['Eduardo Valle', 'Flávia Vasques Bittencourt', 'Micael Carvalho', 'Sandra Avila', 'Michel Fornaciali'] | 2016-04-14 | null | null | null | null | ['melanoma-diagnosis'] | ['computer-vision'] | [ 6.10544205e-01 4.92100120e-01 -1.71509370e-01 3.17900665e-02
-9.46989536e-01 -6.03694379e-01 5.71514726e-01 6.90704942e-01
-9.44784045e-01 6.05943561e-01 3.20488960e-01 -5.69421470e-01
-6.59145117e-02 -5.21217644e-01 -2.77591228e-01 -7.63136804e-01
2.13569015e-01 6.23877287e-01 3.66602421e-01 -3.07920814e-01
1.72988415e-01 3.60872626e-01 -1.34576011e+00 2.24721193e-01
8.66949975e-01 7.22647488e-01 6.33268952e-02 7.21966565e-01
-1.89396799e-01 3.99569362e-01 -5.21573365e-01 -7.40887344e-01
-9.78866816e-02 -3.17929655e-01 -7.23312855e-01 -7.67201632e-02
4.08939064e-01 -3.77878875e-01 7.97242150e-02 1.00735688e+00
3.77682209e-01 -5.39866090e-01 5.75773239e-01 -7.80815601e-01
-4.27125752e-01 2.87932873e-01 -3.68536592e-01 2.96510532e-02
2.75026798e-01 3.57620448e-01 6.99881554e-01 -2.74717808e-01
5.83032429e-01 9.66550589e-01 1.07948840e+00 1.05566299e+00
-8.77765477e-01 -2.61549115e-01 2.31595963e-01 -8.20290968e-02
-1.13513052e+00 -3.66516441e-01 -9.79296640e-02 -6.30917549e-01
7.47354031e-01 7.24556684e-01 9.77361441e-01 1.05241883e+00
1.91048831e-01 4.06164974e-01 9.98558521e-01 -5.45258284e-01
1.59104198e-01 2.81028360e-01 2.12801665e-01 8.34725559e-01
7.58693635e-01 -1.69289127e-01 -4.01214272e-01 -2.81880051e-01
3.89767408e-01 1.66673899e-01 -3.79409313e-01 -3.54112708e-03
-1.03254366e+00 7.14022577e-01 2.73169219e-01 3.26981336e-01
-2.93688715e-01 -4.35196720e-02 3.42860997e-01 -4.56552841e-02
6.75702512e-01 4.83408362e-01 -2.58317471e-01 -2.64617592e-01
-1.09838474e+00 -1.70435309e-02 7.02703238e-01 2.17000812e-01
1.56506583e-01 -4.91247714e-01 -4.72815096e-01 4.97037500e-01
3.83835196e-01 4.01304066e-01 5.10555744e-01 -1.63812324e-01
-1.36445656e-01 1.09894621e+00 1.70281529e-01 -5.70893407e-01
-6.09859943e-01 -6.41521454e-01 -9.35886860e-01 5.07640466e-02
6.50684118e-01 -1.93819761e-01 -1.25031698e+00 1.21441710e+00
3.39323357e-02 -2.82797161e-02 -1.87582165e-01 6.93212688e-01
5.43849945e-01 1.80036798e-01 4.61354226e-01 -7.80247599e-02
1.54705751e+00 -7.72416711e-01 -5.44074655e-01 -4.29485828e-01
7.72718728e-01 -4.43906099e-01 6.80016577e-01 6.87216222e-01
-8.31340432e-01 -7.94882104e-02 -8.34097564e-01 1.11675315e-01
-5.73527515e-01 -2.19026469e-02 8.71376336e-01 1.28029132e+00
-1.44280708e+00 5.30487239e-01 -9.67780709e-01 -9.77093101e-01
5.94428241e-01 5.19529104e-01 -5.28698623e-01 -1.68642759e-01
-9.46141183e-01 1.08765566e+00 9.19363052e-02 2.66019255e-01
-7.39732504e-01 -7.13849068e-01 -6.00812674e-01 -3.06861252e-01
1.79771110e-01 -7.77018011e-01 1.13714695e+00 -8.76593769e-01
-1.07819808e+00 1.16095507e+00 -2.15596646e-01 -6.37614667e-01
7.95225978e-01 -1.90414533e-01 -2.82743692e-01 1.45781755e-01
-1.81663245e-01 7.46165812e-01 7.19226003e-01 -7.91412771e-01
-8.76111031e-01 -4.12963569e-01 -2.91655026e-02 -2.46556234e-02
-6.86811149e-01 8.73336792e-02 -4.84106123e-01 2.99312267e-02
-2.33724132e-01 -9.56972480e-01 -7.40108013e-01 2.33041048e-01
-4.52768564e-01 -1.96202323e-01 2.12505117e-01 -9.02014256e-01
1.26012456e+00 -1.87525058e+00 1.07593626e-01 -2.03816012e-01
6.24670506e-01 6.78180158e-01 8.43826234e-02 5.17388463e-01
9.34454799e-02 4.47707325e-01 -2.48274982e-01 -3.31713974e-01
-2.46019945e-01 -1.23116516e-01 1.24213956e-01 5.48569739e-01
5.87522507e-01 8.32326055e-01 -1.13977158e+00 -2.84350514e-01
1.72041118e-01 6.74239635e-01 -1.86040103e-01 -2.13674739e-01
1.13283936e-02 2.53470123e-01 -2.07749501e-01 9.32926655e-01
4.72770512e-01 -5.47497332e-01 1.58925295e-01 1.84096783e-01
-1.92340434e-01 -1.36128545e-01 -6.70301914e-01 1.50032282e+00
1.13435672e-03 6.06819332e-01 9.49098542e-02 -5.14007211e-01
4.52951252e-01 1.52587220e-01 2.96221763e-01 -2.94729710e-01
6.66168928e-02 2.63540298e-01 2.04196051e-01 -6.07933879e-01
3.05434436e-01 -1.60942748e-01 8.09212103e-02 -7.98196197e-02
-3.29643339e-01 5.57423383e-02 1.07436948e-01 2.98898935e-01
1.58259809e+00 -2.30269283e-01 3.21160853e-01 -1.64565723e-02
4.31494355e-01 4.18792397e-01 1.95600674e-01 7.74532795e-01
-5.09355724e-01 5.24946749e-01 6.85065866e-01 -4.24163193e-01
-7.12722421e-01 -8.54231358e-01 -1.83405370e-01 7.88025379e-01
-1.90400347e-01 -3.71034443e-01 -9.27992880e-01 -8.17174494e-01
1.09831974e-01 2.93749779e-01 -9.44929540e-01 -2.05436885e-01
1.06227044e-02 -1.12926221e+00 5.66154599e-01 1.68771893e-01
2.33469322e-01 -4.62759972e-01 -8.76892686e-01 7.34138638e-02
7.91608095e-02 -6.61192775e-01 4.83800918e-02 2.34256294e-02
-5.95154762e-01 -1.27624941e+00 -1.24211764e+00 -4.54461455e-01
7.15748787e-01 7.68211111e-02 1.08603740e+00 4.68407840e-01
-6.41280591e-01 3.03849757e-01 -3.90918553e-01 -7.72524476e-01
-6.10445082e-01 1.10944390e-01 -2.09528074e-01 2.61127502e-02
5.55869162e-01 1.75914764e-01 -8.69301677e-01 -1.94862589e-01
-8.80936384e-01 1.18427210e-01 1.00776494e+00 7.67552316e-01
3.76173794e-01 -3.12718153e-01 3.15946877e-01 -1.12387061e+00
4.91045147e-01 -4.04623836e-01 -2.94406563e-01 1.74967393e-01
-5.39924562e-01 -2.81301111e-01 -2.64118910e-02 -2.81821400e-01
-6.04054809e-01 1.94467887e-01 -3.72394711e-01 1.63940474e-01
-3.59070659e-01 3.50193232e-01 3.18580955e-01 -1.89792290e-01
7.24379420e-01 6.25859126e-02 2.41806030e-01 -3.50090653e-01
-8.23131576e-03 7.68578231e-01 3.10495108e-01 3.70698571e-01
5.17931461e-01 8.57252836e-01 4.10805494e-02 -1.00538981e+00
-9.26441491e-01 -7.38892436e-01 -4.41868484e-01 -3.50973874e-01
9.19966519e-01 -8.63081396e-01 -5.41049302e-01 6.04635239e-01
-1.03482234e+00 -2.39394665e-01 -9.48410183e-02 8.26383308e-02
1.61279723e-01 5.00164151e-01 -4.16622818e-01 -1.07803667e+00
-5.29663920e-01 -1.04469800e+00 1.41309047e+00 1.61879703e-01
-4.63131875e-01 -1.02313161e+00 1.38910219e-01 4.10589844e-01
7.21272230e-01 4.13561404e-01 6.70025468e-01 -7.11313844e-01
-1.45841360e-01 -9.30192530e-01 -3.69701475e-01 2.35864788e-01
2.44192883e-01 2.28859499e-01 -1.14610147e+00 -2.40011320e-01
-4.40849453e-01 -4.78066280e-02 1.23794568e+00 3.56481731e-01
8.41383755e-01 1.32665783e-01 -8.26226711e-01 1.96496755e-01
1.40404761e+00 2.00514402e-02 5.07006347e-01 3.32626492e-01
3.68097782e-01 7.88672626e-01 1.57953590e-01 9.73954201e-02
3.04691792e-01 2.32381672e-01 7.82166958e-01 -3.03656101e-01
-2.29840547e-01 -6.89023286e-02 1.09229371e-01 3.07294279e-01
-1.99008122e-01 -3.04112375e-01 -1.39200068e+00 6.57491148e-01
-1.56146169e+00 -5.01483858e-01 -5.54989517e-01 2.47393417e+00
6.04305029e-01 4.61711138e-01 2.85388768e-01 6.79526180e-02
6.24227762e-01 -2.81071842e-01 -2.69217819e-01 -2.83081800e-01
-3.41784991e-02 1.48793161e-01 6.52934492e-01 3.30681801e-01
-1.16085100e+00 6.96502388e-01 6.92277861e+00 6.47540689e-01
-1.16318369e+00 1.40697733e-01 8.13696146e-01 -2.27525737e-02
-4.01306376e-02 -1.17618978e-01 -9.78547752e-01 4.51892197e-01
1.26475811e+00 1.80426776e-01 -1.33200496e-01 4.00060833e-01
1.64274171e-01 -5.20328045e-01 -1.07487607e+00 8.04436743e-01
2.32639626e-01 -1.33206987e+00 -1.51369497e-02 1.64265841e-01
3.17780018e-01 9.67789367e-02 4.74967696e-02 8.69114399e-02
2.75762714e-02 -1.27606642e+00 3.00853848e-01 8.51822913e-01
8.66044343e-01 -3.60648066e-01 1.15487146e+00 4.02261615e-01
-4.37732309e-01 -2.84287632e-02 -9.60333571e-02 -2.94076443e-01
-1.03534028e-01 6.85116470e-01 -1.16738141e+00 2.83860534e-01
4.88475651e-01 4.61912125e-01 -1.10819566e+00 1.44420850e+00
-3.60542089e-02 6.15059912e-01 -7.47209713e-02 -5.13869107e-01
3.80500436e-01 2.66029209e-01 2.18186006e-01 1.34412968e+00
2.13315621e-01 -2.09777281e-01 -8.64549801e-02 2.00780630e-01
3.22785974e-01 1.83320642e-01 -5.29448569e-01 -3.97102863e-01
1.40022356e-02 1.42196679e+00 -9.07841504e-01 -2.26182580e-01
-3.94172907e-01 1.18843770e+00 2.21750706e-01 -5.68378568e-02
-5.89941025e-01 -2.14716375e-01 3.58740419e-01 2.76848197e-01
-1.87829867e-01 2.64110357e-01 -1.30510688e-01 -8.71854246e-01
-1.80772990e-01 -6.90925002e-01 4.61062193e-01 -4.35851336e-01
-1.14225125e+00 5.58556855e-01 -5.90325832e-01 -8.70113015e-01
-1.00784108e-01 -1.07667160e+00 -4.25587088e-01 8.89207900e-01
-1.52820158e+00 -1.18549335e+00 -4.72884536e-01 9.13372934e-02
3.37927788e-01 1.74642622e-01 1.18514037e+00 -4.61804532e-02
-7.08867788e-01 4.54735100e-01 -2.64277235e-02 6.46988600e-02
7.52696574e-01 -1.29620111e+00 2.44859278e-01 5.97659469e-01
-3.66160780e-01 3.85501295e-01 7.16598868e-01 -7.63366282e-01
-1.53847075e+00 -8.85633826e-01 1.11727476e+00 -8.03954244e-01
8.46800089e-01 -3.40161622e-01 -7.37267971e-01 2.52122849e-01
1.43044323e-01 -4.34602171e-01 8.49336088e-01 2.17168897e-01
-2.31796086e-01 -6.50930926e-02 -1.14565921e+00 7.32658505e-01
7.10121453e-01 -1.96804836e-01 -2.76642114e-01 8.37947607e-01
5.13399899e-01 -2.57788539e-01 -5.18701255e-01 3.27267021e-01
6.98583782e-01 -1.14381695e+00 6.75966382e-01 -5.29458582e-01
4.05586064e-01 1.76052570e-01 2.89776981e-01 -1.03338015e+00
-2.04591542e-01 -6.00276053e-01 3.55050176e-01 7.21877456e-01
6.53774202e-01 -6.60033643e-01 1.21252477e+00 8.54851604e-01
2.87812818e-02 -9.15059090e-01 -9.97160912e-01 -5.26678205e-01
1.08818673e-01 -3.43665570e-01 4.14278135e-02 6.89980567e-01
8.78348500e-02 1.73740029e-01 -3.51238810e-02 1.24092653e-01
4.71221954e-01 -4.99547124e-01 5.02632082e-01 -1.55800951e+00
-1.97088704e-01 -6.09418452e-01 -6.63358033e-01 -2.05384314e-01
-4.33216721e-01 -7.10320652e-01 -8.85320008e-02 -1.84008515e+00
4.82825637e-01 -1.69382116e-03 -3.40093076e-01 5.73881567e-01
-2.14406163e-01 4.18576539e-01 -1.08454861e-01 -1.19633980e-01
-5.69525719e-01 -3.68134022e-01 9.81700599e-01 -2.99714267e-01
2.77068410e-02 6.80610240e-02 -1.03724611e+00 9.22849476e-01
7.39873350e-01 -2.36340389e-01 -4.06506993e-02 -4.06747818e-01
5.16903579e-01 -9.81621519e-02 5.63250959e-01 -1.04157281e+00
4.35738385e-01 9.26683620e-02 3.65858853e-01 -3.21593821e-01
4.27837968e-01 -6.43948555e-01 9.46234837e-02 1.16110110e+00
-1.53423309e-01 -1.48539022e-01 4.38739330e-01 7.52319634e-01
1.48985907e-01 -3.53553206e-01 5.64451516e-01 -2.63007402e-01
-6.14476204e-01 8.70073438e-02 -6.08102679e-01 -3.72832447e-01
1.07008266e+00 -2.16734841e-01 -5.59348643e-01 -1.50598288e-01
-7.00734556e-01 1.24467939e-01 4.25992399e-01 2.56322145e-01
4.94299471e-01 -5.16016543e-01 -8.93469870e-01 5.18594049e-02
3.18636864e-01 -2.14755729e-01 4.64558393e-01 1.13884807e+00
-7.61218905e-01 5.92816770e-01 1.49578616e-01 -5.79468727e-01
-1.46551383e+00 4.68984246e-01 3.85213852e-01 -5.14118016e-01
-4.68488395e-01 1.22565532e+00 -1.66035354e-01 2.06899241e-01
4.69918966e-01 -3.30702394e-01 -3.46945882e-01 2.28269801e-01
7.86316752e-01 2.52775103e-01 3.45757842e-01 -1.27285123e-01
-6.12121880e-01 3.44122857e-01 -3.72347414e-01 4.85995822e-02
1.16985226e+00 3.07022065e-01 -4.32930440e-01 3.28632534e-01
4.38642740e-01 1.23065412e-01 -8.36094081e-01 3.90085042e-01
1.32873833e-01 -2.34252959e-01 2.18321443e-01 -1.27337897e+00
-5.55838346e-01 1.03586626e+00 8.66443813e-01 5.37519693e-01
1.10656857e+00 -1.29572451e-01 4.63881671e-01 2.63985574e-01
2.11733729e-01 -6.14255667e-01 -3.67934108e-01 1.76395953e-01
4.58100319e-01 -1.40324759e+00 7.18387663e-02 -4.56460029e-01
-3.08717698e-01 9.87690806e-01 2.48672456e-01 3.10401972e-02
4.43144739e-01 1.66452244e-01 1.73919380e-01 -2.56626934e-01
-7.74096370e-01 -3.82350296e-01 1.66538790e-01 7.21585631e-01
5.29685140e-01 1.53756395e-01 -5.14528573e-01 5.55189788e-01
2.24422187e-01 8.38595629e-02 5.12808859e-01 7.44167626e-01
-6.14719868e-01 -1.12722969e+00 -4.04384017e-01 8.41508210e-01
-7.02714801e-01 -3.56615782e-01 -8.12887490e-01 8.75455201e-01
4.42499638e-01 1.06580997e+00 5.05800135e-02 -2.26557136e-01
2.99389482e-01 2.00653851e-01 3.32295954e-01 -7.38981366e-01
-7.18442678e-01 2.45599784e-02 2.32831180e-01 -2.92815894e-01
-3.39469910e-01 -6.74135923e-01 -7.10307777e-01 -9.39535629e-03
-2.99717247e-01 3.54831927e-02 9.14443254e-01 8.67561936e-01
2.62284040e-01 4.82112259e-01 4.11724597e-02 -5.11215031e-01
-5.73554873e-01 -8.58065724e-01 -4.81529057e-01 7.57486448e-02
5.57154298e-01 -2.88677126e-01 -3.06049019e-01 -1.21115953e-01] | [15.605868339538574, -2.9316909313201904] |
075a43fd-620b-4b90-a708-3fe41279accd | clustering-macroeconomic-time-series | 1807.04004 | null | http://arxiv.org/abs/1807.04004v2 | http://arxiv.org/pdf/1807.04004v2.pdf | Clustering Macroeconomic Time Series | The data mining technique of time series clustering is well established in
many fields. However, as an unsupervised learning method, it requires making
choices that are nontrivially influenced by the nature of the data involved.
The aim of this paper is to verify usefulness of the time series clustering
method for macroeconomics research, and to develop the most suitable
methodology.
By extensively testing various possibilities, we arrive at a choice of a
dissimilarity measure (compression-based dissimilarity measure, or CDM) which
is particularly suitable for clustering macroeconomic variables. We check that
the results are stable in time and reflect large-scale phenomena such as
crises. We also successfully apply our findings to analysis of national
economies, specifically to identifying their structural relations. | [] | 2018-07-18 | null | null | null | null | ['time-series-clustering'] | ['time-series'] | [-1.99747354e-01 -4.85275209e-01 -4.22848538e-02 -2.35169366e-01
-2.44502038e-01 -6.47042572e-01 8.81063402e-01 5.59001267e-01
-3.99201691e-01 5.46414435e-01 2.17062503e-01 -7.94102788e-01
-7.79687822e-01 -8.83550942e-01 -7.38785118e-02 -8.00969362e-01
-3.51449192e-01 5.29544711e-01 -1.85222179e-01 -2.59014815e-01
6.29818797e-01 6.55022919e-01 -1.76784801e+00 -1.80316791e-01
8.99830878e-01 7.27790534e-01 3.42647769e-02 3.21729869e-01
-1.53595507e-01 4.98124033e-01 -6.22919917e-01 -1.70122579e-01
1.70778796e-01 -5.71102440e-01 -8.20054293e-01 1.98136657e-01
-6.44382656e-01 5.53072751e-01 2.20611721e-01 9.49135303e-01
3.40191215e-01 4.99783486e-01 1.02321184e+00 -1.12197065e+00
-2.11912647e-01 6.75089717e-01 -3.73164237e-01 6.04193807e-01
2.95249492e-01 -2.52070278e-01 7.22841501e-01 -4.81248021e-01
6.27887011e-01 1.01804268e+00 5.74665844e-01 -2.31516704e-01
-1.16079366e+00 -4.79169577e-01 -1.99763164e-01 4.13602591e-01
-1.33910894e+00 -5.10607123e-01 9.62764502e-01 -5.84061027e-01
8.28083932e-01 3.08462918e-01 5.26309252e-01 5.83386600e-01
2.70984411e-01 -7.31950849e-02 1.65622783e+00 -8.90395284e-01
5.00426114e-01 -7.22584687e-03 1.97588786e-01 -8.13784171e-03
3.50706100e-01 4.54873629e-02 9.66304168e-02 -1.15297318e-01
4.76816952e-01 -3.26349996e-02 1.10201240e-01 3.02380952e-03
-1.01136625e+00 1.07707393e+00 -1.91339687e-01 1.12277901e+00
-4.51411366e-01 -4.79014456e-01 4.42493081e-01 9.29935455e-01
7.62925982e-01 4.04838204e-01 -6.10323310e-01 -4.91586208e-01
-9.43915427e-01 4.17976119e-02 8.74975264e-01 2.87211537e-01
4.84423667e-01 1.98022500e-02 4.74130809e-01 7.30741620e-01
1.09508924e-01 3.34982067e-01 9.70027268e-01 -8.28157842e-01
2.69781590e-01 6.57102764e-01 -1.40339747e-01 -1.36215711e+00
-5.96557498e-01 -2.57089674e-01 -9.81908441e-01 1.85717642e-01
6.35644794e-01 -2.99881250e-01 -3.51936132e-01 1.53922772e+00
2.33811498e-01 -1.65337652e-01 1.57662153e-01 3.57809693e-01
1.97091997e-01 4.69731778e-01 1.71967093e-02 -1.05273139e+00
1.09515488e+00 -2.64763117e-01 -7.93596506e-01 5.29186010e-01
7.84559608e-01 -8.16632390e-01 8.32814634e-01 5.78712702e-01
-9.07317340e-01 -6.60144389e-01 -4.36175734e-01 5.69996357e-01
-6.53089404e-01 -3.06245744e-01 5.89671016e-01 7.39393353e-01
-1.10950541e+00 7.90177882e-01 -8.84917080e-01 -5.88587701e-01
-4.12477583e-01 3.60575825e-01 -2.82488883e-01 6.47767186e-01
-1.10037911e+00 9.09290016e-01 5.49578786e-01 -2.57455945e-01
4.74294163e-02 -1.12969138e-01 -5.99791110e-01 1.03554204e-01
-1.43168727e-02 -2.58522719e-01 8.26606810e-01 -1.10825598e+00
-1.14945936e+00 7.73743629e-01 -1.36879772e-01 -6.68431580e-01
3.63842636e-01 6.42648578e-01 -9.23873246e-01 2.15054497e-01
1.37849078e-01 -1.79353774e-01 5.99624693e-01 -8.56587648e-01
-6.14538133e-01 -4.75975633e-01 -4.29319233e-01 -1.22585163e-01
-5.43525934e-01 4.06130344e-01 1.32100090e-01 -9.55399513e-01
3.04768294e-01 -7.67087579e-01 -4.14563447e-01 -1.17050552e+00
4.84570786e-02 -6.72040582e-01 5.51602364e-01 -6.64439619e-01
1.75210106e+00 -2.17225742e+00 -4.38303798e-02 8.02165091e-01
-1.36103764e-01 -2.52111107e-01 3.98693264e-01 8.70272756e-01
-5.00924885e-01 1.93573833e-01 -3.37996572e-01 -2.41038769e-01
1.24439649e-01 3.26066971e-01 -2.17472017e-01 7.99841046e-01
-5.10809235e-02 4.21727538e-01 -6.72155321e-01 -6.37539744e-01
4.82485145e-01 -5.09549230e-02 -2.63466567e-01 -1.75545439e-01
2.85211295e-01 4.38416272e-01 -3.58787686e-01 5.94159126e-01
4.48232144e-01 2.05175504e-02 3.87054771e-01 1.52512252e-01
-7.55785108e-01 2.59428024e-01 -1.22458375e+00 1.01463413e+00
-6.43682182e-02 7.57518947e-01 -4.70140845e-01 -1.80340636e+00
1.23656118e+00 5.00768542e-01 9.66769576e-01 -9.15828049e-01
3.83840501e-01 2.16595098e-01 2.99816102e-01 -6.56543612e-01
4.62754697e-01 -1.99027076e-01 -9.07540619e-02 6.71118438e-01
-1.69393182e-01 9.59939882e-02 6.65757477e-01 -2.22481385e-01
8.26401711e-01 -3.91452402e-01 4.85899597e-01 -8.09309542e-01
5.17152369e-01 1.49665013e-01 4.56331372e-01 3.57111782e-01
-1.49476394e-01 2.02810988e-01 6.41337991e-01 -3.72256041e-01
-1.18110406e+00 -6.25017405e-01 -4.19641435e-01 6.83564186e-01
-4.53733474e-01 -2.30929241e-01 -6.28740728e-01 -1.65808722e-01
-5.94907291e-02 5.01908004e-01 -5.97703934e-01 2.39959449e-01
-5.04757106e-01 -1.07454085e+00 1.82605192e-01 4.91275974e-02
9.32224542e-02 -1.26309848e+00 -7.73837149e-01 4.56590712e-01
-1.06484972e-01 -5.15132427e-01 1.90998584e-01 4.34239060e-01
-1.21759200e+00 -1.22556055e+00 -3.44465762e-01 -6.98219001e-01
4.80188876e-01 2.67396748e-01 1.22112894e+00 9.92780644e-03
1.79540545e-01 3.41551036e-01 -7.47173667e-01 -4.69554454e-01
-5.50848782e-01 2.97088712e-01 3.11696082e-01 5.44707365e-02
7.20272660e-01 -9.44408059e-01 -2.72198677e-01 2.26477697e-01
-1.03850842e+00 -6.81060553e-01 1.57201648e-01 4.23266858e-01
2.93139666e-01 1.05652583e+00 9.08329010e-01 -5.85703075e-01
9.93254542e-01 -7.88586020e-01 -6.62123024e-01 1.27535596e-01
-1.19414377e+00 -2.32588500e-01 6.68363690e-01 -3.63404185e-01
-8.76366377e-01 -1.58624962e-01 5.84077835e-02 -4.46064807e-02
-4.23024476e-01 1.08611953e+00 2.58734018e-01 1.43623114e-01
4.31335658e-01 1.38407201e-01 1.53970018e-01 -7.49912798e-01
5.33659533e-02 7.21992731e-01 4.88586426e-01 -6.00573421e-01
7.51171172e-01 5.32343268e-01 -1.33775165e-02 -9.41228688e-01
7.63600692e-02 -7.71921873e-01 -7.87121534e-01 -4.07520384e-01
7.15349615e-01 -6.54607594e-01 -7.25727081e-01 3.66201520e-01
-5.44791520e-01 -7.34371394e-02 -1.56119213e-01 7.58476973e-01
-4.58836257e-01 3.59703928e-01 -4.30327684e-01 -9.75098312e-01
1.30099222e-01 -7.83016384e-01 2.88873225e-01 1.10441111e-01
-4.39884156e-01 -1.65497255e+00 5.67650855e-01 8.93629193e-02
2.86904961e-01 7.49125123e-01 9.32685256e-01 -7.75326908e-01
2.34175086e-01 -4.17709053e-02 2.59072691e-01 2.25428026e-02
4.31402713e-01 5.61949015e-01 -6.02845013e-01 -3.18964601e-01
5.54318964e-01 1.47496849e-01 7.18769431e-01 5.77021718e-01
8.74551296e-01 -1.69523880e-01 -1.50464818e-01 2.89691836e-01
1.48623848e+00 7.61127710e-01 5.11170089e-01 8.74320090e-01
1.17547855e-01 1.10480380e+00 5.89551687e-01 7.49395192e-01
4.14202571e-01 3.58515292e-01 3.71320210e-02 -2.21133247e-01
6.30686164e-01 3.43066216e-01 1.66183859e-01 1.37275469e+00
-5.69165945e-01 1.21662043e-01 -1.13966739e+00 7.13875175e-01
-1.83900809e+00 -1.41017914e+00 -4.78699207e-01 2.07650638e+00
7.88883209e-01 2.72495091e-01 7.46879816e-01 9.17717874e-01
5.48995018e-01 -3.23431976e-02 1.13095477e-01 -8.04614723e-01
-3.70330930e-01 4.62196738e-01 3.80777806e-01 2.76998132e-01
-1.02803302e+00 3.72916460e-01 6.87866640e+00 6.48574293e-01
-9.90337253e-01 -2.14907527e-01 7.48271644e-01 3.28024924e-01
-2.93310940e-01 -5.86214736e-02 -1.11426666e-01 7.41765320e-01
1.30473483e+00 -6.03215158e-01 2.37457573e-01 4.47247595e-01
9.50202584e-01 -2.11415097e-01 -5.39001226e-01 6.48044586e-01
-3.20394129e-01 -9.48861301e-01 -2.99979150e-01 2.83986270e-01
8.37216914e-01 -3.33924979e-01 2.90969461e-02 1.06263198e-01
2.42373273e-01 -9.31150198e-01 4.70265210e-01 4.65734035e-01
2.25969464e-01 -1.15214467e+00 6.85418963e-01 2.61491984e-01
-1.24570537e+00 -2.00122625e-01 -2.56705135e-01 -7.20041871e-01
3.05012520e-02 8.25837016e-01 -3.37458104e-01 7.42534637e-01
7.86766529e-01 7.52396464e-01 -5.74753046e-01 9.73761618e-01
3.61542135e-01 8.68862033e-01 -2.00434223e-01 1.14833966e-01
3.96889716e-01 -8.74159813e-01 1.80742890e-01 1.19418466e+00
6.13898516e-01 1.74609184e-01 -1.56537861e-01 3.62761497e-01
4.81305063e-01 4.23135608e-01 -7.67092705e-01 -1.79823071e-01
6.55838609e-01 9.80441868e-01 -1.28318822e+00 -3.81584555e-01
-5.22732198e-01 3.18599671e-01 4.44741547e-02 1.76171556e-01
-4.70873743e-01 -2.56713361e-01 3.92831177e-01 -3.56041379e-02
2.12709755e-01 -5.73322117e-01 -4.07498807e-01 -1.08530545e+00
-3.75192799e-02 -1.06111562e+00 6.51145697e-01 -1.50456935e-01
-1.27171338e+00 4.11223739e-01 3.24819475e-01 -1.39077401e+00
-6.72935009e-01 -2.90616959e-01 -8.28879535e-01 6.43057644e-01
-1.18572080e+00 -4.14829731e-01 2.22391009e-01 9.45306659e-01
3.02644730e-01 -3.64424437e-01 7.51702607e-01 4.77540195e-01
-5.75274527e-01 1.26191139e-01 7.89661884e-01 -2.25442611e-02
4.94039476e-01 -1.36077714e+00 1.71304554e-01 8.79117608e-01
9.23665687e-02 6.99455023e-01 8.94377470e-01 -5.18819928e-01
-8.49630117e-01 -5.76413989e-01 1.58536696e+00 -2.44532183e-01
1.10609210e+00 2.65783537e-03 -8.55489254e-01 2.50658005e-01
3.48864704e-01 -7.22243845e-01 8.76462579e-01 2.41104439e-01
1.41706467e-01 -5.18390574e-02 -1.17913198e+00 4.15992290e-01
5.92883527e-01 -4.44292009e-01 -1.00671029e+00 2.69968748e-01
2.99857944e-01 3.26964200e-01 -1.29328895e+00 1.14732943e-01
2.52985597e-01 -1.30101645e+00 8.08595538e-01 -6.49617672e-01
2.57264078e-01 -6.97848201e-02 -5.12335189e-02 -1.19569933e+00
-5.27842224e-01 -7.48795450e-01 2.75091887e-01 1.51868916e+00
2.82912612e-01 -9.13675368e-01 4.94381785e-01 4.22363341e-01
5.43398485e-02 -3.72329175e-01 -1.01627767e+00 -9.27403688e-01
3.20999742e-01 -4.22034353e-01 7.81610250e-01 1.78754687e+00
5.66055238e-01 -8.06912705e-02 -4.84542772e-02 -2.36392736e-01
6.74865901e-01 3.98533046e-01 5.05306125e-01 -1.69166481e+00
-9.09799859e-02 -9.24983084e-01 -2.84172416e-01 -1.09735586e-01
1.54294088e-01 -4.19359177e-01 -5.16993761e-01 -9.56485152e-01
-2.88413346e-01 -5.93115568e-01 -5.24925053e-01 -7.64492974e-02
6.42513409e-02 1.86934918e-01 4.10917178e-02 5.44248343e-01
-1.08437695e-01 1.56564519e-01 5.47280610e-01 2.22821429e-01
-4.82286632e-01 2.75571913e-01 -5.89806199e-01 7.38796771e-01
1.13889980e+00 -3.43736678e-01 -4.59339976e-01 7.84970447e-02
2.05038890e-01 2.17028931e-01 -2.55083330e-02 -9.04651880e-01
2.20508665e-01 -4.82009679e-01 1.45847008e-01 -6.82280660e-01
-3.78185570e-01 -1.01881421e+00 5.40950775e-01 5.79900384e-01
-1.04090862e-01 9.55566108e-01 -9.02378410e-02 2.06858695e-01
-5.30547261e-01 -2.64452636e-01 5.39126575e-01 -1.80349380e-01
-5.72395980e-01 -3.74023579e-02 -7.20527112e-01 -1.02056675e-01
9.93631124e-01 -3.95584494e-01 1.58537865e-01 -2.97049671e-01
-7.69001186e-01 3.63866165e-02 6.71936452e-01 2.04597369e-01
9.80205983e-02 -1.28403795e+00 -5.48142672e-01 -8.35925434e-03
-2.54518092e-01 -6.09533966e-01 -1.27889782e-01 1.06273341e+00
-5.40844798e-01 6.67958677e-01 -2.35429883e-01 -2.97148764e-01
-1.21257663e+00 9.07858789e-01 3.57868820e-02 -2.63152927e-01
-6.26174808e-01 3.90022136e-02 -4.78947341e-01 6.66291863e-02
1.07455596e-01 -4.34390128e-01 -6.58742309e-01 7.42214322e-01
2.30724871e-01 6.20148659e-01 1.85233012e-01 -7.88442492e-01
-3.71373892e-01 5.32897532e-01 5.93013942e-01 -3.05270404e-01
1.63908994e+00 -4.88996983e-01 -4.18952048e-01 7.41629839e-01
1.12004268e+00 -7.60293603e-02 -7.25335121e-01 1.02590779e-02
8.06793511e-01 -3.12179983e-01 -2.21370980e-01 -1.45248577e-01
-9.28504705e-01 3.15013826e-01 4.47323173e-01 1.32785916e+00
1.54368508e+00 -2.50679135e-01 2.85247535e-01 1.81161687e-01
2.47373998e-01 -1.35286903e+00 -4.18947905e-01 3.30332786e-01
3.86333287e-01 -1.12269270e+00 -6.09483868e-02 1.34490207e-01
-2.49154150e-01 1.25184298e+00 -2.25430146e-01 -1.82375222e-01
9.99508202e-01 1.85495555e-01 1.29183650e-01 -2.99673557e-01
-7.52080262e-01 -3.75649363e-01 -1.19972872e-02 6.09048843e-01
5.06121457e-01 2.38050342e-01 -1.12763929e+00 1.52412608e-01
-6.30157351e-01 -3.22037846e-01 3.88590276e-01 8.61710370e-01
-5.12605190e-01 -1.33428192e+00 -7.27764249e-01 4.18251723e-01
-6.80382729e-01 1.00360058e-01 -4.50845689e-01 1.15450096e+00
1.95886061e-01 1.40663505e+00 3.19981515e-01 -2.94117868e-01
1.35844842e-01 1.12656571e-01 2.75282841e-03 -7.72783309e-02
-8.48855317e-01 4.32949454e-01 1.17262006e-02 -1.03276990e-01
-1.21752357e+00 -1.29101503e+00 -9.33291972e-01 -8.25749099e-01
-2.32345641e-01 7.56408393e-01 6.15955591e-01 1.22437108e+00
-8.12398419e-02 2.38138616e-01 1.28463936e+00 -5.09784162e-01
-2.39416435e-01 -1.05387318e+00 -9.01528537e-01 6.62238896e-01
8.62512887e-02 -5.99173784e-01 -5.62452137e-01 3.93040359e-01] | [7.235757827758789, 3.410902976989746] |
447da6b9-47c1-4e65-bcf1-ac0784ec1934 | uni6d-a-unified-cnn-framework-without | 2203.14531 | null | https://arxiv.org/abs/2203.14531v2 | https://arxiv.org/pdf/2203.14531v2.pdf | Uni6D: A Unified CNN Framework without Projection Breakdown for 6D Pose Estimation | As RGB-D sensors become more affordable, using RGB-D images to obtain high-accuracy 6D pose estimation results becomes a better option. State-of-the-art approaches typically use different backbones to extract features for RGB and depth images. They use a 2D CNN for RGB images and a per-pixel point cloud network for depth data, as well as a fusion network for feature fusion. We find that the essential reason for using two independent backbones is the "projection breakdown" problem. In the depth image plane, the projected 3D structure of the physical world is preserved by the 1D depth value and its built-in 2D pixel coordinate (UV). Any spatial transformation that modifies UV, such as resize, flip, crop, or pooling operations in the CNN pipeline, breaks the binding between the pixel value and UV coordinate. As a consequence, the 3D structure is no longer preserved by a modified depth image or feature. To address this issue, we propose a simple yet effective method denoted as Uni6D that explicitly takes the extra UV data along with RGB-D images as input. Our method has a Unified CNN framework for 6D pose estimation with a single CNN backbone. In particular, the architecture of our method is based on Mask R-CNN with two extra heads, one named RT head for directly predicting 6D pose and the other named abc head for guiding the network to map the visible points to their coordinates in the 3D model as an auxiliary module. This end-to-end approach balances simplicity and accuracy, achieving comparable accuracy with state of the arts and 7.2x faster inference speed on the YCB-Video dataset. | ['Liwei Wu', 'Rui Zhao', 'Ye Zheng', 'Hao Chen', 'Donghai Li', 'Xiaoke Jiang'] | 2022-03-28 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Jiang_Uni6D_A_Unified_CNN_Framework_Without_Projection_Breakdown_for_6D_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Jiang_Uni6D_A_Unified_CNN_Framework_Without_Projection_Breakdown_for_6D_CVPR_2022_paper.pdf | cvpr-2022-1 | ['6d-pose-estimation-1'] | ['computer-vision'] | [ 1.26542151e-01 1.08427003e-01 -3.17063145e-02 -4.99066472e-01
-5.22396088e-01 -4.89085257e-01 2.17918977e-01 -2.07967237e-01
-5.39557397e-01 2.46785864e-01 -5.66560067e-02 -2.97955066e-01
3.67587268e-01 -9.22780275e-01 -9.42055762e-01 -6.06652677e-01
3.26882988e-01 1.57064274e-01 3.32290381e-01 -1.06009841e-01
2.62200445e-01 9.34805572e-01 -1.55751884e+00 8.65026563e-02
2.63486475e-01 1.65971410e+00 -1.32255390e-01 5.48412442e-01
-7.65730813e-02 1.74144909e-01 -4.95086044e-01 -3.13640833e-01
8.88875604e-01 -1.32606644e-02 -5.08564055e-01 4.28883694e-02
6.38188779e-01 -8.54372978e-01 -5.77879071e-01 8.25371623e-01
6.95127845e-01 -1.01241998e-01 2.40406707e-01 -1.17444980e+00
-2.87096977e-01 -1.88958138e-01 -7.96456814e-01 -3.95342559e-01
5.29653609e-01 2.19238058e-01 6.49489582e-01 -8.71858120e-01
6.78921640e-01 1.13895500e+00 7.68082738e-01 5.16893268e-01
-9.93561983e-01 -5.43033838e-01 3.39785963e-02 -1.06359065e-01
-1.43041086e+00 -1.01035014e-01 9.85188365e-01 -1.71061262e-01
1.18838453e+00 2.67481178e-01 9.14119244e-01 7.85019040e-01
2.67156035e-01 4.30586874e-01 8.75959694e-01 -3.06046307e-01
2.10507646e-01 -3.23790818e-01 -2.14693904e-01 7.91453123e-01
1.31273434e-01 9.55523998e-02 -7.15960026e-01 1.33751750e-01
1.22438705e+00 3.47298443e-01 -3.31067860e-01 -7.61628032e-01
-1.04306674e+00 5.08307576e-01 9.22016621e-01 -2.77959228e-01
-2.36522719e-01 4.17053401e-01 -1.03830136e-01 1.11886337e-01
3.89446169e-01 1.31216615e-01 -7.56872714e-01 -1.18399039e-01
-6.89036489e-01 2.77701557e-01 4.46216226e-01 8.83440793e-01
1.09953558e+00 -5.25911808e-01 2.59147167e-01 3.56404513e-01
4.65530843e-01 5.70220053e-01 3.50173801e-01 -1.08171225e+00
6.31531596e-01 9.83714938e-01 6.08030446e-02 -1.02941501e+00
-6.27395749e-01 -2.84662008e-01 -6.57345235e-01 6.22201502e-01
3.92299414e-01 1.09890856e-01 -1.24872363e+00 1.48800790e+00
6.16910219e-01 9.80821252e-03 -1.18189037e-01 1.12129915e+00
8.15675914e-01 4.34699506e-01 -6.99790955e-01 2.81184554e-01
1.21966863e+00 -7.33060658e-01 -2.48361975e-01 -2.50645518e-01
4.88620818e-01 -6.03793859e-01 8.61570179e-01 2.36214221e-01
-1.12341344e+00 -5.81020057e-01 -1.32205904e+00 -7.73896873e-01
-5.24282455e-01 2.50626709e-02 6.22722924e-01 4.26106662e-01
-1.07401216e+00 6.20456398e-01 -9.71264899e-01 -2.33032197e-01
3.37946981e-01 7.44279623e-01 -7.32340932e-01 -1.82791829e-01
-8.11109304e-01 7.97958255e-01 1.77310035e-01 3.50418597e-01
-3.45510423e-01 -6.40496910e-01 -1.01070511e+00 -3.45549062e-02
2.13565513e-01 -7.29957700e-01 1.11384404e+00 -6.20662987e-01
-1.76334763e+00 1.00240099e+00 -6.29030764e-02 -1.16839826e-01
5.57595909e-01 -3.31661016e-01 2.58612901e-01 1.65386036e-01
-1.58180714e-01 8.75199616e-01 9.20061469e-01 -1.26062357e+00
-5.85847437e-01 -9.55156028e-01 2.16072485e-01 3.72780085e-01
-1.03482962e-01 -4.67298031e-01 -1.02087045e+00 -3.20740461e-01
9.97110903e-01 -9.70215738e-01 -1.31697908e-01 6.18857682e-01
-4.39315110e-01 2.28689268e-01 9.41389263e-01 -4.48927790e-01
8.28601003e-01 -2.33872366e+00 2.14033559e-01 3.65133613e-01
4.59184408e-01 2.85791345e-02 1.13720097e-01 2.08356008e-02
-8.19048658e-02 -7.71194100e-02 -4.85840626e-02 -9.14866567e-01
-9.18458551e-02 2.50578016e-01 -1.69825584e-01 6.63368523e-01
3.17385852e-01 9.18374479e-01 -6.98756218e-01 -1.37181059e-01
5.43005705e-01 9.04799163e-01 -6.40720844e-01 1.87214896e-01
3.70323695e-02 4.07061219e-01 -4.05649781e-01 8.55700135e-01
1.11746323e+00 -1.52785316e-01 -1.13545530e-01 -5.90407431e-01
-2.88600117e-01 3.00708055e-01 -1.19576502e+00 2.24741650e+00
-2.27221742e-01 4.64734614e-01 1.13632381e-01 -4.41884607e-01
1.07393730e+00 -2.75404658e-02 6.10435665e-01 -8.63923192e-01
3.70547324e-01 2.55508870e-01 -4.18138832e-01 -1.27264768e-01
5.21354795e-01 1.47737101e-01 -6.40867576e-02 2.57207870e-01
-3.28756459e-02 -4.84173357e-01 -5.53360462e-01 -1.91198140e-01
9.97057319e-01 5.67257047e-01 7.40736723e-02 2.91739851e-01
2.58650184e-01 -8.64942223e-02 6.51576102e-01 3.09389144e-01
-4.10696715e-02 1.07796919e+00 5.54973662e-01 -6.06451690e-01
-9.78356063e-01 -1.08035874e+00 -3.10303848e-02 2.62747973e-01
3.56621593e-01 -3.18069905e-01 -6.48540020e-01 -5.05544960e-01
3.03307742e-01 1.39847964e-01 -6.73342168e-01 -2.11643785e-01
-6.71234310e-01 -3.82862866e-01 2.81376749e-01 6.33334637e-01
8.37250113e-01 -4.93784010e-01 -1.15526545e+00 -1.34448099e-04
1.08065255e-01 -1.16528356e+00 -1.23002857e-01 6.46156788e-01
-1.10764134e+00 -1.02449560e+00 -5.72293818e-01 -3.51636291e-01
8.31656694e-01 3.76056224e-01 7.33137429e-01 -2.70353314e-02
-1.12126991e-01 7.87850618e-02 -3.98084313e-01 -2.37575978e-01
3.65174562e-01 4.88353409e-02 -7.84331858e-02 -1.41149685e-01
2.44935378e-01 -6.30245864e-01 -9.52483654e-01 2.47818694e-01
-8.88877690e-01 2.64130741e-01 5.02824605e-01 4.55632687e-01
7.47076094e-01 -2.41598383e-01 -4.43078250e-01 -8.50693062e-02
-1.62258625e-01 4.63291779e-02 -7.40303218e-01 -2.65722692e-01
-2.37187594e-01 -8.27942342e-02 2.73601979e-01 -8.14478174e-02
-6.86953902e-01 6.52137101e-01 -3.05840790e-01 -7.79588938e-01
-6.49080351e-02 2.17592105e-01 -3.88866365e-01 -2.89289415e-01
3.89109880e-01 -3.37828062e-02 1.92537874e-01 -6.97268784e-01
3.01782131e-01 5.77349305e-01 6.10540569e-01 -1.62375510e-01
8.51233959e-01 9.92697656e-01 3.48136634e-01 -5.33891022e-01
-8.17734659e-01 -2.79480070e-01 -1.09115958e+00 -1.74904794e-01
1.10391259e+00 -1.01450455e+00 -8.90308082e-01 7.74884760e-01
-1.39893544e+00 -2.81091750e-01 -2.38152310e-01 4.70684826e-01
-4.23379689e-01 7.68234655e-02 -5.63905478e-01 -3.46629262e-01
-2.10246816e-01 -1.45599222e+00 1.55443585e+00 2.56053180e-01
4.96851765e-02 -4.36276108e-01 -6.87079728e-02 2.63969481e-01
1.12754866e-01 5.89332700e-01 7.05737293e-01 1.55086488e-01
-9.25928533e-01 -3.26953232e-01 -4.18735653e-01 3.12417865e-01
9.59072486e-02 -1.98569242e-02 -1.21395612e+00 -1.61248893e-01
7.91642219e-02 -1.21379361e-01 6.74442410e-01 3.20915401e-01
1.18541908e+00 1.09216981e-01 -1.40575334e-01 1.40001845e+00
1.59827685e+00 9.01679769e-02 7.19260097e-01 5.55572212e-01
1.04515505e+00 4.12324071e-01 3.77219558e-01 3.25028300e-01
4.90800828e-01 8.73930097e-01 1.02838922e+00 -4.05467749e-01
-9.30209532e-02 -3.51023227e-01 2.10850477e-01 5.70037961e-01
-1.70277730e-01 3.15419369e-04 -8.74877393e-01 -2.98812869e-03
-1.73950219e+00 -3.27421963e-01 9.05710924e-03 2.38542819e+00
6.51081741e-01 2.34841362e-01 -2.69935280e-01 2.22436026e-01
2.80904442e-01 1.22129247e-01 -7.75966585e-01 -2.18372345e-01
-7.47947842e-02 4.38882947e-01 8.91688168e-01 4.20521826e-01
-8.95548940e-01 8.73713613e-01 5.40032339e+00 3.25197697e-01
-1.56654119e+00 -2.32369035e-01 5.33944666e-01 -3.43397617e-01
7.99334943e-02 -3.27109452e-03 -8.31619382e-01 1.79378316e-01
2.56498367e-01 6.43455863e-01 2.80196160e-01 8.01837325e-01
1.11107431e-01 -3.12208146e-01 -1.19196081e+00 1.45175731e+00
9.29374546e-02 -1.18515110e+00 -6.53570518e-02 3.73938173e-01
4.75123405e-01 1.70815021e-01 6.72820583e-02 -3.07483137e-01
-2.70327210e-01 -9.48447704e-01 1.05011952e+00 4.01691586e-01
1.02734971e+00 -7.33475208e-01 7.20748603e-01 1.95450217e-01
-1.04381669e+00 1.25398412e-02 -3.84903938e-01 -3.29387397e-01
1.44980893e-01 5.82294583e-01 -5.76888204e-01 6.04944587e-01
1.03432643e+00 6.62727654e-01 -5.65737367e-01 7.31878221e-01
-3.39710832e-01 -1.92140922e-01 -5.26043534e-01 3.14618051e-01
1.69653609e-01 -8.49773958e-02 2.03691170e-01 5.70277333e-01
5.35576522e-01 1.29681841e-01 -2.49995872e-01 7.27475762e-01
-1.14748180e-01 -4.01089519e-01 -4.58987653e-01 4.37091678e-01
3.28703403e-01 1.13227987e+00 -9.01119828e-01 -1.33533269e-01
-4.23710823e-01 1.25227380e+00 2.32222676e-01 2.14173079e-01
-8.20196807e-01 -4.12948430e-01 1.05628181e+00 1.06244750e-01
5.74674606e-01 -5.90884745e-01 -7.49308407e-01 -1.02579105e+00
3.23912472e-01 -5.55480719e-01 -8.98320675e-02 -1.03574622e+00
-8.25033367e-01 4.84038174e-01 -1.79059669e-01 -1.38665593e+00
-7.76327914e-03 -1.09058940e+00 -1.78839847e-01 9.05761361e-01
-1.68159580e+00 -1.03329396e+00 -8.36892366e-01 8.07798505e-01
2.32667308e-02 4.26899374e-01 6.72156632e-01 2.83392012e-01
-4.48184818e-01 4.12130117e-01 -2.86874652e-01 3.28041226e-01
6.67592049e-01 -1.05702424e+00 6.03490114e-01 5.76177835e-01
-2.14749083e-01 6.30848408e-01 2.74704069e-01 -5.32212794e-01
-1.94177926e+00 -8.30158710e-01 7.28452027e-01 -6.07554555e-01
1.49687037e-01 -6.19249463e-01 -4.96778727e-01 6.17201507e-01
-3.35007846e-01 2.76942074e-01 3.20901215e-01 -3.03261876e-01
-4.45928156e-01 -3.49200159e-01 -1.21581686e+00 5.63908577e-01
1.24988461e+00 -5.78595161e-01 -3.77709776e-01 -3.26430239e-02
1.00971270e+00 -1.16894627e+00 -8.36682379e-01 5.16415298e-01
7.59250045e-01 -1.36260748e+00 1.22941196e+00 1.62964575e-02
5.43153465e-01 -6.86699569e-01 -4.69320059e-01 -9.13195431e-01
-1.95055772e-02 -4.52939659e-01 -2.37237960e-01 7.52238989e-01
1.80508867e-01 -5.98450899e-01 1.04992259e+00 7.87499905e-01
-2.18652070e-01 -1.05076873e+00 -1.16121995e+00 -3.35016131e-01
-2.78497607e-01 -7.42823064e-01 1.00445914e+00 5.85909665e-01
-4.01235968e-01 1.48162708e-01 3.00200805e-02 3.79004091e-01
4.68475789e-01 1.46765038e-01 1.18300653e+00 -1.09619462e+00
2.83909235e-02 -3.67446125e-01 -7.07554519e-01 -1.54941964e+00
-3.49770010e-01 -4.85870183e-01 -5.44413216e-02 -1.43502355e+00
-2.49658182e-01 -4.70556647e-01 9.45567247e-03 6.28945231e-01
1.80945590e-01 5.82468271e-01 2.54482180e-01 1.00746006e-01
-1.21271558e-01 5.16769171e-01 1.39649308e+00 1.75253391e-01
-4.56419051e-01 -1.88922480e-01 -3.42143059e-01 7.80238092e-01
6.32780492e-01 -1.98400542e-01 -1.29296184e-01 -8.39154422e-01
4.91787910e-01 -6.10340796e-02 6.39227152e-01 -1.07035625e+00
2.95796931e-01 1.51244402e-01 9.12667572e-01 -1.05129838e+00
9.30043221e-01 -1.09502387e+00 4.37635891e-02 3.53679508e-01
2.55935282e-01 3.03091079e-01 7.24099576e-02 1.78665742e-01
-1.15089886e-01 1.57678291e-01 6.22570634e-01 -2.45842233e-01
-5.37907720e-01 6.32310688e-01 3.42442721e-01 -4.91957098e-01
9.73452508e-01 -8.02489877e-01 -2.00498611e-01 -2.32050523e-01
-3.48806977e-01 -3.51491719e-02 1.02678275e+00 3.18225801e-01
8.77567053e-01 -1.24254477e+00 -1.93878934e-01 6.56910598e-01
5.75987212e-02 9.11668837e-01 1.39069811e-01 6.77223921e-01
-9.61108327e-01 3.43396634e-01 -2.46789545e-01 -9.30225432e-01
-1.08623350e+00 2.80722052e-01 4.15853083e-01 5.41118383e-02
-6.56524539e-01 9.95264232e-01 -2.38014758e-02 -5.95285416e-01
3.42882097e-01 -8.01228702e-01 3.37185264e-01 -3.83354649e-02
3.87059391e-01 1.75981268e-01 3.77405405e-01 -6.73940301e-01
-4.02161956e-01 1.07706070e+00 3.00485939e-01 -1.99067146e-01
1.48066974e+00 -5.99217601e-02 -3.06253284e-01 2.39317015e-01
1.60243571e+00 -6.79387227e-02 -1.77471566e+00 -1.46928867e-02
-5.00286758e-01 -7.75025070e-01 3.21968228e-01 -5.07240176e-01
-1.29818356e+00 9.43524599e-01 6.01490676e-01 -9.04456899e-02
1.27206481e+00 -3.23581472e-02 8.51640582e-01 2.32757211e-01
5.35927951e-01 -1.02124417e+00 -2.02365895e-03 6.08000457e-01
7.79496670e-01 -1.19833088e+00 2.88293779e-01 -5.64447105e-01
-1.21655658e-01 1.27983773e+00 6.93427682e-01 -1.36847004e-01
6.25293553e-01 1.92067310e-01 2.03290820e-01 -3.24996710e-01
-1.35959849e-01 -3.82274799e-02 2.95612782e-01 4.75845248e-01
3.15358043e-02 -1.44342899e-01 2.06483498e-01 2.33811989e-01
-4.75955278e-01 -1.92394510e-01 5.05354106e-02 1.19889402e+00
-1.40800267e-01 -1.02776539e+00 -6.01438880e-01 6.73158318e-02
-9.59787220e-02 7.85200819e-02 -4.66020375e-01 8.67686212e-01
4.91483986e-01 5.68122745e-01 4.53189254e-01 -7.00692236e-01
5.17451644e-01 -1.27760023e-01 7.24246919e-01 -5.32962680e-01
-4.89851683e-01 4.65057828e-02 -4.68679190e-01 -1.33916545e+00
-4.11569923e-01 -5.32403946e-01 -1.42395747e+00 -3.39531451e-01
-1.56767458e-01 -6.37086868e-01 1.22007287e+00 7.77143955e-01
5.84212959e-01 2.92699188e-01 7.18035758e-01 -1.40053189e+00
-1.83753416e-01 -5.72335482e-01 -4.38553989e-01 5.28280139e-02
6.09852016e-01 -6.46281898e-01 -3.37595612e-01 -1.96324199e-01] | [7.97607946395874, -2.7779479026794434] |
b1c3733d-cc9c-4fa6-82c8-6d6338a28986 | self-supervised-neural-topic-modeling | null | null | https://aclanthology.org/2021.findings-emnlp.284 | https://aclanthology.org/2021.findings-emnlp.284.pdf | Self-Supervised Neural Topic Modeling | Topic models are useful tools for analyzing and interpreting the main underlying themes of large corpora of text. Most topic models rely on word co-occurrence for computing a topic, i.e., a weighted set of words that together represent a high-level semantic concept. In this paper, we propose a new light-weight Self-Supervised Neural Topic Model (SNTM) that learns a rich context by learning a topic representation jointly from three co-occurring words and a document that the triple originates from. Our experimental results indicate that our proposed neural topic model, SNTM, outperforms previously existing topic models in coherence metrics as well as document clustering accuracy. Moreover, apart from the topic coherence and clustering performance, the proposed neural topic model has a number of advantages, namely, being computationally efficient and easy to train. | ['Carsten Eickhoff', 'Martin Jaggi', 'Seyed Ali Bahrainian'] | null | null | null | null | findings-emnlp-2021-11 | ['topic-models'] | ['natural-language-processing'] | [-4.79281694e-02 1.94932714e-01 -4.70711261e-01 -4.90818977e-01
-6.89456403e-01 -9.20494646e-02 9.16723549e-01 6.08133912e-01
-9.46643874e-02 5.23706555e-01 6.66140556e-01 -1.52487420e-02
-3.51692468e-01 -1.01252115e+00 -2.99913228e-01 -8.81842017e-01
-2.67260909e-01 5.99009573e-01 1.13811120e-01 2.49085665e-01
3.96184713e-01 -4.60295916e-01 -1.57653451e+00 8.87446105e-02
1.02880311e+00 9.04415488e-01 6.54712141e-01 8.36924240e-02
-5.88730156e-01 4.61191475e-01 -7.03786373e-01 -1.23509625e-02
-5.72432399e-01 -2.75171161e-01 -8.00511479e-01 3.38458747e-01
2.05735475e-01 4.03922684e-02 -2.86739357e-02 8.81638885e-01
3.03231347e-02 5.74719846e-01 8.14105630e-01 -1.26803684e+00
-5.80703676e-01 9.97485101e-01 -6.22709572e-01 3.32466438e-02
-1.50210485e-01 -6.03167117e-01 1.53810370e+00 -9.09617782e-01
7.33621120e-01 1.25104022e+00 4.70113367e-01 4.41277400e-02
-1.10472977e+00 -6.75325811e-01 5.14979243e-01 1.94963813e-01
-1.44600070e+00 1.87023655e-01 1.11713660e+00 -5.27306437e-01
8.80256772e-01 -3.22594523e-01 7.44528770e-01 1.19822514e+00
4.68242168e-01 6.48710787e-01 8.20777774e-01 -1.98887452e-01
6.26718879e-01 1.63285449e-01 7.53293395e-01 3.54174912e-01
4.83389169e-01 -7.42726147e-01 -8.05282295e-01 -5.25958002e-01
3.77051800e-01 1.45755440e-01 -1.06175542e-01 -5.55564880e-01
-1.25266445e+00 1.35887527e+00 2.57967770e-01 7.47777104e-01
-7.44588196e-01 3.68166476e-01 5.22954464e-01 -2.61572391e-01
1.14575422e+00 2.30443388e-01 -2.16523439e-01 -9.54896659e-02
-1.32951200e+00 2.90201068e-01 6.48404241e-01 7.59518564e-01
9.41224933e-01 6.03105268e-03 -2.86480486e-02 8.67932200e-01
6.46078765e-01 2.61532962e-01 9.24125910e-01 -6.06003702e-01
2.83570856e-01 5.94512165e-01 -6.48114383e-02 -1.42273939e+00
-3.90896231e-01 -5.73534310e-01 -7.82667279e-01 -3.99810702e-01
-3.24687809e-01 -1.34996936e-01 -6.73722744e-01 1.87022769e+00
1.32113054e-01 3.73213977e-01 1.80182904e-01 6.04713321e-01
9.29063201e-01 1.17947412e+00 5.32895684e-01 -3.26060116e-01
1.46253300e+00 -6.83976889e-01 -1.11919737e+00 2.20819991e-02
3.95496428e-01 -5.64185500e-01 9.10079956e-01 3.56621385e-01
-7.72862077e-01 -3.53092223e-01 -9.40959513e-01 -4.03676815e-02
-6.40873373e-01 9.90997851e-02 9.21059251e-01 3.99263829e-01
-1.22888911e+00 2.75054872e-01 -9.01159108e-01 -7.02302456e-01
3.17336261e-01 2.07380846e-01 -4.58542593e-02 1.68259323e-01
-1.25000191e+00 5.70748985e-01 9.98571217e-01 -4.65127289e-01
-9.17953193e-01 -6.82251573e-01 -9.11375761e-01 6.24691904e-01
3.56431961e-01 -5.84161580e-01 9.76323903e-01 -5.96343219e-01
-1.23299837e+00 5.01563966e-01 -6.38347864e-01 -7.27069855e-01
-4.45226997e-01 -4.50948596e-01 -3.95069599e-01 1.91960320e-01
2.79132783e-01 8.21990132e-01 8.40031922e-01 -1.51463532e+00
-9.38943148e-01 -2.97927320e-01 -3.78588915e-01 2.88736463e-01
-1.07796359e+00 -3.99242016e-03 -3.80357802e-01 -8.25263679e-01
2.89722800e-01 -5.89480996e-01 -2.53306981e-02 -3.97349834e-01
-5.80724895e-01 -1.03667557e+00 1.37221563e+00 -4.85649616e-01
1.40580916e+00 -1.97966635e+00 1.29262209e-01 3.36125314e-01
5.36245465e-01 -3.12711924e-01 2.77414978e-01 5.89029014e-01
9.87828430e-03 1.84577435e-01 -2.80844420e-01 -4.43007022e-01
4.83707376e-02 6.10667933e-03 -8.49315405e-01 2.69650221e-01
3.23651992e-02 5.74645042e-01 -9.95213211e-01 -6.67877197e-01
1.37983590e-01 6.80313468e-01 -3.04785639e-01 6.25299523e-04
-6.02688849e-01 -2.07119107e-01 -5.46986163e-01 1.38900295e-01
2.86606491e-01 -4.83788341e-01 3.43781203e-01 1.45899534e-01
1.09254140e-02 6.40676141e-01 -9.56977308e-01 1.98476887e+00
-3.94133598e-01 1.05582774e+00 -3.84350121e-01 -1.09646344e+00
1.26560855e+00 7.43373871e-01 7.48907924e-01 -7.94337839e-02
1.28397018e-01 -1.14206769e-01 -2.73345679e-01 -1.39404982e-01
1.16357923e+00 -2.42334306e-01 -1.61661610e-01 1.10280943e+00
3.86169761e-01 3.57969478e-02 1.51447311e-01 5.42262793e-01
4.87720549e-01 -3.08795720e-01 3.62970173e-01 -7.96019018e-01
1.21576954e-02 -4.93057212e-03 3.53296608e-01 3.93867999e-01
2.94740677e-01 3.90306801e-01 5.54602981e-01 -1.71420142e-01
-1.00326824e+00 -1.03821754e+00 -2.32606709e-01 1.19163477e+00
1.18982397e-01 -7.98831522e-01 -6.94335759e-01 -2.41089657e-01
-2.68955827e-01 9.54323351e-01 -7.44514525e-01 -2.64343657e-02
-2.74554282e-01 -5.88912129e-01 1.70206428e-01 5.95767796e-01
2.92675048e-01 -8.49922061e-01 -5.70618987e-01 3.19440514e-01
-5.28859496e-01 -8.80099416e-01 -3.12996119e-01 1.80183887e-01
-1.32471883e+00 -7.26179898e-01 -7.36329377e-01 -8.56470227e-01
4.65810865e-01 7.75832772e-01 1.10384238e+00 -2.83132106e-01
9.27210897e-02 1.47373274e-01 -3.08511794e-01 -4.53799725e-01
1.90629855e-01 4.72194046e-01 1.08184017e-01 -1.35173155e-02
6.99020386e-01 -7.03185081e-01 -3.10191512e-01 -4.68797609e-02
-9.60209966e-01 1.18448570e-01 2.22503424e-01 7.26125121e-01
3.98042679e-01 6.78059757e-01 7.41987526e-01 -9.00398850e-01
1.05798280e+00 -9.86579359e-01 -2.23147973e-01 1.56339526e-01
-7.99129367e-01 -7.21150339e-02 1.64595127e-01 -3.43220770e-01
-1.35350919e+00 -4.76786375e-01 4.59840655e-01 -1.50968328e-01
-1.48375466e-01 9.00660872e-01 5.07472679e-02 9.14310932e-01
3.50389510e-01 3.20236504e-01 -3.89650404e-01 -3.12008291e-01
6.93883717e-01 6.04262292e-01 4.00774866e-01 -7.60319769e-01
5.03777087e-01 7.96549857e-01 -4.63792950e-01 -9.98591900e-01
-8.07284415e-01 -9.31774795e-01 -4.90784109e-01 -6.64637238e-02
1.02623796e+00 -1.12391078e+00 -2.44377837e-01 2.45456561e-01
-1.34122360e+00 1.60331160e-01 -1.93645656e-01 7.20597148e-01
-5.07320940e-01 2.68139690e-01 -3.49912107e-01 -7.56164908e-01
-6.49716258e-01 -5.55192232e-01 1.17589164e+00 2.72397041e-01
-5.61170101e-01 -1.60734212e+00 2.43433058e-01 -8.49997178e-02
4.88879889e-01 2.29622096e-01 1.21156800e+00 -8.86141360e-01
-2.48241827e-01 -7.80945458e-03 -3.42206150e-01 -1.06606185e-01
4.70264465e-01 -8.66503716e-02 -9.48990703e-01 -2.40812108e-01
7.71057159e-02 -1.33268982e-01 1.15320027e+00 7.60093987e-01
1.02039838e+00 -2.38306731e-01 -7.35432982e-01 8.00099503e-03
1.41108131e+00 2.14987963e-01 2.32035935e-01 3.37909073e-01
5.09940863e-01 7.77231574e-01 4.38524544e-01 5.69354296e-01
6.12394989e-01 4.27167743e-01 1.34885579e-01 -1.40812948e-01
2.80114621e-01 -2.49373943e-01 8.69745389e-02 1.15618169e+00
9.09576789e-02 -4.05904770e-01 -9.95326698e-01 1.04508066e+00
-1.98665619e+00 -9.45419431e-01 -2.96263993e-02 1.82309771e+00
8.25011253e-01 1.59713775e-01 2.96641421e-02 2.85778567e-02
9.81266677e-01 4.71088380e-01 -2.82627165e-01 -1.35268062e-01
-1.35735512e-01 3.31426971e-02 -2.17518672e-01 3.23283613e-01
-1.16887176e+00 1.25102973e+00 6.04341602e+00 8.97833347e-01
-9.81232703e-01 4.16631877e-01 4.62625802e-01 3.44339907e-02
-6.66295409e-01 -1.40269712e-01 -7.31034338e-01 4.18732822e-01
9.32210624e-01 -7.34415531e-01 -4.61525768e-01 1.22790897e+00
2.81782031e-01 -1.36737630e-01 -7.78539538e-01 5.33427179e-01
3.93912107e-01 -1.34744287e+00 3.73147517e-01 2.78438121e-01
1.13136578e+00 -1.25303909e-01 2.82832474e-01 1.56453937e-01
6.46079063e-01 -8.55851829e-01 4.49630558e-01 2.01556429e-01
2.74352610e-01 -8.29617381e-01 8.47876072e-01 1.17214322e-01
-1.14884567e+00 2.63882309e-01 -6.43745720e-01 1.95561886e-01
2.74995387e-01 1.07871532e+00 -8.53610814e-01 4.71036941e-01
7.05718637e-01 9.09860313e-01 -7.56386518e-02 9.39847350e-01
-1.87785685e-01 9.79144037e-01 -2.34988347e-01 -1.07774183e-01
6.28522933e-01 -1.81026623e-01 5.07240057e-01 1.37403190e+00
5.35720110e-01 -1.10494956e-01 1.70708016e-01 1.02349007e+00
-1.02850206e-01 2.76189089e-01 -4.83889759e-01 -1.10033996e-01
5.87714076e-01 1.27612889e+00 -1.26235700e+00 -6.52970374e-01
-2.41074428e-01 4.46005076e-01 1.39280513e-01 4.28346217e-01
-5.64803481e-01 -4.24499840e-01 7.74421573e-01 -2.47840270e-01
5.42850018e-01 -4.82110649e-01 -5.99592686e-01 -1.01843262e+00
-3.40842046e-02 -1.98912129e-01 3.86513650e-01 -6.77536428e-01
-1.00888193e+00 8.42785060e-01 3.64714175e-01 -9.37727690e-01
-4.12447304e-01 -9.45956707e-02 -1.09111607e+00 5.85638762e-01
-1.47302997e+00 -1.16593480e+00 -3.24513853e-01 4.63966042e-01
9.21002746e-01 -2.55010933e-01 9.01310742e-01 -3.22455466e-01
-2.27441713e-01 -4.70360555e-02 3.73675495e-01 -3.02188605e-01
5.31816065e-01 -1.41648459e+00 4.48230475e-01 5.39468706e-01
4.51084912e-01 1.09854162e+00 8.00683498e-01 -7.54146576e-01
-7.41241336e-01 -1.20652485e+00 1.20149910e+00 -4.36157286e-02
7.69349873e-01 -5.88608503e-01 -1.13088393e+00 6.04383707e-01
6.65007412e-01 -8.99496377e-01 1.16262972e+00 8.75835598e-01
-4.33218837e-01 1.71014324e-01 -6.40791953e-01 3.82919401e-01
4.13180619e-01 -4.45217103e-01 -1.13655543e+00 4.17060614e-01
1.20972133e+00 2.41432235e-01 -9.01288450e-01 -2.26303130e-01
2.12108031e-01 -5.06549418e-01 8.01708579e-01 -4.10143048e-01
6.90967679e-01 -5.01746833e-02 -1.51103720e-01 -1.59959781e+00
-2.04924554e-01 -2.85663545e-01 -4.08086479e-01 1.49074960e+00
2.91541457e-01 -4.69285518e-01 8.61910403e-01 1.26344666e-01
-4.96959761e-02 -6.49198353e-01 -8.62530649e-01 -6.26919866e-01
2.47945309e-01 -6.76457047e-01 5.59107602e-01 1.23799801e+00
4.85746980e-01 5.99264801e-01 -3.38626146e-01 -2.62207221e-02
6.97947085e-01 2.84452796e-01 3.26386243e-01 -1.80207348e+00
2.36523494e-01 -5.17672598e-01 -2.00544626e-01 -8.51639271e-01
4.33868676e-01 -6.98469579e-01 1.11009084e-01 -1.90145850e+00
5.49794734e-01 -3.29869032e-01 -4.42768574e-01 3.59612346e-01
-2.89093763e-01 -1.63065448e-01 -5.23136817e-02 5.50421834e-01
-9.20630515e-01 9.60202098e-01 4.44973916e-01 -1.90852031e-01
-2.98222452e-01 -1.49155155e-01 -8.36691618e-01 8.13030243e-01
9.14817035e-01 -6.92599356e-01 -5.85190177e-01 -3.90375704e-01
1.14860818e-01 -4.13318574e-01 6.63770479e-04 -8.22112083e-01
5.26195347e-01 1.31934732e-01 3.74370515e-02 -1.09546268e+00
3.97822082e-01 -6.88031077e-01 -2.75584131e-01 3.54902148e-01
-6.47895038e-01 -1.81302190e-01 -9.94325709e-03 8.38138402e-01
-5.58213472e-01 -1.67385176e-01 2.67787874e-01 5.21978503e-03
-6.47465229e-01 1.39827564e-01 -8.11260939e-01 -7.46731088e-02
9.37330961e-01 -2.49281283e-02 -3.93708408e-01 -6.52072668e-01
-4.25927848e-01 3.05504322e-01 1.82578489e-01 7.99991071e-01
6.46024764e-01 -1.33294356e+00 -5.75818658e-01 -2.84785956e-01
3.11902046e-01 1.29395053e-01 1.52734995e-01 1.74005881e-01
1.42927721e-01 1.12595856e+00 8.53764713e-02 -6.66307390e-01
-9.89508688e-01 5.13935208e-01 -4.10009176e-01 -3.40774149e-01
-8.99402440e-01 6.52160347e-01 5.44073761e-01 -1.84072748e-01
5.26196599e-01 -5.18795013e-01 -4.60977942e-01 2.68898875e-01
4.81597900e-01 2.60188848e-01 -2.61922359e-01 -6.15856588e-01
-7.14426711e-02 5.17870843e-01 -2.78805912e-01 -3.03476155e-01
1.46580851e+00 -1.40181690e-01 -2.97292918e-01 9.51728225e-01
1.10763025e+00 -7.27322161e-01 -8.63118112e-01 -7.32399583e-01
5.01615286e-01 -6.02315068e-02 4.45449919e-01 -2.62624294e-01
-8.08698237e-01 9.18754160e-01 3.15614760e-01 4.24354136e-01
8.50463510e-01 4.47222292e-01 9.30552840e-01 5.47619641e-01
-2.00825799e-02 -1.35684311e+00 2.92868614e-01 4.40161675e-01
6.55957520e-01 -1.12003779e+00 1.16673499e-01 -4.57834810e-01
-5.94669461e-01 1.11169028e+00 4.48476225e-01 -1.24327712e-01
9.21296000e-01 -2.27423564e-01 -1.56903371e-01 -7.31381118e-01
-1.09936428e+00 -1.10892594e-01 4.69313204e-01 6.31327689e-01
7.54975677e-01 1.49598822e-01 -4.89052206e-01 7.74419844e-01
-5.61214864e-01 -3.35400194e-01 4.00833637e-01 6.42380059e-01
-9.82407391e-01 -6.72740459e-01 -3.16353232e-01 4.76742595e-01
-3.80811691e-01 -2.21417695e-01 -2.67760336e-01 6.21292889e-01
-1.39953598e-01 1.05054390e+00 5.99297106e-01 -3.03075671e-01
-4.64196324e-01 2.87007719e-01 -5.45932293e-01 -9.06017840e-01
-3.15536976e-01 3.30798894e-01 -2.47026920e-01 -1.98916197e-01
-6.70634568e-01 -6.81739211e-01 -1.29837239e+00 8.08917824e-03
-6.93186283e-01 7.09790170e-01 1.12269139e+00 1.25547647e+00
3.35221738e-01 6.71008170e-01 6.12377226e-01 -4.85783756e-01
-9.54954140e-03 -1.43810821e+00 -6.91718698e-01 6.35822043e-02
-4.70397510e-02 -7.71896601e-01 -2.99205258e-02 2.99866259e-01] | [10.403322219848633, 6.977249622344971] |
3ea8c976-c580-4390-99f0-ca68975a406c | abstractive-text-summarization-for-resumes | 2306.13315 | null | https://arxiv.org/abs/2306.13315v1 | https://arxiv.org/pdf/2306.13315v1.pdf | Abstractive Text Summarization for Resumes With Cutting Edge NLP Transformers and LSTM | Text summarization is a fundamental task in natural language processing that aims to condense large amounts of textual information into concise and coherent summaries. With the exponential growth of content and the need to extract key information efficiently, text summarization has gained significant attention in recent years. In this study, LSTM and pre-trained T5, Pegasus, BART and BART-Large model performances were evaluated on the open source dataset (Xsum, CNN/Daily Mail, Amazon Fine Food Review and News Summary) and the prepared resume dataset. This resume dataset consists of many information such as language, education, experience, personal information, skills, and this data includes 75 resumes. The primary objective of this research was to classify resume text. Various techniques such as LSTM, pre-trained models, and fine-tuned models were assessed using a dataset of resumes. The BART-Large model fine-tuned with the resume dataset gave the best performance. | ['Senem Tanberk', 'Aysu Deliahmetoglu', 'Sena Nur Cavsak', 'Öykü Berfin Mercan'] | 2023-06-23 | null | null | null | null | ['abstractive-text-summarization', 'text-summarization'] | ['natural-language-processing', 'natural-language-processing'] | [ 3.23360041e-02 1.01302914e-01 -3.22653353e-01 -2.32751906e-01
-7.58386910e-01 -2.78926998e-01 7.38889515e-01 9.74407434e-01
-5.66590965e-01 1.09453380e+00 1.07805800e+00 7.39069842e-03
-1.03014044e-01 -6.04189754e-01 -5.97485304e-01 -6.97549582e-02
1.49964377e-01 9.27258730e-02 1.74916163e-02 -6.19998097e-01
8.94784212e-01 1.20573625e-01 -1.49826610e+00 7.82403290e-01
1.43625343e+00 8.08372498e-01 2.65808403e-01 1.16407275e+00
-6.08663857e-01 1.24608529e+00 -1.33123684e+00 -4.31936860e-01
-4.30756152e-01 -7.35547423e-01 -8.52927566e-01 -2.86140084e-01
6.18427277e-01 -2.39778459e-01 -4.91614819e-01 7.17348039e-01
9.67111468e-01 7.84556925e-01 4.62834299e-01 -7.40453601e-01
-7.90274739e-01 1.20950425e+00 -2.76481003e-01 6.44012094e-01
6.66431725e-01 -3.60835493e-01 6.75543249e-01 -3.37793142e-01
8.06385636e-01 9.61031675e-01 6.98777974e-01 3.79769683e-01
-5.25812328e-01 -4.43263143e-01 -3.38305011e-02 1.57286182e-01
-5.20171463e-01 -3.72166395e-01 5.75046122e-01 -2.33562320e-01
1.40260434e+00 4.42533880e-01 6.47592485e-01 1.30922580e+00
1.11840057e+00 1.07981789e+00 7.87313044e-01 -2.72637784e-01
4.05450277e-02 1.48955673e-01 7.76360869e-01 5.31788588e-01
4.45101887e-01 -6.23917222e-01 -1.00177979e+00 4.06192727e-02
2.24465523e-02 1.90719202e-01 -1.51808426e-01 5.21059990e-01
-1.26926804e+00 6.94105804e-01 3.47261399e-01 4.22375202e-01
-5.63996017e-01 -1.75914705e-01 1.15893006e+00 5.76169312e-01
9.70201850e-01 9.47133303e-01 -3.14529002e-01 -4.74320590e-01
-1.24014282e+00 2.41593346e-01 1.25090587e+00 1.08859122e+00
1.60339341e-01 3.19366068e-01 -8.44136119e-01 1.05191755e+00
-4.59999233e-01 3.76519144e-01 1.19607186e+00 -6.12466812e-01
9.08150673e-01 7.14376330e-01 -1.18796922e-01 -1.15524888e+00
-4.77601916e-01 -5.72999895e-01 -1.30323315e+00 -7.07048357e-01
-3.82697195e-01 -5.69091141e-01 -7.82393098e-01 1.12852621e+00
-2.32977435e-01 -4.64429349e-01 3.99440020e-01 1.35499015e-01
2.03332305e+00 1.33332729e+00 -1.27801642e-01 -4.32792127e-01
9.34590638e-01 -1.44547427e+00 -1.32587802e+00 -1.52411640e-01
4.66500580e-01 -8.39331388e-01 7.39221632e-01 4.45354372e-01
-1.42339337e+00 -6.94079518e-01 -1.25570869e+00 -4.84298676e-01
-6.85176015e-01 3.86538535e-01 3.69767487e-01 1.34021372e-01
-1.09515965e+00 1.04406214e+00 -4.33142751e-01 -8.37339759e-01
1.07394293e-01 1.12202398e-01 -3.41932654e-01 1.79091200e-01
-1.13045490e+00 1.02042949e+00 8.45182300e-01 -1.07971132e-01
-4.72322494e-01 -4.56753790e-01 -9.26845372e-01 4.21475381e-01
2.64242381e-01 -7.30949104e-01 1.63657391e+00 -6.66412890e-01
-1.73563886e+00 5.42183101e-01 1.32932708e-01 -7.82704592e-01
3.53065342e-01 -5.30201495e-01 -3.79257739e-01 2.79512703e-02
1.11484662e-01 2.68263966e-01 3.34864080e-01 -6.79561555e-01
-8.08872640e-01 -1.91450939e-01 -3.73745203e-01 3.38717937e-01
-4.93092895e-01 5.51443994e-02 -7.93099403e-02 -6.04403377e-01
-3.98363739e-01 -5.04009962e-01 -1.09061472e-01 -1.37129092e+00
-7.19496131e-01 -6.29223943e-01 5.24411738e-01 -1.28881609e+00
1.69018960e+00 -1.59544277e+00 7.24413469e-02 -4.60500926e-01
2.18038708e-01 4.49844748e-01 -1.00889117e-01 1.17772937e+00
3.15434128e-01 1.48685187e-01 9.61952731e-02 -3.38839769e-01
-5.49410703e-04 -2.21333057e-01 -4.14327025e-01 -4.38428596e-02
-2.65304834e-01 9.32580113e-01 -1.00148094e+00 -6.85728431e-01
4.89576682e-02 -1.08299837e-01 -1.11491963e-01 4.00530368e-01
-2.26747334e-01 -4.28920165e-02 -5.83635986e-01 4.02981430e-01
8.66421312e-02 -1.43018171e-01 -4.30034339e-01 1.01970032e-01
-4.86752182e-01 3.46945405e-01 -5.22369504e-01 1.92041016e+00
-5.18204033e-01 1.18036664e+00 -1.92576069e-02 -7.79667377e-01
1.05070233e+00 4.06632632e-01 2.75968611e-01 -8.02811086e-01
3.58291268e-01 1.62936881e-01 -4.13324326e-01 -8.40644836e-01
1.61567724e+00 4.06165838e-01 -6.13879442e-01 5.17648637e-01
3.35743099e-01 -4.10107821e-01 7.18718231e-01 6.73142791e-01
1.17343903e+00 -8.15250203e-02 5.14453411e-01 -2.50985533e-01
3.09853792e-01 2.82856703e-01 1.41660750e-01 9.12307262e-01
-4.71734907e-03 5.38109064e-01 5.31250596e-01 -3.15303385e-01
-8.40711176e-01 -3.37577522e-01 4.55266416e-01 1.46477664e+00
-2.48421937e-01 -6.85770452e-01 -8.67584705e-01 -4.53841388e-01
-2.94186562e-01 1.15540683e+00 -4.86913890e-01 -4.54552740e-01
-4.69484538e-01 -3.30817193e-01 5.04559159e-01 2.98767447e-01
9.82081652e-01 -1.62115037e+00 -6.28157139e-01 4.50147033e-01
-5.92200994e-01 -6.25548422e-01 -6.70677960e-01 2.33376637e-01
-9.37425911e-01 -5.54651558e-01 -8.55872869e-01 -8.99791479e-01
2.24977806e-01 1.96362928e-01 1.17226958e+00 -3.19069684e-01
5.28624170e-02 3.40429336e-01 -5.38999498e-01 -8.10308397e-01
-6.12389922e-01 9.41574037e-01 -1.99690133e-01 -5.20784020e-01
2.92606056e-02 -2.54620403e-01 -2.81445861e-01 -4.38586652e-01
-9.44607973e-01 3.18404466e-01 7.62622058e-01 7.04513609e-01
2.54506558e-01 -2.45925546e-01 1.22112262e+00 -1.00917017e+00
1.75900376e+00 -5.91636240e-01 1.69124559e-01 4.63752955e-01
-3.91337156e-01 -2.26221178e-02 8.82141352e-01 -3.13233525e-01
-1.30180264e+00 -6.68515503e-01 -1.93927303e-01 1.37645632e-01
2.09089234e-01 1.10091317e+00 3.75545353e-01 4.82552171e-01
8.22174847e-01 5.37377238e-01 -2.42035568e-01 -5.20192802e-01
1.49043471e-01 1.14775765e+00 8.25577259e-01 -3.00941229e-01
-1.67354509e-01 -3.17376941e-01 -3.99104059e-01 -1.19389689e+00
-1.24522829e+00 -4.84696448e-01 -3.96845400e-01 -4.29761618e-01
6.62590206e-01 -6.41953826e-01 -4.72143143e-01 5.47446430e-01
-1.34620297e+00 -1.96077973e-01 -5.70540130e-01 3.71453375e-01
-2.61685014e-01 3.23311359e-01 -8.93858790e-01 -4.54374284e-01
-1.41775990e+00 -6.38986886e-01 6.59435213e-01 8.24129224e-01
-4.74421352e-01 -7.74040520e-01 1.14726834e-01 4.73588914e-01
6.34869337e-01 5.64098656e-01 7.34010100e-01 -1.38419235e+00
1.21190839e-01 -5.75240791e-01 1.36748418e-01 6.00922823e-01
1.35517299e-01 1.52303159e-01 -4.13315117e-01 -2.60654926e-01
1.01863675e-01 -7.52630651e-01 1.26303279e+00 6.29126310e-01
1.06447959e+00 -7.23879457e-01 -2.62030382e-02 1.08650938e-01
8.42933953e-01 1.33254096e-01 2.96866357e-01 3.40907335e-01
5.87664127e-01 6.25583410e-01 5.72017372e-01 5.53741813e-01
5.86044729e-01 -2.57502854e-01 -1.47881776e-01 3.59216034e-01
-1.42685086e-01 -5.13537884e-01 5.24601042e-01 1.89607227e+00
4.52653365e-03 -5.49509525e-01 -6.84775710e-01 5.16834676e-01
-1.95309889e+00 -1.10913002e+00 -1.01782449e-01 1.84938622e+00
1.05040312e+00 1.99688062e-01 -1.18543126e-01 -1.87023625e-01
5.85999787e-01 4.96264994e-01 -6.28347516e-01 -1.09471500e+00
-4.85639237e-02 2.04429235e-02 3.95141691e-01 1.93188876e-01
-7.65223145e-01 8.43551040e-01 6.19897413e+00 1.10813177e+00
-1.18983722e+00 -3.76889408e-02 6.23104095e-01 -2.72193104e-01
-4.09223922e-02 -6.10449731e-01 -9.80464697e-01 5.21433294e-01
1.51169324e+00 -9.75835621e-01 -5.29800244e-02 6.45417988e-01
2.25614026e-01 -4.60057795e-01 -9.66876268e-01 6.08729005e-01
7.19382286e-01 -1.56204975e+00 2.91773856e-01 -3.73699814e-01
1.30026186e+00 1.45548195e-01 -3.33341867e-01 1.09215081e+00
3.88266593e-01 -9.40593123e-01 5.40403187e-01 7.84899890e-01
3.94969642e-01 -8.36603463e-01 1.12013972e+00 8.71813536e-01
-6.29204214e-01 -1.04998443e-02 -4.26472336e-01 -2.56832033e-01
-1.00772351e-01 3.94775748e-01 -5.80093324e-01 8.52967083e-01
5.36646724e-01 9.81270015e-01 -8.60275805e-01 1.02810943e+00
-1.06311671e-01 6.13370359e-01 -2.42486261e-02 -7.46337116e-01
4.83951598e-01 -2.15431601e-01 5.87628126e-01 1.59122920e+00
3.10573608e-01 9.21268091e-02 2.36502975e-01 2.82006055e-01
-4.60281163e-01 4.86499637e-01 -5.74648440e-01 -4.22219962e-01
2.52655774e-01 1.32619834e+00 -7.10494578e-01 -8.88571858e-01
-4.76095127e-04 6.89165056e-01 1.36903122e-01 3.41379553e-01
-4.70800549e-01 -1.02921045e+00 -2.84271300e-01 -2.23993808e-01
-2.29314789e-01 6.84036240e-02 -2.79615492e-01 -1.05471563e+00
-1.84953481e-01 -9.53193069e-01 4.43595976e-01 -8.96127641e-01
-9.25989568e-01 8.20383191e-01 1.76650479e-01 -7.23337293e-01
-4.52678323e-01 5.03766648e-02 -1.00196576e+00 4.99003828e-01
-1.01310587e+00 -8.88693810e-01 -5.86439013e-01 8.03810507e-02
1.46669245e+00 -4.56748068e-01 5.49340487e-01 1.01735756e-01
-9.20742452e-01 2.62486607e-01 7.04617620e-01 -1.21358804e-01
8.77154052e-01 -1.28174424e+00 4.35044885e-01 6.74928010e-01
-4.39420074e-01 5.83708405e-01 9.09977794e-01 -1.03298485e+00
-1.13573599e+00 -1.17388725e+00 1.29476833e+00 -9.34369639e-02
3.75193238e-01 -8.76886398e-02 -7.57844746e-01 7.65624762e-01
1.11197436e+00 -8.77810657e-01 6.33788943e-01 -7.49477893e-02
5.32258928e-01 -2.00951830e-01 -9.44290221e-01 5.79129338e-01
5.32973945e-01 -1.49099186e-01 -1.04310918e+00 5.30414999e-01
1.08989716e+00 -6.80521429e-01 -9.38547075e-01 1.32380843e-01
3.18074375e-01 -7.50730217e-01 4.62556988e-01 -6.95056379e-01
1.21358454e+00 6.65033281e-01 2.67181575e-01 -1.95232475e+00
2.19025426e-02 -7.72406995e-01 -4.24932063e-01 1.63384974e+00
2.38095924e-01 -4.80490774e-01 4.99680370e-01 4.11559433e-01
-8.67988646e-01 -9.53482628e-01 -5.06803751e-01 -3.18672031e-01
5.39316945e-02 3.69041473e-01 3.63422006e-01 7.91398883e-01
2.82394171e-01 8.62381220e-01 -4.92172539e-01 -8.86747062e-01
1.30242988e-01 2.08006606e-01 8.76193941e-01 -1.24581599e+00
5.10481954e-01 -7.42858589e-01 1.19295411e-01 -9.57694411e-01
2.34860614e-01 -9.32017028e-01 2.99591385e-02 -2.51916599e+00
4.37923133e-01 5.62002599e-01 -3.12394947e-02 2.41908967e-01
-1.65507823e-01 -4.61645007e-01 1.12079278e-01 -9.11053941e-02
-1.10582161e+00 8.80159020e-01 1.41953480e+00 -4.44243848e-01
-4.57216620e-01 3.69807810e-01 -9.39802289e-01 5.05570292e-01
9.35131013e-01 -3.12154949e-01 -2.56450176e-01 -2.55613506e-01
3.69744837e-01 5.32072902e-01 -2.98833758e-01 -1.12488413e+00
7.99367249e-01 2.68005934e-02 4.69707340e-01 -1.33669734e+00
-2.28722580e-02 1.10461460e-02 -3.11777592e-01 4.36858147e-01
-1.10926628e+00 3.76802653e-01 3.15587252e-01 2.22378105e-01
-5.28220296e-01 -5.50262153e-01 3.62603098e-01 -3.85865539e-01
-4.78493035e-01 -2.35457439e-02 -6.93044841e-01 2.17794091e-01
6.05613470e-01 -1.12268299e-01 -5.77097774e-01 -7.56147087e-01
-3.68512630e-01 6.31549239e-01 -6.33992106e-02 7.60780513e-01
6.27605796e-01 -1.02774978e+00 -1.08330667e+00 -2.57541955e-01
-2.37930492e-01 1.92179620e-01 4.59420294e-01 4.39816564e-01
-7.27195799e-01 9.18038666e-01 -4.92541462e-01 -2.51788627e-02
-1.35984337e+00 1.36708170e-01 -1.18882842e-01 -7.53773510e-01
-6.17388606e-01 6.80336237e-01 -4.46475357e-01 -4.70281363e-01
5.24329960e-01 -8.94020915e-01 -8.60348761e-01 6.51524961e-01
5.93687236e-01 1.03926754e+00 2.97242522e-01 -3.60796571e-01
3.82078677e-01 1.50970249e-02 -4.22086686e-01 2.10312396e-01
1.38553059e+00 -1.55879259e-01 -4.02307868e-01 7.99240887e-01
1.21935546e+00 2.64483374e-02 -5.47237992e-01 -2.25854874e-01
1.81483418e-01 1.42776251e-01 1.05241232e-01 -9.35401738e-01
-6.43080175e-01 6.75995708e-01 -2.35190094e-02 4.96694952e-01
8.36563408e-01 -4.34120834e-01 1.23957348e+00 8.48864317e-01
-9.63723436e-02 -1.55726421e+00 4.29824710e-01 9.53594983e-01
1.21805155e+00 -1.23271489e+00 3.05546284e-01 4.34517831e-01
-7.45415032e-01 1.06852925e+00 7.94221520e-01 -9.89938974e-02
2.21565785e-03 -2.92900026e-01 -3.70153546e-01 -2.31000587e-01
-9.80542004e-01 3.94777060e-01 4.21377152e-01 3.77048180e-02
7.56111920e-01 -5.85741699e-02 -6.59289896e-01 1.04795825e+00
-8.47214282e-01 1.18391261e-01 1.06426656e+00 1.08515871e+00
-8.32510710e-01 -4.31978732e-01 -1.32576630e-01 1.25606883e+00
-8.79642904e-01 3.65808271e-02 -7.81397939e-01 5.99573016e-01
-5.19139588e-01 1.19301975e+00 -1.29974917e-01 -3.26724708e-01
6.07439518e-01 4.53375690e-02 1.04915306e-01 -8.23769569e-01
-1.43071449e+00 -5.20020247e-01 4.14559215e-01 -1.67696148e-01
-5.49561232e-02 -3.63518536e-01 -1.10987854e+00 -6.42517567e-01
-3.17177773e-01 7.13462532e-01 9.48436201e-01 7.83461094e-01
4.85948414e-01 1.09982097e+00 4.12913889e-01 -9.22539055e-01
-8.11861455e-01 -1.71890354e+00 -4.12654757e-01 5.73118180e-02
3.91785979e-01 1.31875366e-01 -8.04925933e-02 3.49281519e-03] | [12.574074745178223, 9.570383071899414] |
81c258b9-478a-4978-9b01-434b366a97db | information-plane-analysis-of-deep-neural | 1909.11396 | null | https://arxiv.org/abs/1909.11396v1 | https://arxiv.org/pdf/1909.11396v1.pdf | Information Plane Analysis of Deep Neural Networks via Matrix-Based Renyi's Entropy and Tensor Kernels | Analyzing deep neural networks (DNNs) via information plane (IP) theory has gained tremendous attention recently as a tool to gain insight into, among others, their generalization ability. However, it is by no means obvious how to estimate mutual information (MI) between each hidden layer and the input/desired output, to construct the IP. For instance, hidden layers with many neurons require MI estimators with robustness towards the high dimensionality associated with such layers. MI estimators should also be able to naturally handle convolutional layers, while at the same time being computationally tractable to scale to large networks. None of the existing IP methods to date have been able to study truly deep Convolutional Neural Networks (CNNs), such as the e.g.\ VGG-16. In this paper, we propose an IP analysis using the new matrix--based R\'enyi's entropy coupled with tensor kernels over convolutional layers, leveraging the power of kernel methods to represent properties of the probability distribution independently of the dimensionality of the data. The obtained results shed new light on the previous literature concerning small-scale DNNs, however using a completely new approach. Importantly, the new framework enables us to provide the first comprehensive IP analysis of contemporary large-scale DNNs and CNNs, investigating the different training phases and providing new insights into the training dynamics of large-scale neural networks. | ['Michael Kampffmeyer', 'Sigurd Løkse', 'Kristoffer Wickstrøm', 'Shujian Yu', 'Robert Jenssen', 'Jose Principe'] | 2019-09-25 | null | null | null | null | ['information-plane'] | ['methodology'] | [-1.20235030e-02 1.92163572e-01 2.11380683e-02 -7.70130306e-02
9.89772156e-02 -6.31550968e-01 5.07116318e-01 -3.70822139e-02
-5.59202611e-01 5.42601347e-01 -2.94818670e-01 -3.95616770e-01
-6.72982872e-01 -7.76364684e-01 -7.11786628e-01 -9.94312108e-01
-5.11472166e-01 1.77555770e-01 2.63531506e-01 -8.63950048e-03
-2.26230198e-03 9.02385116e-01 -1.53359759e+00 -5.53123578e-02
5.55973530e-01 1.47706866e+00 1.49325624e-01 6.75837219e-01
9.17442217e-02 7.84415543e-01 -3.31550747e-01 -5.39266407e-01
2.00951055e-01 -6.47596270e-02 -8.69803667e-01 -1.89929798e-01
4.65397835e-01 -6.47155344e-02 -6.10439539e-01 1.20375109e+00
1.55842990e-01 4.21472862e-02 7.97106564e-01 -1.21231127e+00
-4.68837082e-01 7.07824886e-01 -8.33435636e-03 5.01203656e-01
-4.97401804e-01 4.61878106e-02 1.34169328e+00 -6.09625161e-01
5.85340977e-01 9.20486212e-01 6.93291605e-01 5.01211107e-01
-1.44636500e+00 -4.76093888e-01 -4.71875630e-03 2.24883929e-01
-1.23868012e+00 -1.74145967e-01 8.35639477e-01 -6.91909552e-01
7.77786255e-01 -4.80495691e-02 7.11635113e-01 1.11600184e+00
5.63466065e-02 8.23491514e-01 9.69095647e-01 -1.98445633e-01
3.06999415e-01 1.89742982e-01 3.49818826e-01 6.45132065e-01
4.78034765e-01 -4.47099507e-02 -2.48647168e-01 1.34847209e-01
7.74145722e-01 5.58198197e-03 -2.40360007e-01 -5.83197653e-01
-1.27945268e+00 8.50222588e-01 7.33600020e-01 6.72353804e-01
-2.10966423e-01 3.46120477e-01 5.01234412e-01 3.71900409e-01
4.16839540e-01 2.81120121e-01 -7.61671722e-01 -5.23292385e-02
-7.62614071e-01 1.80117086e-01 9.44930613e-01 6.92316473e-01
9.63563561e-01 -2.59760842e-02 3.70086312e-01 5.03497899e-01
3.26514632e-01 1.88628823e-01 2.45160967e-01 -9.21792626e-01
7.20357478e-01 7.59185493e-01 -4.43394065e-01 -1.10542393e+00
-6.94640100e-01 -7.75141537e-01 -1.41672552e+00 2.68905610e-01
6.79551125e-01 -2.69304365e-01 -2.54331440e-01 1.95276368e+00
1.14665926e-02 -1.15902163e-01 2.49036402e-01 5.38343728e-01
4.98494476e-01 2.61901289e-01 -2.19497681e-01 1.31340131e-01
1.28057241e+00 -3.54248166e-01 -3.38425815e-01 5.33342399e-02
9.38989162e-01 -2.32687101e-01 6.11267745e-01 3.16488206e-01
-8.39374363e-01 -3.26660037e-01 -1.11537099e+00 7.99840018e-02
-6.17468894e-01 3.34167689e-01 6.93689108e-01 5.99576116e-01
-1.31970990e+00 1.07706046e+00 -9.76974607e-01 -3.07661235e-01
8.63184273e-01 6.07651651e-01 -6.51403666e-01 1.65912047e-01
-1.27455509e+00 7.28955388e-01 6.36471927e-01 4.17073876e-01
-6.42405748e-01 -7.02785194e-01 -6.11807108e-01 2.22026765e-01
4.38637808e-02 -4.69020009e-01 8.82254779e-01 -9.16394711e-01
-1.43695319e+00 3.99538219e-01 2.61351824e-01 -4.56977904e-01
3.68055165e-01 6.10194467e-02 -9.39113125e-02 3.18793148e-01
-3.50707114e-01 6.49998844e-01 6.61814690e-01 -8.36738110e-01
-2.26818085e-01 -5.87789893e-01 1.37053773e-01 -2.89313763e-01
-7.36336291e-01 -2.91013300e-01 1.88533917e-01 -4.44089323e-01
2.37234473e-01 -9.15895820e-01 -9.70226899e-02 2.39981964e-01
-5.75744808e-01 -2.10115939e-01 6.56203508e-01 -2.71151274e-01
1.02129161e+00 -2.22799206e+00 6.20870054e-01 2.24895999e-01
7.00426280e-01 4.53697145e-01 -7.03003407e-02 4.66623724e-01
-3.43709797e-01 1.91522896e-01 -2.62308031e-01 -2.83171088e-01
1.24050692e-01 3.37832302e-01 -1.69022843e-01 6.48249805e-01
4.60010022e-01 9.07775640e-01 -6.05904400e-01 -2.02990606e-01
1.19052276e-01 6.87801838e-01 -5.75072527e-01 -2.47779518e-01
-1.30234227e-01 2.39316106e-01 -3.52090269e-01 5.90152517e-02
6.30363047e-01 -4.57710057e-01 1.28622591e-01 -3.59435141e-01
-1.64601862e-01 3.33199017e-02 -9.96135116e-01 1.27794707e+00
-3.37736994e-01 1.21495700e+00 7.73483515e-02 -1.60509849e+00
6.53286815e-01 3.69298458e-01 4.80028003e-01 -2.22198576e-01
3.28733504e-01 2.79740274e-01 3.63986135e-01 -3.45194012e-01
7.62750348e-03 -6.68653548e-02 2.41706237e-01 2.02217683e-01
5.22625327e-01 4.51835662e-01 2.25320131e-01 1.51164085e-01
1.24469531e+00 -5.15765607e-01 -2.80316584e-02 -5.37774265e-01
5.45118928e-01 -4.57897633e-01 1.05827883e-01 5.18979669e-01
-1.35464355e-01 2.30127975e-01 1.14036644e+00 -3.20254326e-01
-1.10886705e+00 -1.14912796e+00 -5.87246776e-01 6.11952007e-01
-4.36328858e-01 -8.33113343e-02 -8.50701928e-01 -5.30400395e-01
-7.52303924e-04 8.47291760e-03 -7.38939583e-01 -1.93490565e-01
-2.25396961e-01 -8.23557794e-01 8.69230151e-01 5.17144918e-01
7.35602677e-01 -8.71701419e-01 -4.03844923e-01 9.54868793e-02
2.51441419e-01 -1.48747027e+00 2.34986022e-01 6.01965070e-01
-1.30075192e+00 -1.17791593e+00 -8.39328647e-01 -6.18016303e-01
6.62722886e-01 -2.31378272e-01 8.28703284e-01 -2.39022300e-01
-3.02399158e-01 2.31600046e-01 -3.63866761e-02 -1.04262210e-01
-2.62870193e-01 6.36075795e-01 2.21081257e-01 2.23660707e-01
1.31101459e-01 -9.54016447e-01 -6.02547884e-01 3.39806974e-01
-1.36139143e+00 -1.26617849e-01 8.16102266e-01 6.60944641e-01
6.31480664e-02 3.76773655e-01 4.12930161e-01 -5.45592606e-01
4.51209038e-01 -6.00815177e-01 -7.60165334e-01 9.62281972e-02
-4.05116916e-01 4.17242348e-01 1.00403476e+00 -5.39489806e-01
-6.20636404e-01 -1.53538987e-01 -2.55670249e-01 -2.61911213e-01
-2.33306661e-01 5.58813989e-01 -8.19884017e-02 -4.01500106e-01
4.56990272e-01 2.28467360e-01 2.48273034e-02 -6.44092917e-01
3.49338382e-01 3.83388698e-01 2.59074569e-01 -3.92784059e-01
8.44809890e-01 5.21317959e-01 5.51096797e-01 -1.01913452e+00
-8.84160340e-01 -3.61031801e-01 -1.04257047e+00 -1.43906504e-01
9.11092579e-01 -5.21746576e-01 -1.17898953e+00 7.85032213e-01
-1.11447072e+00 -3.99488449e-01 -2.30731875e-01 7.38711476e-01
-4.93091136e-01 2.30664715e-01 -8.92704964e-01 -7.36716688e-01
2.21674633e-03 -1.08569884e+00 5.83391368e-01 1.48126513e-01
2.92308807e-01 -1.44585049e+00 4.82942536e-02 -5.28676137e-02
5.74980497e-01 1.17171384e-01 1.07140410e+00 -6.97211683e-01
-7.03164220e-01 -1.59374714e-01 -5.72980046e-01 9.18123424e-01
-1.98168263e-01 1.35602862e-01 -1.04132724e+00 -5.49624488e-02
3.60644832e-02 -1.11118250e-01 1.06289029e+00 2.91745394e-01
1.08776617e+00 -4.84960407e-01 5.47906570e-03 7.54228771e-01
1.50731051e+00 -3.87229890e-01 4.73857045e-01 1.79975361e-01
7.77287245e-01 5.81725121e-01 -3.35081577e-01 3.15682024e-01
1.96152061e-01 5.19986451e-01 6.43854260e-01 2.17024460e-01
1.35958001e-01 1.28198728e-01 2.32389808e-01 1.06811297e+00
-3.44858199e-01 -1.16558336e-01 -8.75192583e-01 3.64972532e-01
-1.62780368e+00 -8.79353285e-01 -1.62540853e-01 1.98011470e+00
6.32796407e-01 2.62473494e-01 -2.07868405e-02 4.55919415e-01
3.73148918e-01 2.98835218e-01 -5.68330824e-01 -9.95851159e-02
-2.54795283e-01 1.86976522e-01 6.63116872e-01 1.45482719e-01
-1.18590045e+00 5.83335221e-01 5.75403976e+00 6.60293221e-01
-1.11009991e+00 -4.72022332e-02 4.25962776e-01 1.38530925e-01
-2.52900682e-02 -1.25892356e-01 -9.29172039e-01 1.89749628e-01
1.04990673e+00 2.50675261e-01 4.51768696e-01 8.46752763e-01
-2.47189611e-01 1.05128512e-01 -1.12245154e+00 9.07631576e-01
-3.57427567e-01 -1.44104540e+00 2.72473712e-02 4.72942740e-01
6.30467236e-01 5.79342484e-01 3.66311491e-01 9.37667787e-02
-3.57624441e-02 -8.96116078e-01 4.58795786e-01 7.10105300e-01
4.55984205e-01 -6.84171915e-01 8.27418029e-01 5.12904048e-01
-1.10816884e+00 -1.22411147e-01 -6.27964258e-01 -1.40030906e-01
-3.32659334e-01 1.02597308e+00 -4.84353215e-01 3.01251024e-01
5.28726101e-01 9.58528697e-01 -5.97611487e-01 8.29391778e-01
2.97344774e-01 4.67867434e-01 -5.02777040e-01 -2.33327061e-01
4.08041805e-01 -2.64968485e-01 5.78086436e-01 9.77216721e-01
2.42201880e-01 -3.89130861e-01 -5.80283463e-01 1.11953664e+00
-3.07758451e-01 -1.99433416e-02 -8.06188583e-01 -4.86718506e-01
-1.21291868e-01 1.35274160e+00 -8.00519407e-01 2.04627216e-02
-4.08751011e-01 6.85565770e-01 7.22644448e-01 4.56464559e-01
-4.23616558e-01 -4.37806517e-01 1.17656648e+00 -6.09699488e-02
6.02527320e-01 -7.13505328e-01 -1.42736807e-01 -1.28355098e+00
2.53374517e-01 -3.10824007e-01 -1.12151101e-01 -1.07477963e-01
-1.37903607e+00 6.48094654e-01 2.47802623e-02 -1.02247775e+00
6.64626155e-03 -1.41221619e+00 -3.14730734e-01 6.85448229e-01
-1.56670678e+00 -5.73745847e-01 1.27423331e-01 5.99739254e-01
-1.69578716e-01 -1.66481212e-01 7.49921799e-01 4.88864958e-01
-9.17592704e-01 4.83212292e-01 5.69822252e-01 5.60666442e-01
-7.41633624e-02 -1.31272292e+00 2.50628144e-01 4.89635915e-01
3.04256380e-01 7.42301285e-01 4.26254779e-01 -8.83357897e-02
-1.50265348e+00 -8.46641004e-01 6.80607319e-01 -4.01703089e-01
1.12960279e+00 -6.91974044e-01 -9.44854856e-01 5.33399403e-01
-2.04098865e-01 5.56266725e-01 5.13468087e-01 7.28599578e-02
-4.96936411e-01 -3.91052693e-01 -7.68762767e-01 4.70490485e-01
1.00823247e+00 -7.37921476e-01 -1.69346139e-01 2.08493710e-01
5.96694469e-01 9.59496945e-02 -1.25063491e+00 4.30341601e-01
6.24383092e-01 -1.18757248e+00 1.03509653e+00 -6.31909490e-01
4.31416452e-01 6.94840923e-02 -2.46764198e-01 -1.12540603e+00
-1.02687076e-01 -4.04031456e-01 -2.17008367e-01 1.02443850e+00
5.47373593e-01 -9.41854298e-01 7.42420018e-01 4.88588363e-01
9.90567729e-02 -9.90163624e-01 -1.21407413e+00 -1.00132811e+00
3.02532434e-01 -7.65865743e-01 1.58480048e-01 6.29790902e-01
-1.63298417e-02 4.29249369e-02 2.40553729e-02 2.10315704e-01
7.27970541e-01 -4.79169309e-01 1.97965592e-01 -1.62559259e+00
-3.19503844e-01 -9.83121336e-01 -1.07133305e+00 -8.47042441e-01
4.05744374e-01 -1.05606771e+00 -3.41956645e-01 -1.10721362e+00
4.87400964e-02 -4.66726989e-01 -4.35777426e-01 1.75276250e-01
3.45462918e-01 2.76717633e-01 1.56021982e-01 2.32485682e-01
-4.46155101e-01 5.18113136e-01 1.15683222e+00 8.03681910e-02
1.63961202e-01 2.59354144e-01 -4.00445879e-01 9.04977798e-01
9.16391134e-01 -4.09373045e-01 -3.96759659e-01 -2.76602536e-01
7.68527448e-01 -3.35319638e-01 7.89052725e-01 -1.34192669e+00
2.67199606e-01 3.42359185e-01 3.12780410e-01 -2.98378557e-01
1.44843563e-01 -9.33369279e-01 -8.44565704e-02 3.21536064e-01
-4.59203333e-01 -1.09596156e-01 9.54010338e-02 6.82693839e-01
-2.90705174e-01 -3.33225250e-01 7.56303549e-01 -9.84581932e-02
-3.39305311e-01 6.03114903e-01 -4.75649416e-01 1.76237300e-01
8.00097346e-01 1.87724784e-01 -3.89010698e-01 2.96208058e-02
-8.03829253e-01 -2.93864369e-01 1.53074756e-01 4.12404723e-02
4.28847671e-01 -1.27423978e+00 -2.47975945e-01 3.33542854e-01
1.81172937e-02 6.83564842e-02 3.11531067e-01 1.18474150e+00
-5.79042077e-01 7.33509541e-01 -2.13175833e-01 -7.02246904e-01
-8.06414664e-01 4.49228883e-01 5.39866686e-01 -4.05430555e-01
-5.10738969e-01 7.81253517e-01 2.31959417e-01 -4.57474798e-01
3.17330331e-01 -7.63532639e-01 -3.06847006e-01 2.67304093e-01
3.61233324e-01 2.33636558e-01 1.15809053e-01 -5.86390078e-01
-2.94877350e-01 5.94192266e-01 2.80851759e-02 -1.71572398e-02
1.43491316e+00 -3.54080684e-02 -2.43852094e-01 7.48061597e-01
1.98107862e+00 -6.43121362e-01 -1.28260720e+00 -4.13705528e-01
2.34055042e-01 1.05472259e-01 2.13520795e-01 -2.03940704e-01
-1.51709080e+00 1.36241174e+00 4.13290888e-01 5.56974292e-01
9.90401804e-01 3.99551332e-01 5.81282675e-01 8.05105150e-01
1.46143645e-01 -8.45876515e-01 9.60429907e-02 6.94103062e-01
5.57644248e-01 -1.06716609e+00 -4.60868031e-01 -2.69423667e-02
-2.16688409e-01 1.59165311e+00 1.16694517e-01 -2.57193834e-01
1.26133263e+00 2.24994883e-01 -3.46293837e-01 -3.03951532e-01
-7.03294635e-01 -2.15609923e-01 3.89083236e-01 4.71401483e-01
2.47435853e-01 1.14813343e-01 2.68369436e-01 2.55054981e-01
-1.41174838e-01 -1.29321337e-01 2.64907867e-01 4.58223164e-01
-2.46274531e-01 -9.90940928e-01 2.14029476e-01 5.68021238e-01
-6.16683424e-01 -2.00592667e-01 -1.31692007e-01 8.36357653e-01
8.61102063e-03 3.70293260e-01 2.66923346e-02 -5.31592190e-01
1.27545044e-01 1.35386556e-01 5.58393836e-01 -3.03842127e-01
-2.26285592e-01 -6.33853793e-01 -3.09904784e-01 -5.01007438e-01
-6.19082987e-01 -7.63860762e-01 -8.63925278e-01 -3.57515037e-01
-4.00474608e-01 -1.62444785e-01 9.13874626e-01 1.17694986e+00
1.11807577e-01 4.57917809e-01 5.78242779e-01 -9.79944885e-01
-6.09647334e-01 -8.65540862e-01 -9.75453496e-01 1.40957579e-01
6.35760546e-01 -6.61534607e-01 -7.84425318e-01 -3.83473605e-01] | [7.923905849456787, 3.6034035682678223] |
d1628128-5eb4-4b04-a5ed-dba8980eaf16 | conquer-contextual-query-aware-ranking-for | 2109.10016 | null | https://arxiv.org/abs/2109.10016v1 | https://arxiv.org/pdf/2109.10016v1.pdf | CONQUER: Contextual Query-aware Ranking for Video Corpus Moment Retrieval | This paper tackles a recently proposed Video Corpus Moment Retrieval task. This task is essential because advanced video retrieval applications should enable users to retrieve a precise moment from a large video corpus. We propose a novel CONtextual QUery-awarE Ranking~(CONQUER) model for effective moment localization and ranking. CONQUER explores query context for multi-modal fusion and representation learning in two different steps. The first step derives fusion weights for the adaptive combination of multi-modal video content. The second step performs bi-directional attention to tightly couple video and query as a single joint representation for moment localization. As query context is fully engaged in video representation learning, from feature fusion to transformation, the resulting feature is user-centered and has a larger capacity in capturing multi-modal signals specific to query. We conduct studies on two datasets, TVR for closed-world TV episodes and DiDeMo for open-world user-generated videos, to investigate the potential advantages of fusing video and query online as a joint representation for moment retrieval. | ['Wing Kwong Chan', 'Chong-Wah Ngo', 'Zhijian Hou'] | 2021-09-21 | null | null | null | null | ['moment-retrieval', 'corpus-video-moment-retrieval'] | ['computer-vision', 'computer-vision'] | [ 1.32245183e-01 -6.75854087e-01 -5.86922646e-01 -2.17543021e-01
-1.81835604e+00 -6.00292206e-01 7.73889303e-01 2.35120878e-02
-5.60491800e-01 3.49577278e-01 6.52872264e-01 3.91777515e-01
-3.73608410e-01 -2.84960449e-01 -7.03189194e-01 -6.60517812e-01
-2.93510556e-01 1.17957637e-01 3.32022935e-01 -7.70948678e-02
3.28174472e-01 4.86550421e-01 -1.87712681e+00 5.91164529e-01
3.68073940e-01 1.23396838e+00 5.38952410e-01 1.08808589e+00
1.91345334e-01 9.30032849e-01 -6.26941204e-01 -8.10948983e-02
1.56772807e-01 -2.43401185e-01 -7.06340492e-01 1.28310919e-01
6.62788570e-01 -5.78872442e-01 -8.29569936e-01 8.12487900e-01
8.37930262e-01 7.01705039e-01 6.05313778e-01 -1.15293646e+00
-3.44035804e-01 5.67785680e-01 -5.93691111e-01 8.21785271e-01
1.06918335e+00 -3.06503922e-01 1.24210703e+00 -8.77758682e-01
9.71830308e-01 1.19171643e+00 8.26423839e-02 1.84240386e-01
-8.06297064e-01 -5.05793750e-01 1.38428614e-01 6.26101196e-01
-1.77262366e+00 -7.97488987e-01 1.19058502e+00 -4.03822780e-01
8.68615925e-01 4.91612017e-01 5.95383048e-01 1.03831303e+00
8.94146934e-02 1.23540139e+00 2.96426892e-01 -2.78206199e-01
-1.91922352e-01 -5.63777834e-02 -1.41252145e-01 3.81506503e-01
-4.02528793e-01 -2.27775261e-01 -1.16207409e+00 -2.09105209e-01
7.69682705e-01 2.37626612e-01 -4.12879199e-01 -1.60496250e-01
-1.63465881e+00 7.88189769e-01 1.54956058e-01 5.67392409e-01
-5.23947537e-01 4.88570333e-01 6.17202044e-01 6.10148609e-01
4.12427276e-01 2.45567784e-01 -2.94041842e-01 -4.89934206e-01
-1.17298222e+00 4.05267686e-01 3.72025311e-01 1.02206755e+00
6.29786968e-01 -5.12262322e-02 -5.43745220e-01 1.04457664e+00
3.56988579e-01 7.18516231e-01 7.63497353e-01 -1.13204920e+00
5.36110520e-01 3.20128053e-02 4.98284735e-02 -1.21111012e+00
1.59357898e-02 -3.71662863e-02 -2.48737559e-01 -6.57737911e-01
3.65549512e-02 1.00572087e-01 -5.62954724e-01 1.86212611e+00
2.94998556e-01 3.78965497e-01 -1.96537629e-01 1.05277705e+00
1.02287507e+00 9.70622420e-01 4.16901074e-02 -5.58273315e-01
1.58956671e+00 -6.96609676e-01 -7.88197935e-01 4.06477392e-01
4.80103821e-01 -8.85980368e-01 6.43854201e-01 3.60918224e-01
-1.30882835e+00 -6.70653462e-01 -8.07275653e-01 -4.08397824e-01
-2.47670621e-01 1.76373050e-01 5.47999144e-01 -7.16948733e-02
-1.11974621e+00 3.22661668e-01 -5.34967363e-01 -4.03610528e-01
2.13949345e-02 3.27849299e-01 -5.52642643e-01 -2.02637941e-01
-1.27010214e+00 3.73937488e-01 4.63853061e-01 -2.12589204e-01
-9.46285486e-01 -5.98287702e-01 -8.26859713e-01 -4.01549004e-02
5.13459086e-01 -5.93012631e-01 1.32922184e+00 -7.59499073e-01
-1.31004703e+00 6.10770226e-01 -5.20398915e-01 -2.19364211e-01
1.23844557e-01 -5.91759324e-01 -6.28470838e-01 9.77760136e-01
3.03083807e-01 7.99949348e-01 1.36582947e+00 -9.30362105e-01
-8.02066445e-01 -1.66086406e-01 2.21153706e-01 5.84922850e-01
-3.93341631e-01 1.96339518e-01 -1.12299061e+00 -9.60663438e-01
1.79235473e-01 -7.47185707e-01 3.15447122e-01 -2.61910170e-01
9.12902281e-02 -4.26389128e-01 1.03935528e+00 -6.75299823e-01
1.72185421e+00 -2.17061853e+00 4.15930718e-01 2.78808624e-01
9.14162472e-02 -1.54970095e-01 -4.44680572e-01 5.41799545e-01
-7.74381608e-02 -1.23238124e-01 6.65740728e-01 -1.33297458e-01
-1.80682316e-01 -1.36720970e-01 -5.33492625e-01 4.20057863e-01
-1.10826083e-01 8.57045293e-01 -1.10265040e+00 -9.44846749e-01
2.59064108e-01 6.96976781e-01 -6.47490501e-01 1.99998289e-01
-3.73012881e-04 2.79285550e-01 -7.18061686e-01 9.80830789e-01
1.64132491e-02 -1.08778141e-01 -1.32987022e-01 -7.80468822e-01
1.15330942e-01 -1.98054895e-01 -1.14961755e+00 2.30956125e+00
-3.45617712e-01 1.01572907e+00 -1.30945086e-01 -7.43493915e-01
5.44615567e-01 6.87032878e-01 1.18149614e+00 -8.44158649e-01
6.52346835e-02 -2.39034351e-02 -5.68915009e-01 -8.35040808e-01
1.14837480e+00 2.93945134e-01 -2.36811921e-01 3.12453628e-01
4.07061666e-01 8.75172298e-03 5.60950518e-01 5.77955782e-01
1.06420183e+00 2.12809324e-01 1.80427358e-01 1.62209272e-01
6.13445282e-01 -3.91144395e-01 2.70582974e-01 7.84402668e-01
-1.77067369e-01 6.80434644e-01 2.34391227e-01 -8.20201263e-02
-6.21142328e-01 -1.05365229e+00 1.55688390e-01 1.84845400e+00
2.09088713e-01 -6.44418597e-01 -1.14642285e-01 -3.30161899e-01
-1.58966333e-01 3.20338666e-01 -3.90004098e-01 -1.41640812e-01
-5.97519159e-01 -2.75159478e-01 2.52065688e-01 3.73312384e-01
2.15658054e-01 -9.48453069e-01 -4.27656591e-01 2.51866698e-01
-8.67249668e-01 -1.01877582e+00 -1.10630989e+00 -2.87798166e-01
-7.39324152e-01 -1.03022480e+00 -1.06158543e+00 -6.06341720e-01
2.99418926e-01 7.92541683e-01 1.08508337e+00 -1.00441568e-01
-1.99441686e-01 1.23411727e+00 -6.91669405e-01 1.98474199e-01
1.46294132e-01 6.02636160e-03 1.43230245e-01 2.74188519e-01
1.79607078e-01 -5.45668423e-01 -7.89736152e-01 2.22792909e-01
-1.13317633e+00 -3.58468115e-01 5.91391921e-01 6.56721175e-01
7.55520999e-01 -1.22292817e-01 5.41077971e-01 -2.71432132e-01
7.18439937e-01 -7.31935084e-01 -1.68643206e-01 2.95419484e-01
1.39680028e-01 -1.06732361e-01 -2.47699916e-02 -8.94907951e-01
-9.22903180e-01 6.21114857e-03 9.61568952e-02 -9.52777922e-01
1.76607221e-01 7.98240900e-01 -7.13827014e-02 5.92627898e-02
4.46756512e-01 3.41625810e-01 -3.93251240e-01 -3.75814885e-01
4.96127337e-01 5.37868917e-01 6.92246914e-01 -7.71467149e-01
5.76602459e-01 3.57298970e-01 -1.82756379e-01 -1.21684313e+00
-5.06700039e-01 -1.20544326e+00 -5.00458717e-01 -6.57054663e-01
8.07992399e-01 -1.32829380e+00 -7.96282768e-01 2.42357645e-02
-1.03772306e+00 1.72563925e-01 -1.94825336e-01 7.95263827e-01
-8.57687533e-01 6.13514066e-01 -5.82370579e-01 -5.62847316e-01
-2.75843740e-01 -1.12393951e+00 1.62377584e+00 3.78218666e-02
-1.02161586e-01 -7.79429317e-01 2.72070795e-01 3.53106707e-01
5.75762950e-02 1.72635585e-01 3.78939599e-01 -7.79197991e-01
-8.30084980e-01 -5.78249037e-01 -2.03831241e-01 -1.30008399e-01
-1.37608364e-01 2.98449602e-02 -8.97391498e-01 -4.81894702e-01
-2.03541830e-01 -4.04304743e-01 9.13435519e-01 5.76416075e-01
1.12390018e+00 -1.32379025e-01 -2.21279517e-01 5.23951054e-01
1.25833941e+00 2.33023286e-01 5.59552908e-01 6.43721968e-02
6.44128144e-01 2.26882771e-01 9.93727565e-01 6.91721022e-01
2.78839856e-01 8.21666062e-01 3.63789089e-02 4.46137547e-01
2.02161800e-02 -1.11127034e-01 6.49595439e-01 1.28160954e+00
-3.17752063e-01 -4.08222109e-01 -4.22295809e-01 5.73761463e-01
-1.98737383e+00 -1.62171280e+00 5.41456103e-01 2.13061309e+00
5.11529863e-01 -2.71784902e-01 1.75684020e-01 -8.24023262e-02
6.00868464e-01 5.96946716e-01 -3.63115996e-01 2.22698137e-01
3.15401405e-02 -2.78432429e-01 9.62748230e-02 2.35101655e-01
-1.33140898e+00 7.51819551e-01 5.96634293e+00 1.27294564e+00
-1.04200709e+00 1.29413024e-01 1.42226294e-01 -4.65196133e-01
-3.12947810e-01 -2.63244271e-01 -5.07611871e-01 3.01927298e-01
9.36636746e-01 -3.12864304e-01 4.97903347e-01 7.78181553e-01
2.66608119e-01 -2.64935941e-01 -1.34441304e+00 1.61758900e+00
4.87182617e-01 -1.31318700e+00 3.79715055e-01 -5.43466695e-02
6.06637478e-01 9.81758907e-02 1.17717557e-01 5.08043110e-01
-4.45358962e-01 -3.76048476e-01 8.58754456e-01 1.04143918e+00
8.77345085e-01 -7.57876098e-01 4.99645829e-01 4.25286666e-02
-1.72948921e+00 -1.69090763e-01 -1.31188557e-01 4.99888241e-01
3.78457993e-01 1.53509811e-01 -4.77294952e-01 6.64200366e-01
6.66149795e-01 9.33197856e-01 -6.24378145e-01 1.07567239e+00
4.41110134e-01 7.20584020e-02 -3.37840259e-01 2.65184104e-01
7.30997091e-03 3.57016688e-03 8.13617647e-01 1.29288578e+00
5.39035022e-01 1.48647770e-01 4.03641135e-01 2.20419895e-02
-3.02402943e-01 2.36986533e-01 -6.15285635e-01 -3.15179408e-01
4.73304182e-01 1.24607301e+00 -7.38124490e-01 -5.52324772e-01
-4.42058474e-01 9.93146122e-01 5.98344132e-02 6.44025624e-01
-9.43608880e-01 -5.25595844e-01 5.22621989e-01 -9.47930738e-02
2.65926480e-01 -2.16787830e-01 8.17386687e-01 -1.56281328e+00
-8.02295357e-02 -7.44127631e-01 6.24779642e-01 -1.13202477e+00
-1.04722941e+00 4.43079829e-01 5.53998590e-01 -1.55805910e+00
-8.05571139e-01 -1.54873922e-01 -3.18174183e-01 2.93551087e-01
-1.29001439e+00 -1.21603322e+00 -7.20228851e-02 1.15080237e+00
8.75185847e-01 -1.59762368e-01 4.02813882e-01 8.58411610e-01
-1.25977635e-01 5.40533841e-01 8.55606869e-02 3.74729224e-02
9.55151498e-01 -8.63263309e-01 -4.40106124e-01 7.03131497e-01
4.47891861e-01 7.13850856e-01 4.95600373e-01 -6.11401796e-01
-1.88639557e+00 -7.93666661e-01 7.46172428e-01 -5.71387172e-01
7.38941431e-01 2.94558089e-02 -5.70335329e-01 4.96264786e-01
-2.27981620e-02 -1.14122525e-01 8.43201458e-01 -9.23214629e-02
-2.75298387e-01 -3.43090683e-01 -7.50569582e-01 5.67766428e-01
8.18778694e-01 -1.23806810e+00 -6.16822124e-01 4.01785940e-01
8.98582995e-01 -2.30111584e-01 -1.18343115e+00 2.10136369e-01
8.80632877e-01 -6.43992126e-01 1.36834836e+00 -3.06379467e-01
1.42834947e-01 -3.13289016e-01 -6.38070226e-01 -7.91839242e-01
-2.52149533e-02 -9.28949714e-01 -6.83186531e-01 1.25052297e+00
-7.86303654e-02 9.45328474e-02 2.81764060e-01 3.32438707e-01
-3.25129442e-02 -4.41213489e-01 -9.58297372e-01 -3.93970191e-01
-6.95818186e-01 -8.49243999e-01 3.59952956e-01 6.86765909e-01
1.15909673e-01 3.24877411e-01 -8.03200841e-01 -9.01473034e-03
4.60150540e-01 1.42330244e-01 7.40571916e-01 -9.95015025e-01
-3.33454043e-01 -3.94520849e-01 -5.80880940e-01 -1.38397527e+00
3.77504528e-03 -6.95669711e-01 9.65788402e-03 -1.25737369e+00
4.25671965e-01 3.04686725e-01 -6.08223379e-01 -1.09765492e-01
1.17237635e-01 4.88769621e-01 3.89084697e-01 5.75806856e-01
-1.40716207e+00 3.57091784e-01 1.27619994e+00 -1.61259711e-01
-4.27148402e-01 2.25930829e-02 -3.21240008e-01 5.46480000e-01
1.32376865e-01 -1.92533687e-01 -7.00705111e-01 -3.59261036e-01
3.56398821e-01 8.36261272e-01 2.36755818e-01 -9.27830994e-01
4.67460662e-01 -4.63793986e-02 4.00313407e-01 -1.09196246e+00
8.03339660e-01 -8.79515469e-01 1.62943110e-01 -1.45708650e-01
-5.55212379e-01 1.95577413e-01 -1.70333445e-01 9.19481039e-01
-5.94279289e-01 -6.03180602e-02 3.24337125e-01 -3.67731042e-03
-1.11422312e+00 6.24503374e-01 -3.68047595e-01 -5.83726773e-03
7.62962341e-01 -2.74518639e-01 -1.56745277e-02 -9.51432168e-01
-1.01131809e+00 1.21591344e-01 6.73624873e-03 7.17591405e-01
9.90290225e-01 -1.58868289e+00 -4.53863770e-01 -1.71362218e-02
2.98359036e-01 -5.64728022e-01 5.68325102e-01 7.38201439e-01
-2.97998399e-01 5.11236548e-01 3.15437466e-02 -7.96618581e-01
-1.28149688e+00 7.03782976e-01 -2.67072886e-01 -1.39827684e-01
-3.50914448e-01 7.26143360e-01 2.10649818e-02 3.56658608e-01
4.23314214e-01 -2.64687121e-01 -4.01583910e-01 6.75987482e-01
8.10729444e-01 4.15020078e-01 -2.54248261e-01 -1.23796475e+00
-1.93736523e-01 6.89406812e-01 -1.86856851e-01 -4.11817282e-01
1.21136963e+00 -6.37585044e-01 7.71471113e-02 6.51018798e-01
1.75484133e+00 3.12316176e-02 -9.58815455e-01 -4.14984673e-01
-4.87683341e-02 -6.99710011e-01 1.96857944e-01 -2.14487895e-01
-1.04782796e+00 5.22465646e-01 6.38665378e-01 4.43391502e-02
1.31855237e+00 1.83866113e-01 7.95630157e-01 7.84777164e-01
3.99507612e-01 -1.26008487e+00 6.75252259e-01 3.52045953e-01
1.01261330e+00 -1.20358241e+00 2.99541593e-01 1.09316081e-01
-7.03708112e-01 1.04361272e+00 2.91138619e-01 -3.83754820e-02
7.80684948e-01 -4.55953032e-01 -2.88004220e-01 -2.98254699e-01
-8.71586204e-01 -3.70354444e-01 9.63344634e-01 3.68782073e-01
4.02595997e-01 -2.74274111e-01 -9.50009841e-03 3.79341513e-01
1.47755831e-01 -7.72147030e-02 1.43501043e-01 1.01271653e+00
-4.39991236e-01 -6.73132181e-01 -4.09206301e-01 5.15740097e-01
-6.50067806e-01 6.89867735e-02 1.94235891e-01 7.00342059e-01
-2.26993114e-01 9.61925149e-01 6.81362376e-02 -5.07206798e-01
1.14424586e-01 7.98712149e-02 4.67567921e-01 -3.87615174e-01
-2.98821390e-01 8.27483654e-01 -1.30537748e-01 -9.94737089e-01
-9.11768675e-01 -7.26751864e-01 -7.87582874e-01 -5.66086285e-02
-3.77455771e-01 2.93513596e-01 5.46381176e-01 7.24269569e-01
3.89089108e-01 5.70185781e-01 6.79866493e-01 -1.38863075e+00
-1.07030712e-01 -7.96899855e-01 -7.16401696e-01 6.78933680e-01
4.65254307e-01 -7.83265531e-01 -2.57255524e-01 3.94955218e-01] | [10.174355506896973, 0.7909061312675476] |
1df31a27-f755-485b-b1de-a207503fb14a | a-rank-corrected-procedure-for-matrix | 1210.3709 | null | http://arxiv.org/abs/1210.3709v3 | http://arxiv.org/pdf/1210.3709v3.pdf | A Rank-Corrected Procedure for Matrix Completion with Fixed Basis Coefficients | For the problems of low-rank matrix completion, the efficiency of the
widely-used nuclear norm technique may be challenged under many circumstances,
especially when certain basis coefficients are fixed, for example, the low-rank
correlation matrix completion in various fields such as the financial market
and the low-rank density matrix completion from the quantum state tomography.
To seek a solution of high recovery quality beyond the reach of the nuclear
norm, in this paper, we propose a rank-corrected procedure using a nuclear
semi-norm to generate a new estimator. For this new estimator, we establish a
non-asymptotic recovery error bound. More importantly, we quantify the
reduction of the recovery error bound for this rank-corrected procedure.
Compared with the one obtained for the nuclear norm penalized least squares
estimator, this reduction can be substantial (around 50%). We also provide
necessary and sufficient conditions for rank consistency in the sense of Bach
(2008). Very interestingly, these conditions are highly related to the concept
of constraint nondegeneracy in matrix optimization. As a byproduct, our results
provide a theoretical foundation for the majorized penalty method of Gao and
Sun (2010) and Gao (2010) for structured low-rank matrix optimization problems.
Extensive numerical experiments demonstrate that our proposed rank-corrected
procedure can simultaneously achieve a high recovery accuracy and capture the
low-rank structure. | ['Shaohua Pan', 'Defeng Sun', 'Weimin Miao'] | 2012-10-13 | null | null | null | null | ['quantum-state-tomography'] | ['medical'] | [ 3.87217790e-01 1.35078445e-01 -1.42946452e-01 -6.27601072e-02
-1.03846943e+00 -3.53456110e-01 2.45371372e-01 -7.03941807e-02
-4.50605005e-01 1.09231591e+00 1.86113775e-01 -1.68106139e-01
-7.17707813e-01 -5.48500240e-01 -6.24755740e-01 -1.01912212e+00
5.68682998e-02 1.40005589e-01 -3.04421604e-01 -2.31755152e-01
2.12384582e-01 3.72618109e-01 -9.93124127e-01 -2.80892253e-01
1.23386073e+00 9.49145138e-01 -2.10690815e-02 2.26555392e-01
5.45482278e-01 4.64915842e-01 1.57325491e-01 -3.11982334e-01
5.65708160e-01 -4.77061987e-01 -5.45277774e-01 2.29867995e-01
2.65734583e-01 -4.14205641e-02 -5.56343198e-01 1.71052969e+00
3.61397356e-01 3.36977005e-01 7.02533245e-01 -8.15120935e-01
-4.58268970e-01 6.96264267e-01 -8.17080319e-01 -5.08838259e-02
3.32239926e-01 -3.21619153e-01 1.28104889e+00 -1.37072790e+00
5.10859728e-01 9.10701513e-01 4.68876898e-01 2.26510957e-01
-1.38423085e+00 -6.40435815e-01 -3.93607318e-01 1.36301830e-01
-1.68941188e+00 -6.85094416e-01 7.26594448e-01 -6.59696341e-01
9.25965011e-02 3.37855577e-01 2.79995680e-01 6.51367068e-01
-4.19009477e-02 3.89587611e-01 1.36676335e+00 -6.06153905e-01
8.65922719e-02 -6.76371306e-02 3.15358728e-01 7.16537118e-01
9.44094181e-01 2.57463723e-01 -5.51377416e-01 -3.81460935e-01
7.13640034e-01 -2.40469098e-01 -6.14740252e-01 -6.20347500e-01
-1.53735578e+00 9.14703965e-01 2.88183689e-01 3.85537595e-01
-5.06059647e-01 -2.35290565e-02 1.87149763e-01 9.00396854e-02
1.97792441e-01 3.85075808e-01 1.38373777e-01 1.73928097e-01
-9.77682590e-01 1.97950944e-01 8.18440378e-01 9.00343835e-01
9.12815988e-01 2.36517042e-01 -2.67697453e-01 4.70679790e-01
3.67267936e-01 6.84238613e-01 -7.24442229e-02 -1.10147321e+00
5.59194803e-01 -1.46435037e-01 3.34796518e-01 -1.10007191e+00
-2.52852738e-01 -7.79206455e-01 -1.44472182e+00 -2.15836853e-01
7.61376858e-01 7.39389136e-02 -1.82704613e-01 1.84786415e+00
3.18189949e-01 2.74463147e-01 1.59960270e-01 1.27617741e+00
1.79910958e-01 5.98644733e-01 -4.36511189e-01 -9.16738153e-01
1.04164982e+00 -4.58566606e-01 -9.65468705e-01 -2.47616097e-02
5.97487867e-01 -6.95252359e-01 7.01185226e-01 5.50013483e-01
-1.08241308e+00 -8.14062133e-02 -1.17063391e+00 2.82939225e-01
4.96316075e-01 1.94860801e-01 6.61669254e-01 6.79408252e-01
-6.53902531e-01 7.46815145e-01 -4.64743853e-01 -1.45206064e-01
-6.64708614e-02 3.08592796e-01 -5.50231397e-01 -2.52858996e-01
-1.15031946e+00 7.23151326e-01 3.97179872e-01 6.04725122e-01
-5.08966565e-01 -5.82253635e-01 -6.52841091e-01 4.55274880e-02
5.67621589e-01 -5.61196029e-01 7.96809316e-01 -5.06910205e-01
-1.35446835e+00 5.77891827e-01 -2.89091747e-02 -2.85638988e-01
3.87680411e-01 -1.87316537e-01 -1.29888341e-01 1.38190880e-01
2.34034300e-01 -3.75221372e-01 7.74536669e-01 -8.76639128e-01
1.68725215e-02 -4.05531198e-01 -1.21403851e-01 7.05163926e-02
-2.64574856e-01 -1.01119585e-01 -9.76231843e-02 -4.99566793e-01
6.89273953e-01 -1.00399077e+00 -6.07452393e-01 -2.72141486e-01
-3.54504347e-01 1.40471846e-01 1.69726857e-03 -6.92791760e-01
1.19664836e+00 -2.20926857e+00 6.14064813e-01 6.84795380e-01
2.56317526e-01 -6.38210624e-02 -8.78011901e-03 5.28765142e-01
-3.97825509e-01 -3.64821292e-02 -6.05749428e-01 4.11482975e-02
2.85257120e-02 -7.68309608e-02 -2.86537886e-01 1.15540469e+00
-9.10128001e-03 4.19729501e-01 -9.79897678e-01 -2.97734618e-01
5.72162354e-03 2.97028393e-01 -8.09083641e-01 -1.31575868e-01
3.59065235e-01 7.04785824e-01 -5.81120789e-01 4.24466640e-01
7.86337614e-01 -3.53955328e-01 4.63270038e-01 -7.52399862e-01
-1.78620622e-01 -1.26762018e-02 -1.77485907e+00 1.50063646e+00
-1.66784197e-01 2.26838142e-01 7.22695291e-01 -1.31376350e+00
6.70438707e-01 2.37670034e-01 6.60359085e-01 -4.87291276e-01
1.39273837e-01 7.05517352e-01 7.94394314e-02 -1.43985584e-01
7.06834435e-01 -6.50761843e-01 -3.30937624e-01 1.41361654e-01
-2.17838854e-01 1.13632701e-01 5.07145941e-01 3.79494488e-01
1.01112723e+00 -2.97725827e-01 5.77000439e-01 -7.28082001e-01
9.32852149e-01 -5.54789245e-01 8.01249504e-01 8.29910398e-01
-9.06962380e-02 4.61033642e-01 4.88988250e-01 2.27252766e-01
-1.06917322e+00 -9.05037940e-01 -5.31363487e-01 4.48140651e-01
-1.52364774e-02 -5.11623263e-01 -4.63520288e-01 -6.92708194e-02
-1.76734641e-01 2.06901446e-01 -2.98259974e-01 -2.40395382e-01
-4.89527017e-01 -1.06992209e+00 3.47980380e-01 -2.11083586e-03
5.66835880e-01 -2.90186971e-01 2.93074369e-01 1.89676195e-01
-3.28548998e-01 -1.25817835e+00 -4.41125423e-01 1.07804194e-01
-8.72307718e-01 -1.09396517e+00 -7.29747534e-01 -3.66308570e-01
7.93143749e-01 2.80041546e-01 5.66159070e-01 -1.64892182e-01
1.66914403e-01 3.23799849e-01 -2.61279285e-01 3.18392366e-01
-3.14002275e-01 -2.29694247e-01 6.09099388e-01 4.18257713e-01
-2.54840851e-01 -5.84361494e-01 -3.86858493e-01 3.44400585e-01
-9.70472872e-01 -1.25109583e-01 6.94099426e-01 1.16408134e+00
7.29773045e-01 1.54538989e-01 5.20042539e-01 -7.99994111e-01
7.15365708e-01 -4.75788295e-01 -1.05263877e+00 2.47646403e-02
-7.56679296e-01 3.90593797e-01 7.31244266e-01 -2.74602205e-01
-8.20850134e-01 2.50678748e-01 2.04056233e-01 -2.28097826e-01
7.47280121e-01 9.89255786e-01 -2.08335876e-01 -3.88697088e-01
6.21737540e-01 2.17867613e-01 -5.66687807e-02 -4.62287664e-01
4.94476318e-01 3.65997970e-01 5.62381744e-01 -9.96178448e-01
1.32877326e+00 5.60589314e-01 6.88191712e-01 -9.05782104e-01
-1.16612446e+00 -4.85224098e-01 -4.78438377e-01 -8.40638503e-02
5.77356637e-01 -9.85175431e-01 -8.18594694e-01 7.62751922e-02
-9.02946770e-01 1.30704015e-01 -2.99460649e-01 1.11774731e+00
-7.24657536e-01 9.41134572e-01 -6.11978889e-01 -1.12088931e+00
2.55286437e-03 -1.02305710e+00 5.65055609e-01 3.20486422e-03
2.44538426e-01 -8.48451495e-01 2.00584069e-01 3.86106253e-01
1.19208418e-01 2.72094131e-01 5.78591645e-01 -1.62162974e-01
-6.55171990e-01 -1.21558905e-01 -4.99678165e-01 3.77204388e-01
-3.41131419e-01 -4.78789330e-01 -4.37007666e-01 -4.91654158e-01
3.72450083e-01 -7.98974335e-02 8.46281171e-01 2.83889621e-01
8.20462406e-01 -3.67768258e-01 6.67481869e-02 6.78756833e-01
1.52615488e+00 -3.45230341e-01 6.15322709e-01 3.54158431e-02
7.13403344e-01 5.23046613e-01 8.09598446e-01 8.20530653e-01
-3.68768573e-02 8.27456594e-01 2.56209344e-01 2.61595160e-01
3.73549223e-01 -3.24866921e-02 3.19902599e-01 1.55282223e+00
-3.85390818e-01 2.47504741e-01 -5.30737638e-01 2.89868206e-01
-2.14580846e+00 -1.17198372e+00 -7.03235626e-01 2.84332895e+00
8.96664798e-01 -3.07050645e-01 -6.72910511e-02 3.05379868e-01
8.89231861e-01 1.12908743e-01 -2.59761602e-01 1.39041897e-02
-4.21617389e-01 2.04826072e-01 9.27118778e-01 8.48117292e-01
-8.14168155e-01 6.17596745e-01 6.27380896e+00 1.04201496e+00
-4.70964760e-01 3.12184334e-01 3.42797115e-02 1.57261625e-01
-3.31267238e-01 4.38308269e-01 -6.04666412e-01 3.59017909e-01
7.95374870e-01 -4.46206063e-01 7.24631488e-01 6.35760128e-01
3.51413429e-01 -4.64625843e-02 -8.46014798e-01 1.40490842e+00
4.30648848e-02 -1.13428223e+00 -4.07437056e-01 7.06431448e-01
9.20451939e-01 -3.36240441e-01 -5.70704453e-02 1.66654363e-01
-1.31195605e-01 -8.35812092e-01 5.36201000e-01 7.33131826e-01
8.85883093e-01 -7.69759238e-01 5.14500141e-01 3.25616956e-01
-1.15487576e+00 -2.29717270e-02 -6.64775372e-01 -2.11619988e-01
5.06385863e-01 1.42127740e+00 -1.11212909e-01 9.34753060e-01
5.28873429e-02 8.38950157e-01 -3.74684483e-02 1.10005057e+00
-8.81511867e-02 5.39589167e-01 -4.09994364e-01 2.04667822e-01
6.90397099e-02 -1.13409019e+00 9.13450360e-01 7.47926176e-01
6.46872640e-01 3.87440145e-01 2.57741213e-02 8.43337476e-01
-1.14943497e-01 2.87528604e-01 -6.66162133e-01 -2.04399675e-01
2.90546536e-01 1.28077245e+00 -4.05296057e-01 -6.69303611e-02
-3.73659909e-01 7.70399570e-01 1.06320366e-01 4.99220341e-01
-7.49309063e-01 -2.39222243e-01 3.99648577e-01 8.88820291e-02
1.99355409e-01 -5.46000242e-01 -1.79841816e-01 -1.69127834e+00
3.42493266e-01 -8.39601159e-01 1.32334098e-01 -3.07566345e-01
-1.30631483e+00 7.03490227e-02 3.02988458e-02 -1.33003139e+00
-2.47293916e-02 -6.42549574e-01 -2.87352890e-01 7.60448992e-01
-1.26700413e+00 -5.57435334e-01 7.13310838e-02 5.27482212e-01
-3.88795942e-01 -1.83899477e-02 4.96572375e-01 8.65489364e-01
-7.43942738e-01 3.03594530e-01 6.41215742e-01 7.55610364e-03
5.49263895e-01 -1.07980275e+00 -5.78071475e-01 1.23114192e+00
9.67858476e-04 8.72615159e-01 8.14703882e-01 -5.38794577e-01
-1.83776093e+00 -7.74866343e-01 6.28838480e-01 -4.02353108e-02
1.06175160e+00 -2.68787324e-01 -7.21499324e-01 4.31801736e-01
-2.40425870e-01 1.57372192e-01 6.29539430e-01 2.47034460e-01
-3.24473649e-01 -2.35080704e-01 -9.32075799e-01 4.34161067e-01
9.70950067e-01 -6.78829551e-01 -3.43139946e-01 6.34963274e-01
2.59605169e-01 -1.89880282e-01 -1.04171968e+00 5.92599690e-01
5.02366543e-01 -8.88626039e-01 1.03680778e+00 -3.58975172e-01
2.23609746e-01 -5.67727685e-01 -7.18405545e-01 -9.79242980e-01
-5.57389438e-01 -1.01320457e+00 -2.85777718e-01 1.00333178e+00
9.83761027e-02 -5.67288876e-01 6.48607433e-01 5.07550180e-01
-7.71524087e-02 -5.04160702e-01 -1.23134220e+00 -1.11573160e+00
5.75638153e-02 -3.66931349e-01 -7.58642033e-02 1.03442669e+00
1.49324700e-01 4.22247171e-01 -9.30416703e-01 2.95789272e-01
1.19103575e+00 -1.45483434e-01 5.98942280e-01 -1.19454134e+00
-6.37109280e-01 -2.10036874e-01 -3.26579154e-01 -1.19821215e+00
2.65869856e-01 -1.26557183e+00 -1.33359721e-02 -1.20968461e+00
4.90010709e-01 -5.37018776e-01 -4.22899067e-01 -1.57700151e-01
4.15661633e-02 2.27928832e-01 3.77059817e-01 4.84544516e-01
-6.05465710e-01 8.20216715e-01 1.30132711e+00 4.42219675e-02
2.30931491e-02 -5.64683676e-02 -8.85484934e-01 5.25372803e-01
3.79464507e-01 -5.47497272e-01 -1.01208828e-01 2.13419482e-01
9.73584592e-01 4.32362288e-01 2.75897413e-01 -9.74312842e-01
1.93032175e-01 -2.07574904e-01 -2.25649610e-01 -3.12535912e-01
3.94496500e-01 -6.44219697e-01 2.45178193e-01 4.20040488e-01
-3.68165560e-02 -4.71649170e-01 -4.10775870e-01 8.86703134e-01
-1.68248698e-01 -7.58168817e-01 8.38160098e-01 1.53975129e-01
-1.51238054e-01 2.37537250e-01 -3.07984412e-01 1.78411603e-01
5.76633394e-01 1.38188982e-02 -2.02098295e-01 -5.91411352e-01
-7.86397934e-01 4.14115414e-02 2.88341582e-01 -3.41282278e-01
5.33105791e-01 -1.67398500e+00 -7.70169675e-01 -1.04812704e-01
2.70026717e-02 -2.77340293e-01 2.49170244e-01 1.68541360e+00
-1.30068108e-01 4.80647266e-01 1.93061218e-01 -4.29823667e-01
-6.80185556e-01 3.80170584e-01 1.76401019e-01 -4.59847361e-01
-2.95209795e-01 2.76160330e-01 3.54228556e-01 -2.88465381e-01
-1.90450549e-01 -1.19850777e-01 2.05935404e-01 -4.21553329e-02
5.62475920e-01 6.39602005e-01 -1.00545615e-01 -8.90523434e-01
-3.41032296e-01 7.21799195e-01 3.00714225e-01 -3.99680614e-01
1.25334549e+00 -1.80545360e-01 -4.77460980e-01 3.02684724e-01
1.28748751e+00 5.93973160e-01 -9.80486691e-01 -3.89218956e-01
-1.61787588e-02 -4.01948422e-01 2.13520318e-01 2.90653855e-02
-9.68744814e-01 4.99555588e-01 2.70518303e-01 1.02836989e-01
9.40963089e-01 -1.85015321e-01 5.48014402e-01 7.65671551e-01
6.69558108e-01 -9.90955114e-01 -5.75462878e-02 5.70089221e-01
9.35009599e-01 -1.07310128e+00 4.21319187e-01 -7.06297278e-01
-2.48194098e-01 1.01332772e+00 9.98272598e-02 -3.50965261e-01
5.44524491e-01 -2.99606144e-01 -8.31232071e-01 -1.07252769e-01
-2.11914703e-01 -4.01353449e-01 3.93830776e-01 2.93661505e-01
4.50890034e-01 1.94311380e-01 -1.07787430e+00 6.37752593e-01
-3.93489078e-02 -2.47378036e-01 9.88913357e-01 3.79784822e-01
-3.38304758e-01 -1.24345875e+00 -6.01027668e-01 3.99957925e-01
-4.03719962e-01 -2.16175020e-01 1.86302885e-02 5.29960394e-01
-3.26030910e-01 9.69085693e-01 -6.07721806e-01 -1.77136421e-01
1.68635309e-01 -2.29139939e-01 7.21436560e-01 -6.39595926e-01
2.48391670e-03 3.74374747e-01 2.12081492e-01 -5.10022759e-01
-7.44547248e-01 -7.56579816e-01 -1.00195658e+00 -3.47490877e-01
-6.72345519e-01 4.43494201e-01 5.48230767e-01 8.83666039e-01
-6.23519644e-02 1.24964938e-01 7.73297369e-01 -6.58978045e-01
-1.05933237e+00 -8.27770650e-01 -1.11331904e+00 3.40071887e-01
6.42551407e-02 -7.46353149e-01 -8.48250985e-01 -3.79803061e-01] | [6.908297061920166, 4.6042351722717285] |
17f667d7-4709-4957-96ef-25689555c1b0 | stylegan2-distillation-for-feed-forward-image | 2003.03581 | null | https://arxiv.org/abs/2003.03581v2 | https://arxiv.org/pdf/2003.03581v2.pdf | StyleGAN2 Distillation for Feed-forward Image Manipulation | StyleGAN2 is a state-of-the-art network in generating realistic images. Besides, it was explicitly trained to have disentangled directions in latent space, which allows efficient image manipulation by varying latent factors. Editing existing images requires embedding a given image into the latent space of StyleGAN2. Latent code optimization via backpropagation is commonly used for qualitative embedding of real world images, although it is prohibitively slow for many applications. We propose a way to distill a particular image manipulation of StyleGAN2 into image-to-image network trained in paired way. The resulting pipeline is an alternative to existing GANs, trained on unpaired data. We provide results of human faces' transformation: gender swap, aging/rejuvenation, style transfer and image morphing. We show that the quality of generation using our method is comparable to StyleGAN2 backpropagation and current state-of-the-art methods in these particular tasks. | ['Evgeny Kashin', 'Vladimir Ivashkin', 'Yuri Viazovetskyi'] | 2020-03-07 | stylegan2-distillation-for-feed-forward-image-1 | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/3992_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123670171.pdf | eccv-2020-8 | ['image-morphing'] | ['computer-vision'] | [ 6.26705825e-01 6.02055252e-01 1.33057281e-01 -3.90413672e-01
-5.50077021e-01 -6.58917725e-01 9.88212585e-01 -6.12417400e-01
-2.37282336e-01 8.06209087e-01 1.99532494e-01 -1.70766547e-01
4.68345284e-01 -8.15569580e-01 -9.49511826e-01 -7.09095180e-01
1.39360234e-01 5.37044525e-01 -4.11038280e-01 -1.52811378e-01
1.26635900e-03 4.94086355e-01 -1.39147377e+00 3.10212702e-01
4.06788349e-01 6.00291431e-01 -1.80457398e-01 1.01711333e+00
3.01427636e-02 4.94798273e-01 -6.10392451e-01 -7.60238469e-01
5.06974697e-01 -9.62535858e-01 -6.26251757e-01 1.69895872e-01
7.54863739e-01 -5.01678765e-01 -2.18429282e-01 8.95869732e-01
5.14366448e-01 -3.35820764e-01 6.42270565e-01 -1.47777939e+00
-1.16756117e+00 3.96417707e-01 -4.61851358e-01 -4.25481170e-01
2.28428841e-01 4.39364344e-01 6.88248575e-01 -7.99863815e-01
1.03086805e+00 1.49129927e+00 5.09779453e-01 1.17273879e+00
-1.78478205e+00 -8.84252548e-01 -2.04373002e-01 -4.66237068e-01
-9.46602345e-01 -7.42624819e-01 6.45790696e-01 -5.13999462e-01
6.24784887e-01 3.28575015e-01 6.38455629e-01 1.50093555e+00
1.36692272e-02 5.37456691e-01 1.39984167e+00 -6.28422081e-01
1.33731849e-02 2.63145447e-01 -7.16654718e-01 8.24933529e-01
-2.15862542e-02 3.30592901e-01 -5.53867996e-01 1.87915023e-02
1.22137702e+00 -1.86141968e-01 -4.78488989e-02 -4.81202513e-01
-1.34314454e+00 9.31612015e-01 3.00663471e-01 -2.42091734e-02
-1.08179808e-01 6.83366001e-01 1.39986083e-01 4.81377691e-01
5.96666753e-01 6.09860837e-01 -6.82198778e-02 -5.83834387e-03
-1.13383377e+00 5.10262907e-01 5.91168344e-01 1.01011646e+00
8.15501869e-01 3.43272060e-01 -4.84801918e-01 5.26967168e-01
2.55019099e-01 4.16010231e-01 3.15229207e-01 -1.33665597e+00
2.71209985e-01 3.40466350e-01 -9.65320542e-02 -7.41177261e-01
2.68727154e-01 -6.73689023e-02 -9.06846523e-01 8.35557640e-01
3.89606088e-01 -1.54712066e-01 -1.36602294e+00 1.97618151e+00
3.99092995e-02 6.72373846e-02 -1.97160654e-02 4.41120356e-01
5.76553047e-01 6.01603329e-01 -6.73902333e-02 -6.93618134e-02
1.21370614e+00 -1.11491883e+00 -7.83011973e-01 -2.77596742e-01
1.38095260e-01 -7.85966098e-01 1.08259344e+00 2.39977732e-01
-1.46073687e+00 -6.17735147e-01 -1.01130354e+00 -2.63883978e-01
-4.39357847e-01 1.23923540e-01 8.71658087e-01 9.43000138e-01
-1.45928419e+00 7.13715315e-01 -7.62851417e-01 -3.08450282e-01
7.10418522e-01 5.70605516e-01 -8.48093271e-01 9.69338045e-02
-8.57727826e-01 8.07090461e-01 2.12582871e-01 -2.96485680e-03
-1.23061574e+00 -8.63376617e-01 -1.02979863e+00 -2.24162057e-01
-7.51134530e-02 -1.00622141e+00 9.92952347e-01 -1.17637968e+00
-2.07843566e+00 1.44755828e+00 -9.97760892e-03 -3.64703745e-01
9.00898695e-01 -1.94549978e-01 -2.49803379e-01 3.96639109e-04
-9.08800773e-03 1.45452499e+00 1.39112282e+00 -1.36839354e+00
4.89760749e-02 -4.48189974e-02 5.10391183e-02 -7.65101090e-02
-2.01638415e-01 1.45443693e-01 -4.16512370e-01 -8.22161853e-01
-3.57966900e-01 -1.12576759e+00 -8.38280916e-02 4.25215393e-01
-6.20261669e-01 3.20909888e-01 9.37566459e-01 -7.68154323e-01
7.10229635e-01 -2.07899261e+00 5.20539463e-01 -3.50425914e-02
3.65415066e-01 2.93371648e-01 -4.63496596e-01 4.82635975e-01
-4.60408241e-01 4.67509836e-01 -5.19175589e-01 -9.91550028e-01
1.17964037e-01 1.99431777e-01 -4.37063217e-01 3.86209369e-01
3.97061110e-01 1.26840782e+00 -7.21355855e-01 -3.68818015e-01
3.60118859e-02 6.54985368e-01 -9.00982082e-01 4.93193895e-01
-3.16357464e-01 8.42261612e-01 2.92297125e-01 3.83764356e-01
7.11107194e-01 1.42544106e-01 1.03820272e-01 -1.52888373e-01
-1.23377517e-02 -4.46120389e-02 -5.62361240e-01 2.01656246e+00
-5.07038057e-01 8.30723345e-01 -2.08816081e-02 -5.11739492e-01
7.61578619e-01 3.81424189e-01 -2.28305385e-02 -3.36393744e-01
4.82596196e-02 7.53160790e-02 -2.01477006e-01 -2.14949846e-01
3.32168639e-01 -4.59801912e-01 -4.69497032e-02 4.81452554e-01
5.64821899e-01 -4.98488754e-01 2.07683012e-01 2.40146950e-01
7.95172393e-01 5.55426240e-01 -1.13650285e-01 -1.39561668e-01
1.76139534e-01 -4.93863344e-01 1.76912978e-01 4.59810823e-01
1.92174926e-01 9.78680670e-01 7.88415432e-01 -4.77802873e-01
-1.58773839e+00 -1.15162539e+00 3.39054734e-01 7.83618450e-01
-5.30435324e-01 -4.67558950e-01 -1.19855583e+00 -7.14940727e-01
-2.60167032e-01 7.93811321e-01 -9.94714320e-01 -1.82983562e-01
-6.41315460e-01 -4.84414697e-01 8.58964980e-01 2.72851199e-01
6.11553967e-01 -1.25565314e+00 -3.06883872e-01 -8.64294469e-02
1.87057614e-01 -9.16772604e-01 -4.85101372e-01 -1.33075997e-01
-7.71513045e-01 -7.91600049e-01 -9.01114285e-01 -6.72539651e-01
1.10228634e+00 -3.47015470e-01 1.34819365e+00 2.20766962e-02
-3.68924856e-01 2.53455132e-01 7.95581713e-02 -2.78955013e-01
-8.00031900e-01 7.28039593e-02 -2.58913804e-02 1.74040403e-02
-6.95935935e-02 -9.54311132e-01 -6.72420204e-01 9.31734219e-02
-1.18197799e+00 5.11278093e-01 5.56622982e-01 1.03819954e+00
2.06417397e-01 -1.33749247e-01 6.21815138e-02 -1.33633888e+00
5.56382596e-01 -2.56658476e-02 -7.07505882e-01 6.79062083e-02
-5.47587335e-01 3.17955196e-01 4.45050448e-01 -3.15156966e-01
-1.10864151e+00 1.81362793e-01 -1.52455345e-01 -4.06583071e-01
-3.14649731e-01 -7.95686916e-02 -3.86134177e-01 -8.01754519e-02
6.40086412e-01 8.91870912e-03 2.76303262e-01 -2.51990676e-01
8.90615284e-01 2.47272804e-01 7.18531370e-01 -4.56578463e-01
1.27103698e+00 6.54788554e-01 3.74731869e-02 -4.92817283e-01
-4.58872944e-01 3.40317100e-01 -8.00853014e-01 7.35216886e-02
1.23522580e+00 -8.99325371e-01 -5.07382929e-01 7.41021633e-01
-1.26528597e+00 -6.69181049e-01 -5.72948635e-01 -2.72597764e-02
-7.22274601e-01 -7.99321905e-02 -6.56825006e-01 -4.63316649e-01
-2.36673281e-01 -1.16134644e+00 1.21353400e+00 1.75588932e-02
-3.24988961e-01 -1.10835803e+00 3.17329198e-01 3.53478640e-01
5.11687219e-01 8.08340073e-01 9.12770748e-01 4.88994503e-03
-5.83146393e-01 5.45546114e-02 -8.62976387e-02 4.53075022e-01
2.72615075e-01 1.98708341e-01 -1.17925823e+00 -4.59452480e-01
-2.12729990e-01 -5.24359286e-01 9.55514610e-01 1.22587919e-01
1.05594862e+00 -6.29575551e-01 -2.32782051e-01 1.07041848e+00
1.31042755e+00 -8.27779993e-02 1.12234032e+00 3.25381495e-02
9.46925342e-01 5.91434777e-01 -1.48484692e-01 -7.29804635e-02
9.88426581e-02 6.62389696e-01 4.24322248e-01 -5.84368944e-01
-3.28216642e-01 -5.61343729e-01 6.06992245e-01 5.10456324e-01
-9.72364917e-02 -2.01826528e-01 -5.16119659e-01 4.77054745e-01
-1.54189122e+00 -1.00768232e+00 1.64438784e-01 1.88249874e+00
1.05857062e+00 1.40156612e-01 3.14082354e-02 3.02743423e-03
5.76964319e-01 1.56944290e-01 -3.53904754e-01 -6.37494802e-01
-7.54296482e-02 5.90975285e-01 5.42596877e-01 5.99957287e-01
-7.83460319e-01 1.15230310e+00 7.31261826e+00 6.64900959e-01
-1.24256957e+00 2.42756322e-01 8.64320576e-01 -8.47084001e-02
-7.50907183e-01 1.26476958e-01 -5.41487515e-01 2.91990042e-01
8.46266448e-01 3.31511974e-01 7.88283825e-01 6.08063579e-01
-1.32433951e-01 2.34021604e-01 -1.38439786e+00 1.06709540e+00
2.59500623e-01 -1.50076520e+00 3.38359267e-01 2.23754808e-01
1.04347122e+00 -5.57189524e-01 5.36842883e-01 1.33090794e-01
5.10735869e-01 -1.55517209e+00 7.92799294e-01 4.55264837e-01
1.56854379e+00 -7.79658854e-01 2.89033145e-01 -5.45200109e-02
-6.12570226e-01 3.27085614e-01 -1.24042630e-01 1.27296269e-01
2.44495228e-01 4.93703157e-01 -8.20337296e-01 1.67904571e-01
3.35603088e-01 4.22807366e-01 -8.18735600e-01 2.07794324e-01
-6.50179207e-01 5.81749499e-01 -6.03234768e-03 7.21628129e-01
4.79018129e-02 -2.56328195e-01 4.55328017e-01 1.06927240e+00
4.46923494e-01 -2.56438315e-01 -3.57575446e-01 1.34433377e+00
-3.64547610e-01 -3.91924828e-01 -9.46726501e-01 -4.70638752e-01
4.59667906e-04 1.38940108e+00 -7.64777660e-01 -3.25444549e-01
3.53017710e-02 1.69362354e+00 2.28885189e-01 3.02464396e-01
-7.43844748e-01 -4.68892217e-01 6.97992802e-01 2.29798749e-01
3.09092909e-01 -2.82569021e-01 -1.09965079e-01 -1.22347915e+00
-3.03429306e-01 -9.58558083e-01 -1.24205373e-01 -1.03625381e+00
-1.16286826e+00 7.89764702e-01 9.97886583e-02 -7.97739029e-01
-5.48527002e-01 -7.24929750e-01 -7.36946762e-01 1.03378952e+00
-1.25544846e+00 -1.84463024e+00 -3.51955622e-01 5.75433433e-01
2.80781806e-01 -4.82145309e-01 1.23541582e+00 2.18559027e-01
-2.59166092e-01 9.62522805e-01 -2.73901194e-01 1.47302270e-01
8.96666050e-01 -1.45169866e+00 1.02767944e+00 9.30014133e-01
7.72330314e-02 8.57913554e-01 8.20346057e-01 -5.80546379e-01
-1.18587077e+00 -1.11854100e+00 7.17430115e-01 -5.94773233e-01
1.86942413e-01 -1.09126854e+00 -3.69848579e-01 1.05576921e+00
8.93381596e-01 2.05554590e-02 7.52907574e-01 -2.21162960e-01
-5.13170421e-01 -2.45384667e-02 -1.19217873e+00 9.48421776e-01
1.22049141e+00 -8.27307284e-01 -1.43897757e-01 2.72097260e-01
8.54218781e-01 -4.18968469e-01 -6.32290781e-01 5.67259938e-02
6.58326805e-01 -1.09940934e+00 9.67303574e-01 -6.15752041e-01
8.57821643e-01 -3.84346098e-01 1.35270223e-01 -1.62571335e+00
-3.44269305e-01 -1.25182748e+00 2.68980265e-02 1.64831269e+00
3.68141890e-01 -5.08170485e-01 9.77567017e-01 4.26670551e-01
1.52216569e-01 -2.87561446e-01 -5.24469435e-01 -4.91554290e-01
2.25942969e-01 2.06483956e-02 7.74472058e-01 9.48171437e-01
-5.13822556e-01 3.27596545e-01 -8.43255222e-01 -2.91742086e-01
6.53423250e-01 -1.56073079e-01 1.13689554e+00 -7.66528904e-01
-5.63544989e-01 -2.48659953e-01 -4.23808575e-01 -6.15905643e-01
4.11579996e-01 -8.85065734e-01 -1.05494261e-01 -1.25559866e+00
1.12256967e-01 -1.92640826e-01 1.14168383e-01 6.24588072e-01
-3.94762717e-02 9.44059193e-01 3.61557961e-01 -1.24220895e-02
-1.48280263e-02 5.37267268e-01 1.44475830e+00 -1.56213745e-01
1.06758110e-01 -4.26315308e-01 -8.07627022e-01 4.95075017e-01
8.18278611e-01 -5.55655062e-01 -5.13903201e-01 -4.71841305e-01
5.43908119e-01 -1.28306910e-01 5.20810783e-01 -8.51756752e-01
-1.93983972e-01 2.73881797e-02 6.13204896e-01 -1.65423974e-01
4.33769792e-01 -6.51322603e-01 6.38618827e-01 3.73588562e-01
-4.71183389e-01 2.80564398e-01 2.68004477e-01 1.79989502e-01
-2.04890132e-01 -1.89018957e-02 8.40909183e-01 -5.01978159e-01
-2.24964455e-01 2.77783364e-01 -2.43232548e-01 -7.44073838e-02
7.83517540e-01 -2.04502985e-01 -2.01949492e-01 -6.79221809e-01
-6.36562705e-01 -3.07293773e-01 9.07309711e-01 3.44478577e-01
5.39921582e-01 -1.71297455e+00 -8.35165203e-01 5.93542814e-01
-5.10198027e-02 -1.31092444e-01 1.43200085e-02 3.91908497e-01
-8.39006722e-01 5.85741103e-02 -8.36341381e-01 -4.46370095e-01
-1.26249444e+00 5.78438520e-01 1.68549106e-01 -4.06578928e-01
-8.78031626e-02 1.11408937e+00 4.67500210e-01 -7.68183708e-01
-2.52611250e-01 -5.50996065e-02 1.61978915e-01 -3.98161709e-02
3.41284513e-01 1.03034437e-01 -1.41416386e-01 -5.36039531e-01
6.98524788e-02 5.84871352e-01 8.12574923e-02 -6.67231321e-01
1.41929972e+00 1.05503537e-01 -4.97078508e-01 3.01619321e-01
1.37261713e+00 2.51082659e-01 -1.56949747e+00 4.20617908e-01
-6.19739294e-01 -5.60746312e-01 -2.11655736e-01 -7.38513291e-01
-1.21892250e+00 1.14330649e+00 7.87915409e-01 -1.61702707e-01
1.04150808e+00 -4.15919214e-01 6.25142574e-01 -6.50203675e-02
1.27368972e-01 -7.31692135e-01 3.18135411e-01 1.64973333e-01
1.12799883e+00 -1.15127552e+00 9.32752788e-02 -2.46277750e-01
-4.92659807e-01 9.89624977e-01 5.40519178e-01 -2.78916001e-01
4.95772451e-01 2.72463709e-01 1.80443808e-01 -1.76215932e-01
-5.74465334e-01 2.31363952e-01 2.73916423e-01 7.22724617e-01
4.66790885e-01 -3.46403942e-02 1.22690499e-01 -1.92759588e-01
-6.10863209e-01 -7.21926540e-02 5.00647068e-01 7.94101715e-01
4.35106784e-01 -1.72488308e+00 -2.46543095e-01 -5.73087437e-03
-5.72570086e-01 -2.13394329e-01 -4.13370132e-01 7.29024410e-01
5.04236162e-01 4.53108817e-01 7.25578740e-02 -2.18889162e-01
-5.37794717e-02 2.05812410e-01 9.70045745e-01 -7.06108391e-01
-6.55661702e-01 -1.05844893e-01 -5.34375422e-02 -5.57827771e-01
-4.36100274e-01 -5.73492825e-01 -6.78870022e-01 -2.79936105e-01
-1.01570047e-01 -7.48263374e-02 7.72381544e-01 5.93329191e-01
4.59497780e-01 5.38242877e-01 3.68366808e-01 -1.25127828e+00
7.47974589e-03 -9.16494548e-01 -3.33664030e-01 6.84137821e-01
3.94224077e-01 -4.23908353e-01 -2.87556291e-01 7.37559140e-01] | [11.910778045654297, -0.312527596950531] |
aa10e2ee-3180-484f-b0a4-8a896f60fc37 | end-to-end-chinese-landscape-painting | 2011.05552 | null | https://arxiv.org/abs/2011.05552v1 | https://arxiv.org/pdf/2011.05552v1.pdf | End-to-End Chinese Landscape Painting Creation Using Generative Adversarial Networks | Current GAN-based art generation methods produce unoriginal artwork due to their dependence on conditional input. Here, we propose Sketch-And-Paint GAN (SAPGAN), the first model which generates Chinese landscape paintings from end to end, without conditional input. SAPGAN is composed of two GANs: SketchGAN for generation of edge maps, and PaintGAN for subsequent edge-to-painting translation. Our model is trained on a new dataset of traditional Chinese landscape paintings never before used for generative research. A 242-person Visual Turing Test study reveals that SAPGAN paintings are mistaken as human artwork with 55% frequency, significantly outperforming paintings from baseline GANs. Our work lays a groundwork for truly machine-original art generation. | ['Alice Xue'] | 2020-11-11 | null | null | null | null | ['chinese-landscape-painting-generation'] | ['computer-vision'] | [ 7.69013464e-01 4.65943336e-01 3.29621792e-01 -1.83766469e-01
-1.03170061e+00 -9.19216335e-01 1.03010166e+00 -1.07234395e+00
2.69673049e-01 1.23842692e+00 3.83302510e-01 -3.28124404e-01
5.64177871e-01 -1.25242281e+00 -9.73337770e-01 -3.99386704e-01
7.34597564e-01 6.93812966e-01 -3.66601199e-01 -1.54887080e-01
1.12952568e-01 2.42919683e-01 -9.84129548e-01 9.14532065e-01
7.32922912e-01 4.80196536e-01 -1.56618096e-02 1.11315417e+00
-3.99844438e-01 7.98342109e-01 -1.16691267e+00 -1.25270772e+00
3.91033918e-01 -1.33074021e+00 -6.11009896e-01 2.78657265e-02
7.02635109e-01 -6.05858028e-01 -1.89769492e-01 6.58610940e-01
4.56823409e-01 -3.65187973e-01 1.00773382e+00 -1.36105251e+00
-1.63231730e+00 6.89285040e-01 -4.52886105e-01 -5.05289853e-01
2.99195766e-01 6.40584648e-01 9.01562810e-01 -7.22022176e-01
1.14677608e+00 1.28099597e+00 5.31543016e-01 1.05450916e+00
-1.45510006e+00 -5.76004803e-01 -2.78430074e-01 -3.88772577e-01
-1.08820438e+00 -2.20925048e-01 8.90548348e-01 -3.00843388e-01
5.52342892e-01 4.70959663e-01 1.09580064e+00 1.87236655e+00
-7.07058758e-02 9.11651611e-01 1.36965215e+00 -4.44209397e-01
2.53375113e-01 -1.60954386e-01 -9.24193561e-01 4.77771580e-01
2.11383015e-01 1.77231222e-01 -6.29957378e-01 2.34353226e-02
1.51324153e+00 -3.62969071e-01 -1.10586680e-01 2.60047525e-01
-1.18144906e+00 7.23233104e-01 4.58520323e-01 1.13306329e-01
-5.88342011e-01 8.02009225e-01 -2.65526235e-01 1.77827626e-01
3.90379041e-01 7.72530019e-01 1.78612858e-01 -3.93209636e-01
-9.76865590e-01 5.14639199e-01 5.93486428e-01 1.37055409e+00
4.93254751e-01 4.13746357e-01 -7.38070548e-01 4.25645739e-01
-8.54848772e-02 8.03401291e-01 1.00723103e-01 -1.01469839e+00
4.74377692e-01 4.65110600e-01 1.72957450e-01 -5.30576527e-01
5.92140079e-01 -1.59726143e-01 -8.63988578e-01 6.09970689e-01
3.04787487e-01 -3.29636097e-01 -1.43306386e+00 1.43207610e+00
-1.98298395e-01 -1.68679506e-01 -4.68164450e-03 7.56089568e-01
6.83508992e-01 7.56015062e-01 1.37898639e-01 4.47316974e-01
1.16709399e+00 -1.19746411e+00 -7.25052536e-01 -1.70261264e-01
-3.23716775e-02 -9.71710026e-01 1.52540743e+00 2.56673753e-01
-1.51555538e+00 -5.90398610e-01 -8.69141400e-01 -2.11600393e-01
-2.84903377e-01 1.77852452e-01 5.76982796e-01 9.52965260e-01
-1.37763906e+00 3.64755303e-01 -3.92352670e-01 -1.74218431e-01
1.00823379e+00 -2.57342279e-01 -1.14833847e-01 -4.47451249e-02
-7.07069099e-01 7.78929055e-01 1.28911689e-01 4.39668857e-02
-1.09496808e+00 -7.88190424e-01 -4.37080145e-01 -1.35224581e-01
-6.63607419e-02 -1.39915204e+00 1.30535054e+00 -1.33098304e+00
-1.95840132e+00 8.70149910e-01 -1.38864100e-01 -3.33001822e-01
1.09801781e+00 -2.53865540e-01 -3.11793804e-01 1.15192793e-01
-2.30512768e-02 1.10064340e+00 1.21534705e+00 -1.66660655e+00
-2.05846265e-01 2.90942430e-01 -6.34352118e-02 1.06760427e-01
2.19359487e-01 -4.50404316e-01 -3.81252348e-01 -1.12948513e+00
-4.66023445e-01 -7.47570157e-01 -3.43683362e-02 3.45372371e-02
-8.84945869e-01 6.51571825e-02 6.81959748e-01 -7.72172272e-01
6.82544947e-01 -1.81362081e+00 1.03283249e-01 1.91573068e-01
7.98870698e-02 4.71895821e-02 -4.96929944e-01 7.63333440e-01
7.91596323e-02 6.52733684e-01 -4.24472570e-01 -6.18170619e-01
2.48797894e-01 2.19708476e-02 -9.66608703e-01 -3.47887188e-01
6.73757374e-01 1.92118526e+00 -8.21478844e-01 -4.02114332e-01
1.11626364e-01 7.30530143e-01 -2.96371162e-01 4.59065586e-01
-5.10864079e-01 7.36489296e-01 -1.93071157e-01 8.77541840e-01
6.09836698e-01 -1.44668102e-01 -4.35112342e-02 6.05518818e-02
2.98324972e-01 9.54950415e-03 -3.10714155e-01 1.88442779e+00
-4.11588907e-01 9.17258441e-01 -5.21894097e-01 2.07658052e-01
1.05197167e+00 2.24498749e-01 -4.10428822e-01 -6.51779473e-01
9.67640430e-04 1.71847060e-01 -4.46508795e-01 -6.70793280e-02
7.23776937e-01 -3.65220785e-01 -7.09339306e-02 7.83809662e-01
-2.78204501e-01 -1.09073257e+00 4.36366871e-02 2.21184328e-01
1.08228016e+00 7.89682329e-01 -2.80551612e-01 1.25543788e-01
-1.15151979e-01 1.35339662e-01 -7.10685179e-02 9.15951550e-01
4.88450468e-01 1.39091372e+00 6.80307090e-01 -5.59152961e-01
-1.61528277e+00 -1.62208438e+00 5.43087363e-01 4.60889578e-01
-2.63067782e-01 -3.19657326e-01 -1.32470572e+00 -6.90303326e-01
-3.48574370e-01 1.29452062e+00 -9.26128030e-01 -9.10886843e-03
-6.51279628e-01 -3.05598944e-01 9.99887288e-01 5.94969630e-01
6.63460255e-01 -1.59361935e+00 -7.86305189e-01 -5.14429100e-02
-9.58374590e-02 -7.31494129e-01 -5.46780944e-01 -4.95883286e-01
-6.16978407e-01 -8.34869504e-01 -1.34265745e+00 -6.23949230e-01
9.29035962e-01 -1.72728941e-01 1.65673292e+00 1.76916778e-01
-3.13621014e-01 4.03044939e-01 -3.55844885e-01 -5.73696434e-01
-8.70521367e-01 1.16952062e-01 -6.22489810e-01 -1.11383796e-01
-6.44107312e-02 -6.77201271e-01 -6.67813182e-01 2.27431636e-02
-1.01766634e+00 8.84671032e-01 1.03405190e+00 9.01338279e-01
5.91158092e-01 -3.33283365e-01 3.64291489e-01 -1.12130225e+00
8.59972358e-01 3.37253883e-02 -4.51040745e-01 5.34309983e-01
-2.76448607e-01 -3.35551868e-03 4.58301455e-01 -3.95642459e-01
-1.41191137e+00 -1.26526758e-01 3.72157395e-02 -2.79343933e-01
-2.90335953e-01 -1.31087273e-01 -1.33961275e-01 2.54961729e-01
6.77172065e-01 4.51126277e-01 -3.22429806e-01 -2.70767391e-01
9.28388596e-01 2.80218631e-01 9.46614563e-01 -7.12165952e-01
1.11281991e+00 4.77169514e-01 -1.20086432e-01 -5.78763783e-01
-3.29044908e-01 8.12554538e-01 -4.31294054e-01 -2.97477812e-01
1.12213922e+00 -7.33084559e-01 -3.91590059e-01 5.23538411e-01
-1.45086157e+00 -1.09285963e+00 -8.65932465e-01 -3.02134961e-01
-6.70414925e-01 -2.18836457e-01 -5.96206248e-01 -9.90023971e-01
-6.53179407e-01 -6.10249639e-01 1.42478061e+00 3.36628526e-01
-4.66732651e-01 -5.94190300e-01 1.29494846e-01 2.76658207e-01
6.29829586e-01 7.96511948e-01 9.30450797e-01 3.56505394e-01
-9.69523907e-01 -1.35818765e-01 -3.62374365e-01 1.03317901e-01
1.72777444e-01 2.73173571e-01 -1.02389443e+00 -1.67315118e-02
-4.95264918e-01 -3.36912632e-01 7.80183613e-01 1.05487153e-01
9.23029721e-01 -4.23353374e-01 -6.93692118e-02 7.39513576e-01
1.39651895e+00 3.55800509e-01 1.40636861e+00 6.64871857e-02
9.54672158e-01 2.43765399e-01 -2.30469137e-01 3.96855846e-02
9.09990221e-02 2.07468897e-01 3.15270483e-01 -4.09811169e-01
-8.60798836e-01 -1.17470753e+00 3.27558905e-01 3.05380821e-01
-4.88788992e-01 -7.95003533e-01 -5.78958392e-01 6.31003022e-01
-1.42108130e+00 -1.09220731e+00 -1.70668989e-01 1.74711227e+00
8.17663491e-01 1.16705380e-01 6.21284246e-02 -2.68052906e-01
6.87479973e-01 -1.25721246e-02 -5.27604461e-01 -4.31742847e-01
-5.11460662e-01 8.50994110e-01 2.13614941e-01 2.71582991e-01
-5.50915897e-01 1.23820567e+00 7.14388895e+00 5.97902000e-01
-7.69722342e-01 8.24265331e-02 1.05804181e+00 -2.29085639e-01
-1.02149522e+00 1.67774498e-01 -1.56094462e-01 5.11757016e-01
4.42453086e-01 1.76333249e-01 8.71089637e-01 6.99161649e-01
-3.58191639e-01 -3.84885073e-02 -9.47847009e-01 1.08664048e+00
1.42879933e-01 -1.79471171e+00 6.48904562e-01 9.51617509e-02
1.33292830e+00 -5.40910006e-01 4.53316510e-01 1.36307552e-01
9.20795918e-01 -1.61048722e+00 1.10584128e+00 8.79066765e-01
1.35586452e+00 -6.81750894e-01 3.30403715e-01 -2.21046329e-01
-7.94013619e-01 4.77987766e-01 -1.47048250e-01 -8.68961811e-02
7.10752368e-01 3.14202964e-01 -1.07685935e+00 5.23579061e-01
4.14367288e-01 3.05401862e-01 -6.47684455e-01 5.26079476e-01
-9.98833001e-01 7.30997384e-01 -2.22581588e-02 -1.22680746e-01
1.76563963e-01 -2.50355959e-01 3.46892059e-01 1.07515728e+00
7.00244665e-01 4.36347462e-02 -6.52104080e-01 1.86817324e+00
-4.31035995e-01 -4.93766546e-01 -8.30323994e-01 -5.44247985e-01
3.86237472e-01 1.13807893e+00 -9.06237006e-01 -5.55929601e-01
2.84450680e-01 1.84048343e+00 1.75008133e-01 7.42780209e-01
-9.30522621e-01 -6.03137612e-01 3.28412533e-01 6.12452403e-02
4.32390988e-01 -2.44540811e-01 -8.69050622e-01 -9.22902107e-01
2.19251178e-02 -7.89441407e-01 -1.91065699e-01 -1.63057041e+00
-1.58663356e+00 1.00601649e+00 -3.65403831e-01 -8.25596213e-01
-3.48294586e-01 -3.38224620e-01 -1.15229225e+00 1.10244679e+00
-9.93541121e-01 -1.94345307e+00 -5.84394753e-01 3.43832970e-01
5.24207592e-01 -1.81191608e-01 1.01085341e+00 -1.67661980e-01
-9.85452682e-02 6.70929790e-01 -3.33524555e-01 3.84458601e-01
5.18023133e-01 -1.39369798e+00 1.45965087e+00 9.65990722e-01
5.44137299e-01 5.01757085e-01 5.08378029e-01 -1.08343863e+00
-1.33003974e+00 -1.14700878e+00 6.13371253e-01 -1.04447627e+00
5.14753833e-02 -6.35546863e-01 -3.82817209e-01 7.03945160e-01
9.24062371e-01 -6.42342210e-01 3.35582435e-01 -4.90722090e-01
-4.27949965e-01 2.92841822e-01 -1.16033769e+00 1.08627892e+00
1.39349592e+00 -6.84840500e-01 -5.90119183e-01 1.64336011e-01
7.82153249e-01 -4.00981575e-01 -4.76466894e-01 -1.59263045e-01
8.51927578e-01 -9.41148400e-01 8.57324004e-01 -4.99770164e-01
1.20094943e+00 -3.51978004e-01 1.10895276e-01 -1.40544057e+00
-2.39867017e-01 -1.22848511e+00 9.26979855e-02 1.34248900e+00
5.56904614e-01 -3.02074373e-01 1.00234699e+00 6.28933251e-01
-1.09822735e-01 -1.94761023e-01 -5.71395755e-01 -7.47654438e-01
1.08728863e-01 -2.82277405e-01 1.12447762e+00 8.04715455e-01
-6.75133824e-01 3.97035599e-01 -5.89918733e-01 -5.60138583e-01
5.97076714e-01 2.34071806e-01 1.27001965e+00 -6.64419293e-01
-5.18044710e-01 -7.08010018e-01 -1.50079746e-02 -7.29641259e-01
-2.74995416e-01 -6.81814492e-01 4.32699323e-02 -1.97246778e+00
1.56563183e-03 -1.59816593e-01 4.55489546e-01 5.20240307e-01
-5.70195913e-02 1.10273814e+00 4.91757333e-01 1.43845767e-01
2.80788206e-02 6.36347473e-01 1.66952968e+00 -2.77228177e-01
-1.49542943e-01 -4.60579395e-01 -8.82570267e-01 3.40472460e-01
7.16776550e-01 -3.04517388e-01 -4.18881714e-01 -8.81212115e-01
5.37865400e-01 -2.90017426e-01 7.06871629e-01 -9.15799737e-01
-2.13464141e-01 -1.84200451e-01 9.21934962e-01 -6.53022885e-01
5.17678320e-01 -4.27897841e-01 1.02169287e+00 4.16214317e-01
-3.29393923e-01 2.64831871e-01 3.95601764e-02 4.54498380e-01
1.72867417e-01 2.14848280e-01 3.39819640e-01 -4.95873302e-01
-3.09747279e-01 -5.84852472e-02 -1.75070047e-01 1.06205091e-01
8.01779926e-01 -3.35418791e-01 -6.89112067e-01 -7.42238343e-01
-3.65247428e-01 -4.28967863e-01 9.12276149e-01 3.82033706e-01
7.50163078e-01 -1.74550939e+00 -1.03279209e+00 2.08198100e-01
-1.67719439e-01 1.52314410e-01 1.91550225e-01 5.58954338e-03
-1.01591241e+00 4.36152890e-02 -5.14505506e-01 -1.49192765e-01
-6.95817828e-01 3.73846143e-01 -1.08043864e-01 -2.59447787e-02
-7.61432528e-01 1.10938358e+00 3.11952174e-01 -3.08198389e-02
-3.05723727e-01 -2.24568158e-01 6.98156059e-01 -4.89171416e-01
5.46089470e-01 2.40862042e-01 -4.36906338e-01 -2.14036077e-01
9.54393595e-02 4.85589743e-01 3.33186030e-01 -8.84714246e-01
1.22670102e+00 4.43581641e-01 -1.64340407e-01 3.51851046e-01
6.06148362e-01 1.08386829e-01 -1.48622262e+00 5.03292620e-01
-4.83313084e-01 -5.87490678e-01 -5.32201588e-01 -1.45870173e+00
-9.32167351e-01 8.79096806e-01 1.62844658e-01 -7.06512183e-02
1.03756654e+00 -1.01554498e-01 1.09017158e+00 1.77285388e-01
5.11188984e-01 -8.06062341e-01 3.50776702e-01 8.30205232e-02
1.39397597e+00 -7.33850300e-01 -2.33986944e-01 -1.12447858e-01
-1.00435543e+00 1.04013801e+00 5.91195226e-01 -4.26727861e-01
-7.64547586e-02 1.66902095e-01 7.40557238e-02 -2.07538277e-01
-5.98500371e-01 5.98536246e-02 4.24358934e-01 9.89794135e-01
2.42186978e-01 3.83816332e-01 8.63326937e-02 6.17719054e-01
-7.95319855e-01 1.89924806e-01 3.86199653e-01 8.64695072e-01
3.21715176e-01 -1.14935136e+00 -3.28666061e-01 2.27335453e-01
-5.09645455e-02 -3.13437939e-01 -1.19733334e+00 8.27083945e-01
2.33580783e-01 6.33078992e-01 2.65515983e-01 -4.28660005e-01
1.62697822e-01 2.32503057e-01 7.64305353e-01 -2.77826667e-01
-7.37130761e-01 -1.77241966e-01 2.00611800e-01 -3.31820667e-01
-1.54030040e-01 -2.19993427e-01 -6.62262440e-01 -4.00425583e-01
2.24809647e-01 -1.47567019e-01 7.05926895e-01 5.26707292e-01
5.72273254e-01 7.71437705e-01 6.82668239e-02 -9.05607522e-01
1.94792315e-01 -9.85750020e-01 -2.40892738e-01 4.86573696e-01
-4.94034514e-02 6.27928153e-02 -2.29343921e-02 5.26939332e-01] | [11.643706321716309, -0.3147982656955719] |
eeed8cb3-7fc4-4396-b2c6-8e4dc204ce90 | dao-shi-ping-jie-wang-shu-ju-yi-zai-githubbei | null | null | https://kgco.github.io/RateMySupervisor/html/index.html | https://github.com/kgco/RateMySupervisor | 导师评价网数据已在Github备份，可直接检索导师个人信息！ | 永久免费开源的导师评价数据、数据爬虫、无需编程基础的展示网页以及新信息补充平台。从导师评价网获取的原始数据，本平台开源存储数据以备各种你懂得的不可抗力因素。 项目名称：RateMySupervisor，链接： https://github.com/kgco/RateMySupervisor
在线浏览：可以直接访问这个GitHub Pages页面，打开即可在线浏览数据。导师评价信息 https://kgco.github.io/RateMySupervisor/html/index.html 网页前端加载出来之后，要加载一个20MB左右的js数据文件，由于网络原因可能速度比较慢，所以会有一小段时间下拉列表里没有数据，请耐心等待。
离线浏览：点击右上方Code按钮中的Download ZIP，下载文件，然后打开html/index.html即可浏览数据。由于调用了bootstrap和jquery，所以打开的时候最好保持网络连接（不打开也行啦，就是UI可能有点乱）。 | ['导师评价网'] | 2020-08-25 | null | null | null | null | ['2d-human-pose-estimation'] | ['computer-vision'] | [-1.01411295e+00 -4.14208382e-01 -1.54846027e-01 3.28896344e-01
-1.15597939e+00 -6.90958261e-01 7.78430626e-02 1.42859125e+00
-1.23165540e-01 7.69456089e-01 5.56540310e-01 -2.23837733e-01
-2.11709693e-01 -1.18935001e+00 -7.84280717e-01 -7.98518300e-01
-8.43119562e-01 1.80020773e+00 8.07940006e-01 -7.58795083e-01
2.02205658e-01 1.80815235e-01 -9.77557898e-01 -3.77624370e-02
5.91515005e-01 8.52833092e-01 7.98166633e-01 5.07649183e-01
1.99908465e-01 6.13558948e-01 -1.02528024e+00 1.06897607e-01
-1.17508680e-01 3.34166080e-01 -6.78391576e-01 4.84329909e-01
-4.67660666e-01 -1.21933186e+00 -1.30361092e+00 9.07155454e-01
3.76907438e-01 -1.01745605e-01 7.77185619e-01 -1.30270016e+00
-7.26480246e-01 5.48000455e-01 8.87946486e-02 9.10138488e-01
8.84226337e-02 -2.65841991e-01 6.42515600e-01 -4.35628295e-01
5.56381881e-01 1.00174534e+00 1.40941310e+00 -2.56809205e-01
-7.99283564e-01 -8.60856712e-01 3.93930167e-01 -2.20594034e-02
-1.65769005e+00 4.22355503e-01 -8.06556344e-02 2.72057503e-01
9.11737800e-01 4.83008057e-01 9.64911878e-01 1.51131916e+00
5.17810464e-01 1.29034007e+00 1.08349395e+00 2.25815341e-01
1.84374779e-01 7.41042256e-01 -1.55188009e-01 1.79088041e-02
7.32710004e-01 -1.46190181e-01 2.52890259e-01 2.14408174e-01
7.35253930e-01 4.24440593e-01 8.64624381e-02 1.11869228e+00
-1.36409223e+00 1.54908764e+00 3.06669693e-03 1.02594125e+00
5.02063194e-03 -6.86335802e-01 -1.48180962e-01 6.50770247e-01
5.63386858e-01 -4.96627390e-01 -1.10784590e-01 1.13386452e-01
-4.88670841e-02 1.62607983e-01 1.33393300e+00 1.83693731e+00
1.53375685e+00 1.48528710e-01 -1.79363057e-01 1.11703718e+00
1.05498517e+00 6.73040152e-01 1.03011084e+00 -8.87683153e-01
-2.77149349e-01 3.89639169e-01 -7.37892687e-01 -1.19982159e+00
-5.69736421e-01 -1.20869565e+00 -3.69233727e-01 -6.32862091e-01
-1.52849659e-01 -8.76263380e-01 -3.43562067e-01 1.02714717e+00
7.22745001e-01 -6.74121678e-01 2.21297264e-01 3.70101243e-01
7.37170696e-01 1.52738297e+00 -5.46306610e-01 -4.71038520e-01
1.32335997e+00 -5.91116250e-01 -1.39384016e-01 -4.32865880e-02
-2.67788738e-01 -1.85307205e+00 1.35496032e+00 1.43967047e-01
-1.08742535e+00 2.25881130e-01 -1.16475856e+00 -3.35726142e-02
-7.10816532e-02 -5.76524198e-01 4.73767072e-01 1.10491312e+00
-1.47185957e+00 -6.52140975e-02 -9.39379215e-01 -1.12739003e+00
1.01310961e-01 3.50111812e-01 2.80456483e-01 -5.62684298e-01
-1.19423985e+00 1.66850656e-01 -4.13918078e-01 -1.27414420e-01
-1.95171714e+00 -5.86757064e-01 -2.53098086e-02 1.14241101e-01
1.39610946e+00 -1.40253201e-01 1.77728117e+00 -8.28440726e-01
-8.43513370e-01 5.65743089e-01 -3.56813461e-01 -1.07845917e-01
5.30846834e-01 7.16631591e-01 -1.37096569e-01 5.52988827e-01
6.06750071e-01 5.68051815e-01 4.61385965e-01 -2.21534133e+00
-1.20461321e+00 -8.67884815e-01 5.01823068e-01 6.31429628e-02
-5.57728052e-01 1.16337538e+00 -8.71296048e-01 -7.26713240e-01
-6.63493425e-02 -1.62710512e+00 2.15345711e-01 -8.86164427e-01
-6.92523539e-01 3.25259477e-01 8.49622309e-01 -1.56386614e-01
1.45018768e+00 -2.11041307e+00 -6.71242356e-01 1.04582930e+00
3.47893268e-01 1.46657810e-01 2.44919598e-01 2.13774514e+00
6.77020788e-01 1.18233585e+00 -7.09020436e-01 6.30854309e-01
-7.50256330e-02 -9.89888757e-02 7.76217461e-01 -2.24201083e-01
-1.03948784e+00 -3.80854532e-02 -7.05517828e-02 -3.84003252e-01
1.19948827e-01 6.67845309e-01 4.15287435e-01 5.18466175e-01
6.25721589e-02 -1.58116341e-01 -2.19705001e-01 1.24817407e+00
8.28024507e-01 -6.83601439e-01 4.82740968e-01 1.97009653e-01
-3.87821466e-01 -1.55756667e-01 -2.07763839e+00 1.51718986e+00
-3.75860214e-01 3.56334478e-01 7.99473226e-01 -5.53625524e-01
9.23130035e-01 1.30176961e+00 1.30779850e+00 2.62881607e-01
8.20954368e-02 5.14421642e-01 1.54841110e-01 -9.80696976e-02
7.95140922e-01 -4.43111122e-01 6.36004865e-01 1.23614430e+00
-5.09549201e-01 2.24958241e-01 5.28949976e-01 -3.16069067e-01
1.57712388e+00 -5.75596452e-01 -1.16100550e-01 -3.40313256e-01
-4.94611263e-02 2.24293303e-02 2.73921758e-01 2.49413867e-02
-4.48152274e-01 1.77977681e-01 -6.22716486e-01 1.48952395e-01
-1.93467177e-02 -1.62045288e+00 -5.71803562e-02 1.88100541e+00
-2.02421471e-01 -1.50983822e+00 -1.17634177e-01 -4.72443581e-01
-6.46738708e-01 4.35270965e-01 4.52458411e-01 6.68537915e-01
-3.48353297e-01 -6.56395674e-01 8.54949728e-02 -3.64348173e-01
1.33654559e+00 -1.22169983e+00 2.16608852e-01 -4.56513941e-01
-3.88240695e-01 -7.10295796e-01 -6.28301203e-01 -3.04748356e-01
-1.37469971e+00 -5.85046351e-01 -2.12115020e-01 -1.32440698e+00
1.18608844e+00 7.44301021e-01 6.75418615e-01 7.38711357e-01
4.31739613e-02 1.07849503e+00 -5.23319423e-01 -1.35222280e+00
3.59193496e-02 8.85250628e-01 2.29772463e-01 5.47176123e-01
7.69520521e-01 -5.57090163e-01 -1.42400491e+00 7.26186991e-01
-1.57838380e+00 -1.65762097e-01 1.26746893e-01 -1.59501895e-01
8.44083786e-01 5.19369543e-01 -9.78987217e-02 -6.74020231e-01
1.08444107e+00 -1.66961432e-01 -1.10181630e+00 -7.70261437e-02
-1.73345447e+00 -1.53697813e+00 1.04263818e+00 -4.14313018e-01
-7.01742172e-01 -9.85606730e-01 -6.29400671e-01 -3.53700101e-01
2.50167340e-01 4.13903445e-01 -4.06226099e-01 7.67366409e-01
-4.33701336e-01 5.22954389e-02 6.02844596e-01 -7.62223899e-01
-2.34523803e-01 8.47780049e-01 1.42043486e-01 -8.28041196e-01
7.05064237e-01 4.56256658e-01 -4.13978577e-01 -3.11047286e-01
5.54222167e-01 -3.29182416e-01 1.21304087e-01 7.00860202e-01
8.11538219e-01 -1.49717748e+00 -1.80688679e-01 5.97720265e-01
-2.76579857e-01 3.58289806e-03 -4.98003393e-01 1.41823280e+00
-4.60524559e-01 -1.19484738e-01 -1.52906036e+00 1.71188846e-01
-1.81190223e-01 -2.36638093e+00 2.25377202e-01 1.73425484e+00
2.47942179e-01 -9.59604919e-01 9.54610407e-02 8.88981879e-01
5.21713376e-01 -6.01627588e-01 7.91572928e-01 -7.54266679e-01
-1.54489541e+00 -3.96743268e-01 7.18842000e-02 1.06318808e+00
6.70973897e-01 8.89160156e-01 1.10130288e-01 -1.01283121e+00
-6.26307726e-01 1.69887379e-01 -4.97922413e-02 8.33640218e-01
2.89467037e-01 -1.00224900e+00 -4.92895603e-01 1.04295373e+00
2.02175617e+00 7.33908117e-01 7.38128603e-01 2.27567601e+00
-1.40743405e-01 1.43314630e-01 4.06805396e-01 9.30103719e-01
5.40956676e-01 1.07940662e+00 1.80796891e-01 9.42245603e-01
-6.82889402e-01 1.71737671e-02 1.25047851e+00 1.64667559e+00
5.29639460e-02 3.91759686e-02 -1.96571425e-01 8.25540245e-01
-2.23898768e+00 -1.45822656e+00 3.38480353e-01 2.61941743e+00
6.29453361e-01 -4.55487818e-01 1.02564208e-01 -7.29815364e-01
1.29819715e+00 2.98936516e-01 -5.96384779e-02 -7.51165986e-01
8.49720597e-01 8.40275049e-01 6.70340359e-01 3.71586323e-01
-5.82203746e-01 -7.76417628e-02 1.78891408e+00 8.48686635e-01
-2.11801624e+00 2.07310423e-01 4.46236461e-01 -1.45816457e+00
-4.88709360e-01 5.81634343e-01 -1.16086435e+00 1.73836815e+00
1.10214400e+00 -1.37382388e+00 2.40808353e-01 5.53250849e-01
3.26999396e-01 2.13097647e-01 -1.38288647e-01 1.60285497e+00
-1.57994442e-02 -1.93969572e+00 -7.50361264e-01 5.19855559e-01
8.09717357e-01 3.56195033e-01 -3.59543383e-01 -4.88235168e-02
8.26534331e-01 -5.41811109e-01 1.11574411e+00 -6.44553602e-01
8.62546027e-01 -7.06548810e-01 1.12134349e+00 -1.55148253e-01
-1.32613194e+00 -4.36402231e-01 -6.18332560e-05 -6.25416040e-02
1.04505205e+00 3.41906905e-01 -1.40534341e-01 7.71813273e-01
3.87198031e-01 5.17785847e-02 -5.39399862e-01 2.29183912e+00
-6.49096966e-01 1.75930604e-01 -2.37381682e-01 -1.23948646e+00
3.01947951e-01 -4.28104341e-01 8.46384883e-01 1.41404951e+00
6.91281080e-01 3.96377832e-01 1.38492778e-01 6.34698451e-01
-6.33389771e-01 2.32904583e-01 -1.75504163e-01 8.93580019e-01
9.68392313e-01 1.26054525e+00 -1.41137815e+00 -5.95094681e-01
-4.13776338e-01 -1.24948323e-01 -7.14654028e-01 -5.16736135e-02
-1.30176544e-01 -7.14340210e-01 1.74893007e-01 6.36546493e-01
-2.76620746e-01 -7.04035163e-01 9.59919691e-01 -5.52516043e-01
-8.41967613e-02 -1.25214064e+00 4.98469561e-01 -7.07519472e-01
-1.25085962e+00 2.82096177e-01 4.01121497e-01 -1.31302905e+00
1.38505369e-01 -8.77529502e-01 1.49941996e-01 4.38718438e-01
1.05774313e-01 -7.40496278e-01 -5.91544807e-02 4.90132868e-02
1.07376432e+00 1.48860261e-01 6.56605005e-01 8.00931215e-01
-9.71109629e-01 1.98570460e-01 1.65350068e+00 7.33978629e-01
4.17095214e-01 -1.48203278e+00 -1.29619992e+00 3.73563051e-01
8.04985344e-01 -4.46152650e-02 -4.33755703e-02 -6.66266859e-01
-1.14983773e+00 -7.80095875e-01 6.64052844e-01 5.05320609e-01
6.92979634e-01 -5.24250031e-01 -3.00751776e-01 1.74361050e+00
7.26718307e-01 3.91378522e-01 8.91076028e-01 -5.73725939e-01
3.34499031e-01 -1.68969035e+00 -1.47201502e+00 8.42320085e-01
5.05428553e-01 -3.72902244e-01 -2.41704565e-02 7.06915975e-01
4.26725805e-01 -9.94050503e-01 -1.67238319e+00 9.99397505e-03
1.14303517e+00 -6.08036935e-01 -2.67906755e-01 2.33050942e-01
3.47251117e-01 -1.47457048e-01 -3.18798512e-01 -1.00905824e+00
3.78789186e-01 -1.10995138e+00 1.03781092e+00 1.46514165e+00
3.51366013e-01 -1.11416996e+00 1.14238012e+00 1.40655541e+00
-5.64324021e-01 -6.03922069e-01 -1.97172773e+00 -1.16814077e+00
6.54394403e-02 -2.62710899e-01 3.55725288e-01 -1.81511134e-01
-1.70440033e-01 1.16865478e-01 5.99162281e-01 -3.06074172e-01
1.74246728e-01 -5.97634673e-01 5.00753045e-01 -7.36388266e-01
3.94081444e-01 -1.36750847e-01 -3.13071400e-01 -6.58964157e-01
-1.32696342e+00 -8.24604928e-01 -2.01315713e+00 -2.07248259e+00
2.83681184e-01 -7.08000422e-01 2.81484686e-02 -8.87063965e-02
1.50045204e+00 -5.90259917e-02 7.34896779e-01 1.11172521e+00
-2.19177082e-01 -6.67987287e-01 1.21303105e+00 -4.49026562e-02
-4.69339877e-01 4.60310161e-01 -9.91273344e-01 5.60418844e-01
3.92911047e-01 -3.81507456e-01 -3.63328695e-01 -5.91394842e-01
7.02340662e-01 -9.11991715e-01 -8.30647767e-01 -6.40819073e-01
-1.18473746e-01 1.52329743e-01 -4.41969156e-01 9.02596533e-01
3.42583477e-01 -5.56567907e-01 3.67075711e-01 1.92842588e-01
5.48552096e-01 1.46688128e+00 6.30533323e-02 -4.02443349e-01
-1.95356786e-01 -1.07731116e+00 1.83814690e-01 -6.30863607e-01
-1.32659515e-02 -2.75812410e-02 -7.81414270e-01 -7.73480535e-01
1.17272878e+00 2.11454853e-01 -1.32152140e+00 -3.75794530e-01
-7.63941646e-01 -3.25122178e-01 5.97017646e-01 4.64800656e-01
9.25735772e-01 -1.66663408e+00 -1.44235921e+00 -1.14351019e-01
-4.09046382e-01 4.67850536e-01 8.84386599e-02 1.41732025e+00
-7.38033414e-01 6.02665722e-01 -5.83172679e-01 -5.50578415e-01
-2.38777089e+00 7.89772451e-01 -5.96895874e-01 -4.32026833e-01
-1.51689813e-01 5.24655700e-01 -5.88343322e-01 4.85302150e-01
3.42002898e-01 -2.43708193e-01 1.29783869e-01 3.25912446e-01
2.57633209e-01 6.21267855e-01 2.93332428e-01 -1.08638060e+00
-4.08010095e-01 1.16626108e+00 -1.00226879e+00 -6.54756963e-01
9.92326856e-01 -6.21469438e-01 -1.07888031e+00 -3.03654131e-02
1.41850817e+00 5.20882130e-01 -6.80045635e-02 6.75835371e-01
-1.04067457e+00 -3.17887902e-01 -8.97168994e-01 -1.56035304e+00
-8.17547202e-01 2.60269940e-01 2.25760594e-01 -5.09968579e-01
1.06156588e+00 -3.57266247e-01 1.09190881e+00 1.05796047e-01
2.26711822e+00 -2.05305791e+00 -7.10530519e-01 6.36373937e-01
6.24690592e-01 -1.79901969e+00 4.30461764e-01 -1.09063044e-01
3.42330754e-01 1.13993907e+00 5.09719551e-01 -1.75968140e-01
1.41829622e+00 7.91670009e-02 -1.09939754e-01 -1.99803069e-01
-3.52325499e-01 -4.55491215e-01 -5.18341780e-01 4.81713742e-01
4.39091146e-01 -9.55286175e-02 -1.39570546e+00 4.23403710e-01
-6.94590926e-01 6.48773670e-01 1.35266244e+00 1.46414483e+00
-5.56139469e-01 -1.24068308e+00 -1.75825393e+00 5.85450292e-01
-1.37751305e+00 5.76528847e-01 3.34425628e-01 8.06814134e-01
8.37280005e-02 2.48883915e+00 -6.16211355e-01 -8.45475852e-01
1.00786221e+00 -2.60871202e-01 5.14478981e-01 -6.13240957e-01
-1.37260401e+00 1.06439888e-01 4.36595798e-01 -3.37318256e-02
-3.66522232e-03 9.58261117e-02 -1.00532341e+00 -1.35613477e+00
2.03630161e-02 1.61724985e+00 1.06427538e+00 1.13857701e-01
1.68619752e-01 6.46314025e-01 6.85412347e-01 -4.62769896e-01
-8.30903351e-01 -6.36967421e-01 -1.12654567e+00 -3.04134101e-01
-3.14043015e-01 -1.83137447e-01 -1.17559063e+00 -4.08570290e-01] | [-3.315948724746704, 6.907647132873535] |
d80da344-65c0-4612-8961-d05de6a968ae | efficient-approach-of-using-cnn-based | 2209.13005 | null | https://arxiv.org/abs/2209.13005v1 | https://arxiv.org/pdf/2209.13005v1.pdf | Efficient approach of using CNN based pretrained model in Bangla handwritten digit recognition | Due to digitalization in everyday life, the need for automatically recognizing handwritten digits is increasing. Handwritten digit recognition is essential for numerous applications in various industries. Bengali ranks the fifth largest language in the world with 265 million speakers (Native and non-native combined) and 4 percent of the world population speaks Bengali. Due to the complexity of Bengali writing in terms of variety in shape, size, and writing style, researchers did not get better accuracy using Supervised machine learning algorithms to date. Moreover, fewer studies have been done on Bangla handwritten digit recognition (BHwDR). In this paper, we proposed a novel CNN-based pre-trained handwritten digit recognition model which includes Resnet-50, Inception-v3, and EfficientNetB0 on NumtaDB dataset of 17 thousand instances with 10 classes.. The Result outperformed the performance of other models to date with 97% accuracy in the 10-digit classes. Furthermore, we have evaluated the result or our model with other research studies while suggesting future study | ['Md Sakib Ullah Sourav', 'Md Mostak Shaikh', 'Zubaer Haque', 'Jannatul Nayeem', 'Rejwan Bin Sulaiman', 'Musarrat Saberin Nipun', 'Shabbir Ahmed Shuvo', 'Muntarin Islam'] | 2022-09-19 | null | null | null | null | ['handwritten-digit-recognition'] | ['computer-vision'] | [-3.79266500e-01 -4.71236885e-01 -9.66278985e-02 -4.06399965e-01
-8.56595412e-02 -5.71751893e-01 5.89528263e-01 -3.23973805e-01
-5.41100979e-01 7.16273487e-01 2.91503131e-01 -6.56835496e-01
1.34870157e-01 -8.61581802e-01 -1.97761104e-01 -4.49173003e-01
3.02105725e-01 4.93593514e-01 2.66837716e-01 -3.10712606e-01
6.73450291e-01 1.16233695e+00 -1.06065750e+00 2.78985858e-01
6.20090783e-01 9.34958816e-01 1.26352727e-01 8.40473950e-01
2.09708177e-02 1.51297152e+00 -9.08229291e-01 -5.61046660e-01
4.11461979e-01 -4.45900351e-01 -8.34088266e-01 2.29483753e-01
4.70172554e-01 -8.68589342e-01 -9.93928194e-01 7.67503977e-01
8.36174071e-01 -1.24542930e-04 6.78849041e-01 -4.91918564e-01
-1.41904044e+00 4.64951724e-01 -5.78098118e-01 4.24566895e-01
-1.88979849e-01 -1.14021629e-01 3.52599859e-01 -1.01169336e+00
4.89514947e-01 1.05980027e+00 6.62748516e-01 7.06178904e-01
-4.65675950e-01 -6.98445797e-01 -1.35280624e-01 2.43683115e-01
-1.22828031e+00 -3.28988999e-01 7.19169021e-01 -3.13847601e-01
1.31158805e+00 1.03986830e-01 3.60258818e-01 7.61717916e-01
2.64638513e-01 7.31495917e-01 1.26473582e+00 -6.74823165e-01
-2.21815687e-02 5.00089526e-02 4.22481865e-01 6.43434763e-01
2.85290599e-01 -2.15072602e-01 -1.55873775e-01 3.73195916e-01
1.23192120e+00 1.14458926e-01 1.22602828e-01 3.51885766e-01
-1.08101642e+00 6.38161004e-01 3.97009522e-01 5.46552360e-01
-4.47305828e-01 -1.61586314e-01 4.33964074e-01 4.44726586e-01
2.82009989e-01 3.36546928e-01 -4.82560396e-01 -3.70995671e-01
-9.52439547e-01 1.99259117e-01 8.68353248e-01 9.55907166e-01
1.23407282e-01 5.40185273e-01 -6.71267286e-02 1.41634548e+00
1.78712711e-01 4.28589374e-01 8.92578840e-01 -4.54137892e-01
6.73842371e-01 7.44502962e-01 -2.56507874e-01 -1.19537520e+00
-2.55034920e-02 -3.46372038e-01 -1.32920659e+00 2.21192583e-01
6.75974309e-01 -1.34594172e-01 -1.45227206e+00 8.76547515e-01
-5.68555892e-01 -5.57301879e-01 1.54616311e-01 9.74398255e-01
8.28051031e-01 7.85858035e-01 -4.62608486e-02 3.99323225e-01
1.23536849e+00 -9.03200090e-01 -5.32037497e-01 -2.27269486e-01
5.03980331e-02 -1.01544034e+00 8.80858004e-01 7.86075175e-01
-8.89289737e-01 -7.76035368e-01 -1.26078677e+00 -1.12466238e-01
-4.48685616e-01 7.43629754e-01 6.88202977e-01 9.35100615e-01
-8.31283689e-01 4.35583323e-01 -7.18401968e-01 -6.33313715e-01
8.27580810e-01 2.51852155e-01 -3.03032726e-01 -4.59537208e-01
-6.79791272e-01 1.20811248e+00 2.71376312e-01 1.40179902e-01
-7.09486842e-01 -2.05971152e-01 -2.62921423e-01 -3.40901226e-01
-2.67198026e-01 3.41377109e-01 1.10959065e+00 -1.03291166e+00
-1.68854630e+00 9.61218953e-01 8.05582628e-02 -4.26567048e-01
8.29344571e-01 2.03146390e-03 -9.52773213e-01 -1.60801783e-01
-2.92592645e-01 4.55412507e-01 3.72876912e-01 -6.93270206e-01
-5.78152597e-01 -6.53509140e-01 -1.72673374e-01 5.99535331e-02
-4.26921755e-01 2.71872252e-01 -3.75686944e-01 -1.07026279e+00
4.73716676e-01 -6.00867629e-01 3.01518627e-02 -1.58258200e-01
-3.08792740e-01 -2.51216501e-01 1.17359197e+00 -1.22415781e+00
1.10552394e+00 -2.20576930e+00 -4.05373573e-01 8.86422917e-02
2.15743091e-02 5.57484090e-01 -1.07599430e-01 3.60695034e-01
-1.66642647e-02 7.74836093e-02 2.95934435e-02 2.20783070e-01
-2.03302711e-01 1.80257753e-01 -5.80486178e-01 3.49496543e-01
2.82565683e-01 6.32315397e-01 -4.83240664e-01 -1.60661802e-01
2.87115842e-01 4.17382330e-01 -6.44164383e-02 8.98399353e-02
2.23251984e-01 -7.10879192e-02 -2.28986517e-01 1.34691441e+00
8.33557844e-01 -7.01128617e-02 6.57070801e-02 -8.29116255e-03
-3.05735201e-01 1.81674555e-01 -1.09348416e+00 1.05221474e+00
-1.59910962e-01 1.14776778e+00 -5.31871319e-01 -1.08060873e+00
1.33010936e+00 1.62725657e-01 -2.22192273e-01 -8.08041394e-01
1.76231816e-01 6.07233942e-01 4.31885272e-01 -4.23106492e-01
7.16245294e-01 2.37435281e-01 2.11230054e-01 2.30392456e-01
2.39956286e-02 9.65171158e-02 1.03390701e-01 -8.92153010e-02
8.74136329e-01 -1.78524226e-01 1.51398420e-01 -2.81514555e-01
3.94941330e-01 1.48309723e-01 2.86128342e-01 8.86504352e-01
-2.65498161e-01 8.93450499e-01 2.53265500e-01 -9.24036443e-01
-1.35479569e+00 -7.14908481e-01 -2.34658986e-01 8.28963339e-01
-3.32793623e-01 4.38239723e-01 -5.30039549e-01 -3.28727961e-01
-3.70011218e-02 4.13414955e-01 -3.27485621e-01 4.29021657e-01
-7.59483397e-01 -7.62619913e-01 1.17571437e+00 8.73171687e-01
1.69549096e+00 -1.22980917e+00 -3.73201340e-01 3.62222493e-01
3.57233733e-01 -1.01177263e+00 -2.64761329e-01 1.86123326e-01
-1.02813780e+00 -1.05725408e+00 -1.29414487e+00 -1.35764647e+00
6.15293145e-01 1.36906467e-02 6.98667288e-01 -1.68608248e-01
-6.52480006e-01 -1.01858944e-01 -4.08284456e-01 -6.21497095e-01
-3.47836405e-01 3.00845981e-01 -1.07263930e-01 -4.25853521e-01
8.10235620e-01 -1.28657907e-01 -4.12853777e-01 1.82700142e-01
-6.31807804e-01 -1.86460167e-01 8.73714149e-01 5.97016752e-01
-2.69739553e-02 2.67961413e-01 7.32567132e-01 -5.92477262e-01
7.50817299e-01 -1.36581495e-01 -5.17119646e-01 3.25421929e-01
-4.04315978e-01 -3.59737933e-01 6.54169440e-01 -5.06850183e-01
-9.63925183e-01 -1.27274156e-01 -1.47487313e-01 2.11560234e-01
-3.27200323e-01 5.70504129e-01 3.99450473e-02 -1.47943109e-01
6.73618197e-01 7.54652143e-01 -9.37337205e-02 -4.94494259e-01
-3.32205266e-01 1.46223402e+00 6.82415783e-01 -3.33693653e-01
2.60718495e-01 6.33663833e-02 -4.34499860e-01 -1.23169112e+00
-2.85693437e-01 -4.18695360e-02 -7.36747146e-01 -2.22191483e-01
6.67895794e-01 -7.95883119e-01 -6.26885831e-01 1.60731471e+00
-1.25224292e+00 -1.11917607e-01 2.28395492e-01 5.12010872e-01
1.40038908e-01 4.32308674e-01 -9.39053893e-01 -8.55324268e-01
-2.93200642e-01 -9.03254569e-01 2.56857693e-01 2.84570843e-01
-5.29365689e-02 -8.67167711e-01 -2.15369284e-01 2.81288117e-01
8.26841533e-01 -2.43496392e-02 1.01492226e+00 -7.24775136e-01
-3.77385259e-01 -5.73480844e-01 -6.54213428e-01 9.65104878e-01
6.09095395e-01 1.58258379e-01 -8.16550612e-01 -1.25841513e-01
-2.72932082e-01 -4.84953523e-01 8.93112540e-01 4.08333391e-01
1.31167305e+00 -3.55364412e-01 1.78885743e-01 1.46357775e-01
1.55376804e+00 8.55834901e-01 1.07502556e+00 6.24018312e-01
6.02416754e-01 3.05353571e-02 4.92127910e-02 5.84729731e-01
2.29004741e-01 5.85916005e-02 -1.11616999e-02 1.72104090e-01
-5.07461786e-01 2.07022242e-02 9.09470916e-02 9.73933280e-01
-7.29963422e-01 -3.22270811e-01 -1.36561382e+00 5.71414530e-01
-1.25264382e+00 -1.04386687e+00 -1.73203349e-02 1.88684654e+00
9.98527944e-01 1.43604428e-01 7.77523816e-02 4.07413006e-01
8.09948146e-01 -9.73710325e-04 -6.88835263e-01 -4.30964768e-01
-4.30682749e-01 4.89741415e-01 8.44726801e-01 8.73929858e-02
-1.07174003e+00 9.83158529e-01 6.76014328e+00 3.91105831e-01
-1.61967456e+00 -4.77523595e-01 9.80663896e-01 6.32974088e-01
5.00158548e-01 -6.07485831e-01 -1.02654564e+00 3.33399504e-01
6.05148077e-01 2.14838073e-01 5.82523644e-01 8.47153187e-01
-5.48554100e-02 7.83859491e-02 -8.63759160e-01 1.07687259e+00
3.33300501e-01 -1.44459569e+00 5.23033977e-01 -2.12454908e-02
8.48578393e-01 -3.77942473e-02 1.17570423e-01 3.19655091e-01
3.62681836e-01 -1.41690958e+00 6.58156872e-01 5.18161416e-01
8.94386232e-01 -8.84092748e-01 9.27666664e-01 2.97632545e-01
-6.43890381e-01 -2.11988509e-01 -6.94447637e-01 -3.84156734e-01
-6.38416946e-01 3.06324601e-01 -6.84521437e-01 1.18646780e-02
6.71739399e-01 8.17274213e-01 -5.91876388e-01 9.23362970e-01
1.15396202e-01 9.53073919e-01 -1.60349533e-02 -3.65052313e-01
4.13246542e-01 6.27468005e-02 -1.21663334e-02 1.33001626e+00
4.64818627e-01 3.62944216e-01 -1.80643022e-01 4.24703419e-01
-3.80524963e-01 -7.30324909e-02 -2.97895432e-01 -5.77644825e-01
6.22546732e-01 6.82144403e-01 -9.44255531e-01 -4.30318683e-01
-4.84107167e-01 1.15968633e+00 -5.41825965e-02 2.58743584e-01
-5.48698306e-01 -1.00110972e+00 2.80547827e-01 6.13573678e-02
2.55620509e-01 -5.02975941e-01 -5.24601698e-01 -1.07508981e+00
7.25084683e-03 -1.31604922e+00 4.77148443e-02 -5.08185863e-01
-1.44878924e+00 7.59914219e-01 -5.87803364e-01 -9.77702737e-01
1.66682392e-01 -1.29350221e+00 -3.38859677e-01 1.19723654e+00
-1.26401925e+00 -1.11515927e+00 -4.73013163e-01 5.75418651e-01
1.16612613e+00 -9.12150979e-01 9.28880274e-01 4.67222214e-01
-6.71079218e-01 7.03529060e-01 3.72108310e-01 1.08511949e+00
6.66257560e-01 -9.51405883e-01 5.86031437e-01 7.98820198e-01
-4.18134749e-01 4.98481631e-01 1.20655902e-01 -6.26693845e-01
-1.32929087e+00 -1.20749152e+00 9.67167258e-01 -1.80684894e-01
3.81195217e-01 3.86537984e-02 -7.07907081e-01 6.87104166e-01
2.37071842e-01 -8.66113529e-02 3.97236735e-01 -2.77653009e-01
-4.80856240e-01 -1.85635298e-01 -1.29847300e+00 3.36157084e-01
4.95064914e-01 -4.29358125e-01 -6.42726481e-01 4.90329489e-02
-6.27324060e-02 -4.53320026e-01 -7.23848224e-01 8.45913887e-02
8.73043656e-01 -7.15668142e-01 6.57199144e-01 -6.01727962e-01
9.85484540e-01 1.29223000e-02 -5.42468786e-01 -1.01255524e+00
-4.84012812e-01 1.37013495e-01 1.29177719e-01 1.28080237e+00
2.92406648e-01 -8.15341115e-01 7.77610302e-01 7.29264796e-01
-1.00496955e-01 -5.01367033e-01 -5.80221236e-01 -1.00268102e+00
3.25270981e-01 -1.48686677e-01 6.38157070e-01 1.30111575e+00
-4.64028507e-01 -1.69998527e-01 -4.10145491e-01 -6.89471737e-02
4.09097999e-01 -2.10193977e-01 4.92751926e-01 -1.14380336e+00
-1.41473353e-01 -5.64156115e-01 -8.14099133e-01 -1.01694047e+00
-4.06583697e-01 -5.86019099e-01 -3.15966010e-01 -1.58138001e+00
3.18208605e-01 -4.13171381e-01 -1.46062836e-01 8.41051102e-01
4.33762044e-01 5.60256541e-01 1.05850749e-01 5.71136177e-01
9.48598608e-02 -7.58384541e-02 1.26596415e+00 -5.11443436e-01
-1.28600717e-01 -8.34542215e-02 -6.08025551e-01 4.93333638e-01
9.90211546e-01 -1.84099916e-02 -5.66189140e-02 -8.28065693e-01
-2.41794959e-01 -1.20944977e-01 1.45166129e-01 -1.17554009e+00
3.01139027e-01 -1.46942943e-01 1.18517482e+00 -8.13582301e-01
-1.02670021e-01 -4.73546356e-01 -1.64361328e-01 6.16450667e-01
-3.96122605e-01 8.43578428e-02 3.01308721e-01 -9.73933563e-02
-3.15313905e-01 -1.22010641e-01 9.47806120e-01 -3.33091766e-01
-1.05678296e+00 5.19946925e-02 -7.90505350e-01 -2.93660641e-01
8.65913332e-01 -8.12550008e-01 -5.15785873e-01 -1.36780590e-01
-3.93561810e-01 -3.84462535e-01 8.50922316e-02 5.37972808e-01
7.94041812e-01 -1.27506137e+00 -1.04785252e+00 2.39017606e-01
-9.63005275e-02 -2.04172537e-01 -7.44221509e-02 8.07832778e-02
-1.47064674e+00 7.46174812e-01 -9.12932336e-01 -6.39043450e-02
-1.16598344e+00 -1.21152177e-01 4.06427503e-01 5.39601557e-02
-4.63341504e-01 8.20040405e-01 -7.21760035e-01 -3.75524491e-01
5.85271835e-01 -2.54926473e-01 -2.92700380e-01 -1.40084773e-01
6.83815300e-01 8.18899870e-01 4.34270978e-01 -4.02150095e-01
-3.91302586e-01 3.55029076e-01 -6.77059591e-01 9.97345448e-02
1.47006154e+00 5.52977800e-01 -3.04728955e-01 3.02846432e-01
1.02801788e+00 -2.60663062e-01 -8.83179724e-01 5.63347675e-02
-7.82516450e-02 -5.75873733e-01 6.87132450e-03 -1.29826474e+00
-1.27082777e+00 7.51533389e-01 8.40339065e-01 -5.32921292e-02
1.04749095e+00 -4.93118346e-01 5.28456092e-01 1.07786322e+00
3.65568548e-01 -1.29648435e+00 1.18402287e-01 1.05420840e+00
1.04201472e+00 -1.18395042e+00 -1.34842634e-01 2.40679607e-01
-5.19809186e-01 1.67045987e+00 8.01915348e-01 -6.78311884e-01
8.42202842e-01 2.30569631e-01 3.76230419e-01 2.04362735e-01
-1.71474978e-01 4.80541646e-01 8.63392577e-02 6.41409934e-01
1.03404498e+00 -3.78267616e-02 -3.07120115e-01 4.51270133e-01
-3.20268542e-01 2.49667317e-01 6.51542544e-01 1.12341332e+00
-4.93590891e-01 -1.00570357e+00 -4.22240973e-01 7.67711401e-01
-7.10408509e-01 -1.63325295e-01 -6.89684629e-01 1.00274837e+00
-1.41748920e-01 7.57453978e-01 2.15134725e-01 -3.09605151e-01
4.01392370e-01 5.20666037e-03 7.13234186e-01 -2.94179440e-01
-5.63090026e-01 -5.21930456e-01 -2.10824624e-01 2.30698526e-01
-2.55337179e-01 -4.26597983e-01 -9.93816137e-01 -8.38706434e-01
-2.20752999e-01 -3.77546728e-01 9.15182412e-01 7.51627207e-01
1.14358291e-01 4.36093181e-01 4.69602615e-01 -4.45332021e-01
-8.10688376e-01 -1.30865920e+00 -9.20172811e-01 -1.79754034e-01
1.08754821e-01 2.87932809e-02 5.43463789e-02 2.76396632e-01] | [11.84997844696045, 2.694119453430176] |
52303221-cb38-4f68-b0a2-7a000c27bf8a | mvt-multi-view-vision-transformer-for-3d | 2110.13083 | null | https://arxiv.org/abs/2110.13083v1 | https://arxiv.org/pdf/2110.13083v1.pdf | MVT: Multi-view Vision Transformer for 3D Object Recognition | Inspired by the great success achieved by CNN in image recognition, view-based methods applied CNNs to model the projected views for 3D object understanding and achieved excellent performance. Nevertheless, multi-view CNN models cannot model the communications between patches from different views, limiting its effectiveness in 3D object recognition. Inspired by the recent success gained by vision Transformer in image recognition, we propose a Multi-view Vision Transformer (MVT) for 3D object recognition. Since each patch feature in a Transformer block has a global reception field, it naturally achieves communications between patches from different views. Meanwhile, it takes much less inductive bias compared with its CNN counterparts. Considering both effectiveness and efficiency, we develop a global-local structure for our MVT. Our experiments on two public benchmarks, ModelNet40 and ModelNet10, demonstrate the competitive performance of our MVT. | ['Ping Li', 'Tan Yu', 'Shuo Chen'] | 2021-10-25 | null | null | null | null | ['3d-object-recognition'] | ['computer-vision'] | [-3.82897079e-01 -2.03788400e-01 -1.77121922e-01 -2.53122300e-01
-3.81474227e-01 -5.55748701e-01 7.13423729e-01 -5.99334598e-01
4.63288337e-01 -1.26428396e-01 3.82485092e-01 -2.40233272e-01
5.94117977e-02 -9.43382680e-01 -1.09217846e+00 -4.66839999e-01
3.17606181e-01 1.37602240e-01 1.60176501e-01 7.78211653e-02
-7.02560171e-02 7.22476661e-01 -1.25863874e+00 7.42667675e-01
1.79138243e-01 1.52942920e+00 2.88548142e-01 3.07184994e-01
1.24777341e-02 1.18618989e+00 -2.20881119e-01 -4.60073143e-01
3.95502090e-01 4.01991382e-02 -7.45276153e-01 3.38570237e-01
7.94698298e-01 -5.72930455e-01 -9.01153982e-01 8.35354745e-01
5.00830531e-01 -4.27793294e-01 5.47067225e-01 -1.16092527e+00
-1.21229708e+00 4.02477801e-01 -5.19616604e-01 3.50151099e-02
5.98898888e-01 -1.07005455e-01 1.20756471e+00 -1.54581523e+00
5.09320319e-01 1.37919414e+00 7.26286411e-01 5.60481310e-01
-9.59613264e-01 -5.08705556e-01 6.92678392e-01 3.07090402e-01
-1.15005791e+00 -2.70413250e-01 9.81908560e-01 -1.24108620e-01
1.40009248e+00 1.38818428e-01 1.20929146e+00 1.31546068e+00
2.30814472e-01 1.31998599e+00 1.27655745e+00 7.97644183e-02
-2.94824213e-01 7.72538856e-02 -1.93237618e-01 7.00184464e-01
-1.15838483e-01 2.58205444e-01 -8.38333488e-01 5.71075715e-02
1.07268178e+00 5.39230108e-01 -4.31245834e-01 -7.14054883e-01
-1.27524805e+00 4.55519438e-01 9.94808316e-01 4.11777385e-02
-3.51377666e-01 1.03718720e-01 1.49903893e-01 3.60361367e-01
5.37954390e-01 -8.40096474e-02 -2.57316411e-01 3.67643416e-01
-3.06439310e-01 -2.35978037e-01 7.08630145e-01 1.33727193e+00
6.49131298e-01 3.12164277e-02 1.23015054e-01 8.34008217e-01
5.27574122e-01 8.24288785e-01 -8.03460255e-02 -7.19220519e-01
5.52529514e-01 1.09545743e+00 -4.27700669e-01 -1.38480318e+00
-1.47653893e-01 -6.82964921e-01 -1.31481338e+00 -1.25091925e-01
-1.64503581e-03 6.79705739e-01 -9.32542920e-01 1.28701818e+00
1.81970090e-01 -2.05190945e-02 2.03347549e-01 9.78609979e-01
1.36238885e+00 4.24889237e-01 -5.75684190e-01 2.11781561e-01
1.26617968e+00 -1.15268481e+00 -5.73103465e-02 -3.22778374e-01
2.77262717e-01 -7.43019164e-01 4.73736614e-01 4.22331959e-01
-1.16204953e+00 -6.92192018e-01 -8.56792629e-01 -1.59348622e-01
-2.30575413e-01 -2.55772509e-02 7.10342884e-01 1.28207222e-01
-1.10583639e+00 1.87744200e-02 -7.88625300e-01 -2.64180362e-01
9.67254102e-01 3.40511650e-01 -7.63961852e-01 -5.99105656e-01
-6.47014081e-01 8.05474579e-01 -3.17202300e-01 4.12992150e-01
-1.46091843e+00 -7.69843757e-01 -9.18705344e-01 2.62695327e-02
2.65879687e-02 -1.04606020e+00 9.23516750e-01 -6.91444337e-01
-1.15198576e+00 1.19992995e+00 -2.58057773e-01 -2.44012997e-01
3.51796836e-01 -9.55414996e-02 -2.77008086e-01 3.29244047e-01
2.47198977e-02 6.88921034e-01 8.34316790e-01 -1.54168630e+00
-5.55472493e-01 -7.32156992e-01 6.26971006e-01 2.79040992e-01
-3.62733841e-01 -3.39792162e-01 -7.16809511e-01 -2.58651614e-01
6.86908007e-01 -7.35609174e-01 5.22569288e-03 2.64858931e-01
-5.00658929e-01 -2.72358179e-01 9.54958856e-01 -1.44864529e-01
2.90429473e-01 -2.13191247e+00 2.89169252e-01 6.58698156e-02
7.85027206e-01 -4.93659526e-02 -1.63063183e-01 2.59470373e-01
-1.19343661e-01 -2.22358614e-01 4.02695507e-01 -2.59988397e-01
-1.85935363e-01 3.79863948e-01 -6.72318757e-01 4.57275838e-01
1.21619403e-01 1.45981455e+00 -7.74316311e-01 -1.85835645e-01
3.26631725e-01 6.37850881e-01 -6.82680786e-01 4.66893762e-01
1.83780611e-01 1.87054679e-01 -6.80441022e-01 1.08505702e+00
8.63207936e-01 -9.26702261e-01 2.66458653e-02 -6.82296693e-01
1.72677502e-01 3.14754009e-01 -5.15402257e-01 1.80554354e+00
-7.58170247e-01 6.30457163e-01 -1.28610417e-01 -1.26657295e+00
1.18168664e+00 2.53925800e-01 5.32996118e-01 -1.00352728e+00
-9.73503888e-02 1.07671484e-01 -5.09838462e-01 -2.99341232e-01
9.14932638e-02 4.15717922e-02 3.03314000e-01 2.77692050e-01
3.11632723e-01 -2.30841756e-01 -6.70727432e-01 2.27594137e-01
8.53923142e-01 1.33340821e-01 7.45839700e-02 9.29265656e-03
5.15242815e-01 -2.30312496e-01 3.43958378e-01 7.01770782e-01
-6.41647577e-02 9.33540821e-01 1.85380176e-01 -9.51772690e-01
-6.88521028e-01 -1.24625814e+00 2.28124969e-02 4.86364692e-01
6.96010709e-01 -3.38197082e-01 -1.15394704e-01 -1.05229473e+00
2.34447792e-02 -8.08292404e-02 -7.25303054e-01 -2.37957776e-01
-4.28068936e-01 -2.54521132e-01 4.64441329e-01 8.06072712e-01
1.00370181e+00 -5.37738264e-01 -4.56337959e-01 -1.05006723e-02
-1.86117008e-01 -1.57795811e+00 -4.80909228e-01 9.90426093e-02
-1.10131025e+00 -1.25895369e+00 -8.45753133e-01 -9.84516978e-01
9.33528662e-01 1.19808364e+00 1.41707480e+00 7.11502284e-02
-3.11898359e-04 1.04334176e+00 -3.67882907e-01 -3.26996416e-01
-2.35458352e-02 -7.92697296e-02 1.07892193e-02 1.13654360e-01
4.17839974e-01 -8.80710423e-01 -9.13169444e-01 8.05769086e-01
-6.13131940e-01 2.27110699e-01 6.41860962e-01 1.02171707e+00
6.94830060e-01 -5.98579466e-01 1.27924427e-01 -4.24678594e-01
-1.48172647e-01 -2.81395167e-01 -3.67468476e-01 4.99860078e-01
-3.68670583e-01 -4.28150773e-01 4.83554244e-01 -3.70091647e-01
-7.23832905e-01 7.88044278e-03 -5.63836694e-02 -1.13936234e+00
-1.34876251e-01 1.67004064e-01 -3.57642680e-01 -5.61151922e-01
2.72022724e-01 7.66935408e-01 9.88921225e-02 -4.40046996e-01
2.53441304e-01 3.19842100e-01 1.72537059e-01 -1.87414065e-01
7.24517226e-01 9.19690788e-01 -1.19650140e-01 -6.27088368e-01
-1.27820313e+00 -2.90394127e-01 -4.35608476e-01 -4.51804101e-01
7.81055093e-01 -1.44230187e+00 -1.10274231e+00 7.22599864e-01
-1.46721387e+00 6.42915070e-02 -1.01452693e-01 4.51648861e-01
-5.29000282e-01 1.76095515e-01 -3.94967407e-01 -3.07806998e-01
-2.91986227e-01 -1.13267529e+00 1.22409606e+00 1.15541503e-01
4.14287418e-01 -1.24040401e+00 -4.10874128e-01 5.29351652e-01
3.74711215e-01 -1.93507865e-01 8.01249683e-01 -2.50871211e-01
-1.16316271e+00 -8.88919756e-02 -5.77600837e-01 5.18218637e-01
1.84592754e-01 -5.32047510e-01 -1.32583916e+00 -3.31945330e-01
4.12548780e-01 -4.17669475e-01 1.08369637e+00 3.20923835e-01
1.23164928e+00 -7.12471306e-02 -5.86073220e-01 9.99828339e-01
1.39386833e+00 2.41459031e-02 6.37186110e-01 3.58430706e-02
1.14787948e+00 2.75638044e-01 6.14859350e-02 5.20804971e-02
7.22487092e-01 6.67335212e-01 9.00080323e-01 -3.63734305e-01
-3.06477010e-01 -7.15814233e-01 4.31481987e-01 1.33708000e+00
-2.25611910e-01 -3.18294495e-01 -6.34232998e-01 5.68016291e-01
-1.62484157e+00 -8.57542694e-01 8.49074349e-02 1.59515059e+00
-3.94380726e-02 1.62495956e-01 -2.39433274e-01 -2.16461331e-01
2.90887326e-01 4.43426669e-01 -7.05736816e-01 -1.04752444e-01
-4.90438193e-01 -1.16881177e-01 3.07582080e-01 1.58908237e-02
-8.51784050e-01 6.33822560e-01 6.66301966e+00 6.70374513e-01
-1.25105858e+00 -2.38401797e-02 7.35127270e-01 3.93268839e-02
-5.71145713e-01 -6.28518909e-02 -9.33554232e-01 -2.33066142e-01
-8.26577768e-02 1.72151685e-01 2.75852919e-01 8.07223976e-01
-4.27418441e-01 3.60392034e-01 -1.41049147e+00 1.40235448e+00
4.68122721e-01 -1.69645381e+00 6.87914848e-01 2.49468744e-01
7.65541732e-01 4.62373406e-01 2.93903410e-01 3.80414911e-02
1.14023961e-01 -1.10841608e+00 7.64352500e-01 4.06012595e-01
5.36552489e-01 -3.12675864e-01 5.21747589e-01 3.33936125e-01
-1.37103987e+00 -8.22462142e-02 -7.32954204e-01 -1.19708814e-01
1.12741143e-02 6.49246395e-01 -5.26204586e-01 9.71093297e-01
9.50560808e-01 1.76304495e+00 -3.68170947e-01 6.81004167e-01
-7.61524141e-02 2.20577255e-01 -1.94630995e-01 1.39965624e-01
3.54440540e-01 -1.06760778e-01 6.65428400e-01 6.09732330e-01
2.81736076e-01 3.55355479e-02 -4.69378009e-02 1.32836151e+00
-1.63458869e-01 -3.27505797e-01 -1.35084403e+00 3.19360107e-01
9.63893384e-02 1.08704722e+00 -4.81887907e-01 -1.58316106e-01
-1.01008987e+00 8.84764671e-01 4.79253858e-01 4.69175696e-01
-6.53816164e-01 3.76612574e-01 6.97859108e-01 -1.40919490e-02
8.32182348e-01 -1.93199992e-01 -1.57962590e-01 -1.41538763e+00
3.04336846e-01 -1.00594997e+00 1.57766119e-01 -8.52768004e-01
-1.66309655e+00 8.54629934e-01 -2.61973858e-01 -1.86161494e+00
4.06646192e-01 -8.86950254e-01 -2.75795430e-01 6.06902122e-01
-1.55675566e+00 -1.66468596e+00 -4.16286588e-01 8.60546589e-01
5.73602736e-01 -2.93111324e-01 8.35394442e-01 3.53244096e-01
-5.14145158e-02 5.70344388e-01 -3.69330883e-01 2.95897812e-01
2.58775890e-01 -8.38674545e-01 6.16302907e-01 3.35782528e-01
6.04498804e-01 4.86174703e-01 -1.81146964e-01 -1.77413702e-01
-2.11252737e+00 -1.05104601e+00 5.67621112e-01 -7.24659145e-01
2.59605020e-01 -5.09843826e-01 -7.32535124e-01 7.66872048e-01
2.29406565e-01 2.44944885e-01 3.66752356e-01 1.27862707e-01
-1.06832266e+00 -3.91402096e-01 -5.49560964e-01 3.03741992e-01
1.62379980e+00 -1.08725655e+00 -3.99567455e-01 1.71936661e-01
5.76109231e-01 -6.74369812e-01 -1.04869771e+00 6.19284809e-01
7.88354874e-01 -1.34105790e+00 1.49252582e+00 -5.84856689e-01
7.15791762e-01 -2.87089478e-02 -5.89624703e-01 -1.26899827e+00
-3.78851265e-01 -3.97911482e-02 -1.95678622e-01 9.01822686e-01
3.52613062e-01 -7.22443581e-01 5.78458846e-01 1.74429998e-01
-1.85013562e-01 -1.10398853e+00 -9.99317527e-01 -7.51918018e-01
-2.35553868e-02 -5.54573834e-01 6.34233534e-01 8.07040691e-01
-4.71719772e-01 5.11621296e-01 -4.59050000e-01 4.03929263e-01
7.95197785e-01 8.06132615e-01 8.33882749e-01 -1.13772345e+00
-3.60342085e-01 -3.22698027e-01 -8.48473370e-01 -2.01510453e+00
1.35508031e-01 -1.16927171e+00 -2.79369324e-01 -1.55286050e+00
7.32584953e-01 -3.10280323e-01 -2.83263505e-01 4.59495425e-01
2.86419392e-01 6.84282005e-01 3.46047640e-01 3.38273644e-01
-6.96894944e-01 8.10176253e-01 1.99008548e+00 -6.67248607e-01
3.08511674e-01 9.07660127e-02 -7.16211140e-01 7.73268342e-01
1.41616732e-01 -2.19783142e-01 -2.84667343e-01 -1.07158196e+00
4.62783694e-01 4.14409861e-02 9.78467107e-01 -6.75574481e-01
4.29366976e-01 2.50000983e-01 7.13485897e-01 -9.69336689e-01
5.60766518e-01 -1.05258226e+00 1.80523023e-01 3.15964460e-01
-5.56599163e-02 -4.59149107e-02 5.59824780e-02 7.94957340e-01
-6.08729482e-01 5.65349400e-01 4.73153472e-01 -4.44780827e-01
-6.88684285e-01 8.02067161e-01 9.65938158e-03 -9.94245261e-02
7.21003175e-01 -6.61027312e-01 -3.80255818e-01 -4.60184962e-01
-7.16306865e-01 3.07645708e-01 4.29862320e-01 7.79577076e-01
1.24685454e+00 -1.69416201e+00 -3.61264348e-01 5.61386466e-01
3.49833429e-01 2.57822871e-01 3.58719528e-01 1.05910826e+00
-1.35557994e-01 4.12104309e-01 -1.92555025e-01 -1.21007800e+00
-1.11844516e+00 5.14546692e-01 7.41089225e-01 4.04316299e-02
-9.80658412e-01 9.85978603e-01 1.14571524e+00 -5.71144819e-01
2.65569478e-01 -3.74024332e-01 -4.67818528e-02 -3.40235949e-01
3.15585762e-01 3.07592377e-02 1.10184578e-02 -5.61180055e-01
-5.61888933e-01 1.10555077e+00 -2.12278470e-01 3.11287045e-01
1.55766749e+00 -2.00858712e-01 -2.30929464e-01 5.17529607e-01
1.53265488e+00 -5.17222703e-01 -1.24338567e+00 -5.30269980e-01
-6.29871070e-01 -5.50762355e-01 1.16977140e-01 -3.65970582e-01
-1.52046931e+00 1.13863921e+00 3.78011435e-01 2.13944018e-01
1.16620231e+00 3.82746339e-01 7.97556818e-01 5.60446322e-01
6.77533567e-01 -4.60406572e-01 3.26618969e-01 8.72028172e-01
1.05679572e+00 -1.31758201e+00 -2.17825294e-01 -5.47959030e-01
-3.49979103e-01 1.05276096e+00 7.85089195e-01 -1.20951369e-01
1.03760159e+00 -2.48628911e-02 1.45830810e-01 -7.44279623e-01
-9.95720148e-01 -3.42262313e-02 4.45964187e-01 7.64176190e-01
2.65727174e-02 -1.18017934e-01 6.49389923e-01 3.49273235e-01
1.80016398e-01 -1.18497021e-01 3.54753733e-02 5.97823441e-01
1.50275901e-01 -7.60673940e-01 4.93810475e-02 3.52495790e-01
-1.96559623e-01 7.17093423e-02 -7.31879234e-01 6.18447542e-01
-1.97646961e-01 7.94844091e-01 1.66616708e-01 -7.33513534e-01
6.93130434e-01 -5.11289060e-01 8.15394580e-01 -4.16103214e-01
-5.13586164e-01 5.99610917e-02 -2.91148514e-01 -9.07584488e-01
-9.23248887e-01 -2.34781027e-01 -5.41538417e-01 -2.42987618e-01
-3.47021610e-01 -2.14741543e-01 3.18556130e-01 9.32846665e-01
6.22323871e-01 3.31040591e-01 1.21719587e+00 -8.15974414e-01
-4.33388472e-01 -5.59041917e-01 -5.73218822e-01 1.96375132e-01
3.67798537e-01 -7.11040139e-01 -2.62287468e-01 -2.77288496e-01] | [8.254111289978027, -3.6521849632263184] |
526bc039-7bb7-4bdd-ae2e-195db2d8bc4c | learning-to-cluster-faces-via-confidence-and | 2004.00445 | null | https://arxiv.org/abs/2004.00445v2 | https://arxiv.org/pdf/2004.00445v2.pdf | Learning to Cluster Faces via Confidence and Connectivity Estimation | Face clustering is an essential tool for exploiting the unlabeled face data, and has a wide range of applications including face annotation and retrieval. Recent works show that supervised clustering can result in noticeable performance gain. However, they usually involve heuristic steps and require numerous overlapped subgraphs, severely restricting their accuracy and efficiency. In this paper, we propose a fully learnable clustering framework without requiring a large number of overlapped subgraphs. Instead, we transform the clustering problem into two sub-problems. Specifically, two graph convolutional networks, named GCN-V and GCN-E, are designed to estimate the confidence of vertices and the connectivity of edges, respectively. With the vertex confidence and edge connectivity, we can naturally organize more relevant vertices on the affinity graph and group them into clusters. Experiments on two large-scale benchmarks show that our method significantly improves clustering accuracy and thus performance of the recognition models trained on top, yet it is an order of magnitude more efficient than existing supervised methods. | ['Rui Zhao', 'Dapeng Chen', 'Lei Yang', 'Chen Change Loy', 'Xiaohang Zhan', 'Dahua Lin'] | 2020-04-01 | learning-to-cluster-faces-via-confidence-and-1 | http://openaccess.thecvf.com/content_CVPR_2020/html/Yang_Learning_to_Cluster_Faces_via_Confidence_and_Connectivity_Estimation_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Yang_Learning_to_Cluster_Faces_via_Confidence_and_Connectivity_Estimation_CVPR_2020_paper.pdf | cvpr-2020-6 | ['face-clustering', 'connectivity-estimation'] | ['computer-vision', 'graphs'] | [-2.07793966e-01 3.47889364e-02 -4.15961266e-01 -6.51913106e-01
-4.63678271e-01 -4.72954571e-01 2.84649253e-01 -2.07852628e-02
-1.89187694e-02 2.32429013e-01 -6.72516376e-02 5.36536449e-04
-1.81973651e-01 -7.66584754e-01 -5.65940440e-01 -8.02641511e-01
-2.30989262e-01 5.60687065e-01 8.13238695e-02 4.33410645e-01
-6.25678375e-02 5.10663748e-01 -1.61669266e+00 1.20208547e-01
7.96524346e-01 1.08868194e+00 -1.59705088e-01 6.66714683e-02
-2.91134685e-01 5.87979138e-01 -1.88580185e-01 -4.25448149e-01
8.69085044e-02 -3.77911061e-01 -7.96439588e-01 6.18577361e-01
4.60149586e-01 -2.22999379e-02 -4.43191379e-01 1.16578686e+00
2.44708166e-01 3.24894413e-02 6.34306967e-01 -1.47312009e+00
-5.07697403e-01 7.34556377e-01 -1.04618406e+00 -2.99878269e-01
1.08501628e-01 -2.19838664e-01 8.82140636e-01 -8.63310337e-01
3.95089835e-01 1.28625488e+00 6.47691965e-01 5.51294446e-01
-1.02769804e+00 -9.38402176e-01 4.27424401e-01 2.22346455e-01
-1.87180662e+00 -4.90620017e-01 8.39299083e-01 -2.97818720e-01
5.13873577e-01 9.58054811e-02 4.81601387e-01 5.01234472e-01
-6.27876103e-01 5.84337234e-01 7.51083434e-01 -1.07623875e-01
2.35175893e-01 -1.70918927e-01 1.68441653e-01 1.03279293e+00
4.63397741e-01 -4.01502937e-01 -3.09857428e-01 -4.54453602e-02
7.00281262e-01 3.80430639e-01 -2.82642514e-01 -5.78907847e-01
-6.75334990e-01 8.77019167e-01 5.41912854e-01 2.83924758e-01
-3.39470804e-02 2.45876044e-01 3.14673990e-01 -6.78125918e-02
4.65720773e-01 -1.59647167e-01 -2.37545550e-01 3.76956195e-01
-9.84985530e-01 -5.21250069e-01 7.43688822e-01 1.10049152e+00
1.15077937e+00 -3.10623437e-01 -7.31619522e-02 9.15475905e-01
4.41126525e-01 3.30850452e-01 7.59682283e-02 -8.02415192e-01
2.13713050e-01 1.09508920e+00 -3.98503661e-01 -1.54156792e+00
-3.62800807e-01 -1.60433471e-01 -1.18729222e+00 -3.08776498e-01
2.49891177e-01 -1.58588484e-01 -1.17280269e+00 1.72727954e+00
5.82619965e-01 7.96424091e-01 -3.90913576e-01 8.57583404e-01
8.94351780e-01 4.12681580e-01 3.23105603e-02 -4.34434205e-01
1.19981301e+00 -1.12511039e+00 -6.52607143e-01 -1.67378381e-01
6.26149297e-01 -6.32258296e-01 7.27608621e-01 1.96549714e-01
-8.00522268e-01 -5.00621080e-01 -7.85395801e-01 1.71634853e-01
-1.75278679e-01 3.02395016e-01 1.07975340e+00 7.36925721e-01
-1.23227012e+00 5.68705559e-01 -9.44166064e-01 -3.74122709e-01
8.76294613e-01 7.65165210e-01 -5.49619377e-01 -4.44473147e-01
-5.45738041e-01 2.88598761e-02 3.33477885e-01 2.80239820e-01
-7.32571483e-01 -3.19597691e-01 -9.71890628e-01 3.23797554e-01
6.50367320e-01 -7.19581991e-02 7.41494417e-01 -8.04339468e-01
-1.22556043e+00 8.39202940e-01 -3.23282093e-01 1.05881840e-01
1.07816398e-01 6.20806925e-02 -4.29094344e-01 3.31419557e-01
7.35547720e-03 7.22303152e-01 8.41174543e-01 -1.22575760e+00
-5.87151587e-01 -6.35974050e-01 -2.48180583e-01 -5.76342903e-02
-8.03938746e-01 -6.89649656e-02 -1.48077798e+00 -4.30739880e-01
4.82667714e-01 -9.62952971e-01 -2.69293875e-01 -9.19874758e-02
-6.14585042e-01 -6.03630602e-01 9.26373303e-01 -2.62071282e-01
1.27188504e+00 -2.09430718e+00 5.03325015e-02 6.54279590e-01
6.35545313e-01 1.66464776e-01 -2.34941348e-01 8.14283192e-02
-2.58578770e-02 3.47511619e-01 -2.66251892e-01 -3.00528526e-01
-1.57667920e-01 1.77442178e-01 6.62088692e-02 6.24847710e-01
2.42488846e-01 8.09460223e-01 -7.76513577e-01 -7.31661439e-01
4.21603397e-02 6.16297722e-01 -7.50333667e-01 1.85430571e-01
-1.56186298e-01 3.92821640e-01 -3.89211595e-01 6.49350762e-01
1.00435996e+00 -6.93808496e-01 8.83928776e-01 -2.92356908e-01
4.43735659e-01 -1.79267466e-01 -1.32931232e+00 1.48082948e+00
-1.62587553e-01 4.25419450e-01 2.47984186e-01 -1.21076238e+00
7.18458533e-01 3.45508456e-02 7.06482947e-01 -2.80830890e-01
2.63012648e-01 -5.12449071e-02 -1.70925260e-01 -2.78087020e-01
-6.47688732e-02 4.02987927e-01 1.91221863e-01 5.65245450e-01
1.29901260e-01 5.46993613e-01 2.57898599e-01 4.13484484e-01
9.85615075e-01 -2.62609899e-01 -1.18059956e-01 -2.72532105e-01
6.34489536e-01 -3.95133346e-01 8.14278960e-01 2.42621079e-01
-5.54591790e-02 5.22205710e-01 7.27930903e-01 -3.95714819e-01
-5.43498635e-01 -7.26348400e-01 7.87617192e-02 8.98491919e-01
3.65752548e-01 -8.06500614e-01 -1.16421795e+00 -9.80859578e-01
3.85547690e-02 -2.25635335e-01 -7.16461360e-01 -2.34263122e-01
-5.17295539e-01 -8.02130401e-01 3.15903336e-01 6.18010819e-01
4.49646473e-01 -8.25552940e-01 1.47098392e-01 -8.26933980e-02
-1.42578423e-01 -1.27445257e+00 -8.08684230e-01 -1.55056566e-01
-7.75795519e-01 -1.58449149e+00 -3.27373832e-01 -1.22337329e+00
1.26011717e+00 6.80550039e-01 1.08161330e+00 8.05405915e-01
-4.11076486e-01 3.44653964e-01 -3.14839482e-01 1.20556034e-01
3.53724450e-01 1.10845879e-01 1.12756021e-01 3.97030503e-01
6.81392312e-01 -6.12694323e-01 -5.41551769e-01 4.30487514e-01
-7.78128028e-01 -2.26262823e-01 6.01999342e-01 7.24734843e-01
7.62254953e-01 3.26388866e-01 4.32404310e-01 -1.27972543e+00
1.05543219e-01 -4.86320257e-01 -6.23233080e-01 3.59362930e-01
-8.11652720e-01 -5.67738293e-03 4.79077905e-01 -4.22453821e-01
-6.92551732e-01 5.64264834e-01 1.54697523e-01 -8.26941729e-01
-6.35897964e-02 5.58102190e-01 -7.13794649e-01 -2.80212730e-01
1.34506896e-01 4.77962494e-02 -1.51930779e-01 -4.92063731e-01
4.20836538e-01 5.70562661e-01 4.30955172e-01 -5.25834918e-01
8.57045174e-01 6.27872348e-01 -4.17021625e-02 -7.11360335e-01
-7.81988561e-01 -6.70001388e-01 -7.07554162e-01 -2.59240955e-01
8.17467928e-01 -9.58800018e-01 -1.02394903e+00 2.77323335e-01
-8.95673931e-01 -1.99214250e-01 2.98400074e-01 2.27691814e-01
-8.97077098e-02 6.91295683e-01 -6.71134531e-01 -5.60837209e-01
-1.03281476e-01 -1.13798404e+00 1.01952493e+00 4.00165081e-01
2.97361910e-01 -8.22185576e-01 -2.63267279e-01 4.03336585e-01
-1.00023627e-01 1.51678815e-01 8.97047162e-01 -6.17019355e-01
-6.66281939e-01 -1.64796650e-01 -5.01985610e-01 7.20252618e-02
3.48532766e-01 1.41324148e-01 -9.86661196e-01 -5.52788258e-01
-5.15604019e-01 -4.33630079e-01 1.06998825e+00 1.70966744e-01
1.79607034e+00 -3.38359177e-01 -8.36815894e-01 8.01461518e-01
1.31677938e+00 -4.33486663e-02 5.80468893e-01 -4.91811365e-01
1.14751446e+00 6.90366447e-01 3.52110237e-01 3.54019314e-01
2.84529179e-01 5.46889246e-01 3.52995574e-01 -1.74339518e-01
-1.22371696e-01 -3.14134061e-01 1.61551788e-01 9.70381916e-01
-4.04344406e-03 -1.60414532e-01 -8.69697869e-01 4.29735959e-01
-2.02736616e+00 -6.70090795e-01 -1.32213414e-01 2.14501691e+00
6.51625812e-01 -1.96995541e-01 1.49781555e-01 2.10487377e-02
1.34625030e+00 -3.29861641e-02 -6.09182417e-01 2.46113449e-01
1.71401322e-01 2.00403050e-01 1.56839937e-01 3.24786484e-01
-1.18438148e+00 1.20389926e+00 6.21424341e+00 9.89513755e-01
-9.25795436e-01 -2.00892650e-02 9.71058428e-01 4.91376184e-02
-2.72679310e-02 -7.84516409e-02 -7.31576800e-01 5.43730438e-01
6.28674626e-01 1.13002695e-01 5.55671751e-01 1.02509224e+00
-2.26349011e-01 2.43913040e-01 -1.22756004e+00 1.24123228e+00
2.44337901e-01 -1.12120426e+00 2.40451887e-01 2.32452080e-01
9.09964800e-01 -3.52504939e-01 -1.50185393e-03 2.00450450e-01
3.37319851e-01 -1.34670103e+00 4.51611690e-02 8.61891881e-02
1.01275134e+00 -1.08699322e+00 6.49222553e-01 7.99678266e-02
-1.70301628e+00 6.57365844e-02 -6.16569698e-01 2.77855545e-01
-3.40945601e-01 7.24420428e-01 -4.06898648e-01 3.93019319e-01
8.52626204e-01 7.28917062e-01 -6.77555978e-01 8.91431272e-01
-2.48049974e-01 6.63046300e-01 -3.18678200e-01 1.45481125e-01
5.79569712e-02 -4.48777109e-01 -1.63091362e-01 1.05494332e+00
1.76307127e-01 3.08351576e-01 5.95734954e-01 5.57438135e-01
-7.48293996e-01 2.60028750e-01 -3.48877966e-01 -2.79691696e-01
7.50492990e-01 1.76705933e+00 -1.24494731e+00 -3.46401006e-01
-5.28162360e-01 1.07222736e+00 8.82811964e-01 2.08331645e-01
-9.39794004e-01 -3.98835987e-01 6.55873477e-01 5.12473062e-02
5.29910386e-01 -8.17637369e-02 1.78144470e-01 -1.10833406e+00
1.85631532e-02 -7.93003678e-01 5.98141491e-01 -2.85216182e-01
-1.40423238e+00 6.15976989e-01 -4.15575743e-01 -9.10249591e-01
1.77771524e-01 -5.93948364e-01 -6.93198144e-01 2.61708021e-01
-1.27962518e+00 -1.20797849e+00 -7.64838934e-01 8.80352676e-01
1.40807867e-01 -6.88654929e-02 5.97069860e-01 6.03481889e-01
-1.00292075e+00 8.75503838e-01 -5.87130804e-03 7.38881826e-01
7.58079052e-01 -1.01707828e+00 2.76257284e-02 7.42896974e-01
4.88402545e-01 6.99790001e-01 -1.25034064e-01 -7.14563727e-01
-1.41318250e+00 -1.51705933e+00 5.04684985e-01 -9.25190300e-02
2.63230890e-01 -5.22600889e-01 -8.89355838e-01 5.78913689e-01
-6.60827607e-02 4.65004683e-01 8.41220737e-01 4.03314769e-01
-6.86234295e-01 -2.42323846e-01 -1.04524779e+00 5.10441661e-01
1.43178618e+00 -5.14522851e-01 1.04528368e-01 5.93193471e-01
6.58346772e-01 -6.39695600e-02 -8.26120436e-01 4.43193257e-01
3.10731739e-01 -8.98389339e-01 7.48963058e-01 -6.10162973e-01
2.78183430e-01 -5.00125587e-01 6.50854781e-02 -8.65881205e-01
-5.50772369e-01 -5.42254746e-01 -3.03514600e-01 1.62405014e+00
3.30919743e-01 -4.15560842e-01 1.22400081e+00 5.50911963e-01
1.62728578e-01 -7.57837772e-01 -6.74864292e-01 -6.42353356e-01
-3.01380515e-01 -2.04425976e-01 8.02510083e-01 1.31641316e+00
-2.80112736e-02 4.90443349e-01 -1.69719964e-01 4.42990482e-01
8.50257814e-01 3.90533775e-01 7.83932567e-01 -1.53597987e+00
3.12398020e-02 -3.85793865e-01 -5.94769478e-01 -1.06536651e+00
4.50762391e-01 -1.07312584e+00 8.11319873e-02 -1.39318848e+00
6.16117477e-01 -5.74245989e-01 -2.96004027e-01 8.09885859e-01
-3.94229442e-01 6.37103677e-01 -1.53057396e-01 3.33679408e-01
-1.10974431e+00 4.17010784e-01 8.48978102e-01 -3.01760525e-01
-1.72129974e-01 -2.58865803e-01 -8.07531297e-01 9.16667283e-01
6.87049866e-01 -3.82784426e-01 -4.93519664e-01 -4.72618759e-01
-1.11566149e-01 -2.72952557e-01 8.58695433e-02 -1.03430676e+00
5.93227327e-01 -6.89940825e-02 6.46910429e-01 -4.62706983e-01
1.48515120e-01 -8.90421629e-01 1.24810643e-01 1.84422776e-01
-9.21454504e-02 -1.07810602e-01 1.13549118e-03 9.17649567e-01
-3.16680074e-01 1.97102070e-01 8.95504057e-01 8.64797831e-02
-5.71511507e-01 1.00978994e+00 -2.92883515e-02 -9.60131660e-02
1.23485100e+00 -1.99736618e-02 -6.91423118e-02 -3.13941360e-01
-6.99735522e-01 5.46880960e-01 5.03264427e-01 3.79949659e-01
6.64720833e-01 -1.55831552e+00 -3.62532347e-01 2.41364464e-01
1.13597535e-01 1.36929110e-01 3.10633659e-01 6.67844296e-01
-3.20218116e-01 2.61696190e-01 1.94973782e-01 -7.36779153e-01
-1.44779229e+00 9.55703080e-01 1.64127335e-01 -4.53126319e-02
-4.74698693e-01 9.41677988e-01 3.39198709e-01 -4.46559310e-01
6.02986813e-01 1.23953909e-01 -3.61324817e-01 4.33420762e-03
6.03043556e-01 2.66691923e-01 -1.50027439e-01 -7.85624683e-01
-6.02478564e-01 8.87825727e-01 -2.06526250e-01 4.81193095e-01
1.30727303e+00 -8.62006769e-02 -5.50822914e-01 -2.13849217e-01
1.31426823e+00 5.23447804e-02 -1.11990404e+00 -1.63966686e-01
2.78178096e-01 -4.23414618e-01 -4.77948077e-02 -2.54288793e-01
-1.92976129e+00 9.59523380e-01 4.85242516e-01 8.26953873e-02
1.28152478e+00 2.82534331e-01 7.21489847e-01 5.27683318e-01
3.29615593e-01 -1.06108546e+00 1.71833977e-01 9.98286828e-02
3.26686412e-01 -1.31639540e+00 1.10847421e-01 -1.24977100e+00
-3.03349972e-01 9.28081572e-01 1.04702699e+00 3.38708088e-02
9.08368766e-01 1.22867651e-01 -8.59389976e-02 -4.53570962e-01
-4.69277143e-01 -3.74638230e-01 3.43473971e-01 6.05479062e-01
4.38659668e-01 5.41548990e-02 -6.37196004e-02 7.58522213e-01
2.58750200e-01 -2.76605904e-01 5.75882867e-02 5.46025336e-01
-4.29210633e-01 -1.14476264e+00 -7.48625174e-02 6.48620129e-01
-3.25262368e-01 3.77514623e-02 -6.61524951e-01 5.80707669e-01
2.57333696e-01 1.26470399e+00 2.11048722e-01 -6.99745595e-01
-1.19070485e-01 -1.26883402e-01 3.82023990e-01 -7.66610980e-01
-2.02880368e-01 2.23737031e-01 -1.98811948e-01 -7.54480720e-01
-5.25579333e-01 -3.36288542e-01 -1.42876005e+00 -5.28702974e-01
-5.73276639e-01 4.91055429e-01 2.86698967e-01 8.51174891e-01
6.77123725e-01 2.31061712e-01 9.35246110e-01 -5.56940615e-01
-2.01489367e-02 -7.70192087e-01 -6.96652114e-01 4.82109159e-01
-2.00434148e-01 -7.70000577e-01 -2.51627564e-01 6.53392747e-02] | [13.464945793151855, 1.0667654275894165] |
e62ca7e3-e7d7-43ea-b217-535082cb14e4 | histogram-of-cell-types-deep-learning-for | 2107.02293 | null | https://arxiv.org/abs/2107.02293v2 | https://arxiv.org/pdf/2107.02293v2.pdf | Histogram of Cell Types: Deep Learning for Automated Bone Marrow Cytology | Bone marrow cytology is required to make a hematological diagnosis, influencing critical clinical decision points in hematology. However, bone marrow cytology is tedious, limited to experienced reference centers and associated with high inter-observer variability. This may lead to a delayed or incorrect diagnosis, leaving an unmet need for innovative supporting technologies. We have developed the first ever end-to-end deep learning-based technology for automated bone marrow cytology. Starting with a bone marrow aspirate digital whole slide image, our technology rapidly and automatically detects suitable regions for cytology, and subsequently identifies and classifies all bone marrow cells in each region. This collective cytomorphological information is captured in a novel representation called Histogram of Cell Types (HCT) quantifying bone marrow cell class probability distribution and acting as a cytological "patient fingerprint". The approach achieves high accuracy in region detection (0.97 accuracy and 0.99 ROC AUC), and cell detection and cell classification (0.75 mAP, 0.78 F1-score, Log-average miss rate of 0.31). HCT has potential to revolutionize hematopathology diagnostic workflows, leading to more cost-effective, accurate diagnosis and opening the door to precision medicine. | ['Clinton JV Campbell', 'Hamid R. Tizhoosh', 'Ronan Foley', 'Monalisa Sur', 'Catherine Ross', 'Taher Dehkharghanian', 'Youqing Mu', 'Rohollah Moosavi Tayebi'] | 2021-07-05 | null | null | null | null | ['cell-detection'] | ['computer-vision'] | [ 5.30757606e-02 -2.70674229e-01 -2.19156578e-01 2.56150775e-03
-1.22906566e+00 -6.31492376e-01 2.52237111e-01 1.26868606e+00
-3.21192563e-01 6.63592219e-01 -7.60823563e-02 -4.02661175e-01
1.87506810e-01 -1.09849691e+00 -4.73221168e-02 -1.13633931e+00
1.87623635e-01 1.23019505e+00 4.72826883e-02 5.33940434e-01
2.47594431e-01 1.08172560e+00 -8.45235527e-01 3.94353807e-01
5.85513771e-01 8.78293216e-01 5.06291427e-02 1.13852632e+00
-4.91655618e-01 6.96201026e-01 -6.72447026e-01 -1.79205224e-01
-3.15011829e-01 -4.82944727e-01 -4.84908462e-01 -1.40266389e-01
1.29583582e-01 -3.39174360e-01 -6.21942580e-02 6.85498357e-01
7.99846709e-01 -4.16701347e-01 1.09150088e+00 -7.42459595e-01
-2.69697964e-01 2.85576791e-01 -7.81664848e-01 5.46428978e-01
2.51915574e-01 4.48615789e-01 4.68208462e-01 -6.71647131e-01
5.77164054e-01 6.17273450e-01 9.07028317e-01 2.29431912e-01
-1.26297033e+00 -4.21554536e-01 -9.84004557e-01 -2.64769923e-02
-1.54543316e+00 -1.27852246e-01 6.15964271e-02 -7.20716238e-01
7.95269847e-01 2.29319274e-01 9.23007905e-01 2.54883856e-01
8.79284143e-01 4.44026381e-01 8.50265980e-01 -4.50501174e-01
4.65829581e-01 1.42344097e-02 1.76243961e-01 7.99898863e-01
6.57218754e-01 -1.51978254e-01 -5.33311665e-01 -1.45959631e-01
7.03882992e-01 4.36512679e-01 -1.20043449e-01 -2.27030516e-01
-1.12377810e+00 7.48018503e-01 1.82617784e-01 4.98788953e-01
-4.18476909e-01 4.53727357e-02 4.69707429e-01 -4.77055907e-02
-7.78934034e-03 6.63602650e-02 1.49672404e-01 -1.39558673e-01
-1.09251571e+00 -1.25617981e-01 5.42767584e-01 1.07136458e-01
4.57755566e-01 -2.56523192e-01 -3.16543162e-01 7.82171190e-01
7.14207888e-02 8.55871737e-01 5.13087213e-01 -4.75872070e-01
-4.62517560e-01 9.67113197e-01 -1.68180794e-01 -6.23646855e-01
-9.53558505e-01 -4.91287112e-01 -1.09013176e+00 9.04705748e-03
7.89450705e-01 2.99972236e-01 -8.07600379e-01 9.76405084e-01
5.27828991e-01 -1.09338425e-01 -4.74928081e-01 8.23308051e-01
4.78522241e-01 2.81913817e-01 2.92755902e-01 -1.10081032e-01
1.74887657e+00 -1.31576493e-01 -4.13180709e-01 5.12547612e-01
8.62375677e-01 -7.47936845e-01 7.61426032e-01 4.18053687e-01
-9.13789392e-01 5.35846576e-02 -9.73033130e-01 -1.15074366e-01
-4.65967447e-01 4.68465500e-02 6.67604804e-01 8.16407025e-01
-1.03660607e+00 3.83086711e-01 -9.57577944e-01 -6.34260535e-01
9.85702395e-01 5.26187062e-01 -4.17132407e-01 -2.49469757e-01
-5.47426343e-01 7.13999629e-01 1.76261842e-01 -7.68871605e-02
-7.15756655e-01 -1.26249969e+00 -5.21899760e-01 -5.77665791e-02
-3.53037983e-01 -8.18092465e-01 9.42650437e-01 -1.51980668e-01
-1.34684837e+00 1.29803288e+00 -1.36866525e-01 -3.41010034e-01
5.11453867e-01 4.35622871e-01 -2.38428086e-01 6.20782077e-01
2.38210097e-01 5.41996181e-01 3.52585837e-02 -6.30023539e-01
-8.84387136e-01 -8.16775084e-01 -9.28435266e-01 -2.73068905e-01
8.53448212e-02 -2.89292872e-01 -2.68233627e-01 -3.13526660e-01
1.09105758e-01 -6.07881248e-01 -3.73088964e-03 1.72395214e-01
-2.24067532e-02 -1.57321602e-01 3.08668733e-01 -6.05330288e-01
9.74503458e-01 -1.82454383e+00 -5.46443582e-01 3.92826706e-01
5.43732643e-01 3.20611119e-01 3.65704119e-01 1.46672666e-01
2.75306612e-01 6.06841668e-02 4.84862812e-02 -8.82934313e-03
-6.64512962e-02 -3.37397158e-01 1.52358264e-01 9.93479967e-01
-1.46447364e-02 1.19231009e+00 -1.05019319e+00 -7.88189650e-01
5.89222133e-01 8.66740763e-01 -1.84039816e-01 1.60516381e-01
1.53843388e-01 3.06787521e-01 -3.38954143e-02 1.26940620e+00
6.95936501e-01 -8.26846242e-01 2.92248935e-01 -1.09089576e-02
2.17359424e-01 -5.83685562e-02 -6.67814434e-01 1.32431877e+00
-3.59932452e-01 7.86466420e-01 -7.00554252e-02 -2.97513396e-01
8.43955636e-01 5.87099679e-02 7.06555903e-01 -7.61543989e-01
5.52872300e-01 3.02815497e-01 -1.23407632e-01 -3.94166797e-01
1.17164560e-01 -8.47932756e-01 1.92909926e-01 7.87848175e-01
-9.57412124e-02 -1.52267516e-01 5.86359464e-02 1.22543648e-01
1.29413891e+00 -5.15829444e-01 4.32568818e-01 -3.79831731e-01
5.62418699e-01 1.04479760e-01 1.60979688e-01 6.04027331e-01
-6.32209003e-01 8.73195827e-01 4.23730552e-01 -5.85676134e-01
-7.84780979e-01 -1.45273840e+00 -5.04044116e-01 7.38536954e-01
-1.76498502e-01 3.61472219e-01 -4.25381452e-01 -4.61596936e-01
6.21482968e-01 3.03699732e-01 -7.05074430e-01 5.34885786e-02
-4.53427821e-01 -1.02932489e+00 7.95582771e-01 5.20339072e-01
1.87878937e-01 -8.33294630e-01 -7.57903039e-01 4.51037496e-01
-6.92127645e-02 -5.79723179e-01 -2.16928631e-01 2.72099167e-01
-7.48545289e-01 -1.35647488e+00 -8.96011710e-01 -7.77895987e-01
5.33195376e-01 7.22261667e-02 1.13706422e+00 1.37821779e-01
-1.08995175e+00 4.09601331e-02 4.30436991e-02 -4.75162446e-01
-3.77941251e-01 -9.53897908e-02 -1.71041608e-01 -2.98439320e-02
1.00712526e+00 -2.04597443e-01 -1.09059381e+00 -1.26593426e-01
-7.90074468e-01 -3.54741901e-01 8.17440987e-01 8.48813772e-01
1.01472557e+00 -3.42768341e-01 7.69926369e-01 -1.06691217e+00
2.35347196e-01 -4.68356848e-01 -4.88619864e-01 4.30564761e-01
-6.54062808e-01 -4.38383996e-01 6.67426348e-01 5.99090531e-02
-7.18379498e-01 -2.19602957e-01 -5.71263395e-02 -7.12504983e-02
-2.41153002e-01 2.79229373e-01 4.72140968e-01 -1.13181144e-01
6.45207167e-01 4.59631443e-01 3.81241292e-01 -5.27731888e-03
-2.54914463e-01 7.43311405e-01 6.61510170e-01 1.10321358e-04
1.09501421e-01 9.63008523e-01 5.97098529e-01 -8.98851752e-01
-2.11966783e-01 -8.52001607e-01 -5.88633597e-01 -2.61408299e-01
6.77343011e-01 -6.53077245e-01 -1.29649758e+00 5.26724160e-01
-5.69176555e-01 -2.96790838e-01 -1.97605535e-01 4.78525251e-01
-2.44498923e-01 4.87203449e-01 -1.24757421e+00 -6.74537539e-01
-8.78557920e-01 -9.34843302e-01 1.20094359e+00 4.10104543e-01
-7.05417633e-01 -1.33847117e+00 4.64198768e-01 4.14032668e-01
6.38146818e-01 4.86519068e-01 1.04104781e+00 -6.39981151e-01
-3.49849880e-01 -7.47387171e-01 -3.86100680e-01 -3.34373027e-01
3.27451676e-01 1.25953242e-01 -1.04532743e+00 -2.41390273e-01
-2.56449372e-01 -2.06135720e-01 8.49864602e-01 8.01512241e-01
9.80217755e-01 4.57752734e-01 -6.08227968e-01 7.48292446e-01
1.73928785e+00 4.19162571e-01 4.47293520e-01 1.66553617e-01
4.12174523e-01 3.92649740e-01 1.87595651e-01 6.47777736e-01
2.37468138e-01 4.98874709e-02 -1.75246239e-01 -1.32736623e-01
-2.89995998e-01 -1.44646376e-01 -3.74042511e-01 3.54595363e-01
5.49520135e-01 -2.52891600e-01 -1.53662062e+00 6.82757735e-01
-1.11865377e+00 -6.98186040e-01 -1.25004768e-01 2.12307596e+00
6.05837524e-01 -1.84459686e-01 2.40903422e-01 5.49530536e-02
9.36756074e-01 -6.04799986e-01 -6.62296534e-01 -9.66673568e-02
-2.49130830e-01 5.95801592e-01 4.12812561e-01 5.15743434e-01
-6.61257505e-01 4.94532049e-01 6.86236382e+00 7.62372434e-01
-1.56167352e+00 -1.55050144e-01 1.14568865e+00 -4.60314155e-01
-1.25753552e-01 -3.69468391e-01 -5.57602286e-01 7.20923364e-01
7.54240692e-01 -2.08719134e-01 -1.73968554e-01 1.64538905e-01
1.35765269e-01 -9.10810947e-01 -1.27060735e+00 1.09119666e+00
1.23137616e-01 -1.78616190e+00 1.12212360e-01 3.85110259e-01
2.62525588e-01 -9.94071588e-02 9.20584276e-02 9.42491442e-02
1.88593969e-01 -1.18464577e+00 -9.19869542e-02 5.09877324e-01
1.48122597e+00 -8.84657085e-01 1.26400185e+00 -7.71939084e-02
-8.72309327e-01 4.61850643e-01 -2.67485708e-01 2.07230151e-01
-1.49867982e-01 1.01848423e+00 -1.74121916e+00 -6.93144370e-03
2.53610790e-01 1.67566702e-01 -3.85866612e-01 1.22666526e+00
3.89786422e-01 4.91238356e-01 -4.05763894e-01 -4.23444390e-01
-1.52624428e-01 1.77732315e-02 -2.51747351e-02 1.46344912e+00
1.60021052e-01 1.94194838e-01 -2.39165008e-01 5.14039218e-01
-1.52460322e-01 3.42330933e-01 -1.83520302e-01 1.20319068e-01
7.28992462e-01 1.32219732e+00 -1.64157045e+00 -4.37650323e-01
-3.56839187e-02 5.09345829e-01 2.91604489e-01 -6.96863756e-02
-5.96857905e-01 -4.42879647e-01 1.65910721e-01 3.52230132e-01
-1.22288547e-01 5.93605518e-01 -8.32374692e-01 -8.40991855e-01
-5.16756892e-01 -4.10300970e-01 8.69984806e-01 -4.00033504e-01
-1.33195686e+00 3.92964222e-02 -7.71279097e-01 -9.19441521e-01
-5.26082888e-02 -6.75915003e-01 -6.79688513e-01 7.22319484e-01
-1.39038777e+00 -8.23970318e-01 -3.55437011e-01 1.44027263e-01
4.81076352e-02 -5.11815809e-02 9.49047625e-01 1.18075594e-01
-6.19046867e-01 8.92101347e-01 4.11858886e-01 2.57965386e-01
7.91799724e-01 -1.40716541e+00 -4.24567074e-01 2.44153097e-01
-4.53091919e-01 5.59696674e-01 4.07990932e-01 -5.86580038e-01
-1.65142453e+00 -1.08670914e+00 1.00407839e+00 -4.82451767e-01
5.16261458e-01 2.00911574e-02 -8.09693992e-01 2.68274546e-01
-1.20365553e-01 3.19046788e-02 1.68534350e+00 -1.70572966e-01
-1.59390822e-01 -3.37722480e-01 -1.84378278e+00 3.67982954e-01
2.58471817e-02 -5.00499070e-01 -5.06098531e-02 4.56905127e-01
-1.67179212e-01 -4.56951380e-01 -1.09699452e+00 6.55051619e-02
9.40260947e-01 -1.01883101e+00 6.56900227e-01 -7.35534281e-02
1.71106741e-01 -1.97848350e-01 6.29952848e-02 -8.55379343e-01
-4.85187352e-01 -1.94246724e-01 -5.17009459e-02 8.84047568e-01
2.13068381e-01 -7.53306270e-01 1.49401593e+00 3.77869666e-01
8.57262984e-02 -1.03223503e+00 -1.06486332e+00 -3.51755440e-01
4.24046934e-01 -5.51175140e-02 7.36085057e-01 8.47857594e-01
5.84043145e-01 -3.96839269e-02 5.31041920e-01 -1.95215851e-01
7.12959230e-01 2.68798590e-01 6.83348119e-01 -1.08907104e+00
-2.36740798e-01 -1.21884727e+00 -6.43933296e-01 -3.81842941e-01
-3.91392320e-01 -1.04956853e+00 -3.51024449e-01 -1.55218267e+00
6.71845853e-01 -5.49375951e-01 -5.26340544e-01 1.70938164e-01
-2.72896737e-01 5.82798719e-01 -2.92160034e-01 1.48316681e-01
-6.48404241e-01 -1.71869040e-01 9.23101783e-01 -1.13737106e-01
1.16108038e-01 -1.81401134e-01 -6.47620916e-01 3.84431422e-01
9.70987678e-01 -3.17912370e-01 1.75832629e-01 7.71632046e-02
-3.72222364e-02 3.56875360e-01 2.72545725e-01 -1.09455824e+00
3.42547297e-01 -1.10591397e-01 1.13766825e+00 -1.02875364e+00
-7.24579731e-04 -3.60230625e-01 1.41251236e-01 1.14226425e+00
-9.81281772e-02 -2.30698854e-01 2.43590623e-01 3.89440417e-01
-1.67912871e-01 1.05841123e-01 1.23703158e+00 -1.03694513e-01
-6.96616098e-02 2.76891708e-01 -7.82270074e-01 -4.05264832e-02
1.17631435e+00 -5.86496949e-01 -4.07644778e-01 -5.20871319e-02
-4.71588641e-01 3.20154458e-01 7.66962290e-01 -4.91078347e-01
6.07595026e-01 -1.05245459e+00 -9.15880680e-01 3.23594779e-01
3.59894425e-01 7.89270550e-02 8.42064321e-01 9.46135998e-01
-1.56418073e+00 7.11014569e-01 -1.63428977e-01 -1.29757261e+00
-1.01359057e+00 2.62280107e-01 5.30342817e-01 -5.03096819e-01
-4.47205991e-01 1.17038357e+00 -9.17403847e-02 -6.18753862e-03
2.31491942e-02 -1.64808944e-01 1.42563403e-01 1.68647051e-01
8.99487495e-01 6.23316467e-01 3.60259324e-01 -3.29376787e-01
-6.09566331e-01 5.42582929e-01 -3.13482374e-01 1.47856787e-01
1.04639959e+00 1.66039571e-01 -2.37731278e-01 5.03516197e-01
1.22429991e+00 8.04850236e-02 -7.10713267e-01 -5.06526530e-02
-1.61309004e-01 -5.33234954e-01 1.84252113e-02 -9.46701169e-01
-9.61347103e-01 1.09763980e+00 8.07892442e-01 1.35649130e-01
1.00564575e+00 -5.81311695e-02 8.43676090e-01 -4.01897728e-02
4.90955532e-01 -8.83084595e-01 -1.53971106e-01 1.80180177e-01
8.17963630e-02 -1.26573455e+00 2.89217472e-01 -9.47391316e-02
-2.63716519e-01 1.27804446e+00 3.44747841e-01 1.27569929e-01
6.59334660e-01 6.74937844e-01 3.89476150e-01 -3.19202423e-01
-7.21916735e-01 1.42453060e-01 -4.18771952e-01 8.53900790e-01
8.92465830e-01 5.63160121e-01 -3.92632484e-02 1.47802666e-01
-8.05085152e-02 2.82204568e-01 2.65952826e-01 9.68632042e-01
-8.24512720e-01 -6.48589849e-01 -5.71516335e-01 8.10114205e-01
-7.30773449e-01 2.85969943e-01 -2.98788816e-01 7.78715014e-01
-1.16220899e-01 7.32350588e-01 4.39408988e-01 9.58091095e-02
-1.67060271e-01 1.81052476e-01 5.78777611e-01 -4.15566862e-01
-6.66918933e-01 1.90311208e-01 -5.87302685e-01 -9.31261107e-02
2.67625511e-01 -7.46477604e-01 -1.48275506e+00 -7.68168211e-01
-1.61197856e-01 -8.11313931e-03 4.12946522e-01 6.14210606e-01
3.93789232e-01 4.48207051e-01 5.06470799e-01 -3.12794387e-01
-1.07931420e-01 -8.82348955e-01 -8.74631286e-01 3.21917623e-01
4.57153529e-01 -3.18792790e-01 -2.75924712e-01 -1.10184357e-01] | [15.00773811340332, -3.1477739810943604] |
391811f0-d72a-4eac-a7b2-cf3c1e738377 | do-graph-neural-networks-learn-traditional | 2211.09912 | null | https://arxiv.org/abs/2211.09912v1 | https://arxiv.org/pdf/2211.09912v1.pdf | Do graph neural networks learn traditional jet substructure? | At the CERN LHC, the task of jet tagging, whose goal is to infer the origin of a jet given a set of final-state particles, is dominated by machine learning methods. Graph neural networks have been used to address this task by treating jets as point clouds with underlying, learnable, edge connections between the particles inside. We explore the decision-making process for one such state-of-the-art network, ParticleNet, by looking for relevant edge connections identified using the layerwise-relevance propagation technique. As the model is trained, we observe changes in the distribution of relevant edges connecting different intermediate clusters of particles, known as subjets. The resulting distribution of subjet connections is different for signal jets originating from top quarks, whose subjets typically correspond to its three decay products, and background jets originating from lighter quarks and gluons. This behavior indicates that the model is using traditional jet substructure observables, such as the number of prongs -- energetic particle clusters -- within a jet, when identifying jets. | ['Javier Duarte', 'Raghav Kansal', 'Farouk Mokhtar'] | 2022-11-17 | null | null | null | null | ['jet-tagging'] | ['graphs'] | [-2.61927426e-01 3.41973573e-01 8.20128247e-02 -3.38551551e-01
-2.17211649e-01 -7.96390355e-01 9.99219358e-01 6.41378045e-01
-3.21377844e-01 5.98167360e-01 2.81525493e-01 -5.30413032e-01
-2.12744102e-01 -8.98419976e-01 -6.52195513e-01 -6.84545636e-01
-4.35821682e-01 1.27982354e+00 6.30477786e-01 1.30728912e-02
1.56985253e-01 7.68783271e-01 -9.67421532e-01 7.52564490e-01
2.06333026e-01 9.48202074e-01 -2.15173736e-01 9.47559893e-01
-5.97015083e-01 6.79262459e-01 -3.87378156e-01 -8.67200136e-01
2.67446488e-01 -7.28272080e-01 -7.47706831e-01 -7.03402638e-01
1.81270346e-01 5.13466120e-01 -2.84441769e-01 1.46163535e+00
1.05995014e-01 1.70752749e-01 7.91991174e-01 -1.21885300e+00
-5.28267086e-01 8.71360123e-01 2.02946439e-01 6.21260166e-01
-4.53128338e-01 -5.29828146e-02 1.42026925e+00 -7.45442629e-01
9.73068237e-01 1.58129156e+00 6.74318075e-01 3.34251016e-01
-1.31616676e+00 -4.81783539e-01 -1.43662691e-01 1.25064664e-02
-6.84684098e-01 -7.64583573e-02 7.54482567e-01 -9.36238050e-01
1.10344374e+00 4.06266093e-01 7.67809272e-01 8.75683367e-01
6.93526328e-01 4.96174619e-02 8.43021810e-01 -5.38604856e-01
3.36503893e-01 4.11916316e-01 2.46544063e-01 4.48912710e-01
4.93023723e-01 4.26094353e-01 -4.10799801e-01 -4.19888884e-01
7.87962675e-01 -2.74874836e-01 1.85153157e-01 -6.14041924e-01
-1.13309658e+00 1.09309721e+00 7.30679154e-01 4.66006517e-01
-7.11959898e-02 2.76712120e-01 5.92356920e-01 2.19125059e-02
8.77703547e-01 9.19708133e-01 -8.24992001e-01 2.51741350e-01
-5.49907744e-01 3.22230905e-01 1.03297949e+00 6.58830106e-01
5.01506984e-01 -2.34032333e-01 -3.78494620e-01 3.92032355e-01
4.66438115e-01 -1.40357837e-01 -1.59006253e-01 -6.95792615e-01
5.61016381e-01 5.92552900e-01 1.77751392e-01 -5.37183344e-01
-5.50415337e-01 -6.30504668e-01 -5.74390590e-01 6.48013175e-01
4.71477687e-01 -4.63723183e-01 -9.47224796e-01 1.21170902e+00
5.15381753e-01 -5.23947775e-02 1.23036191e-01 1.02977335e+00
8.79190862e-01 6.39436543e-01 2.34155834e-01 9.54086035e-02
1.35391617e+00 -9.05651689e-01 -5.62222973e-02 -2.60968432e-02
2.21904069e-01 -7.29174495e-01 1.48589954e-01 5.76823354e-02
-9.59860682e-01 -7.24021018e-01 -5.42994082e-01 1.47711575e-01
-5.50352931e-01 -2.02410650e-02 1.07321322e+00 1.83757469e-01
-8.62774491e-01 1.28673363e+00 -7.50932455e-01 -4.74747211e-01
2.27152146e-02 3.28847468e-01 -1.02103867e-01 5.27412713e-01
-9.88369167e-01 8.12654197e-01 8.46235514e-01 1.64178044e-01
-5.85918069e-01 -8.35594535e-01 -3.85818273e-01 2.62759149e-01
-1.32865328e-02 -7.90447116e-01 1.26452196e+00 -6.69287026e-01
-1.26221406e+00 6.20373666e-01 1.38552621e-01 -5.19358635e-01
3.18942130e-01 1.79869533e-01 -9.06400263e-01 1.41057059e-01
2.91133430e-02 5.67929685e-01 6.41947985e-01 -1.37337816e+00
-8.72083843e-01 1.78632829e-02 -9.73554328e-02 -2.38733530e-01
4.64155972e-01 7.84575224e-01 -8.43245834e-02 -2.76969075e-01
2.45960489e-01 -1.00811613e+00 -4.12502915e-01 -3.50375801e-01
-8.05300534e-01 -5.88279486e-01 3.13884676e-01 -4.53960598e-01
3.14278245e-01 -1.91497362e+00 9.92969424e-02 3.10491830e-01
6.37718439e-01 -6.05259985e-02 1.76082551e-01 5.80214739e-01
-3.69480610e-01 3.13319445e-01 2.19710156e-01 -1.09771527e-01
3.46262068e-01 -1.80404782e-02 -2.06148148e-01 1.83397889e-01
3.80140066e-01 8.58475089e-01 -1.02385354e+00 -9.05018765e-03
1.02434032e-01 4.12384331e-01 -3.12966347e-01 1.20096512e-01
-5.98885477e-01 6.93217099e-01 -4.28899914e-01 6.97120056e-02
6.27355397e-01 -1.11287832e-01 9.46534500e-02 -3.22224945e-01
-5.10786235e-01 6.30465209e-01 -6.12006426e-01 1.25695515e+00
1.53380021e-01 8.50944877e-01 3.26192230e-01 -6.87748253e-01
9.61692572e-01 2.40358114e-01 4.49496329e-01 -5.04374027e-01
1.05682552e-01 3.54424715e-02 8.70884061e-01 -2.11282089e-01
6.42666757e-01 -3.75047594e-01 -1.62786499e-01 1.13588333e-01
3.60607266e-01 2.36588612e-01 3.99435937e-01 4.11033154e-01
1.28206813e+00 -4.88593951e-02 -2.77380228e-01 -4.65411156e-01
-1.14162117e-01 3.75089228e-01 6.99005246e-01 8.65655065e-01
1.30613506e-01 3.90744478e-01 9.33462262e-01 -6.26564801e-01
-1.47135210e+00 -1.31464469e+00 -3.08953136e-01 9.38948572e-01
-2.68437378e-02 -6.25590563e-01 -1.74445301e-01 -7.85265684e-01
2.85248518e-01 9.54609156e-01 -3.78817320e-01 -1.78631142e-01
-5.58738351e-01 -8.89292777e-01 3.44780423e-02 5.43195605e-01
-2.81161249e-01 -1.45935607e+00 -1.43138841e-01 4.42631632e-01
6.20243490e-01 -1.02937734e+00 1.05257601e-01 7.86339462e-01
-7.49060631e-01 -8.88906360e-01 -2.66740145e-03 -4.85623449e-01
6.83097064e-01 -5.99111199e-01 1.60103703e+00 -1.51958168e-01
-4.50393021e-01 -2.13156998e-01 -3.59149486e-01 -5.36222935e-01
-9.13420558e-01 -3.66366297e-01 1.58184934e-02 -7.92038664e-02
2.51363665e-01 -2.51102298e-01 -4.90140110e-01 -2.95291096e-01
-2.45251089e-01 -6.09481819e-02 6.57957613e-01 3.07379544e-01
3.47951680e-01 4.70529616e-01 -3.56015302e-02 -8.81247580e-01
4.98429716e-01 -6.36187255e-01 -9.38240647e-01 -5.38291894e-02
-2.40195468e-01 2.96703994e-01 8.93906951e-01 -1.22959718e-01
-8.79605770e-01 -3.28873992e-01 3.03585455e-02 -2.97180533e-01
-6.34111524e-01 7.17718378e-02 -1.47697143e-02 -2.41104811e-01
2.36991301e-01 -2.85519511e-01 -6.26635730e-01 -9.07967269e-01
5.97726941e-01 2.29772925e-02 6.95407569e-01 -5.47292888e-01
7.56150067e-01 2.28888467e-01 3.53649825e-01 -4.31487143e-01
-7.68335879e-01 -6.23031735e-01 -9.95737553e-01 -2.06763908e-01
1.32687700e+00 -5.28949261e-01 -6.97555840e-01 -2.03827992e-01
-1.64928150e+00 3.06153577e-02 -5.49044967e-01 9.69924927e-01
-1.15091866e-02 -1.85899407e-01 -9.16775525e-01 -7.00195193e-01
-2.04077110e-01 -8.50841582e-01 8.82674575e-01 1.68321580e-01
1.61002167e-02 -1.32305443e+00 4.01065856e-01 -1.74986809e-01
1.22516707e-01 4.09666225e-02 1.72914231e+00 -1.37310672e+00
-6.46055400e-01 -1.37319937e-01 -4.71946865e-01 2.27401212e-01
-2.52574980e-01 1.43142790e-01 -5.61193585e-01 -2.89044023e-01
-2.94931144e-01 2.60415763e-01 1.11373079e+00 4.06447142e-01
6.78719640e-01 -8.37249234e-02 -7.70435750e-01 3.11668456e-01
1.38050699e+00 2.44453982e-01 -1.97361577e-02 -2.68237907e-02
6.12533510e-01 4.60716367e-01 7.20884055e-02 -1.46321028e-01
-1.71082854e-01 5.78444004e-01 5.52568436e-01 -2.59136260e-01
-3.45798373e-01 -2.01076806e-01 2.35443264e-01 9.59974229e-01
-2.66314417e-01 -5.94048977e-01 -7.32951224e-01 3.05683285e-01
-1.81196606e+00 -9.49470818e-01 -6.87728941e-01 2.10723472e+00
4.25898999e-01 8.44727993e-01 -7.50959963e-02 -5.50874591e-01
8.72400641e-01 -1.27104864e-01 -2.40750745e-01 -7.99150169e-01
5.92169166e-01 4.97082412e-01 6.25231802e-01 4.54515755e-01
-1.04003787e+00 8.84421766e-01 6.72845554e+00 3.00316632e-01
-9.37387526e-01 2.43999869e-01 1.74130589e-01 -1.13320217e-01
-3.85350287e-01 5.85648954e-01 -1.09259379e+00 7.58505344e-01
1.02101457e+00 3.91731858e-01 2.06257388e-01 5.45012712e-01
4.88520227e-03 6.80394918e-02 -1.39893103e+00 3.00868988e-01
-3.62813503e-01 -1.79891610e+00 -1.31299138e-01 3.03073991e-02
6.14387095e-01 5.89367867e-01 -5.91317594e-01 4.36783880e-01
1.00873232e+00 -7.07492530e-01 1.00434792e+00 8.28538179e-01
2.87868023e-01 -6.51671171e-01 5.75191021e-01 1.23165078e-01
-1.11602974e+00 2.46330485e-01 -8.27108204e-01 -9.60845053e-02
3.59437257e-01 1.14998591e+00 -1.37997723e+00 5.99490583e-01
7.25683749e-01 1.54207572e-01 -5.42562544e-01 1.35687959e+00
-3.04632574e-01 9.18512523e-01 -2.83055902e-01 -1.65059969e-01
2.94588774e-01 -5.27850866e-01 9.31796134e-01 1.36664534e+00
3.07251364e-01 -3.34621429e-01 4.85034108e-01 1.39972389e+00
-1.55596539e-01 -3.29235375e-01 -5.07784009e-01 -3.43206614e-01
2.12171495e-01 1.80595219e+00 -1.60899341e+00 -2.40683109e-01
-3.35412621e-01 6.03710890e-01 4.28975850e-01 4.03361082e-01
-6.19228542e-01 -2.92028874e-01 7.66604543e-01 8.67719278e-02
7.15273440e-01 -4.40575749e-01 1.81408376e-01 -6.04023635e-01
-2.35635042e-01 6.16539717e-02 1.35890573e-01 -7.90021122e-01
-1.64456356e+00 7.22760320e-01 -1.09342501e-01 -6.60129845e-01
-1.00195169e-01 -9.92556095e-01 -1.24905586e+00 9.95837867e-01
-1.11946118e+00 -9.14242566e-01 4.25335258e-01 -2.26368215e-02
1.37400731e-01 -3.69627357e-01 5.18230677e-01 1.43752247e-02
-3.22433531e-01 -1.71117350e-01 3.46007437e-01 3.97290051e-01
3.00079197e-01 -1.67653215e+00 9.99140918e-01 6.77049458e-01
7.50908434e-01 2.45196581e-01 9.04738724e-01 -1.18770766e+00
-1.07575071e+00 -1.36135876e+00 1.06845093e+00 -6.71877027e-01
1.13076055e+00 -9.87606645e-01 -6.00204766e-01 6.30916059e-01
4.46470141e-01 5.96626699e-01 3.30966055e-01 4.01009053e-01
3.03046461e-02 2.28566885e-01 -7.35687852e-01 1.05657965e-01
1.03195167e+00 -5.51315069e-01 -5.22764742e-01 8.39762747e-01
8.86858940e-01 -1.92547098e-01 -8.28851759e-01 3.96317899e-01
-3.03717077e-01 -9.68547702e-01 9.94665384e-01 -1.48819375e+00
4.01238799e-02 -2.34379366e-01 3.07832271e-01 -1.75023746e+00
-1.04571533e+00 -4.19286579e-01 1.97011635e-01 1.21459520e+00
7.02305913e-01 -3.94741207e-01 7.93644190e-01 1.65313587e-01
-4.76810843e-01 -1.90660775e-01 -9.55259919e-01 -7.83729017e-01
1.84790328e-01 -4.00335342e-01 3.40753078e-01 6.96304321e-01
-3.58266413e-01 7.59336054e-01 1.04494601e-01 5.40567875e-01
5.12960136e-01 3.76053154e-01 5.14200032e-02 -1.94906068e+00
-5.45663595e-01 -4.21995759e-01 -5.23362815e-01 -2.34768495e-01
2.02937424e-01 -1.76117873e+00 1.97186753e-01 -1.63426828e+00
5.24603995e-03 -7.24857450e-01 -3.69287461e-01 7.21045434e-02
3.14048618e-01 -2.59946659e-02 3.75943214e-01 -4.03261231e-03
-8.01056206e-01 1.50033012e-01 1.03298843e+00 -1.27033656e-02
2.51645207e-01 1.66715965e-01 -1.58763468e-01 1.20387793e+00
7.60809898e-01 -9.81714070e-01 1.40241668e-01 -3.93011302e-01
8.11168849e-01 1.90891296e-01 8.60832751e-01 -1.19538629e+00
3.01517457e-01 1.97569028e-01 8.20636988e-01 -5.60598373e-01
4.32046920e-01 -6.35913253e-01 3.00348014e-01 1.40490144e-01
-4.38012481e-01 2.11517766e-01 1.97125375e-02 6.00971997e-01
-3.07991076e-02 -7.37492383e-01 3.72302622e-01 -5.32858908e-01
-6.07042611e-01 3.27633888e-01 -3.37720096e-01 1.82835773e-01
1.01377392e+00 5.09403944e-01 -3.87859613e-01 2.85219342e-01
-1.16755891e+00 -1.88153222e-01 4.19971384e-02 4.61237013e-01
1.08868770e-01 -1.42821753e+00 -7.08597898e-01 8.62052888e-02
-1.91280946e-01 -2.67036170e-01 -4.48585212e-01 5.75267375e-01
-2.68573701e-01 6.03959382e-01 1.77997425e-02 -4.66425568e-01
-1.11514199e+00 5.44405401e-01 2.96414554e-01 -2.62092471e-01
-9.16855454e-01 1.29082894e+00 -1.17694084e-02 -6.17426872e-01
-1.61755085e-01 -7.32734740e-01 -1.21973455e-01 -1.18161496e-02
-1.36663035e-01 4.65242974e-02 3.98022592e-01 -3.82657796e-01
-1.82747036e-01 2.20093932e-02 -9.60726291e-03 2.05076024e-01
1.09647310e+00 6.10407948e-01 -6.08297169e-01 5.93984067e-01
1.09013283e+00 3.94921571e-01 -1.23501825e+00 1.03213862e-01
2.56296724e-01 -9.38489288e-02 -2.25359183e-02 -9.46321666e-01
-9.67253029e-01 8.73768806e-01 4.60676670e-01 9.05295610e-01
-1.25973389e-01 8.16175997e-01 6.37872100e-01 2.05691561e-01
3.10378850e-01 -7.59198129e-01 -3.91990066e-01 6.72655106e-01
5.52676737e-01 -1.06211531e+00 -3.58536214e-01 -3.55845779e-01
-1.63550779e-01 1.44146752e+00 4.29771900e-01 -2.08646685e-01
5.17557681e-01 3.98655951e-01 -2.33952552e-01 -7.13908613e-01
-9.93779063e-01 -2.25206316e-01 6.33201838e-01 3.14922184e-01
3.52533758e-01 5.32905817e-01 2.19903588e-02 4.72652912e-01
-4.79833245e-01 -5.98405480e-01 1.61948860e-01 4.05573666e-01
-5.98304033e-01 -1.26568985e+00 -3.36810023e-01 9.20360982e-01
-2.87721962e-01 -3.82570058e-01 -6.70563519e-01 6.99934900e-01
7.60052323e-01 6.72440529e-01 6.44795716e-01 -2.37600088e-01
2.62006577e-02 2.60670483e-01 5.55463970e-01 -9.51235533e-01
-8.35766733e-01 -1.99276045e-01 4.16662574e-01 -3.03258985e-01
-2.00420469e-01 -5.74569285e-01 -1.60308325e+00 5.33098504e-02
-3.66415113e-01 9.76959109e-01 7.88186014e-01 1.09828413e+00
3.12775791e-01 9.58409607e-01 2.03907445e-01 -1.05560577e+00
-3.57923090e-01 -6.92391813e-01 -7.60185182e-01 4.02798980e-01
1.05448075e-01 -5.97195446e-01 -3.17233860e-01 -5.08405566e-01] | [15.706509590148926, 2.9187138080596924] |
7d84eacb-05f6-4994-899f-81d60cf96974 | protodiff-learning-to-learn-prototypical | 2306.14770 | null | https://arxiv.org/abs/2306.14770v1 | https://arxiv.org/pdf/2306.14770v1.pdf | ProtoDiff: Learning to Learn Prototypical Networks by Task-Guided Diffusion | Prototype-based meta-learning has emerged as a powerful technique for addressing few-shot learning challenges. However, estimating a deterministic prototype using a simple average function from a limited number of examples remains a fragile process. To overcome this limitation, we introduce ProtoDiff, a novel framework that leverages a task-guided diffusion model during the meta-training phase to gradually generate prototypes, thereby providing efficient class representations. Specifically, a set of prototypes is optimized to achieve per-task prototype overfitting, enabling accurately obtaining the overfitted prototypes for individual tasks. Furthermore, we introduce a task-guided diffusion process within the prototype space, enabling the meta-learning of a generative process that transitions from a vanilla prototype to an overfitted prototype. ProtoDiff gradually generates task-specific prototypes from random noise during the meta-test stage, conditioned on the limited samples available for the new task. Furthermore, to expedite training and enhance ProtoDiff's performance, we propose the utilization of residual prototype learning, which leverages the sparsity of the residual prototype. We conduct thorough ablation studies to demonstrate its ability to accurately capture the underlying prototype distribution and enhance generalization. The new state-of-the-art performance on within-domain, cross-domain, and few-task few-shot classification further substantiates the benefit of ProtoDiff. | ['Cees Snoek', 'Shengcai Liao', 'Zehao Xiao', 'Yingjun Du'] | 2023-06-26 | null | null | null | null | ['meta-learning', 'few-shot-learning'] | ['methodology', 'methodology'] | [ 2.22031489e-01 -3.26020658e-01 -3.47199410e-01 -2.90288091e-01
-8.52774799e-01 -2.48138279e-01 7.23914504e-01 7.48688262e-03
-2.33712733e-01 4.15870070e-01 8.75404850e-03 -1.19899407e-01
-1.01406097e-01 -6.67724967e-01 -5.91566980e-01 -6.02950931e-01
1.19713627e-01 3.47394884e-01 -1.40403407e-02 -3.92292738e-02
1.15261838e-01 3.74369100e-02 -1.64175010e+00 2.89956778e-01
1.13907564e+00 1.00857341e+00 5.42069733e-01 5.58737516e-01
-1.51346579e-01 3.16712141e-01 -6.61000609e-01 -8.49014893e-02
2.23606035e-01 -4.56771910e-01 -3.10578734e-01 2.96412051e-01
3.88481140e-01 -2.31468379e-01 -3.26320887e-01 8.51303399e-01
5.52869201e-01 5.57018936e-01 9.59228873e-01 -1.21099949e+00
-9.90091324e-01 5.10791481e-01 -5.36008537e-01 2.23217666e-01
-4.05246727e-02 5.09300888e-01 9.75986421e-01 -1.16271770e+00
5.00790060e-01 8.17741632e-01 7.77721703e-01 6.55824244e-01
-1.25536704e+00 -8.58251512e-01 2.57679403e-01 1.00089632e-01
-1.48520422e+00 -5.78759372e-01 9.29276884e-01 -4.85234886e-01
9.28996980e-01 -9.91158485e-02 7.10053980e-01 1.54451799e+00
6.07628338e-02 9.90001738e-01 7.99782991e-01 -4.16802645e-01
7.80865014e-01 1.89879850e-01 3.86016250e-01 6.48152292e-01
2.28969678e-01 2.49772653e-01 -5.21768689e-01 -3.74320239e-01
5.72307706e-01 4.62124377e-01 -2.53059506e-01 -6.16766274e-01
-7.68481553e-01 9.54904675e-01 4.08806771e-01 1.87458888e-01
-5.47387004e-01 -8.22495203e-03 4.83526260e-01 2.57243156e-01
5.40281594e-01 6.88382268e-01 -2.17116535e-01 -2.11185083e-01
-1.26083124e+00 2.11507961e-01 4.96643513e-01 1.06827188e+00
9.35936570e-01 2.82779455e-01 -5.38321912e-01 1.26515114e+00
6.28997907e-02 2.96609610e-01 1.01133430e+00 -6.77777708e-01
4.75298256e-01 5.84485173e-01 -1.15637980e-01 -4.67370540e-01
-2.28912346e-02 -7.98931420e-01 -7.63539910e-01 3.51800472e-02
-7.93947950e-02 -2.48606414e-01 -1.36671770e+00 1.77839589e+00
3.21201235e-01 6.59341693e-01 2.38694325e-02 6.45830989e-01
4.98544365e-01 6.82317972e-01 2.01472536e-01 -1.53894350e-01
9.89803016e-01 -1.18056154e+00 -1.67083472e-01 -4.80619520e-01
6.53982222e-01 -2.04789773e-01 1.45338929e+00 1.88693374e-01
-8.23943853e-01 -8.47018898e-01 -1.17027295e+00 2.46496364e-01
-2.03700230e-01 -1.01067452e-02 6.31623387e-01 6.85573280e-01
-6.23294413e-01 6.52484179e-01 -7.86661208e-01 -1.40569314e-01
9.71673906e-01 -1.37088656e-01 -1.38951406e-01 -5.67685008e-01
-9.09884512e-01 5.26835263e-01 5.14144421e-01 -3.56735259e-01
-1.26505530e+00 -1.22901738e+00 -9.55809891e-01 2.32805058e-01
2.21936345e-01 -1.02058542e+00 1.36173952e+00 -5.73454678e-01
-1.31042099e+00 3.74237984e-01 -1.72348559e-01 -5.11541009e-01
5.22748470e-01 -5.79870567e-02 -3.62545401e-01 -8.23401958e-02
4.18440662e-02 5.85194349e-01 1.41740441e+00 -1.06288922e+00
-6.10122979e-01 -1.25464335e-01 -2.96519011e-01 2.44062111e-01
-6.90111101e-01 -7.09144354e-01 -3.60679507e-01 -7.59886086e-01
-8.13505873e-02 -8.30029488e-01 -1.73293933e-01 -1.20864071e-01
-2.82652229e-01 -3.44820440e-01 7.69117057e-01 -1.00720756e-01
1.28492916e+00 -2.43095827e+00 -2.19868906e-02 5.92390671e-02
2.76827186e-01 5.00778019e-01 -4.61485267e-01 1.59988284e-01
8.81353468e-02 -7.06179291e-02 -4.06407088e-01 -7.38610506e-01
-9.45702344e-02 1.32279873e-01 -4.87326950e-01 1.63322076e-01
2.63495952e-01 1.15783060e+00 -1.12469983e+00 -1.21668220e-01
1.55261621e-01 3.43190551e-01 -4.62248415e-01 2.90635973e-01
-1.90822020e-01 -1.11810211e-02 -4.79692727e-01 6.27708316e-01
4.27068770e-01 -3.89871389e-01 -2.07244307e-01 -5.77414855e-02
2.74668574e-01 -8.23601261e-02 -8.52076888e-01 2.02001929e+00
-8.27675879e-01 4.21836406e-01 -3.98489207e-01 -1.07303333e+00
1.11358654e+00 6.97281137e-02 2.16875553e-01 -6.66352630e-01
9.73033346e-03 1.52163714e-01 -8.21143016e-02 -4.51677650e-01
5.25372386e-01 -3.57296199e-01 6.85405394e-04 7.34156907e-01
3.98301542e-01 -8.33356082e-02 1.25629157e-01 4.47128080e-02
1.20088685e+00 5.99331148e-02 3.36132377e-01 2.04069450e-01
-2.10068658e-01 5.07543348e-02 4.59904492e-01 1.23422635e+00
-2.60094076e-01 7.41871238e-01 -7.48394281e-02 -3.40475768e-01
-1.16181314e+00 -1.26192439e+00 -1.57373235e-01 1.16127717e+00
-1.99320436e-01 -2.10578263e-01 -3.90960068e-01 -8.26774836e-01
2.37414092e-01 1.05969417e+00 -8.79394174e-01 -6.85680628e-01
-1.66831136e-01 -6.71141684e-01 2.13153556e-01 6.01068258e-01
3.73071045e-01 -1.02805269e+00 -7.33060122e-01 4.42932874e-01
1.50949001e-01 -6.85608923e-01 -6.18471980e-01 2.81692266e-01
-1.07853818e+00 -9.35826898e-01 -1.13851106e+00 -6.18035257e-01
7.45117068e-01 7.16149509e-01 8.06509078e-01 -2.48149708e-01
-4.77861613e-01 3.65984231e-01 -3.91963840e-01 -1.92510054e-01
-2.98791617e-01 2.17786938e-01 1.28361344e-01 -1.01763807e-01
3.09574693e-01 -7.35399127e-01 -5.87773442e-01 -5.53778233e-03
-7.61395693e-01 -1.14840046e-01 7.47036576e-01 1.30868304e+00
5.88054419e-01 9.24363434e-02 8.03365111e-01 -1.09031129e+00
1.02526903e+00 -8.88250113e-01 -1.00335293e-01 3.75991434e-01
-6.96318746e-01 3.49061675e-02 7.39502609e-01 -1.09779477e+00
-1.00525999e+00 -1.43330276e-01 2.20417231e-01 -1.14720285e+00
-2.31900457e-02 6.27281904e-01 2.05472112e-01 1.76685333e-01
1.08047307e+00 6.24239266e-01 -8.10330510e-02 -4.11747992e-01
6.71114206e-01 8.51836860e-01 5.03947258e-01 -6.03219450e-01
6.05445325e-01 2.25026205e-01 -4.67578709e-01 -8.70972276e-01
-9.51272607e-01 -7.06683159e-01 -2.86078513e-01 7.24135265e-02
2.73751825e-01 -1.05498028e+00 -5.32990359e-02 2.94365108e-01
-8.01536143e-01 -7.24824607e-01 -8.38417709e-01 4.14846092e-01
-5.43432593e-01 9.94347036e-02 -3.22687745e-01 -7.77271807e-01
-6.05079532e-01 -9.04895604e-01 9.41426396e-01 2.95161873e-01
-3.73127520e-01 -9.67267156e-01 2.84735829e-01 -2.74036787e-02
4.65663284e-01 -5.32539599e-02 1.04241824e+00 -8.30844581e-01
-3.24124508e-02 -5.28500199e-01 -7.54388943e-02 4.27958667e-01
3.00581306e-01 -2.92594224e-01 -9.95923996e-01 -5.52422762e-01
1.45139266e-03 -6.08037114e-01 1.11806703e+00 3.83701950e-01
1.12022507e+00 -5.93484342e-02 -4.63147700e-01 7.12830484e-01
1.21180701e+00 1.09600179e-01 2.47611091e-01 1.84664369e-01
5.54823101e-01 2.25784287e-01 5.84339201e-01 6.70669079e-01
1.43587679e-01 4.03384656e-01 -1.16338551e-01 3.14326227e-01
-3.45167756e-01 -4.86990750e-01 9.39819813e-02 6.99891806e-01
3.78936529e-01 3.53092141e-02 -8.30592811e-01 6.61603928e-01
-1.81351137e+00 -1.09706295e+00 7.03374445e-01 2.29005909e+00
7.25005448e-01 3.18523794e-01 2.00625584e-01 -8.85156542e-02
6.69505596e-01 2.66751766e-01 -1.13044906e+00 -1.48245946e-01
2.98655599e-01 3.87487739e-01 -1.06377274e-01 3.10833212e-02
-1.00378048e+00 9.40803170e-01 6.52739763e+00 9.82404709e-01
-1.28255856e+00 3.88172597e-01 5.11173069e-01 -2.74344146e-01
-3.75957906e-01 -2.30305150e-01 -8.65495265e-01 5.99988163e-01
7.85607457e-01 -5.18683910e-01 4.77163672e-01 1.39418042e+00
1.98287107e-02 2.00922012e-01 -1.04760790e+00 9.52299356e-01
1.81955174e-01 -1.40661192e+00 9.69663784e-02 -1.32128187e-02
1.10283577e+00 2.53814936e-01 4.82159734e-01 1.01049435e+00
3.29933256e-01 -6.76736832e-01 6.52258098e-01 3.32398504e-01
8.95579934e-01 -6.70399129e-01 2.08542794e-01 5.84074557e-01
-1.15426588e+00 -5.74188828e-01 -8.69620681e-01 1.26745000e-01
-4.07893816e-03 7.59111404e-01 -1.25640368e+00 3.04598570e-01
1.72562093e-01 6.48310959e-01 -6.10642731e-01 1.37574768e+00
-1.11818790e-01 6.41387284e-01 -9.67262015e-02 -5.92336729e-02
2.72238672e-01 1.24182686e-01 4.85688835e-01 1.09716773e+00
5.95837295e-01 -1.55147523e-01 3.07478726e-01 1.14893317e+00
-2.18383670e-01 -1.07524537e-01 -6.16713822e-01 -1.15574583e-01
1.01381934e+00 1.13734233e+00 -3.64606589e-01 -5.69648743e-01
-2.89801121e-01 9.60645735e-01 6.91123784e-01 5.83556652e-01
-6.49699748e-01 -5.83242416e-01 7.00694323e-01 -3.54264081e-02
6.17731094e-01 -1.72434226e-01 -2.58168548e-01 -1.14547241e+00
-6.94403723e-02 -7.31942713e-01 3.53712916e-01 -6.25916302e-01
-1.53744936e+00 5.78668118e-01 3.44875678e-02 -1.33755302e+00
-4.88303065e-01 -7.85861239e-02 -1.23692179e+00 8.92968357e-01
-1.42080975e+00 -1.08498406e+00 -4.51961279e-01 3.70145768e-01
1.00233889e+00 -4.74628538e-01 6.92283750e-01 -6.33897483e-02
-7.32021987e-01 8.73073757e-01 2.73810118e-01 -3.17196637e-01
6.29732132e-01 -9.39810932e-01 7.70436466e-01 7.03795135e-01
2.21479431e-01 9.21330750e-01 4.65821445e-01 -7.66053140e-01
-1.22547448e+00 -1.42651463e+00 2.49414533e-01 -5.76406606e-02
6.01158679e-01 -3.63971710e-01 -1.15154588e+00 4.05695885e-01
-4.13236201e-01 2.88036138e-01 7.50455081e-01 3.62139702e-01
-6.91872001e-01 1.99027322e-02 -9.81389165e-01 6.68873370e-01
1.11662865e+00 -7.08940446e-01 -8.62456679e-01 1.76652923e-01
7.38413036e-01 -6.00943789e-02 -6.09660804e-01 1.62234530e-01
6.09906912e-01 -6.93111539e-01 9.40567255e-01 -6.74272478e-01
3.42827857e-01 1.44655369e-02 -1.43751830e-01 -1.82681656e+00
-6.29339755e-01 -4.87135708e-01 -6.97773218e-01 1.17383838e+00
3.58907849e-01 -4.43615079e-01 1.02400708e+00 4.57824379e-01
-4.38616961e-01 -8.86577904e-01 -8.04431379e-01 -1.15610850e+00
9.03106630e-02 -6.58360600e-01 5.56310833e-01 8.37284923e-01
1.56335719e-02 3.11334580e-01 -3.92308414e-01 -2.57726431e-01
7.68785834e-01 1.75514713e-01 8.45988512e-01 -1.16710842e+00
-6.45493627e-01 -4.10872489e-01 -1.71240628e-01 -1.12642634e+00
2.09460184e-01 -9.78811443e-01 3.17596614e-01 -1.35649300e+00
3.39299738e-01 -7.56209671e-01 -5.30705810e-01 4.38745469e-01
-5.92413425e-01 8.88542309e-02 1.72401905e-01 4.80095416e-01
-5.95084786e-01 8.13784659e-01 1.24206865e+00 -3.34470421e-01
-5.96240580e-01 2.84227818e-01 -9.07316744e-01 2.42401689e-01
6.66792929e-01 -5.98450303e-01 -8.57275903e-01 -2.11904407e-01
-3.21007401e-01 -1.96095794e-01 1.72063634e-01 -1.20756280e+00
3.70201617e-01 -8.07316899e-02 5.14477313e-01 -2.80541420e-01
5.62214613e-01 -3.22008461e-01 -8.05597380e-02 4.66677666e-01
-4.12477702e-01 -3.73277068e-01 4.79975753e-02 1.02465844e+00
5.24741262e-02 -5.15652776e-01 7.69593179e-01 -1.57042861e-01
-9.04330969e-01 5.98659456e-01 4.93215322e-02 2.48606414e-01
1.09843493e+00 -4.77907538e-01 -3.07028621e-01 -7.43955746e-03
-6.26417220e-01 1.66042417e-01 5.03016770e-01 4.83733833e-01
9.05815899e-01 -1.40344512e+00 -5.09759605e-01 4.54282045e-01
5.24230063e-01 -5.75850978e-02 6.14334345e-01 5.48342884e-01
2.50188351e-01 6.78835213e-02 -1.19368784e-01 -6.46746874e-01
-7.36726642e-01 7.74795890e-01 6.46702200e-02 -7.68712312e-02
-8.70081902e-01 9.88370776e-01 2.18784377e-01 -2.88222939e-01
1.36195794e-01 -1.73401251e-01 1.33270204e-01 3.96747142e-02
7.89942205e-01 4.45542634e-01 -3.64826098e-02 -1.04828134e-01
-3.49579342e-02 2.26772398e-01 -5.02424300e-01 -7.58153126e-02
1.38541794e+00 2.07625832e-02 7.28404224e-01 6.92017615e-01
1.41985238e+00 -5.48083842e-01 -1.82092917e+00 -5.03725648e-01
-2.23782331e-01 -4.55510050e-01 1.30196273e-01 -6.24569356e-01
-6.22430980e-01 9.77558076e-01 5.21824241e-01 -4.40512560e-02
7.60718524e-01 -1.99014798e-01 9.59413469e-01 5.43104231e-01
4.23721731e-01 -1.08346128e+00 4.93592113e-01 3.85426253e-01
5.61801851e-01 -1.28627288e+00 -1.59828514e-02 8.38700160e-02
-8.61468434e-01 9.56451178e-01 5.07752419e-01 -1.57270119e-01
6.42805696e-01 4.29631770e-03 -2.12944239e-01 1.34715261e-02
-9.28045034e-01 -8.38518068e-02 3.92729044e-01 8.53413641e-01
-3.50936912e-02 5.73693998e-02 3.85697514e-01 7.50128448e-01
1.53842062e-01 2.45858833e-01 1.35002375e-01 1.03286397e+00
-7.20412672e-01 -8.45681727e-01 7.22985789e-02 8.62566829e-01
2.22305298e-01 -2.11989552e-01 -3.41716036e-02 4.83214200e-01
-9.47451517e-02 7.47569740e-01 1.47937894e-01 -5.44849813e-01
4.45586056e-01 3.38121116e-01 4.94017988e-01 -1.05981791e+00
-3.05470765e-01 -1.44452199e-01 -3.67770493e-01 -2.65611976e-01
2.16018215e-01 -4.36758369e-01 -8.46788645e-01 -5.50105534e-02
-3.66840154e-01 2.04517782e-01 5.26975870e-01 1.02948833e+00
7.04908073e-01 6.22519195e-01 8.99771035e-01 -1.04260123e+00
-1.14157355e+00 -1.05416393e+00 -6.03667200e-01 3.99040282e-01
1.91810146e-01 -9.20168757e-01 -3.50528270e-01 -2.76392132e-01] | [9.993277549743652, 2.9673967361450195] |
d4d837b8-8914-43af-9b92-1e237f923798 | impact-of-content-features-for-automatic | 1704.03289 | null | http://arxiv.org/abs/1704.03289v1 | http://arxiv.org/pdf/1704.03289v1.pdf | Impact Of Content Features For Automatic Online Abuse Detection | Online communities have gained considerable importance in recent years due to
the increasing number of people connected to the Internet. Moderating user
content in online communities is mainly performed manually, and reducing the
workload through automatic methods is of great financial interest for community
maintainers. Often, the industry uses basic approaches such as bad words
filtering and regular expression matching to assist the moderators. In this
article, we consider the task of automatically determining if a message is
abusive. This task is complex since messages are written in a non-standardized
way, including spelling errors, abbreviations, community-specific codes...
First, we evaluate the system that we propose using standard features of online
messages. Then, we evaluate the impact of the addition of pre-processing
strategies, as well as original specific features developed for the community
of an online in-browser strategy game. We finally propose to analyze the
usefulness of this wide range of features using feature selection. This work
can lead to two possible applications: 1) automatically flag potentially
abusive messages to draw the moderator's attention on a narrow subset of
messages ; and 2) fully automate the moderation process by deciding whether a
message is abusive without any human intervention. | ['Linares Georges LIA', 'Dufour Richard LIA', 'Labatut Vincent LIA', 'Papegnies Etienne LIA'] | 2017-04-11 | null | null | null | null | ['abuse-detection'] | ['natural-language-processing'] | [ 0.21221358 -0.26755 0.20950758 -0.10596668 -0.44514272 -0.68792635
0.67683786 0.67835885 -0.65040433 0.44812712 0.17784023 -0.44429982
0.02727143 -0.7722462 0.10149755 -0.25157046 -0.11107598 0.2451156
0.5518023 -0.53860694 0.7909761 0.38208765 -1.6993729 0.3889654
1.026556 0.72858 0.06490244 0.69963175 -0.3915815 0.8732516
-0.9610121 -0.68498874 0.05956836 -0.6690669 -0.7794369 0.30368868
0.29393196 -0.32790864 0.09624135 1.2513047 0.37800306 -0.04023534
0.3368435 -1.2025065 0.04708555 0.8558346 -0.47735417 0.4412809
0.60118884 0.29958957 1.0226829 -0.6128026 0.8052672 1.097119
0.54295886 0.20836268 -1.2555214 -0.7853087 0.06534819 0.4906641
-1.487472 -0.43610588 0.7388627 -0.87469816 0.61670375 0.5464766
0.6822339 0.9274236 -0.13142258 0.3737412 1.0855542 -0.5181074
0.20782112 0.49338192 0.36091903 0.5046381 0.29244527 -0.627974
-0.5691438 -0.6092723 0.05681602 -0.23740487 -0.36294907 -0.07197527
-0.7361653 1.1565576 -0.2169207 0.5968824 -0.18775403 -0.23500115
0.83065027 0.47773418 0.49080446 0.6638592 0.02015868 -0.565502
-0.9499982 0.3831786 1.2309738 0.59632546 0.7826093 -0.48871845
0.05708029 1.0043617 0.1977757 0.02030371 0.57245463 -0.78574216
0.40004843 0.86671364 0.05985568 -1.5263008 -0.13058129 -0.11001725
-0.49577695 0.24998759 0.67742866 -0.15405089 0.01404023 1.3222497
0.26372677 -0.3940786 -0.6617755 0.6277721 0.554042 0.32587174
0.05440837 -0.52087 1.449367 -0.40562505 -0.8347384 -0.1520642
0.6025936 -1.1013199 1.1409342 0.57163024 -0.7740182 -0.08142211
-0.9645936 0.12822439 -0.41826022 -0.3632431 0.21353306 1.0898777
-0.70860505 0.7504016 -0.18046936 -0.61982316 -0.09484427 0.2666088
-0.24591777 0.34907487 -1.4030192 0.80996644 0.11863831 -0.2838771
-0.30093595 -0.4373473 -0.67210203 0.13340062 0.8049842 -0.18925522
1.2430735 -1.2388501 -1.3266048 0.87449133 0.02550562 -0.22550519
0.75843173 -0.21918091 -0.31703717 0.2048364 0.23665461 -0.17307915
0.884726 -0.88432074 -0.8136119 -0.28629452 0.14133641 0.01068151
-0.7385314 0.6061882 -0.28087524 -0.56556016 -0.16706552 -0.75668496
-0.13772441 -0.12438823 -0.3210017 -0.27173832 0.76286185 -1.0961366
1.9476706 -2.0679343 -0.09455788 0.60747 0.5007619 0.31736648
0.14438167 0.6438849 0.157542 0.62296164 -0.06729379 -0.07817551
-0.15737587 -0.31782365 0.09547818 0.39294735 -0.01205536 0.13930131
-0.82254964 -0.7831307 -0.09320147 0.1449841 -0.64685524 0.2836603
-0.08160026 0.07379146 -0.2473077 0.2707105 0.3751918 -0.20357332
0.54138637 0.1860363 -0.3348154 0.4999408 -1.3740392 1.0051222
-0.43546346 0.80357206 0.32648313 -0.55158186 0.904306 0.05216303
0.30964097 -0.33898935 0.39300874 0.20662965 0.16316055 -0.71559083
0.6937835 0.05726627 -0.12404456 0.77268034 -0.28806823 0.13822268
0.79660386 0.62483895 1.2176286 -0.44541174 0.83140624 -0.08826312
0.90564185 -0.2262214 0.30931348 0.69971186 -0.43733853 0.29897058
0.9702302 0.02249088 -0.7686229 -0.25331244 0.1817657 1.3396474
-0.02264856 -1.0354307 -0.8372309 -0.71884745 -0.05809738 0.52402586
-0.48660165 -0.14896128 -0.5463666 -0.43457022 0.47729248 -0.20657729
0.40291476 -0.8938376 -0.7082473 0.28494024 -0.6451345 -0.74509656
-0.595683 -0.05285582 -0.4842039 -1.3203386 -0.23768288 -0.29448238
0.15726548 0.4204339 1.0456846 0.7256721 -0.10355532 0.21338077
-0.70088536 -0.0440635 -0.80745274 0.2722266 -0.20366605 0.22876869
0.6607233 -0.59108245 -0.29112065 0.348517 -0.83959997 -0.1991822
0.25839007 0.5383639 -0.18260537 -0.09705493 0.3137945 -1.3346143
1.1599609 -0.6495288 -0.35282516 -0.08522562 -0.65973073 -0.2495111
0.5988671 -0.56042904 -0.84268767 -0.20961152 -0.35650975 0.28242972
-0.18541305 0.58681285 -0.33531624 -0.09351824 0.91596645 -0.09869904
0.23547195 -0.6242189 -0.01196952 1.1117958 -0.1485798 -0.21487561
0.8664868 0.12372505 -0.45581037 -1.0961477 -0.45624506 -0.9375873
-0.4424366 -0.52436864 0.38043106 -0.4500057 -0.94657874 0.29447982
-1.2013093 -0.06321527 0.24533473 -0.03057439 -0.2502051 1.0344214
-0.6281042 -0.89822227 -0.29678476 -0.96277404 0.36458156 0.05391085
-0.88578624 -0.5841856 0.16299495 0.49431586 0.4562805 -0.03724314
0.82439846 -1.0256523 -0.1940432 -0.28104314 -0.13436957 0.2845905
0.08496255 0.33600116 -0.66711175 -0.1565233 -0.08599541 -0.02559689
0.52041686 -0.27655718 0.9067293 -0.49787033 -0.10651956 0.02112668
1.0694784 0.06867112 0.48924625 0.6222437 0.40511918 0.9232277
0.5759456 0.9058957 0.08502629 0.734146 0.43602353 0.45557135
0.14016476 -0.14483717 0.4118978 0.722005 0.02778072 -0.06622917
-0.8965334 0.39221916 -1.7242957 -1.1621542 -0.5340269 2.1698596
1.0159935 0.36434844 0.6006194 0.58012426 1.0253453 0.05784515
0.09363231 -0.50015795 0.17070995 -0.08180261 0.14417173 0.61596507
-0.95839477 0.54561234 5.8060384 0.8678317 -1.1975943 0.15616716
0.53086925 0.06602848 -0.14118302 0.06328822 -0.7043256 0.776167
0.7683213 -0.31240836 0.39725167 0.9451578 0.54075825 -0.39116982
-0.9773983 0.88707775 0.3905206 -1.0483456 -0.10415787 0.1773036
0.22250766 -0.6111115 -0.29289216 0.3698681 -0.11285307 -0.5992417
0.7985218 0.12413137 0.39396006 -0.7035765 0.6992477 0.4700084
-0.831555 -0.04731705 0.01529041 -0.4673463 0.07093622 0.7594972
-0.92572945 -0.07992612 0.60451746 0.29403293 -0.99780434 1.311168
-0.23281643 0.82046866 -0.27656266 -0.58628047 -0.2138786 -0.18552002
0.7139896 1.4841218 0.09856136 -0.12351277 0.08759963 0.53523755
0.02929583 0.8777018 -0.39038467 -0.21379691 0.45285466 1.3888257
-0.8567764 -0.41460025 -0.3357683 0.9238372 0.45149 0.08911298
-0.5691828 -0.7000233 0.7553623 0.8092419 0.23633926 -0.05525111
-0.33978498 -0.94955873 -0.05217617 -1.4290035 0.46775988 -0.33048368
-1.2726692 0.47177294 -0.2506403 -1.0244551 -0.4071164 -0.2802525
-0.6771715 0.7314014 -0.99911803 -0.47494647 -0.36227635 0.35297683
0.5224444 0.0978641 0.49416688 0.62070084 -0.50610393 0.5029745
-0.19484983 0.246269 1.0617913 -1.0809848 0.01538164 0.96321356
-0.12353076 0.91079926 1.1611109 -0.7715977 -0.79907525 -0.5662795
1.4961486 -0.19524634 1.0569043 -0.5128695 -1.0569744 0.27011096
0.21420385 -0.6488386 0.94145715 0.48915294 -0.44392675 0.06527323
-0.99660873 0.8173319 0.8182987 -0.44721967 -0.4718497 0.2150139
0.17372271 0.09858856 -0.4831735 -0.26687384 0.39272082 -1.2007322
0.3562226 -0.5241971 0.5181525 -0.30639386 0.25442067 -1.2142738
-0.35276613 -0.9218447 0.0842289 1.4265155 0.41524193 -0.3197433
0.5948959 0.59262735 0.45811406 -0.20432493 -0.6611437 -0.337703
-0.30554184 -0.38397273 0.24896048 1.0025237 0.691696 0.7430268
-0.47964615 -0.3018439 0.2886791 -0.09125401 1.0867563 -1.4835223
-0.4051757 -0.69326955 -0.44283083 -0.81192315 -0.10421774 -0.5372456
0.028484 -0.98191905 0.4448017 -0.25272235 0.34629533 0.12547098
-0.39746714 0.3734474 0.39447865 0.14560154 -0.7488279 -0.01670454
0.6695135 -0.08421706 -0.3335838 0.20695837 -0.84906304 0.7971907
0.7366223 -0.49196923 0.03571072 0.3109949 0.7299982 -0.26726398
0.02517433 -0.6234285 0.17877556 -0.23962902 -0.2180435 -0.16093332
-0.07699476 -0.699006 -0.07622615 0.54179937 -0.25119534 -0.08286629
-0.16564506 0.42295715 -0.21074353 -0.8345385 0.80624807 -0.29644006
-0.64496565 -0.35019872 -1.2860582 0.0849338 0.95122415 -0.3164957
-0.14255193 -0.8555609 -0.66723174 -0.07193419 0.5702327 0.36637154
0.24272938 -0.7527187 -0.6469988 -0.04384531 0.308072 -0.712984
0.04847871 0.9199353 -0.7754094 -0.13594736 -0.13187434 -0.3133066
-1.9684753 0.58734006 0.05241119 -0.34306854 -0.30516902 0.5322466
-0.30738306 0.10715211 0.3647489 0.03549741 -0.684111 0.6635547
0.8653832 0.6940566 0.28303355 -0.6577596 -0.31825683 -0.03894857
-0.25950614 -0.18973425 1.0604546 -0.33442232 -0.3950222 0.45524392
1.0078211 0.64272106 -0.5115419 -0.13176334 0.4277532 -0.83865976
-0.22761461 -0.5738249 -0.6886012 0.69046736 0.18747173 0.8508362
0.8405499 -0.14702445 0.52284014 0.14849524 0.2013315 -1.2465119
0.21122018 0.61812127 0.7270634 -1.1439077 0.2187854 -0.72020334
-0.5478073 1.1478847 0.5771265 0.10507172 0.65077627 0.09020336
0.19199531 -0.04664173 -0.701979 -0.24138874 -0.04640716 0.4287906
0.73998886 -0.11908461 -1.2748501 0.7074002 -0.19209647 -0.31956375
1.0155162 0.822464 -0.7886545 -1.1193898 -0.6394612 0.5892426
-0.60398316 -0.15425001 -1.0101173 0.56520116 0.25544178 1.2706661
-0.15567864 -0.43366098 0.38418347 0.12843917 -0.04245347 -0.762594
-1.2181255 0.03053857 0.55126673 -0.21659537 -0.16046251 -1.0209044
-0.71356636 -0.76822335 -0.4489758 0.3615572 0.5406966 0.9578223
0.16179961 -0.00687209 0.5755567 -0.5459077 -0.68203896 -1.2870996
-0.39480427 0.82265747 -0.05058919 -0.55543095 -0.68874687 0.11228392] | [8.570852279663086, 10.33877944946289] |
368413c1-e1db-4ba4-b80e-19d3a9ca5587 | cuts-neural-causal-discovery-from-irregular | 2302.07458 | null | https://arxiv.org/abs/2302.07458v1 | https://arxiv.org/pdf/2302.07458v1.pdf | CUTS: Neural Causal Discovery from Irregular Time-Series Data | Causal discovery from time-series data has been a central task in machine learning. Recently, Granger causality inference is gaining momentum due to its good explainability and high compatibility with emerging deep neural networks. However, most existing methods assume structured input data and degenerate greatly when encountering data with randomly missing entries or non-uniform sampling frequencies, which hampers their applications in real scenarios. To address this issue, here we present CUTS, a neural Granger causal discovery algorithm to jointly impute unobserved data points and build causal graphs, via plugging in two mutually boosting modules in an iterative framework: (i) Latent data prediction stage: designs a Delayed Supervision Graph Neural Network (DSGNN) to hallucinate and register unstructured data which might be of high dimension and with complex distribution; (ii) Causal graph fitting stage: builds a causal adjacency matrix with imputed data under sparse penalty. Experiments show that CUTS effectively infers causal graphs from unstructured time-series data, with significantly superior performance to existing methods. Our approach constitutes a promising step towards applying causal discovery to real applications with non-ideal observations. | ['Qionghai Dai', 'Kunlun He', 'Jinli Suo', 'Zongren Li', 'Tingxiong Xiao', 'Runzhao Yang', 'Yuxiao Cheng'] | 2023-02-15 | null | null | null | null | ['causal-discovery', 'irregular-time-series'] | ['knowledge-base', 'time-series'] | [ 3.56383175e-01 4.43880111e-01 -3.85643065e-01 -3.57192785e-01
-2.29711220e-01 -2.52218634e-01 7.42890120e-01 2.32083928e-02
2.95080841e-01 1.13230515e+00 6.48670912e-01 -6.55269921e-01
-5.59698761e-01 -9.72911716e-01 -8.97217512e-01 -6.59794271e-01
-6.80415571e-01 7.32260048e-01 -2.39892215e-01 2.86543727e-01
-5.50026149e-02 2.24281117e-01 -1.08324027e+00 6.74547255e-02
1.09204113e+00 4.20751542e-01 -1.24445781e-01 3.45474929e-01
1.39643773e-01 1.18451345e+00 -1.12619333e-01 -4.23628181e-01
-9.20151621e-02 -4.60132569e-01 -7.28922725e-01 -3.57271969e-01
4.12831530e-02 -2.06399739e-01 -5.25230050e-01 9.82642114e-01
3.03367764e-01 -2.34444037e-01 5.93047380e-01 -1.58566678e+00
-1.06184471e+00 1.31972420e+00 -7.82425046e-01 2.39762545e-01
2.76714265e-01 -1.86243147e-01 1.17205513e+00 -6.32449925e-01
4.59516048e-01 1.58481300e+00 7.92779446e-01 9.08599198e-02
-1.74183393e+00 -9.17180419e-01 1.58218518e-01 3.31852823e-01
-1.03712797e+00 -1.54658467e-01 9.31203783e-01 -5.69202483e-01
6.11706138e-01 1.06282137e-01 6.73655868e-01 1.58499396e+00
4.22164023e-01 6.49103880e-01 8.89436603e-01 7.37515166e-02
3.98038507e-01 -7.02314198e-01 1.69375315e-01 3.96470487e-01
2.02558368e-01 5.44518948e-01 -7.11002052e-01 -5.68144917e-01
1.07954443e+00 2.00121701e-01 -2.55101681e-01 -4.43574756e-01
-1.47553623e+00 1.15564573e+00 6.96369767e-01 -8.69053155e-02
-7.36103892e-01 4.77029741e-01 3.42124075e-01 3.59477818e-01
6.04106843e-01 4.52071168e-02 -3.19804639e-01 2.70254493e-01
-7.07894206e-01 2.89499700e-01 7.28989959e-01 6.25120163e-01
2.14760065e-01 3.04564059e-01 9.69506279e-02 5.99530339e-01
4.41465288e-01 4.36704546e-01 1.36446625e-01 -6.33335948e-01
2.95639634e-01 5.32080650e-01 -1.88756421e-01 -1.28257501e+00
-5.38625360e-01 -5.26628733e-01 -1.66463077e+00 -2.48509958e-01
4.18665290e-01 -2.84637809e-01 -7.81724274e-01 1.98112226e+00
4.26832348e-01 7.43129969e-01 -1.31990135e-01 1.16466689e+00
7.96094060e-01 5.46246886e-01 1.82597905e-01 -5.40520906e-01
1.13890922e+00 -2.21852988e-01 -9.18749571e-01 -2.23814882e-02
5.74093722e-02 -4.12800431e-01 7.04621553e-01 4.04223531e-01
-7.34985352e-01 -2.06558660e-01 -7.32971907e-01 1.97817445e-01
5.97073846e-02 -3.06557685e-01 1.27906382e+00 2.03069806e-01
-9.70043004e-01 5.83874702e-01 -8.04177761e-01 -1.37595668e-01
4.28317755e-01 4.88753110e-01 -3.79510224e-01 -1.00117192e-01
-1.68244457e+00 2.16293961e-01 3.99976581e-01 3.17481816e-01
-1.15702093e+00 -8.55663955e-01 -6.50787652e-01 7.02043101e-02
5.83455086e-01 -1.08563650e+00 8.10221493e-01 -7.15764642e-01
-8.76191139e-01 2.79823095e-01 -4.79947738e-02 -6.34436786e-01
4.26199764e-01 -1.97140768e-01 -6.36876404e-01 -4.62574452e-01
4.15717930e-01 1.61446080e-01 9.92332578e-01 -1.16989684e+00
-1.97881296e-01 -6.61449075e-01 -2.83302397e-01 -2.70561844e-01
2.32429937e-01 -2.85764188e-01 1.57390609e-02 -8.67283165e-01
3.81634951e-01 -6.87296987e-01 -4.03991073e-01 -5.82420230e-01
-9.75101709e-01 -4.21700686e-01 6.86105430e-01 -6.06110692e-01
1.05887091e+00 -1.77346349e+00 2.01392934e-01 2.16458902e-01
8.63074780e-01 -2.27083191e-01 -6.64963052e-02 6.72920823e-01
-8.12960148e-01 -6.62927255e-02 -3.79718274e-01 2.59241667e-02
-9.69371647e-02 1.83121637e-01 -7.33624518e-01 5.21161437e-01
2.99608320e-01 9.94289935e-01 -1.30870414e+00 -1.64021760e-01
-8.05247054e-02 4.30528253e-01 -5.74555218e-01 2.30743423e-01
-2.87691295e-01 7.77256668e-01 -3.57727885e-01 3.90865594e-01
4.85054433e-01 -7.20906138e-01 6.25294089e-01 -7.72210509e-02
7.96108916e-02 3.65600675e-01 -1.38045073e+00 1.49632108e+00
-6.71453625e-02 5.47857821e-01 -6.90738410e-02 -1.45143580e+00
9.17590022e-01 6.24567509e-01 5.91272652e-01 -5.36770046e-01
6.13969341e-02 -6.90935180e-02 6.07024021e-02 -5.27543187e-01
3.71311456e-02 -2.51005083e-01 -4.76414636e-02 4.97522861e-01
4.61460724e-02 4.66964543e-01 -1.79525644e-01 5.18429518e-01
1.62202203e+00 -5.29672429e-02 2.85993844e-01 -2.86837250e-01
-1.30148306e-01 -1.13909028e-01 7.61127770e-01 7.56132841e-01
1.45258278e-01 3.85182291e-01 1.09555590e+00 -7.68434346e-01
-9.57885623e-01 -1.40823221e+00 -1.08852632e-01 7.03763783e-01
-1.53558940e-01 -2.94658821e-02 -3.66454601e-01 -6.41656339e-01
1.98070973e-01 7.38877833e-01 -8.58278275e-01 -1.40311942e-01
-4.50171471e-01 -1.28603423e+00 3.40832740e-01 5.25795937e-01
8.00239444e-02 -1.16897368e+00 -3.42298709e-02 4.38831687e-01
-3.34440559e-01 -6.82023764e-01 -1.27195179e-01 5.02732277e-01
-9.53034937e-01 -1.20237911e+00 -2.64760435e-01 -3.83515894e-01
6.12693965e-01 1.02189638e-01 1.23135042e+00 -2.20124051e-01
-2.74991900e-01 -1.67001128e-01 -2.07369596e-01 -1.86057270e-01
-3.43217731e-01 -2.57619292e-01 2.25010946e-01 2.11474791e-01
4.21863645e-01 -1.23361313e+00 -5.27780712e-01 -2.22222097e-02
-8.02025139e-01 2.62171268e-01 6.07961357e-01 1.15479982e+00
2.46902570e-01 2.74200916e-01 9.84638095e-01 -1.04758835e+00
8.26172292e-01 -1.18183279e+00 -6.34434044e-01 -1.17346697e-01
-6.90983057e-01 2.14516923e-01 6.64615810e-01 -3.64239335e-01
-9.77744460e-01 1.10907033e-01 2.29155287e-01 -5.26375771e-01
-2.13615149e-01 1.05561686e+00 -2.77692050e-01 8.90554130e-01
6.55069411e-01 -2.08187759e-01 2.15857141e-02 -3.27739120e-01
5.63831866e-01 1.99720636e-01 7.53936231e-01 -4.37015027e-01
7.05279768e-01 4.25547689e-01 2.10653111e-01 -3.06876183e-01
-7.43447542e-01 -1.83702871e-01 -7.63155401e-01 -9.42912605e-03
7.75493026e-01 -1.06216669e+00 -1.04518747e+00 1.48930460e-01
-1.34906900e+00 -3.44362468e-01 2.24699080e-02 6.89037085e-01
-3.40491354e-01 5.96203469e-03 -6.61952674e-01 -8.83380532e-01
-1.20197363e-01 -6.93551064e-01 9.78981793e-01 -3.18840861e-01
-4.72064286e-01 -1.21005273e+00 2.88585842e-01 1.68128207e-01
-1.69391319e-01 6.05515599e-01 1.13976407e+00 -4.77965683e-01
-6.08952761e-01 -8.67570564e-02 -6.57553613e-01 -4.36753511e-01
1.91542894e-01 -6.78572506e-02 -8.85861456e-01 -1.58850383e-02
-3.60776305e-01 -1.07520230e-01 8.96688104e-01 8.98574471e-01
1.16709614e+00 -5.87899685e-01 -5.36591887e-01 4.29775625e-01
1.17904389e+00 1.38944507e-01 4.98208880e-01 -2.21376508e-01
1.06898916e+00 7.69929767e-01 2.06774563e-01 5.75119615e-01
5.75806022e-01 3.18564296e-01 9.12472010e-01 -2.92927176e-01
7.97816291e-02 -6.30118906e-01 -4.37742705e-03 6.91327751e-01
-2.15478227e-01 -3.98783296e-01 -1.20728827e+00 7.83088565e-01
-2.41579366e+00 -1.23723006e+00 -1.12942564e+00 2.04379988e+00
8.14094424e-01 -1.33076357e-02 1.90464154e-01 4.50187512e-02
9.06324804e-01 -3.57447304e-02 -7.43374050e-01 1.62889138e-01
-1.18222475e-01 1.38256839e-02 4.45287347e-01 3.35844427e-01
-1.11403465e+00 7.19729722e-01 6.12113762e+00 5.01369178e-01
-9.43291187e-01 1.68033943e-01 6.57076597e-01 7.63950720e-02
-7.27148950e-01 2.74573594e-01 -1.81257829e-01 5.25584519e-01
8.63252401e-01 -1.25076380e-02 4.58025187e-01 3.86681437e-01
6.68463051e-01 3.21312636e-01 -1.23618710e+00 8.57698619e-01
-5.09591401e-01 -1.48807740e+00 5.35691194e-02 2.08982274e-01
7.37136960e-01 1.27534121e-01 -1.78399056e-01 -3.27126458e-02
1.40141499e+00 -1.36675060e+00 4.65875119e-01 7.31504440e-01
6.24605775e-01 -7.52910078e-01 6.43833876e-01 3.66187304e-01
-8.84035051e-01 -1.02852985e-01 -2.92561829e-01 -4.93949533e-01
5.12624502e-01 1.49667275e+00 -1.00302660e+00 7.01686919e-01
7.22297072e-01 1.08148372e+00 -1.96096697e-03 8.65422785e-01
-4.43919957e-01 1.18440425e+00 -5.88729121e-02 1.18641280e-01
-1.58041939e-01 -2.94591755e-01 6.22846007e-01 6.74008012e-01
2.58138567e-01 1.80944934e-01 8.61682966e-02 1.31680465e+00
-1.10320617e-02 -3.17383915e-01 -8.97725701e-01 -1.04693078e-01
5.60477197e-01 9.43661392e-01 -7.81749785e-01 -1.88173652e-01
-2.94219822e-01 7.00808942e-01 5.33675730e-01 5.91193736e-01
-1.07863700e+00 2.01644883e-01 3.95279348e-01 2.12041870e-01
-1.70662716e-01 -1.75446317e-01 -5.56916714e-01 -1.13107729e+00
-2.66721785e-01 -6.68532610e-01 6.92986310e-01 -7.31049120e-01
-1.83086419e+00 2.50471175e-01 -6.85217679e-02 -1.01900339e+00
-3.18987817e-01 -4.93969908e-03 -7.42801368e-01 7.89978445e-01
-1.08469033e+00 -1.24724054e+00 -1.79104373e-01 6.81723356e-01
2.62133539e-01 -5.83389290e-02 6.81134760e-01 4.20076936e-01
-7.32104540e-01 1.36057958e-01 -2.57853754e-02 1.58659011e-01
6.15440011e-01 -1.48963034e+00 7.14767873e-01 8.69685411e-01
2.10810900e-01 9.03459370e-01 8.78957450e-01 -1.27432060e+00
-1.47805810e+00 -1.34004450e+00 8.91693234e-01 -2.97794580e-01
1.26286113e+00 -5.53214848e-01 -1.09959650e+00 8.50611806e-01
1.57611981e-01 -8.15884545e-02 5.84752202e-01 7.66298890e-01
-3.52385074e-01 1.20179310e-01 -6.57710373e-01 6.53957129e-01
1.32428217e+00 -2.95388788e-01 -5.94284236e-01 3.74140352e-01
9.22253251e-01 1.28216157e-02 -8.67449522e-01 4.81096089e-01
3.83585751e-01 -9.04758811e-01 1.00159693e+00 -7.19286740e-01
8.45591664e-01 -3.18519324e-01 2.19743595e-01 -1.59360266e+00
-7.66588211e-01 -7.91499913e-01 -3.37532520e-01 1.20419812e+00
3.48908752e-01 -7.14663684e-01 6.57174349e-01 4.39527690e-01
4.60206158e-02 -1.98741287e-01 -8.18352818e-01 -4.22738075e-01
-1.86993092e-01 -6.98947549e-01 8.21389139e-01 1.69119477e+00
8.96183699e-02 8.94462943e-01 -8.15315068e-01 5.29264212e-01
1.14785206e+00 2.53306270e-01 6.90527558e-01 -1.77520144e+00
-2.98178196e-01 -2.62646437e-01 -1.85544655e-01 -7.81253278e-01
2.50163645e-01 -9.68730211e-01 1.00486930e-02 -1.45012426e+00
3.55149955e-01 -7.06128240e-01 -1.26323342e-01 7.43155122e-01
-3.63610953e-01 -3.72740068e-02 -5.01834929e-01 3.06735456e-01
-1.62892655e-01 6.85385346e-01 1.08564484e+00 -1.38049990e-01
-2.17537612e-01 -8.80543701e-03 -6.22360408e-01 6.54517591e-01
6.48276985e-01 -6.11656785e-01 -7.66187191e-01 -2.72427648e-01
6.09662056e-01 8.43280196e-01 1.08658397e+00 -4.64370787e-01
2.82028407e-01 -3.30567718e-01 3.79503667e-01 -6.53449297e-01
-2.23402694e-01 -6.88352287e-01 8.16151381e-01 3.39570731e-01
-3.44610333e-01 4.53038737e-02 -6.27443194e-02 1.02676153e+00
-2.17305765e-01 4.62899894e-01 1.44193724e-01 9.18364972e-02
-4.34028178e-01 4.81678933e-01 -2.62253553e-01 -1.51989758e-01
6.60739183e-01 1.79383829e-01 -3.50401580e-01 -4.92022187e-01
-1.02660978e+00 2.94154525e-01 -9.00780857e-02 6.53352678e-01
5.57236433e-01 -1.58834374e+00 -9.84589577e-01 3.77186418e-01
-1.15399130e-01 1.04473099e-01 4.00574356e-01 1.08828735e+00
5.81485890e-02 3.72369498e-01 1.64082050e-01 -6.97051167e-01
-8.08942795e-01 8.78638327e-01 -2.07604244e-01 -2.43660375e-01
-8.65718305e-01 7.55525410e-01 6.39461458e-01 -5.51551640e-01
4.94214892e-02 -1.43227503e-01 -2.13620812e-01 2.34604627e-01
2.51513243e-01 2.08920211e-01 -1.03302285e-01 -1.97833836e-01
-1.17859416e-01 -2.50322700e-01 2.91049153e-01 1.74918354e-01
1.68587327e+00 -1.67635083e-02 -5.64524949e-01 6.98474050e-01
8.61756444e-01 -3.09635133e-01 -1.24814224e+00 -2.02461764e-01
9.76071432e-02 -2.43397459e-01 9.23701301e-02 -7.39242852e-01
-1.10209775e+00 6.46343887e-01 2.07656011e-01 7.70250738e-01
9.70511496e-01 1.86326727e-01 5.04145920e-01 -1.75095409e-01
4.18051600e-01 -4.38780308e-01 -1.23687036e-01 3.48067820e-01
1.02885771e+00 -1.22057176e+00 -1.65956929e-01 -4.41187441e-01
-2.01508731e-01 8.36800218e-01 2.17636928e-01 -2.01413766e-01
7.26609230e-01 2.95751274e-01 -2.24567235e-01 -7.29537666e-01
-1.09354591e+00 9.88793299e-02 3.96722317e-01 6.11255407e-01
5.70676565e-01 4.65898335e-01 -2.56067127e-01 6.46420360e-01
-3.40783000e-01 6.54873401e-02 4.29825008e-01 1.10359155e-01
1.68912351e-01 -8.07312012e-01 -5.17763138e-01 7.65134752e-01
-2.89519578e-01 -4.18388098e-01 -3.27912301e-01 7.05186307e-01
-1.07910767e-01 1.02994156e+00 1.33627757e-01 -1.48273155e-01
1.89677831e-02 -2.64471978e-01 -7.75244310e-02 -4.07154799e-01
-4.37297039e-02 2.66237348e-01 1.60706356e-01 -6.57124519e-01
-3.71384084e-01 -9.02444780e-01 -1.31415761e+00 -5.56998551e-01
-1.82719648e-01 1.71074495e-01 1.90125033e-01 1.01175582e+00
3.81966740e-01 9.54920292e-01 4.24247056e-01 -6.09203398e-01
-1.91762820e-01 -8.49499524e-01 -7.17659831e-01 3.62473011e-01
5.32793760e-01 -8.40596020e-01 -3.47128093e-01 3.19307894e-01] | [7.841884136199951, 5.2841691970825195] |
554ad404-c9f2-42f5-9925-e4cbdf386984 | identity-preserving-talking-face-generation | 2305.08293 | null | https://arxiv.org/abs/2305.08293v1 | https://arxiv.org/pdf/2305.08293v1.pdf | Identity-Preserving Talking Face Generation with Landmark and Appearance Priors | Generating talking face videos from audio attracts lots of research interest. A few person-specific methods can generate vivid videos but require the target speaker's videos for training or fine-tuning. Existing person-generic methods have difficulty in generating realistic and lip-synced videos while preserving identity information. To tackle this problem, we propose a two-stage framework consisting of audio-to-landmark generation and landmark-to-video rendering procedures. First, we devise a novel Transformer-based landmark generator to infer lip and jaw landmarks from the audio. Prior landmark characteristics of the speaker's face are employed to make the generated landmarks coincide with the facial outline of the speaker. Then, a video rendering model is built to translate the generated landmarks into face images. During this stage, prior appearance information is extracted from the lower-half occluded target face and static reference images, which helps generate realistic and identity-preserving visual content. For effectively exploring the prior information of static reference images, we align static reference images with the target face's pose and expression based on motion fields. Moreover, auditory features are reused to guarantee that the generated face images are well synchronized with the audio. Extensive experiments demonstrate that our method can produce more realistic, lip-synced, and identity-preserving videos than existing person-generic talking face generation methods. | ['Guanbin Li', 'Liang Lin', 'Gangming Zhao', 'Pengxu Wei', 'Yinqi Cai', 'Chaowei Fang', 'Weizhi Zhong'] | 2023-05-15 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Zhong_Identity-Preserving_Talking_Face_Generation_With_Landmark_and_Appearance_Priors_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Zhong_Identity-Preserving_Talking_Face_Generation_With_Landmark_and_Appearance_Priors_CVPR_2023_paper.pdf | cvpr-2023-1 | ['talking-face-generation', 'face-generation'] | ['computer-vision', 'computer-vision'] | [ 1.76649407e-01 -9.12829563e-02 3.10753882e-02 -3.86385441e-01
-8.83037567e-01 -4.54856515e-01 4.72284466e-01 -6.53546393e-01
1.34112924e-01 4.73357856e-01 4.30383086e-01 5.36097944e-01
3.21758509e-01 -5.13594508e-01 -5.43588758e-01 -7.83764243e-01
2.18022645e-01 -4.51641381e-02 -1.21911746e-02 -5.44673316e-02
2.26352975e-01 4.62834209e-01 -1.90887070e+00 2.43174776e-01
5.13123930e-01 1.06684339e+00 2.66178280e-01 6.27080083e-01
-8.50862265e-02 2.25383058e-01 -7.35099792e-01 -4.09995645e-01
2.19107196e-01 -5.17431498e-01 -3.92198414e-01 4.04912561e-01
5.72251916e-01 -5.21553040e-01 -3.18094879e-01 9.87697899e-01
7.83810258e-01 1.84359640e-01 4.36016291e-01 -1.57710159e+00
-3.74792963e-01 2.38691956e-01 -7.03717232e-01 -3.62982810e-01
9.38620389e-01 1.88025147e-01 4.75856841e-01 -1.00373840e+00
7.76791394e-01 1.58791280e+00 5.62583625e-01 8.31346214e-01
-1.07607448e+00 -1.06435978e+00 1.59776881e-01 1.45184025e-01
-1.82909274e+00 -1.12990582e+00 1.20486104e+00 -3.08421940e-01
-2.17362158e-02 3.91250283e-01 8.04115295e-01 1.13394094e+00
-2.34682322e-01 4.32129771e-01 6.35059357e-01 -4.28464979e-01
-8.31857845e-02 2.90906549e-01 -7.88610697e-01 6.12857401e-01
-3.59121859e-01 -3.35424803e-02 -8.70832205e-01 -1.30649462e-01
9.92543280e-01 -2.52340466e-01 -7.53018975e-01 -5.02092242e-01
-1.25299430e+00 4.98540491e-01 -8.54200646e-02 3.15968782e-01
-2.78516173e-01 1.07574545e-01 3.11367154e-01 -1.83867157e-01
2.42367491e-01 -7.01154023e-02 2.48184144e-01 -1.07977286e-01
-1.19612181e+00 8.05975497e-02 5.40371597e-01 1.08179879e+00
6.83439791e-01 1.89403608e-01 -3.26792955e-01 9.26106334e-01
4.82979923e-01 6.24141634e-01 4.33971196e-01 -1.21563113e+00
2.03762501e-01 2.67500490e-01 2.31144637e-01 -1.34473932e+00
8.52138549e-02 -5.20895980e-02 -6.52164757e-01 -1.02702998e-01
1.58225864e-01 1.75557972e-03 -4.58841264e-01 1.90215480e+00
7.42915154e-01 6.22969031e-01 -4.25459705e-02 9.12814021e-01
9.56857920e-01 7.53311872e-01 -1.16356559e-01 -6.34765506e-01
1.36903048e+00 -6.36558592e-01 -1.08903241e+00 9.63498652e-02
-4.39587422e-02 -9.29755092e-01 1.08137047e+00 1.37983099e-01
-1.17398059e+00 -9.28634048e-01 -7.86419034e-01 5.38920378e-03
2.75355130e-01 4.52240944e-01 1.53375417e-01 8.33580077e-01
-1.16305566e+00 4.24881931e-03 -2.60586977e-01 -4.51309457e-02
2.00857684e-01 2.30006754e-01 -6.40070140e-01 1.59117188e-02
-1.06858492e+00 3.47606689e-01 1.78582191e-01 2.18580201e-01
-1.13607013e+00 -6.29758716e-01 -1.16214192e+00 -1.07859589e-01
1.19834132e-01 -7.04520941e-01 1.19936037e+00 -1.31381392e+00
-1.86789048e+00 9.57176566e-01 -7.45462954e-01 2.01442868e-01
5.29291689e-01 7.63922557e-02 -5.48378408e-01 5.55015385e-01
1.56129718e-01 9.04394567e-01 1.40652239e+00 -1.61289155e+00
-5.09317219e-01 -1.44851521e-01 -5.61124422e-02 4.21096176e-01
-3.50903720e-01 1.50518194e-01 -9.00993705e-01 -7.19525754e-01
9.57686007e-02 -6.49381042e-01 3.84170711e-01 4.07272905e-01
-3.95343781e-01 -5.96102886e-03 1.20191920e+00 -8.30628932e-01
1.05032134e+00 -2.52524662e+00 4.13396105e-04 2.06354275e-01
4.11848538e-02 1.35141462e-01 -3.32940817e-01 2.32101142e-01
-8.04166421e-02 -1.96219951e-01 1.40215680e-01 -6.72234714e-01
-1.87875092e-01 -4.79884893e-02 -3.21120083e-01 5.71673751e-01
1.72646772e-02 5.43364644e-01 -9.10868585e-01 -1.02249622e+00
3.46899778e-01 1.09678459e+00 -5.08307576e-01 3.74401897e-01
7.61883706e-02 7.68853843e-01 -3.64903569e-01 7.17933178e-01
8.33317876e-01 4.08648998e-01 6.87094182e-02 -5.60693324e-01
4.08666246e-02 -1.79597363e-01 -1.43609262e+00 1.82740343e+00
-6.36536956e-01 5.14036834e-01 5.44285476e-01 -5.11393011e-01
1.18472290e+00 6.44794524e-01 5.67398012e-01 -3.49259049e-01
1.37791008e-01 8.18572044e-02 -4.10063684e-01 -6.96467757e-01
1.94925964e-01 -2.06546888e-01 1.47407770e-01 2.57729918e-01
-6.81302771e-02 -4.17740166e-01 6.24706671e-02 5.43325432e-02
2.58743137e-01 3.10105324e-01 -1.31658271e-01 1.52000874e-01
1.09910381e+00 -6.91977084e-01 6.73601806e-01 -1.88264698e-02
-1.87559441e-01 8.71421158e-01 2.16912627e-01 -1.67896524e-01
-7.71989584e-01 -1.14631808e+00 7.58240595e-02 8.75838518e-01
2.99131095e-01 -6.36086941e-01 -1.23325288e+00 -2.88581759e-01
-4.49333400e-01 5.54363132e-01 -5.17317295e-01 -1.28771111e-01
-5.97441971e-01 1.33180693e-01 4.72392261e-01 1.91076696e-01
6.87574983e-01 -1.05678165e+00 -3.52058977e-01 7.57736713e-02
-7.52028465e-01 -1.07755780e+00 -1.10759175e+00 -1.02940214e+00
-3.43708038e-01 -1.02142918e+00 -1.05928576e+00 -1.10011601e+00
9.45180357e-01 3.89809132e-01 6.11247480e-01 -1.51827678e-01
-2.80271500e-01 5.52002430e-01 -8.19010735e-02 -1.06660083e-01
-4.04278666e-01 -5.25067270e-01 2.76320964e-01 7.11154044e-01
-8.46393406e-02 -7.26285100e-01 -7.22367346e-01 6.51442826e-01
-7.69986987e-01 2.43914574e-01 1.99126035e-01 6.73915803e-01
5.64843357e-01 1.48470148e-01 4.97645885e-01 -9.18571092e-03
5.17366886e-01 -8.82566199e-02 -3.42699617e-01 1.67536348e-01
3.08615386e-01 -4.69210356e-01 6.44706130e-01 -8.40316713e-01
-1.32207441e+00 2.75385410e-01 -5.98497503e-02 -7.75488079e-01
-1.53596163e-01 -1.93427622e-01 -8.13575029e-01 -1.34475395e-01
3.02787840e-01 4.59896386e-01 2.42351830e-01 -9.21528712e-02
3.83005708e-01 9.59858716e-01 1.05050564e+00 -6.87425256e-01
9.89908099e-01 6.67533398e-01 -2.55433559e-01 -1.00011909e+00
-4.10159588e-01 -2.23052517e-01 -6.07019842e-01 -7.41093278e-01
7.24087238e-01 -9.61535692e-01 -1.06419516e+00 5.98585844e-01
-1.24404240e+00 9.56435651e-02 -1.57545611e-01 4.27750945e-01
-8.46252084e-01 4.79980707e-01 -2.46364608e-01 -9.67784584e-01
-2.34158859e-01 -1.35440302e+00 1.48055744e+00 4.47797358e-01
-1.69806689e-01 -5.42525947e-01 -7.80866668e-02 4.33511317e-01
3.30433458e-01 2.99011290e-01 4.37353998e-01 1.57182917e-01
-5.32456934e-01 -2.38434061e-01 1.47474213e-02 2.62848675e-01
5.90528727e-01 2.88339257e-01 -1.24248338e+00 -2.26195350e-01
-9.81222186e-03 -8.67919475e-02 1.93213865e-01 4.19681430e-01
1.06542146e+00 -5.93758643e-01 -4.66668129e-01 7.75338233e-01
8.17065477e-01 3.60086501e-01 6.53189838e-01 -1.96285173e-01
5.97104013e-01 9.94542360e-01 7.91824222e-01 4.79224801e-01
3.52106541e-01 1.04026198e+00 1.81220815e-01 1.76864397e-03
-3.14212352e-01 -7.34963655e-01 5.33805609e-01 5.49415946e-01
-8.72407295e-03 7.87668303e-02 -3.06093246e-01 5.16933739e-01
-1.27324545e+00 -1.25351691e+00 4.30944353e-01 2.34483910e+00
9.23219383e-01 -3.64570260e-01 1.93901256e-01 3.39287430e-01
1.26008832e+00 6.64108917e-02 -3.94107819e-01 -5.29670306e-02
-1.71409771e-02 -8.33339468e-02 -4.24803674e-01 5.76017141e-01
-8.68170440e-01 8.46848249e-01 5.53546047e+00 1.03671181e+00
-1.50567698e+00 -6.45182328e-03 5.58912039e-01 -2.94841617e-01
-5.61562717e-01 -1.57285407e-01 -6.38488293e-01 5.40072203e-01
4.90138859e-01 -4.40488696e-01 3.15280139e-01 7.39671469e-01
6.80907071e-01 8.66839066e-02 -9.45531785e-01 1.52626967e+00
4.00793225e-01 -1.13391864e+00 1.23162933e-01 -1.73346415e-01
4.60704893e-01 -9.88332987e-01 2.46710941e-01 -6.00231849e-02
-3.81664008e-01 -9.78633404e-01 9.84397590e-01 6.52521431e-01
1.48760438e+00 -1.01791370e+00 2.58318275e-01 1.04036577e-01
-1.63181007e+00 1.23536885e-01 -2.40453556e-01 4.19668764e-01
5.30108929e-01 2.27547854e-01 -7.55992413e-01 4.91775811e-01
6.75319374e-01 4.71468985e-01 -2.29627743e-01 9.23139393e-01
-1.62784010e-01 1.45352110e-01 -2.12308452e-01 3.85057986e-01
-3.40803951e-01 -1.84419692e-01 7.39935756e-01 9.94827807e-01
5.83000124e-01 7.81493038e-02 1.01737596e-01 8.28837097e-01
2.04278845e-02 4.34510767e-01 -7.79395461e-01 2.07109079e-01
7.81846881e-01 1.35090148e+00 -4.27034885e-01 -2.32048295e-02
-9.96063948e-02 1.02174222e+00 -3.68779719e-01 3.58945101e-01
-1.00444198e+00 -4.37918633e-01 6.80231929e-01 3.55257511e-01
2.94621363e-02 -3.74413915e-02 4.33795959e-01 -1.03719974e+00
2.14919254e-01 -9.43086982e-01 -4.72966023e-03 -1.10377991e+00
-7.17517138e-01 7.87921906e-01 -1.23318965e-02 -1.45740509e+00
-4.15912390e-01 -3.95599790e-02 -8.77466679e-01 8.99746954e-01
-1.17061532e+00 -1.56534016e+00 -6.34883940e-01 9.85688746e-01
5.76370299e-01 -7.69010885e-03 7.99376249e-01 2.98015207e-01
-4.45947587e-01 9.03198361e-01 -3.79481703e-01 2.05647781e-01
9.65358794e-01 -4.29127902e-01 8.06667879e-02 5.82769930e-01
-1.26555115e-02 6.88460290e-01 6.51026487e-01 -4.89558786e-01
-1.28758943e+00 -9.12571013e-01 6.63490534e-01 -4.35873531e-02
8.05635825e-02 -4.84437495e-01 -6.87998831e-01 3.78731251e-01
-9.90529079e-03 4.30976637e-02 5.20667315e-01 -5.05572796e-01
-2.68204808e-01 -5.25026917e-01 -1.28288341e+00 6.99178696e-01
1.06994772e+00 -6.55085325e-01 -2.88105011e-01 1.04798667e-01
4.25970227e-01 -4.53610182e-01 -6.26920938e-01 2.89827198e-01
8.85578930e-01 -1.06903875e+00 1.05868101e+00 1.53325155e-01
4.16017957e-02 -6.47386074e-01 -3.55335139e-02 -1.04076660e+00
1.26773104e-01 -1.20736420e+00 2.20669985e-01 1.84167600e+00
-1.76733762e-01 -4.64601308e-01 5.60817599e-01 3.65642667e-01
3.18729505e-02 -3.95495147e-01 -9.64391410e-01 -5.34565628e-01
-4.78807628e-01 -1.88849121e-01 6.87550187e-01 8.32124174e-01
-2.78809797e-02 3.07944342e-02 -5.96712470e-01 1.30477712e-01
6.87171876e-01 1.09479599e-01 1.17915154e+00 -1.03261602e+00
1.94383129e-01 -2.21309260e-01 -4.58232284e-01 -8.88420343e-01
4.46563870e-01 -4.12937790e-01 2.60051303e-02 -1.28198242e+00
2.24821661e-02 -2.41099820e-01 3.31665933e-01 1.65559560e-01
5.96618354e-02 3.49610001e-01 1.07128039e-01 1.67942226e-01
-1.83015421e-01 7.91517735e-01 1.53100753e+00 -1.10759388e-03
-2.62241185e-01 7.88566992e-02 -4.73401427e-01 7.76489615e-01
5.27041852e-01 -8.45967457e-02 -6.59922004e-01 -2.62201224e-02
-3.29585761e-01 4.59126145e-01 4.89300132e-01 -1.05665660e+00
1.86666250e-01 -2.07404554e-01 4.97305155e-01 -6.27059877e-01
9.15287077e-01 -6.31766737e-01 5.37127554e-01 2.68765628e-01
-1.60049096e-01 -2.58918535e-02 1.36692360e-01 3.09765667e-01
-3.68513227e-01 -6.40214384e-02 1.00135171e+00 -5.61112165e-02
-3.40855539e-01 4.69191968e-01 -6.03374355e-02 -8.77850205e-02
1.23576605e+00 -5.69626987e-01 9.37866420e-02 -1.02458990e+00
-6.10874593e-01 -1.30835995e-02 5.79356015e-01 5.95827818e-01
9.90735829e-01 -1.49916542e+00 -8.43957007e-01 5.87301672e-01
-6.93317056e-02 2.98411609e-03 6.55766308e-01 6.69523299e-01
-4.59275872e-01 1.92280143e-01 -3.44832569e-01 -7.72881866e-01
-1.55530214e+00 6.08610094e-01 4.85638916e-01 5.56726396e-01
-5.48510432e-01 8.21284950e-01 6.07635975e-01 3.27109545e-02
3.31443310e-01 1.76833361e-01 -2.01056853e-01 1.65296331e-01
8.21090519e-01 2.72135794e-01 -4.06389087e-01 -1.22425795e+00
-2.92297125e-01 1.02557003e+00 2.07333505e-01 -3.89785022e-01
9.43779886e-01 -4.89600837e-01 8.33782107e-02 2.39160359e-01
1.19414949e+00 7.21369922e-01 -1.37994003e+00 -5.23666292e-02
-7.08713174e-01 -9.38559949e-01 -2.42233902e-01 -2.56832629e-01
-1.26506209e+00 1.05630124e+00 5.19943178e-01 -3.02323461e-01
1.40282142e+00 -2.75002360e-01 7.74338961e-01 -3.04291099e-01
5.40357411e-01 -8.81351769e-01 3.97143960e-01 -2.41750777e-01
1.18471849e+00 -6.80390298e-01 -4.09271300e-01 -7.87751317e-01
-5.95617533e-01 1.06821787e+00 7.40889430e-01 3.77907395e-01
3.93882304e-01 1.04403049e-01 2.83399045e-01 1.98493227e-01
-2.83523977e-01 4.67984863e-02 3.71084094e-01 1.10749769e+00
1.31916523e-01 -3.28690648e-01 1.40600234e-01 5.12592614e-01
-5.35509765e-01 3.13106342e-03 2.61641175e-01 4.32944506e-01
-3.02488089e-01 -8.90766680e-01 -7.70366549e-01 -3.46674025e-01
-3.26028585e-01 5.08690812e-02 -2.01182589e-01 6.70257211e-01
2.65180975e-01 1.06803870e+00 2.03147605e-02 -1.97429702e-01
2.15404376e-01 9.96866748e-02 5.88944554e-01 -4.30346012e-01
-8.07436556e-02 5.01131892e-01 -1.64535522e-01 -6.11241817e-01
-4.18838203e-01 -5.76978147e-01 -1.09730589e+00 -2.80103594e-01
-2.72276312e-01 3.32199842e-01 5.50867975e-01 4.71304417e-01
4.03446108e-01 2.06256136e-01 9.35037673e-01 -1.42701256e+00
-1.10671639e-01 -7.81011403e-01 -4.49279219e-01 4.42112058e-01
3.67373943e-01 -7.99563050e-01 -4.56325769e-01 4.20141488e-01] | [13.230057716369629, -0.40445783734321594] |
3ef8802b-5710-4f49-bddd-1fa6c07433f5 | learning-dynamical-human-joint-affinity-for | 2109.07353 | null | https://arxiv.org/abs/2109.07353v1 | https://arxiv.org/pdf/2109.07353v1.pdf | Learning Dynamical Human-Joint Affinity for 3D Pose Estimation in Videos | Graph Convolution Network (GCN) has been successfully used for 3D human pose estimation in videos. However, it is often built on the fixed human-joint affinity, according to human skeleton. This may reduce adaptation capacity of GCN to tackle complex spatio-temporal pose variations in videos. To alleviate this problem, we propose a novel Dynamical Graph Network (DG-Net), which can dynamically identify human-joint affinity, and estimate 3D pose by adaptively learning spatial/temporal joint relations from videos. Different from traditional graph convolution, we introduce Dynamical Spatial/Temporal Graph convolution (DSG/DTG) to discover spatial/temporal human-joint affinity for each video exemplar, depending on spatial distance/temporal movement similarity between human joints in this video. Hence, they can effectively understand which joints are spatially closer and/or have consistent motion, for reducing depth ambiguity and/or motion uncertainty when lifting 2D pose to 3D pose. We conduct extensive experiments on three popular benchmarks, e.g., Human3.6M, HumanEva-I, and MPI-INF-3DHP, where DG-Net outperforms a number of recent SOTA approaches with fewer input frames and model size. | ['Yu Qiao', 'Zhe Wang', 'Tianyu Luan', 'Zhipeng Zhou', 'Yali Wang', 'Junhao Zhang'] | 2021-09-15 | null | null | null | null | ['3d-pose-estimation'] | ['computer-vision'] | [-4.37998325e-01 -2.64156222e-01 -1.55782923e-01 1.49282049e-02
-8.17053542e-02 -2.44643673e-01 4.55404557e-02 -3.79180223e-01
-4.92399633e-01 3.53523254e-01 3.93703192e-01 1.63383067e-01
-2.36352161e-01 -4.94910181e-01 -7.50337839e-01 -4.99049425e-01
-5.03307104e-01 5.75770140e-01 4.91412848e-01 -2.11381376e-01
-2.49249712e-01 5.05305409e-01 -1.18346369e+00 -1.01221308e-01
4.21565533e-01 7.81711459e-01 1.70176566e-01 7.38139689e-01
2.35422984e-01 6.18211210e-01 -4.10059541e-01 -1.93792433e-01
2.30539486e-01 -3.20459425e-01 -8.18268180e-01 2.14558631e-01
5.76154292e-01 -5.70184171e-01 -9.91547644e-01 9.27123487e-01
6.90883398e-01 4.97142851e-01 3.86311352e-01 -1.39056373e+00
-4.88421947e-01 2.52687216e-01 -7.97868967e-01 4.39711541e-01
5.85324228e-01 5.19235253e-01 6.03680968e-01 -7.79225111e-01
8.41317892e-01 1.77816677e+00 7.44252026e-01 6.00573182e-01
-8.09123933e-01 -6.25240982e-01 5.09200394e-01 5.05353153e-01
-1.60393691e+00 5.45755327e-02 7.23960698e-01 -4.32181001e-01
9.31539237e-01 2.39780862e-02 1.18394887e+00 1.27601612e+00
3.79696935e-01 1.05650640e+00 5.36049247e-01 7.30202049e-02
-2.21065000e-01 -1.10749149e+00 -2.37402350e-01 1.08695281e+00
-4.39878255e-02 -1.77196283e-02 -8.18135023e-01 1.47085143e-02
1.37480688e+00 4.85827886e-02 -5.58701634e-01 -5.93683898e-01
-1.63995111e+00 5.27669013e-01 6.96621835e-01 3.09384465e-02
-2.60525465e-01 6.43153369e-01 6.58487082e-01 1.22957200e-01
2.61226147e-01 -1.28347844e-01 -5.62287867e-01 -1.81866810e-01
-5.49338758e-01 4.95280504e-01 4.80381578e-01 1.06542790e+00
5.03419995e-01 -3.62001806e-02 -2.84389824e-01 5.55136919e-01
3.72430086e-01 5.97094655e-01 3.35356295e-01 -1.03119063e+00
5.66745460e-01 5.73344648e-01 -8.39645490e-02 -1.53426325e+00
-6.92903399e-01 -2.75145382e-01 -1.13010526e+00 -2.31543094e-01
4.44179624e-01 5.29259183e-02 -1.22826672e+00 1.71509981e+00
6.85082674e-01 5.76243222e-01 -3.59067291e-01 1.49146712e+00
9.76567268e-01 5.24459124e-01 9.09706019e-03 9.61113535e-03
1.30616200e+00 -1.17684758e+00 -5.85529983e-01 -2.21580744e-01
4.52738911e-01 -4.99736458e-01 1.01855600e+00 1.79768637e-01
-9.61252213e-01 -9.13541257e-01 -7.29368746e-01 -2.03018650e-01
1.58343315e-02 5.11261518e-04 6.22599423e-01 1.04442909e-01
-7.81620860e-01 6.44067645e-01 -1.17272449e+00 -3.57531548e-01
1.12804957e-01 3.53064954e-01 -5.50787926e-01 -1.65604472e-01
-1.51333928e+00 4.28960592e-01 4.36233401e-01 6.32711589e-01
-8.73921931e-01 -3.41041893e-01 -1.07295191e+00 -2.07525939e-01
7.74464786e-01 -1.07349944e+00 8.83795202e-01 -4.49976176e-01
-1.32707357e+00 6.07867599e-01 2.31734857e-01 -3.67399603e-01
7.51653016e-01 -6.48725450e-01 -3.66004884e-01 3.63795072e-01
4.20568138e-02 9.44768786e-01 8.33340645e-01 -7.77218640e-01
-2.95942068e-01 -5.69468200e-01 7.03608319e-02 5.03846228e-01
-2.86693964e-02 -2.04559237e-01 -1.36338675e+00 -9.09510195e-01
3.43486845e-01 -1.26350534e+00 -2.55028188e-01 3.75611126e-01
-5.56647301e-01 -3.61222863e-01 8.76202643e-01 -7.98837662e-01
1.37047112e+00 -1.87591696e+00 9.02359009e-01 1.89130485e-01
4.66780633e-01 2.79158354e-01 -8.84143189e-02 6.41563907e-02
1.69639438e-01 -1.66340977e-01 -3.73787358e-02 -1.94338694e-01
-8.55937898e-02 5.42061925e-01 3.50391299e-01 6.85520351e-01
-2.47384720e-02 1.21269333e+00 -9.87684369e-01 -8.41643393e-01
3.14644486e-01 7.44978607e-01 -6.07098341e-01 1.86830670e-01
-2.76289076e-01 5.88683963e-01 -5.91126919e-01 7.21103191e-01
5.52602470e-01 -3.97433043e-01 5.18622138e-02 -6.19365811e-01
2.93795705e-01 -2.40009680e-01 -1.33589411e+00 2.26082945e+00
-1.42473299e-02 3.27578753e-01 -4.86526638e-03 -6.75736725e-01
5.71760476e-01 2.11765870e-01 7.00983465e-01 -5.22090316e-01
2.74364114e-01 -1.43325120e-01 -1.69138089e-02 -8.00022066e-01
2.23248214e-01 4.72195745e-01 -6.04759865e-02 -5.59722399e-03
8.43787044e-02 3.77934963e-01 2.69809604e-01 1.73153341e-01
1.16392338e+00 5.10623455e-01 -5.67840561e-02 -1.52618691e-01
5.68473816e-01 -2.87976742e-01 8.35688174e-01 3.78056914e-01
-4.50357080e-01 7.01537728e-01 5.13973296e-01 -8.21378291e-01
-7.58798718e-01 -1.20340037e+00 4.84906733e-01 8.62732112e-01
6.60952866e-01 -5.87798774e-01 -7.28104472e-01 -7.32538700e-01
8.76112357e-02 -2.41187453e-01 -6.87047184e-01 -2.67396212e-01
-1.08692658e+00 -3.73356879e-01 5.54567277e-01 7.83808351e-01
7.12975740e-01 -1.10679042e+00 -6.02770925e-01 2.48353973e-01
-4.69732076e-01 -1.22442698e+00 -1.10478616e+00 -3.27764243e-01
-7.02584147e-01 -1.31131411e+00 -9.86025691e-01 -6.52903795e-01
4.91544604e-01 2.81473100e-01 1.16024876e+00 2.38148972e-01
-3.71356189e-01 4.99432385e-01 -3.99616331e-01 3.40960830e-01
1.57230929e-01 1.70431435e-01 2.85501182e-01 -1.21353678e-01
9.15098935e-02 -5.80735624e-01 -1.07685483e+00 8.16109121e-01
-8.08771789e-01 3.49798560e-01 4.98791844e-01 7.86256731e-01
7.22334802e-01 8.39664638e-02 -5.60025088e-02 -4.62214708e-01
4.16285813e-01 -3.45970601e-01 -2.14645341e-01 3.00511003e-01
-4.93655317e-02 -7.93160647e-02 2.97084212e-01 -7.01031446e-01
-6.56108141e-01 1.38153613e-01 -1.00248933e-01 -1.24293828e+00
-8.87929928e-05 5.74520051e-01 -3.13191831e-01 -5.08541390e-02
4.82079178e-01 6.27668202e-02 1.62667692e-01 -4.25053358e-01
1.84123740e-01 -3.89411747e-02 8.64983559e-01 -6.65136933e-01
7.58819938e-01 3.40097636e-01 2.22476840e-01 -5.39943814e-01
-7.10945964e-01 -4.79149103e-01 -9.91856158e-01 -5.54438531e-01
1.31698537e+00 -9.85913038e-01 -9.00815845e-01 8.18921745e-01
-1.20262492e+00 -6.01157248e-01 2.23662704e-01 7.23701417e-01
-5.23823023e-01 7.18279064e-01 -9.80726779e-01 -2.44463339e-01
-3.03956062e-01 -1.20160770e+00 1.31570446e+00 8.65095779e-02
-2.02647224e-01 -9.98383641e-01 4.34042281e-03 2.16170728e-01
-7.80684948e-02 6.04856193e-01 5.54290831e-01 1.42977253e-01
-4.98829216e-01 -3.69793251e-02 -2.43911058e-01 1.44718334e-01
1.25399577e-02 -1.18507385e-01 -1.16123170e-01 -5.55402696e-01
-5.04885435e-01 -2.35549614e-01 6.81202710e-01 5.27187705e-01
1.29489124e+00 -3.27401876e-01 -3.77496064e-01 8.05626452e-01
7.83458054e-01 -2.87670612e-01 6.02999687e-01 9.72072035e-02
1.41887283e+00 3.55275959e-01 8.92957389e-01 3.61302078e-01
5.16737461e-01 9.90375400e-01 4.70328927e-01 4.53233831e-02
-3.48358512e-01 -4.35224891e-01 4.23487067e-01 8.92569065e-01
-6.05939567e-01 -3.27387124e-01 -7.86595702e-01 2.15443254e-01
-2.27383447e+00 -6.44653022e-01 -1.14356950e-01 1.77926230e+00
4.73174512e-01 3.67263019e-01 6.00992620e-01 -2.07343772e-01
9.46395397e-01 3.17281961e-01 -7.68986940e-01 4.36936617e-01
6.30650595e-02 -3.41902860e-02 5.02580941e-01 2.05408409e-01
-1.09319127e+00 9.01341915e-01 4.77956009e+00 8.75645399e-01
-9.56530333e-01 -2.30789185e-02 2.72524029e-01 -2.66538113e-01
3.48767899e-02 -2.84974754e-01 -5.05663872e-01 5.12068212e-01
3.17726374e-01 2.16436774e-01 3.77073348e-01 7.57795095e-01
1.40128866e-01 1.05723992e-01 -1.00182939e+00 1.29468071e+00
5.21258032e-03 -1.04174018e+00 1.61519423e-01 1.28196999e-02
4.11707997e-01 -1.14045322e-01 -1.43822148e-01 2.22336963e-01
-7.42655480e-03 -9.86630559e-01 7.57766426e-01 5.77109098e-01
6.88527644e-01 -9.62308943e-01 6.09911263e-01 2.93247551e-01
-1.89337707e+00 2.67700821e-01 -3.39117199e-01 1.08492881e-01
3.47676098e-01 2.66500503e-01 -2.18927622e-01 6.20131850e-01
1.14689600e+00 8.99920702e-01 -5.68207681e-01 8.40217054e-01
-3.01788628e-01 1.16499245e-01 -3.83880794e-01 1.02398664e-01
3.32638562e-01 -1.15149096e-03 6.99249864e-01 1.05729473e+00
3.22203279e-01 2.40183815e-01 6.92796767e-01 5.51322639e-01
1.19659536e-01 -2.25393623e-01 -2.43521929e-01 7.69351572e-02
3.67530406e-01 1.15635073e+00 -8.90768707e-01 -1.51272625e-01
-2.20742598e-01 1.34219098e+00 2.88669258e-01 3.71755272e-01
-1.17162418e+00 -1.03307709e-01 7.83938408e-01 2.13253304e-01
2.19233841e-01 -8.67048562e-01 6.76520169e-01 -1.18803215e+00
2.14094639e-01 -7.60971546e-01 7.71217406e-01 -8.69077563e-01
-1.10585511e+00 4.06402439e-01 3.32908332e-01 -1.26029861e+00
-1.44154012e-01 -5.59817791e-01 -3.64054978e-01 4.31859255e-01
-7.86357582e-01 -1.34650457e+00 -6.69358671e-01 1.00835180e+00
6.12076700e-01 1.57685652e-01 2.97465205e-01 4.28985417e-01
-6.15430653e-01 6.28678203e-01 -7.14763463e-01 5.20261347e-01
6.82423413e-01 -1.16193056e+00 9.04042959e-01 5.90777755e-01
1.11825041e-01 5.34217715e-01 5.16284823e-01 -9.96302903e-01
-1.80716467e+00 -1.26016974e+00 2.84415454e-01 -3.38416010e-01
5.03350735e-01 -2.96816021e-01 -9.19066429e-01 7.07569599e-01
-1.75555319e-01 5.28997600e-01 8.57184529e-02 -5.56003228e-02
-3.13052773e-01 2.03112632e-01 -8.97279322e-01 8.16274583e-01
1.93923104e+00 -3.76789689e-01 -2.43110880e-01 4.08953667e-01
1.13375556e+00 -1.11003399e+00 -1.04816461e+00 6.14164829e-01
6.07076168e-01 -8.86046410e-01 1.24519515e+00 -5.73564470e-01
3.15326542e-01 -6.30500972e-01 6.44848719e-02 -1.08859694e+00
-5.83212912e-01 -6.73266649e-01 -5.95369399e-01 6.98066711e-01
-3.19643885e-01 -8.66826847e-02 1.03105807e+00 1.67831108e-01
-2.85106063e-01 -9.62739050e-01 -1.14764845e+00 -9.29927766e-01
-3.28595519e-01 -3.65105242e-01 5.30373931e-01 8.11159194e-01
-5.27482688e-01 2.01417655e-01 -7.97398627e-01 3.56999516e-01
6.93885624e-01 -1.30234838e-01 1.17172778e+00 -8.67664635e-01
-4.80325341e-01 -3.08684379e-01 -9.23778892e-01 -1.49404502e+00
9.54271331e-02 -7.18648255e-01 2.27599833e-02 -1.34574282e+00
2.39001494e-02 1.34956546e-03 -1.74187481e-01 3.73037100e-01
-2.08489373e-01 2.74535894e-01 9.09889936e-02 1.78889453e-01
-8.78204942e-01 5.81217468e-01 1.87105751e+00 -1.19465075e-01
-2.28701949e-01 -2.39892229e-01 1.52055621e-01 7.97341883e-01
4.05785650e-01 -1.87005728e-01 -5.26368618e-01 -5.31958699e-01
2.41951048e-01 4.41159815e-01 8.91979098e-01 -1.29129601e+00
3.71559322e-01 -3.05620402e-01 6.19685531e-01 -9.26440001e-01
3.39826643e-01 -6.54555798e-01 5.34402788e-01 7.84628689e-01
-3.80396172e-02 4.74858493e-01 8.23849160e-03 9.45434451e-01
-9.26163122e-02 4.40488875e-01 5.04757583e-01 -4.20666337e-01
-1.21210694e+00 1.09166610e+00 -3.95267829e-02 2.11036623e-01
8.48859668e-01 -3.11669439e-01 9.59936753e-02 -1.99256182e-01
-1.08497941e+00 5.60678244e-01 4.67186272e-01 7.53909409e-01
9.16293383e-01 -1.79092050e+00 -4.41939145e-01 -5.83112799e-02
1.39885232e-01 4.92728114e-01 8.74240100e-01 9.39371049e-01
-8.89839172e-01 6.19570836e-02 -2.96642274e-01 -9.97591078e-01
-1.26964939e+00 5.87879956e-01 5.42140722e-01 -1.82770252e-01
-9.93207216e-01 1.03914070e+00 3.36320072e-01 -2.75005072e-01
3.83336335e-01 -3.71719986e-01 1.38622314e-01 -2.67585844e-01
1.52116135e-01 5.47829211e-01 -2.69747704e-01 -8.29337001e-01
-5.14139831e-01 8.53982687e-01 9.29182395e-02 1.81196451e-01
1.00575042e+00 -2.09389389e-01 -5.59711605e-02 2.84291953e-01
1.27088344e+00 -4.06651199e-01 -1.60403812e+00 -2.80557394e-01
-4.33077186e-01 -6.79660857e-01 -3.14381361e-01 -1.51321873e-01
-1.47936642e+00 7.43581653e-01 6.10892057e-01 -1.72066286e-01
8.89610708e-01 7.47416243e-02 1.34745491e+00 3.96109581e-01
4.59354818e-01 -1.13885820e+00 6.86865032e-01 4.84220445e-01
1.03525925e+00 -9.36897695e-01 2.35648796e-01 -5.28380513e-01
-4.56057519e-01 1.13196206e+00 1.27366030e+00 -2.07037330e-01
7.14332402e-01 -1.73376873e-01 -2.05980465e-01 -6.00228786e-01
-3.68860513e-01 -1.56436905e-01 6.76893890e-01 5.78341722e-01
2.76952296e-01 2.10849401e-02 -1.19329810e-01 3.49287510e-01
-5.59401400e-02 -1.07132643e-01 -6.35291263e-02 8.62855554e-01
-2.24050820e-01 -7.28031456e-01 -2.30618149e-01 2.16941535e-01
-7.24008903e-02 3.50350887e-01 -3.27180058e-01 1.20039654e+00
3.35120797e-01 4.85805988e-01 4.49564010e-02 -7.31043696e-01
5.04371643e-01 -4.00348842e-01 7.85279751e-01 -3.85555297e-01
-2.31756285e-01 4.27198648e-01 -3.23439129e-02 -9.79698658e-01
-5.80846369e-01 -6.01226389e-01 -1.56780720e+00 -5.30069113e-01
-1.64103419e-01 -4.11356390e-02 -1.01584997e-02 8.52646351e-01
2.75741577e-01 6.70666456e-01 7.37736821e-02 -1.23445189e+00
-1.68597311e-01 -8.05749297e-01 -5.47118127e-01 5.32592356e-01
7.06659555e-02 -1.18825603e+00 -7.21671656e-02 -1.17823116e-01] | [7.241910457611084, -0.4355258643627167] |
0f398cff-9da0-4b8b-81f2-5c927aeeb02f | prior-based-sampling-for-adaptive-lidar | 2304.07099 | null | https://arxiv.org/abs/2304.07099v1 | https://arxiv.org/pdf/2304.07099v1.pdf | Prior based Sampling for Adaptive LiDAR | We propose SampleDepth, a Convolutional Neural Network (CNN), that is suited for an adaptive LiDAR. Typically,LiDAR sampling strategy is pre-defined, constant and independent of the observed scene. Instead of letting a LiDAR sample the scene in this agnostic fashion, SampleDepth determines, adaptively, where it is best to sample the current frame.To do that, SampleDepth uses depth samples from previous time steps to predict a sampling mask for the current frame. Crucially, SampleDepth is trained to optimize the performance of a depth completion downstream task. SampleDepth is evaluated on two different depth completion networks and two LiDAR datasets, KITTI Depth Completion and the newly introduced synthetic dataset, SHIFT. We show that SampleDepth is effective and suitable for different depth completion downstream tasks. | ['Shai Avidan', 'Amit Shomer'] | 2023-04-14 | null | null | null | null | ['depth-completion'] | ['computer-vision'] | [ 4.01115030e-01 8.77063349e-02 -1.77114293e-01 -6.64656758e-01
-4.75247651e-01 -3.10791761e-01 5.83733439e-01 1.12869851e-01
-7.59667456e-01 6.25977814e-01 -2.27717414e-01 -1.13721967e-01
-5.20805828e-02 -1.20577884e+00 -7.53060043e-01 -3.98493916e-01
7.98432231e-02 9.86999989e-01 4.28534091e-01 9.16895717e-02
3.86012226e-01 1.16477609e+00 -1.64390147e+00 1.71697542e-01
4.83856916e-01 1.15587282e+00 4.57736254e-01 9.65501487e-01
-1.47358179e-01 4.09512699e-01 -3.89301628e-01 -1.16334848e-01
5.73168993e-01 -9.11626592e-02 -7.08272517e-01 1.97652131e-01
9.79357481e-01 -5.72193682e-01 -3.07411462e-01 5.59184134e-01
5.37509978e-01 1.51204720e-01 7.29529321e-01 -1.10560143e+00
2.21675672e-02 5.62489927e-01 -2.97235548e-01 9.84496698e-02
-2.42030114e-01 4.21679646e-01 8.03152561e-01 -1.19038057e+00
7.43155718e-01 1.38124669e+00 9.06260192e-01 4.11182314e-01
-1.34814227e+00 -6.25672638e-01 -6.12199977e-02 -2.63401913e-03
-1.29756653e+00 -3.52520049e-01 7.72294879e-01 -3.96836549e-01
5.01597762e-01 -3.32953930e-01 9.42076623e-01 9.14425313e-01
3.28786343e-01 5.24660528e-01 7.93959796e-01 -2.24240109e-01
7.00540841e-01 -3.01935136e-01 -1.54769540e-01 5.02377093e-01
3.26379687e-02 4.17901129e-01 -7.77911663e-01 1.80230856e-01
8.74582350e-01 -8.11420605e-02 -1.04860306e-01 -6.21168435e-01
-1.13684726e+00 8.13593149e-01 4.44137663e-01 -3.45419526e-01
-2.63718128e-01 5.69622755e-01 5.21568418e-01 1.33277327e-01
3.13917756e-01 3.90381694e-01 -6.20444775e-01 3.45591903e-02
-1.25753403e+00 4.20655429e-01 5.61292171e-01 9.52801347e-01
1.52172160e+00 8.31097662e-02 -1.64764076e-01 4.09456611e-01
-1.82845090e-02 5.66909611e-01 4.49804440e-02 -1.58831406e+00
3.69401246e-01 2.68005103e-01 -3.12925912e-02 -2.45188251e-01
-3.51992428e-01 -1.19154967e-01 -7.31495440e-01 7.77159631e-01
6.13650382e-01 -2.51000017e-01 -1.00079966e+00 1.43243921e+00
4.24670190e-01 2.62831300e-01 -5.66276573e-02 6.20938182e-01
5.97084403e-01 5.03304303e-01 -2.60625899e-01 7.39640817e-02
9.37858701e-01 -7.49036908e-01 -5.90853877e-02 -5.15729249e-01
3.53619248e-01 -4.38916445e-01 1.06272757e+00 7.63774812e-01
-1.02454507e+00 -8.53722870e-01 -1.23016047e+00 -2.97144860e-01
-1.84072837e-01 2.23191351e-01 3.29569072e-01 2.87923753e-01
-1.02565527e+00 9.01405036e-01 -7.59208143e-01 -1.80977046e-01
6.00889325e-01 3.58833015e-01 -1.67289868e-01 -2.09048450e-01
-7.69219220e-01 5.32163262e-01 3.45767319e-01 1.44492155e-02
-1.50555670e+00 -9.40284848e-01 -8.96042228e-01 4.30033766e-02
3.58347058e-01 -7.77658224e-01 1.61736870e+00 -5.12973905e-01
-1.55121565e+00 7.65361428e-01 -1.62380248e-01 -1.01539230e+00
7.76166677e-01 -2.10129023e-01 3.03329974e-01 2.76263684e-01
2.84788400e-01 1.29898584e+00 1.17846894e+00 -1.11674249e+00
-6.92278147e-01 -2.86233157e-01 6.07164167e-02 -1.18125621e-02
1.15382299e-01 -6.68536901e-01 -4.43030834e-01 -6.08478822e-02
-4.05809842e-02 -7.62206197e-01 -5.52715838e-01 5.93730390e-01
-3.96752775e-01 -5.29003032e-02 1.18173969e+00 2.60482758e-01
5.01087546e-01 -2.04827905e+00 -8.52118209e-02 -4.06972989e-02
1.86621770e-01 1.62732258e-01 -2.69387901e-01 3.48094314e-01
7.82560632e-02 -1.55584529e-01 -5.44265926e-01 -8.22644055e-01
-2.76732117e-01 5.26046097e-01 -2.01924950e-01 3.89760137e-01
3.54546368e-01 7.31176019e-01 -8.58741462e-01 -5.84356189e-01
8.00647676e-01 4.85767603e-01 -6.79854512e-01 1.83014214e-01
-7.22518682e-01 5.92133582e-01 -1.70622125e-01 4.61023837e-01
1.02594244e+00 1.06201410e-01 -2.04617023e-01 -1.22791357e-01
-4.40015972e-01 1.76232457e-01 -1.12434137e+00 1.89098024e+00
-7.26277769e-01 1.00613463e+00 2.07111314e-02 -8.15786302e-01
1.10497248e+00 -1.66441172e-01 5.23329377e-01 -3.78120989e-01
9.32870731e-02 1.75341502e-01 -2.49595731e-01 -2.73780644e-01
7.96932280e-01 -5.34741320e-02 7.37944245e-02 1.97197780e-01
2.27346383e-02 -8.70641232e-01 1.22218780e-01 4.71933745e-02
1.03377295e+00 3.56624156e-01 3.75097215e-01 -3.08941931e-01
4.71451551e-01 5.11459708e-02 4.94662523e-01 7.46636510e-01
-7.45422691e-02 1.12392902e+00 3.44427466e-01 -6.96303487e-01
-1.19990802e+00 -1.17749000e+00 -4.27865684e-01 6.23403847e-01
9.48812142e-02 -2.58862734e-01 -6.81461692e-01 -5.94449461e-01
2.12366194e-01 6.20507717e-01 -7.52172709e-01 1.93237618e-01
-8.17028403e-01 6.41428530e-02 5.55964649e-01 5.54003298e-01
7.65420258e-01 -1.29135251e+00 -1.23192978e+00 3.10865670e-01
4.17525202e-01 -1.21729910e+00 -4.44901079e-01 7.04630613e-01
-1.00689280e+00 -1.20384789e+00 -3.15842718e-01 -2.31870860e-01
2.62900203e-01 2.72488832e-01 1.37666690e+00 -1.02145582e-01
-2.99989760e-01 2.69484013e-01 -4.06638533e-02 -4.59198147e-01
-2.91792929e-01 5.43449938e-01 -3.62429351e-01 -1.49292797e-01
1.45856723e-01 -7.61428833e-01 -6.58662558e-01 -6.53233826e-02
-9.99373436e-01 -3.25804874e-02 3.73988032e-01 7.76417851e-01
6.61817431e-01 -6.40516952e-02 1.60467193e-01 -8.28758359e-01
6.74598217e-02 -9.80364829e-02 -8.44638050e-01 -3.89717102e-01
-5.24400532e-01 5.86627573e-02 9.60105717e-01 -2.94264764e-01
-7.75682688e-01 5.13537347e-01 -2.48034835e-01 -7.84422815e-01
-4.01703775e-01 4.37776409e-02 -1.99867815e-01 2.65964288e-02
7.22565949e-01 5.26790284e-02 -1.81626424e-01 -3.54536891e-01
3.90439600e-01 2.77884036e-01 7.93350577e-01 -7.37193942e-01
8.88268948e-01 9.05100524e-01 5.36545575e-01 -1.04733157e+00
-8.81678522e-01 -2.57057011e-01 -1.12939382e+00 -2.67005771e-01
9.15914834e-01 -8.09640884e-01 -6.89533353e-01 4.33730006e-01
-1.42185974e+00 -9.15929794e-01 -9.72950399e-01 1.50874630e-01
-7.21023202e-01 1.63846001e-01 -1.97540939e-01 -8.40704560e-01
-7.46004954e-02 -1.36563528e+00 1.65100551e+00 -4.16338369e-02
-4.10165591e-03 -9.61189926e-01 -3.44351493e-02 -1.32178277e-01
7.18723089e-02 3.68006200e-01 7.46689677e-01 -1.28258318e-01
-1.16547906e+00 1.94037765e-01 -1.87169939e-01 5.46571910e-01
2.13211384e-02 4.09135193e-01 -1.30973232e+00 -4.95400615e-02
-1.78731218e-01 -4.08097267e-01 1.06772268e+00 5.43189704e-01
1.60623038e+00 1.57116756e-01 -1.66530088e-01 1.04740882e+00
1.61279631e+00 1.28272325e-01 7.23624229e-01 3.75207633e-01
5.98739326e-01 4.76137757e-01 7.30935574e-01 5.91257513e-01
5.09226739e-01 6.30689442e-01 9.88814950e-01 9.24921632e-02
-2.84974784e-01 -3.73976856e-01 1.35407388e-01 -1.16677126e-02
2.47060522e-01 -1.72681600e-01 -9.10798788e-01 6.06443465e-01
-1.50119495e+00 -8.41454804e-01 -1.79704487e-01 2.25210786e+00
5.21769702e-01 4.28902417e-01 -2.65991427e-02 2.43381590e-01
3.79985929e-01 3.32179129e-01 -8.27894330e-01 -4.68742102e-01
-1.09557189e-01 8.74501050e-01 6.88999891e-01 7.80498564e-01
-8.03510010e-01 1.15729809e+00 6.36430693e+00 7.77426362e-01
-1.25716245e+00 -3.35539505e-02 5.04790723e-01 -5.36818206e-02
-2.47925565e-01 2.91484565e-01 -1.08633316e+00 2.37505451e-01
7.61700392e-01 8.24053958e-02 2.36139089e-01 1.01771879e+00
4.78830934e-01 -4.45957810e-01 -1.47174788e+00 6.95345581e-01
-3.84137660e-01 -1.63689673e+00 8.98203999e-02 6.46396503e-02
5.10707617e-01 2.73391098e-01 -1.54242203e-01 3.10532898e-01
1.73600659e-01 -1.01425683e+00 8.30329895e-01 2.83526003e-01
9.36088741e-01 -9.98864174e-01 3.35056722e-01 5.68596900e-01
-1.45500028e+00 -1.25154098e-02 -5.16341031e-01 -2.60977477e-01
1.46197855e-01 8.89971316e-01 -1.11590683e+00 2.95686096e-01
6.36364043e-01 8.45933259e-01 -6.24868989e-01 8.38081241e-01
-3.32745671e-01 3.06009144e-01 -2.13892221e-01 3.92536670e-01
3.93699348e-01 -3.44506502e-01 4.74663407e-01 1.29823458e+00
4.03313011e-01 -4.08900946e-01 3.73062700e-01 8.02224457e-01
-1.37100726e-01 -3.58186930e-01 -9.30503488e-01 4.84133959e-01
7.50324965e-01 1.27979267e+00 -6.21224403e-01 -3.73631895e-01
-2.08279863e-02 7.86863148e-01 4.18066055e-01 2.06411764e-01
-6.94475830e-01 -2.64941752e-01 7.41130471e-01 5.31369209e-01
4.96636957e-01 -5.15244544e-01 -4.04774666e-01 -7.03264952e-01
-2.45094270e-01 -3.67170513e-01 -4.23633028e-04 -1.06538904e+00
-9.46994007e-01 3.65479350e-01 2.58231878e-01 -1.36377108e+00
-3.23975503e-01 -7.66101658e-01 -6.62429631e-01 7.26848722e-01
-1.72030437e+00 -9.84514117e-01 -7.08072126e-01 4.11154509e-01
1.05507588e+00 1.10219888e-01 3.62135768e-01 -7.69052515e-03
-3.01427126e-01 1.56494994e-02 -4.67760056e-01 8.03763196e-02
5.71339011e-01 -1.49100149e+00 6.96747005e-01 7.49255776e-01
-1.87489033e-01 2.05071107e-01 5.89069128e-01 -5.17506003e-01
-1.11299169e+00 -1.55139780e+00 5.20244062e-01 -1.47950649e-01
4.24979895e-01 -4.08962190e-01 -6.22626662e-01 6.65596187e-01
2.24210262e-01 2.61302888e-01 -2.70271245e-02 -3.98605019e-01
-1.22371934e-01 -4.87033129e-01 -1.30889666e+00 5.27073503e-01
1.23719811e+00 -4.79680002e-01 -2.35807210e-01 2.28230774e-01
9.43433404e-01 -6.42226398e-01 -7.29780316e-01 5.50709546e-01
4.68555391e-01 -1.60522854e+00 1.06797183e+00 5.18207112e-03
6.79496288e-01 -2.81484336e-01 -2.26683304e-01 -1.24330974e+00
6.48901761e-02 -5.95043004e-01 -1.10150594e-02 8.30388546e-01
1.05587758e-01 -4.33282137e-01 1.31620538e+00 -1.67919826e-02
-4.99393642e-01 -8.57961774e-01 -1.12386596e+00 -7.20635116e-01
2.15040088e-01 -8.20364296e-01 8.63413692e-01 4.39780682e-01
-1.06494701e+00 4.87261504e-01 -1.04562134e-01 2.86944449e-01
8.08391452e-01 1.03778720e-01 1.27392864e+00 -1.53616858e+00
-2.27898061e-01 -2.47804806e-01 -2.71747828e-01 -1.42225802e+00
2.38414392e-01 -5.99163473e-01 1.61013648e-01 -1.38051260e+00
-3.70183885e-01 -5.53723395e-01 5.00542700e-01 1.74991295e-01
2.73844749e-01 2.85760015e-01 1.45104095e-01 -6.19140603e-02
-1.37130041e-02 5.94589233e-01 1.48263228e+00 -2.21972004e-01
-2.26124734e-01 2.72391379e-01 -1.21096037e-01 7.86274135e-01
7.82263577e-01 -3.65357488e-01 -7.30558157e-01 -5.04068911e-01
1.27841294e-01 2.05480248e-01 5.76043010e-01 -1.47247410e+00
9.89507511e-02 -3.33305776e-01 5.24177909e-01 -1.31823313e+00
8.40026200e-01 -7.35495687e-01 -1.79513600e-02 5.14756143e-01
-2.23112553e-01 1.20368779e-01 8.07483643e-02 2.73125738e-01
-3.61199565e-02 -5.92937827e-01 1.24628437e+00 -2.93547988e-01
-7.23593235e-01 8.74934733e-01 -3.27255100e-01 -1.65700875e-02
6.90319002e-01 -5.28405130e-01 1.55714154e-01 -2.17550114e-01
-5.76926708e-01 1.47933066e-01 6.87832534e-01 -5.20264357e-02
8.11911583e-01 -1.29437292e+00 -4.69754338e-01 3.83176178e-01
1.94702551e-01 1.03470886e+00 -7.32002780e-02 3.54077339e-01
-6.92739129e-01 7.98580050e-02 3.00277919e-02 -1.16734409e+00
-8.15455616e-01 3.53433907e-01 5.80995381e-01 -1.37196317e-01
-8.34312916e-01 7.10515797e-01 2.44496763e-01 -6.25539482e-01
1.22320890e-01 -9.16209817e-01 1.56293258e-01 7.34269395e-02
2.74307728e-01 2.78767109e-01 1.68754280e-01 -2.05013186e-01
-3.35560180e-02 7.01672196e-01 3.69972020e-01 -3.43301505e-01
1.18259501e+00 -5.45032807e-02 9.52876210e-02 6.11404002e-01
1.10296500e+00 -6.96258768e-02 -1.89419591e+00 -9.95786712e-02
-1.68441266e-01 -5.06286621e-01 1.46731362e-02 -9.76954326e-02
-1.16445005e+00 1.22837472e+00 3.07839632e-01 6.20529018e-02
1.02045166e+00 -4.63640928e-01 6.77876949e-01 6.14101291e-01
5.12823582e-01 -8.99041653e-01 3.54988962e-01 8.49547386e-01
7.13685691e-01 -1.08586729e+00 2.62948908e-02 -1.76010400e-01
-1.00392506e-01 1.50113821e+00 7.78065920e-01 -3.35488021e-01
6.22924984e-01 6.17183983e-01 -1.66138738e-01 -2.63723075e-01
-6.46984696e-01 -3.31676424e-01 -1.56598583e-01 8.04019153e-01
1.14702769e-01 -2.59830087e-01 4.01237309e-01 -4.45477188e-01
-5.69385767e-01 2.91964948e-01 7.20513105e-01 7.90986061e-01
-6.92676544e-01 -1.24464703e+00 -4.60904717e-01 2.55596727e-01
2.92936593e-01 4.96479124e-02 -2.96493143e-01 1.19265294e+00
7.34054446e-01 5.07704377e-01 3.24886829e-01 -1.63790375e-01
1.89360037e-01 -1.59539074e-01 5.87714136e-01 -9.43239391e-01
-3.46767247e-01 -2.07791582e-01 -4.70109247e-02 -6.10406876e-01
-3.87932092e-01 -7.78206706e-01 -1.18171084e+00 -1.81180134e-01
7.40751773e-02 -1.13133512e-01 6.06726348e-01 8.93468320e-01
1.86804961e-02 4.61312085e-01 7.24576712e-01 -1.55822909e+00
-2.36175120e-01 -9.50893342e-01 -3.99658620e-01 -1.43139020e-01
6.23446882e-01 -5.87550402e-01 -2.74432898e-01 -2.39907563e-01] | [8.395977973937988, -2.558716297149658] |
e75be01b-71f7-4f48-96cb-fbec2800d9c7 | improving-response-diversity-through | null | null | https://aclanthology.org/2022.rocling-1.37 | https://aclanthology.org/2022.rocling-1.37.pdf | Improving Response Diversity through Commonsense-Aware Empathetic Response Generation | Due to the lack of conversation practice, the main challenge for the second-language learners is speaking. Our goal is to develop a chatbot to encourage individuals to reflect, describe, analyse and communicate what they read as well as improve students’ English expression skills. In this paper, we exploit COMMET, an inferential commonsense knowledge generator, as the background knowledge to improve the generation diversity. We consider two approaches to increase the diversity of empathetic response generation. For nonpretrained models, We apply AdaLabel (Wang et al., 2021) to Commonsense-aware Empathetic model (Sabour et al., 2022) and improve Distinct-2 score from 2.99 to 4.08 on EMPATHETIC DIALOGUES (ED). Furthermore, we augment the pretrained BART model with various commonsense knowledge to generate more informative empathetic responses. Not only has the automatic evaluation of distinct-2 scores improved from 9.11 to 11.21, but the manual case study also shows that CE-BART significantly outperform CEM-AdaLabel. | ['Chia-Hui Chang', 'Tzu-Hsien Huang'] | null | null | null | null | rocling-2022-11 | ['empathetic-response-generation'] | ['natural-language-processing'] | [-6.91219121e-02 7.23603606e-01 2.07943127e-01 -3.17608446e-01
-7.57625043e-01 -6.66377068e-01 7.00853527e-01 -7.28950649e-02
-2.99566299e-01 1.27854621e+00 7.81641245e-01 -1.39026716e-02
2.56310731e-01 -7.89407611e-01 1.74804516e-02 -3.89743447e-01
8.12854230e-01 4.95791584e-01 2.24693143e-03 -9.86715972e-01
3.48474771e-01 3.16732496e-01 -1.14319730e+00 9.02350068e-01
1.37881005e+00 3.78370792e-01 7.25108013e-02 7.39255965e-01
-3.07228446e-01 1.73426068e+00 -1.23614788e+00 -1.08375776e+00
-4.96412367e-01 -1.00185692e+00 -1.54151154e+00 -4.64106053e-01
4.83071283e-02 -3.28042626e-01 1.89617916e-03 9.36867714e-01
7.70243347e-01 6.01985693e-01 8.08264077e-01 -1.06323564e+00
-1.10078359e+00 1.48004937e+00 1.06188999e-02 -2.43666887e-01
7.88257360e-01 8.68886635e-02 7.78975785e-01 -6.03243649e-01
6.55170619e-01 1.29637074e+00 7.69849956e-01 1.32416022e+00
-8.64787877e-01 -7.10375369e-01 -4.35903966e-01 6.11639500e-01
-8.27883244e-01 -3.64421517e-01 1.08654642e+00 -2.48247981e-01
7.43916154e-01 4.44552928e-01 7.51590014e-01 1.72885215e+00
-4.74626899e-01 1.00940347e+00 1.51701283e+00 -7.31887102e-01
-1.32485712e-02 5.14245570e-01 2.00243503e-01 2.69957334e-01
-3.89845967e-01 -2.12008119e-01 -7.42859721e-01 1.10882558e-01
3.80531698e-01 -4.90632087e-01 -2.26589710e-01 6.27926111e-01
-9.22261298e-01 1.21455121e+00 3.18715990e-01 4.55147952e-01
-3.14301401e-01 -1.81303531e-01 5.55464089e-01 5.56219816e-01
5.44687733e-02 1.28869224e+00 3.75849679e-02 -9.93938088e-01
-2.79022664e-01 3.21272075e-01 1.17402196e+00 8.53018343e-01
1.69906840e-01 -3.99454348e-02 -4.46244895e-01 1.46308887e+00
9.90847275e-02 2.67035723e-01 9.51763809e-01 -1.41130757e+00
2.82416642e-01 6.22385502e-01 2.19333563e-02 -6.85125947e-01
-2.24158078e-01 -1.42340675e-01 -6.77663565e-01 2.75731236e-02
4.73415464e-01 -4.90239024e-01 4.90302071e-02 1.82039261e+00
1.70267373e-01 -3.86619985e-01 6.77933574e-01 6.61571443e-01
1.56264734e+00 7.94957995e-01 3.00682038e-01 -2.24679574e-01
1.09551001e+00 -1.38111889e+00 -9.40651894e-01 -1.32628262e-01
1.15843809e+00 -1.03949082e+00 1.31623304e+00 6.31146550e-01
-1.33085215e+00 -2.93847919e-01 -7.05607176e-01 -2.86119968e-01
-2.02973679e-01 1.93574458e-01 4.92294163e-01 6.28297806e-01
-6.20495141e-01 4.09520745e-01 -1.35065347e-01 -2.06804976e-01
7.90740475e-02 -2.29255959e-01 -1.74696371e-01 1.09967217e-01
-1.83183908e+00 1.46348548e+00 3.65886003e-01 -4.12638992e-01
-3.05427164e-01 -6.61063969e-01 -7.31690884e-01 3.69666368e-02
-2.79704720e-01 -4.36739832e-01 1.81890547e+00 -9.53044295e-01
-2.51970911e+00 1.06404650e+00 4.86410171e-01 -2.30836824e-01
7.09219158e-01 -2.40515605e-01 -1.84253473e-02 2.56199628e-01
6.24236502e-02 7.41648853e-01 -6.16713762e-02 -1.04111195e+00
-4.14930731e-01 7.29881525e-02 2.66195893e-01 6.95328832e-01
-4.64408875e-01 2.80552894e-01 4.91906315e-01 -7.06818044e-01
-3.33339036e-01 -1.01405466e+00 1.58102781e-01 -4.75883037e-01
-3.32990348e-01 -7.73589075e-01 3.58786702e-01 -8.52842391e-01
1.13092732e+00 -1.81555963e+00 -8.53467658e-02 -1.16224036e-01
6.01457544e-02 4.59764361e-01 6.48549199e-02 7.29965210e-01
2.76254974e-02 -9.55990404e-02 -8.52712989e-02 -4.06765848e-01
1.75306275e-01 1.95298433e-01 -3.39826971e-01 -1.88508973e-01
-2.30624415e-02 7.58729517e-01 -1.17935479e+00 -5.92853248e-01
8.01179633e-02 2.54462659e-01 -4.91040617e-01 6.87244594e-01
8.31068158e-02 4.84307706e-01 -3.16670507e-01 1.95242709e-03
1.93858936e-01 9.31092277e-02 -7.87786543e-02 4.58049029e-01
-9.52941626e-02 8.81173193e-01 -7.59280324e-01 1.45613539e+00
-9.35904503e-01 7.92719781e-01 -2.91364580e-01 -6.23818338e-01
1.27189028e+00 6.80257797e-01 2.81842109e-02 -4.78517026e-01
2.13618651e-01 9.93326828e-02 2.94446766e-01 -6.94433868e-01
7.00160444e-01 -6.90291703e-01 -4.16964293e-01 8.90761614e-01
-1.60144478e-01 -6.59439445e-01 -2.27573551e-02 3.16341579e-01
9.05468643e-01 -1.25970870e-01 4.87633109e-01 3.00407000e-02
7.73909092e-01 2.25988150e-01 2.21024260e-01 5.53565502e-01
-4.10377353e-01 3.52623701e-01 5.34842074e-01 -4.31228243e-02
-4.22217786e-01 -7.78859794e-01 1.69152603e-01 1.44594455e+00
-2.40502805e-01 -1.80790886e-01 -8.59138727e-01 -6.17097139e-01
-4.73676950e-01 1.65080535e+00 -4.83314902e-01 -5.06477475e-01
-7.13425398e-01 -2.85990715e-01 1.22463155e+00 4.72176403e-01
7.63025880e-01 -1.71008921e+00 -5.70639312e-01 4.13487524e-01
-9.29347456e-01 -9.80791569e-01 -1.81242034e-01 -1.53119192e-01
-2.35992074e-01 -6.18777394e-01 -5.49992323e-01 -8.30618978e-01
2.25919768e-01 -2.08425540e-02 1.00218928e+00 1.55372158e-01
1.65365130e-01 1.08070336e-01 -9.10697877e-01 -4.07589972e-01
-1.24143338e+00 -4.63876538e-02 -3.05759937e-01 -6.44864261e-01
2.94337928e-01 -5.60577333e-01 -1.19670518e-01 6.85896575e-02
-2.19204009e-01 5.04908144e-01 2.95954645e-01 1.02318060e+00
-2.48160765e-01 -4.90698069e-01 1.00002837e+00 -8.74008954e-01
1.35323954e+00 -3.00715923e-01 3.74131590e-01 1.90361679e-01
-3.52860034e-01 -7.13074878e-02 9.88377750e-01 -7.70649314e-01
-1.66379952e+00 -4.39624697e-01 -4.84533131e-01 -8.63895472e-03
-2.22015113e-01 3.05250585e-01 -1.91091765e-02 2.07365472e-02
1.20804000e+00 2.29847863e-01 8.05387422e-02 -2.01017141e-01
5.50216973e-01 1.07233572e+00 7.77580202e-01 -1.16958225e+00
2.33286649e-01 -3.29875112e-01 -4.64483976e-01 -6.52807057e-01
-1.24598038e+00 -2.70703405e-01 -1.63662910e-01 -4.75623041e-01
7.62002826e-01 -8.12467635e-01 -1.08233416e+00 4.48938102e-01
-1.63620090e+00 -8.32433045e-01 -3.51174504e-01 6.36005521e-01
-8.42464745e-01 2.89798796e-01 -1.14816785e+00 -7.81867325e-01
-6.32385254e-01 -8.31539452e-01 2.65937686e-01 3.45770568e-01
-1.09728098e+00 -1.12155366e+00 3.96456242e-01 1.01776433e+00
4.30450082e-01 2.00543508e-01 8.87691677e-01 -1.23148441e+00
5.71687818e-01 -3.14941294e-02 5.26465327e-02 5.25788844e-01
-1.37090310e-01 -8.65874253e-03 -1.00998735e+00 5.37291050e-01
2.18980744e-01 -1.08543003e+00 1.32005662e-01 -3.82794708e-01
9.09802198e-01 -8.56688023e-01 3.25271666e-01 7.28162006e-02
3.82902086e-01 2.51972377e-01 6.64213538e-01 3.55240911e-01
4.90227282e-01 1.03403115e+00 6.68817997e-01 4.84525770e-01
7.69864976e-01 7.67502487e-01 -1.65982187e-01 4.10400182e-01
-3.78949344e-01 -4.72910166e-01 7.06791043e-01 1.30464053e+00
-2.36478642e-01 3.57979611e-02 -8.93989503e-01 4.71930057e-01
-1.77816081e+00 -1.49741745e+00 -5.00647187e-01 1.57988238e+00
1.87237811e+00 -3.49522144e-01 1.73431560e-01 -7.83189610e-02
7.54154623e-01 -1.49539366e-01 1.59179971e-01 -1.12614632e+00
-3.95469330e-02 3.41807991e-01 -3.93357873e-01 8.52694333e-01
-3.77988935e-01 1.27153778e+00 5.25574589e+00 7.44143605e-01
-8.96538496e-01 1.47113755e-01 2.59887755e-01 1.09038286e-01
-2.50168681e-01 -3.33236456e-01 -5.53040862e-01 4.98387516e-01
9.28638399e-01 -4.64711517e-01 3.10083628e-01 9.84532773e-01
-6.10406883e-02 2.86129028e-01 -1.01221979e+00 9.53802526e-01
2.36250833e-01 -1.01918125e+00 -1.24195941e-01 -3.95554811e-01
9.18934822e-01 -7.93358505e-01 -2.29275987e-01 1.02807927e+00
8.93582940e-01 -1.15539205e+00 6.72937930e-01 2.64824778e-01
2.12452203e-01 -8.39379430e-01 9.80564535e-01 6.37287915e-01
-2.49082789e-01 1.18702076e-01 -1.45941138e-01 -5.70122659e-01
3.18887681e-02 -6.73376769e-02 -1.35923088e+00 -6.30312860e-02
3.55920523e-01 3.66833180e-01 -3.53961021e-01 5.45718610e-01
-1.18497479e+00 7.33425915e-01 -2.45014839e-02 -7.41833806e-01
2.41345346e-01 -7.05718761e-03 3.95760894e-01 1.37796986e+00
1.79610580e-01 7.19932199e-01 -4.56092134e-02 1.09350872e+00
-2.21861809e-01 3.19436133e-01 -2.99674332e-01 7.58239441e-03
1.04565382e+00 1.28806508e+00 6.74908906e-02 -3.96303743e-01
1.10421240e-01 1.03678513e+00 6.28661096e-01 -8.15122947e-02
-7.03738928e-01 -5.44735372e-01 3.84049088e-01 -3.30024481e-01
-3.79045337e-01 3.44085634e-01 -5.76917827e-01 -6.38389766e-01
-5.19289017e-01 -1.15398276e+00 3.57673883e-01 -1.17461944e+00
-1.62560332e+00 3.77889633e-01 1.64556615e-02 -8.98343563e-01
-6.87797427e-01 -4.04010743e-01 -1.28851926e+00 9.24716532e-01
-8.15804362e-01 -1.19675958e+00 -4.88890231e-01 4.96471047e-01
5.95879495e-01 -2.32169554e-01 1.15798604e+00 -7.66204596e-02
-4.19783920e-01 9.34265494e-01 -3.83910209e-01 3.41608405e-01
1.11102223e+00 -1.41491091e+00 -3.81008014e-02 3.62947047e-01
-2.94884950e-01 6.89622462e-01 9.01788533e-01 -3.63145292e-01
-5.28316557e-01 -7.59453535e-01 1.16283274e+00 -4.17808086e-01
8.57909620e-01 2.26908743e-01 -1.04831946e+00 5.18266499e-01
6.66132331e-01 -7.11174667e-01 1.27988350e+00 1.03672139e-01
-3.20665658e-01 3.95528704e-01 -1.39721549e+00 9.41535234e-01
8.06928635e-01 -4.08965886e-01 -1.34587634e+00 3.77528518e-01
6.65963352e-01 -5.75024784e-01 -1.13611686e+00 -1.81685090e-01
3.01785320e-01 -1.04669058e+00 6.44060493e-01 -8.02015960e-01
1.14449811e+00 3.48713636e-01 1.10081412e-01 -1.82718730e+00
-1.23899598e-02 -9.46447253e-01 2.33196001e-02 1.52843630e+00
3.16952407e-01 -5.45910835e-01 4.74363297e-01 9.16914046e-01
-5.85286915e-01 -5.07077813e-01 -6.92264974e-01 -4.52543557e-01
6.45690084e-01 -1.94061697e-01 4.19561297e-01 1.63979506e+00
1.13206458e+00 8.54038417e-01 -1.98067069e-01 -4.71137822e-01
2.01218456e-01 1.22644097e-01 8.77797961e-01 -1.09033382e+00
-4.12294596e-01 -8.55143249e-01 6.72810972e-02 -9.27800059e-01
6.20375395e-01 -9.53185022e-01 3.05343211e-01 -1.51492107e+00
9.34881940e-02 -4.74364966e-01 3.04588974e-01 6.86723650e-01
-4.35365409e-01 1.22228995e-01 2.59029835e-01 -3.01232208e-02
-3.70471090e-01 7.84879148e-01 1.53548539e+00 3.16078156e-01
-3.69239926e-01 1.72856112e-03 -1.07718539e+00 8.37551415e-01
1.25354791e+00 -3.95860672e-01 -4.15812761e-01 -2.13813558e-02
1.57095522e-01 2.45880157e-01 2.69811153e-01 -7.15227246e-01
2.76905954e-01 -5.86411059e-01 -1.51572274e-02 -1.45893157e-01
4.04573083e-01 -2.16365516e-01 -2.51054406e-01 3.44821483e-01
-9.57125008e-01 -4.09120657e-02 6.14013486e-02 -4.12989378e-01
-2.69874185e-01 -8.89585137e-01 9.93502021e-01 -2.39790231e-01
-4.07852113e-01 -7.35249460e-01 -5.45202255e-01 5.78514695e-01
1.03040981e+00 -1.26937896e-01 -8.99412453e-01 -9.13468659e-01
-4.99943465e-01 1.46655709e-01 2.47668132e-01 2.21027344e-01
6.27713799e-01 -1.47433937e+00 -9.65786397e-01 -4.32593912e-01
5.95135242e-02 4.10073623e-02 3.60315770e-01 8.72972190e-01
-7.31791914e-01 1.66045517e-01 -5.44948518e-01 1.77754760e-01
-1.43592668e+00 -7.17686936e-02 3.67112815e-01 -3.65940034e-01
-5.41944802e-01 1.21779466e+00 -2.90762901e-01 -7.87277281e-01
9.18158069e-02 1.25929639e-01 -5.81551850e-01 1.60834223e-01
6.77952647e-01 8.19961488e-01 -2.67593682e-01 -5.23152173e-01
1.97816100e-02 -8.68721083e-02 -1.79112092e-01 -4.42625761e-01
1.17853367e+00 1.94691658e-01 -1.67788714e-01 4.46006179e-01
1.00631380e+00 4.21518147e-01 -5.90947449e-01 1.00267529e-02
-2.66018271e-01 -1.11880869e-01 -4.32070196e-01 -1.36817575e+00
-2.57605076e-01 9.71690953e-01 -3.27346534e-01 2.25801215e-01
5.87243140e-01 -2.15185378e-02 8.82332206e-01 7.37891972e-01
-8.74285772e-02 -1.36525834e+00 4.46102709e-01 1.04227364e+00
1.27492654e+00 -1.01877141e+00 -1.64969161e-01 -2.94417143e-01
-1.48958552e+00 1.30006886e+00 1.15818918e+00 2.86531568e-01
-2.62529433e-01 1.91664770e-02 4.22977030e-01 5.63405603e-02
-9.53997374e-01 2.55786151e-01 1.25753814e-02 6.30843043e-01
9.89509702e-01 3.48827779e-01 -5.80108225e-01 1.21054149e+00
-1.21388137e+00 -2.08507419e-01 1.13767934e+00 5.67196727e-01
-5.40610373e-01 -9.86021817e-01 -4.25405264e-01 -1.42028064e-01
-1.49326906e-01 -1.97510913e-01 -1.25430155e+00 7.73778141e-01
-2.66375363e-01 1.25178730e+00 -3.07963222e-01 -3.63203764e-01
3.94825935e-01 3.98689657e-01 4.54086423e-01 -8.47355068e-01
-1.21369028e+00 -8.33434522e-01 7.40347087e-01 9.14685801e-02
-3.76084179e-01 -5.09995759e-01 -1.64863384e+00 -6.59076452e-01
-1.67427123e-01 5.22140563e-01 2.72520959e-01 1.06763458e+00
-1.82312593e-01 3.38662207e-01 5.82170606e-01 -2.49855161e-01
-1.11458206e+00 -1.51814711e+00 -9.04955119e-02 6.19248867e-01
-2.31747106e-01 -5.01058280e-01 -5.92786491e-01 -5.75048439e-02] | [13.11286735534668, 7.694706916809082] |
ffffb61a-0ecc-4576-851e-23c433a7b8cf | pygsl-a-graph-structure-learning-toolkit | 2211.03583 | null | https://arxiv.org/abs/2211.03583v1 | https://arxiv.org/pdf/2211.03583v1.pdf | pyGSL: A Graph Structure Learning Toolkit | We introduce pyGSL, a Python library that provides efficient implementations of state-of-the-art graph structure learning models along with diverse datasets to evaluate them on. The implementations are written in GPU-friendly ways, allowing one to scale to much larger network tasks. A common interface is introduced for algorithm unrolling methods, unifying implementations of recent state-of-the-art techniques and allowing new methods to be quickly developed by avoiding the need to rebuild the underlying unrolling infrastructure. Implementations of differentiable graph structure learning models are written in PyTorch, allowing us to leverage the rich software ecosystem that exists e.g., around logging, hyperparameter search, and GPU-communication. This also makes it easy to incorporate these models as components in larger gradient based learning systems where differentiable estimates of graph structure may be useful, e.g. in latent graph learning. Diverse datasets and performance metrics allow consistent comparisons across models in this fast growing field. The full code repository can be found on https://github.com/maxwass/pyGSL. | ['Gonzalo Mateos', 'Max Wasserman'] | 2022-11-07 | null | null | null | null | ['graph-structure-learning'] | ['graphs'] | [-3.52842152e-01 9.15369317e-02 -5.67168117e-01 -2.27633864e-01
-4.24945623e-01 -7.59101748e-01 5.26695967e-01 3.80146295e-01
-2.60212123e-01 4.40551937e-01 1.76556751e-01 -6.92320943e-01
6.25265241e-02 -1.09046626e+00 -6.15126073e-01 -4.04528737e-01
-4.29294467e-01 5.13049603e-01 3.69007885e-01 -3.80567089e-02
2.88081139e-01 5.28413355e-01 -1.25419319e+00 2.33464465e-01
3.69697750e-01 3.49351823e-01 1.14853114e-01 1.07847202e+00
-2.30898187e-01 7.50564098e-01 -6.25542924e-02 -3.54905277e-01
2.72157937e-01 -4.31307107e-02 -8.94479930e-01 -1.25592321e-01
8.34506392e-01 -3.36760759e-01 -4.57335830e-01 7.20812619e-01
4.25115556e-01 6.49135560e-02 3.29836845e-01 -1.45522702e+00
-2.20963493e-01 6.22889102e-01 -4.97835726e-01 3.52861851e-01
2.39381105e-01 2.96976686e-01 1.20401609e+00 -6.69088840e-01
7.59892166e-01 1.25716317e+00 9.50504243e-01 1.77725822e-01
-1.48189068e+00 -5.93809962e-01 1.90219864e-01 2.17339337e-01
-9.52931583e-01 -3.92085940e-01 5.08960009e-01 -5.37986815e-01
1.32094717e+00 2.08431557e-01 8.71575177e-01 1.04552281e+00
8.46570805e-02 8.56655657e-01 8.05369139e-01 -3.27407777e-01
1.98601820e-02 -1.35166161e-02 5.51771700e-01 1.43854761e+00
2.41804212e-01 8.29315558e-02 -5.36129236e-01 -4.23489153e-01
8.66797805e-01 -7.29246214e-02 -5.77534810e-02 -6.97489619e-01
-1.17343676e+00 8.77888083e-01 5.87859333e-01 -7.46319145e-02
-8.01591426e-02 5.70093811e-01 7.56792665e-01 4.55652684e-01
4.34085965e-01 2.94046730e-01 -4.62204665e-01 -4.76689488e-01
-8.70524645e-01 2.35157028e-01 1.23610556e+00 1.03390229e+00
1.11047494e+00 -8.50192979e-02 4.87688091e-03 5.95279694e-01
2.78450161e-01 -3.07171047e-01 -5.01911044e-02 -9.93538916e-01
3.09070289e-01 7.49766588e-01 -4.29091066e-01 -9.07131612e-01
-6.21216893e-01 -6.51672840e-01 -5.44262826e-01 2.22795501e-01
4.96740729e-01 -7.44957551e-02 -5.23448706e-01 1.38411129e+00
5.41739464e-01 3.96733791e-01 -4.11958158e-01 5.49588382e-01
7.44148612e-01 4.92897958e-01 1.64068975e-02 1.80438310e-01
1.26477885e+00 -1.47917354e+00 -3.28713432e-02 -2.70971209e-01
1.26076901e+00 -7.38658130e-01 1.47894037e+00 3.16062629e-01
-9.23426449e-01 -1.07402071e-01 -8.72573853e-01 -4.39891040e-01
-6.58750594e-01 1.11781554e-02 1.31433737e+00 6.94261611e-01
-1.62521386e+00 1.04026616e+00 -1.47122097e+00 -6.92775965e-01
4.69440609e-01 3.47540736e-01 -2.61552006e-01 -5.26539609e-02
-6.87959194e-01 7.72110164e-01 4.13710713e-01 -1.48512498e-01
-8.58876944e-01 -9.53499436e-01 -6.90339923e-01 3.57190855e-02
4.61426407e-01 -1.02149022e+00 1.23439884e+00 -6.09872878e-01
-1.22390020e+00 8.76262963e-01 6.61074370e-02 -3.45158041e-01
7.51174033e-01 7.34774247e-02 1.48791015e-01 -1.08724602e-01
-6.79439232e-02 4.75244850e-01 6.40642643e-01 -5.92464983e-01
-1.52854264e-01 -3.39835882e-01 7.20810667e-02 2.56565064e-01
-4.66163605e-01 -8.78523588e-02 -7.55730867e-01 -3.69191408e-01
-2.46913746e-01 -9.51290250e-01 -2.91384906e-01 3.32721800e-01
-4.00444090e-01 -2.23788649e-01 8.70981812e-01 -5.13526618e-01
1.23739314e+00 -1.69808054e+00 4.54155318e-02 3.12064588e-01
6.59611166e-01 1.01453975e-01 -2.65576005e-01 1.00176096e+00
-3.24378684e-02 6.32176101e-02 -1.54602872e-02 -4.33971047e-01
9.38999429e-02 7.55619854e-02 8.65846798e-02 6.31791413e-01
-8.63393247e-02 9.74134505e-01 -1.00993931e+00 -3.20250243e-01
3.11908096e-01 3.98155659e-01 -6.89600945e-01 -2.80169584e-02
-3.91457200e-01 2.26997405e-01 -3.62333804e-01 5.62678576e-01
4.74160433e-01 -8.23855817e-01 4.96410280e-01 -8.56884420e-02
-1.61629379e-01 6.41129732e-01 -1.22974479e+00 1.73103642e+00
-5.80812871e-01 7.20025420e-01 3.05849433e-01 -9.02691841e-01
5.28654933e-01 -1.54184073e-01 1.59141779e-01 -1.57219410e-01
-8.74353293e-03 -5.71931824e-02 -2.51199216e-01 -1.12827167e-01
4.62789357e-01 6.00522161e-01 4.30885285e-01 9.10449266e-01
7.48209059e-02 -4.46290299e-02 5.90051889e-01 6.25468850e-01
1.51375079e+00 2.46361598e-01 3.15350831e-01 -3.71997982e-01
1.53252646e-01 1.29800320e-01 1.00008361e-02 8.15546155e-01
5.33612445e-02 -5.65648824e-02 7.16160476e-01 -5.90861380e-01
-1.08937621e+00 -1.04382825e+00 -7.43330941e-02 1.73543286e+00
-3.66442412e-01 -1.03683364e+00 -6.03775561e-01 -5.26754200e-01
1.93792060e-01 3.96757245e-01 -4.44985032e-01 1.53669611e-01
-4.17829096e-01 -6.51430488e-01 5.56335330e-01 5.70603848e-01
3.79980385e-01 -7.99135149e-01 -1.44185409e-01 2.34610856e-01
3.27955246e-01 -8.10268641e-01 -4.86018032e-01 8.47255588e-02
-1.18627560e+00 -1.23295045e+00 -3.69853348e-01 -7.71038294e-01
5.74968815e-01 2.80105978e-01 1.50823617e+00 5.56444228e-01
-5.39948523e-01 6.57857716e-01 -6.40033260e-02 -3.81693728e-02
-4.09936070e-01 6.93264425e-01 -2.31951639e-01 -6.66618168e-01
1.65144742e-01 -7.73983479e-01 -5.91354966e-01 1.01215020e-01
-5.60205340e-01 4.92697001e-01 1.25286236e-01 6.45573735e-01
4.87566203e-01 -2.82561064e-01 9.86565053e-02 -1.23779547e+00
9.16312993e-01 -6.32691681e-01 -1.05084443e+00 1.89948186e-01
-9.92298901e-01 2.54353881e-01 6.77873850e-01 -1.39039353e-01
-3.85080069e-01 -7.23693073e-02 -6.42219707e-02 -1.60209492e-01
2.99173966e-02 6.58632815e-01 4.46749747e-01 -3.78447860e-01
7.40938902e-01 5.30820303e-02 3.27977926e-01 -5.34574926e-01
6.84492469e-01 3.35086495e-01 2.26205796e-01 -7.10173607e-01
5.79302132e-01 2.59195924e-01 1.97346658e-01 -9.39628065e-01
-5.37444949e-01 -6.68360949e-01 -3.95600736e-01 -1.95759714e-01
3.58607680e-01 -8.62620056e-01 -8.72274518e-01 4.64236945e-01
-8.27105343e-01 -9.98620510e-01 1.69919252e-01 2.32536554e-01
-5.29808223e-01 4.17500854e-01 -1.02600098e+00 -2.10730135e-01
-5.27850509e-01 -1.14312863e+00 9.68911350e-01 6.39894232e-02
-2.18539596e-01 -1.69556022e+00 2.46643066e-01 2.66978890e-01
5.94157875e-01 1.85061976e-01 9.25921619e-01 -4.38312411e-01
-9.68960583e-01 -5.91814183e-02 -4.25859541e-01 5.03519326e-02
-1.38570338e-01 3.50869417e-01 -7.08281875e-01 -7.91090727e-01
-7.02060461e-01 -4.83585179e-01 9.77262974e-01 3.27863127e-01
1.22459364e+00 -4.55122471e-01 -6.52739644e-01 1.04602683e+00
1.30564451e+00 -7.54113555e-01 3.68169814e-01 3.94369423e-01
9.70344961e-01 3.13516259e-01 1.88420489e-01 2.46698245e-01
7.20259309e-01 5.09126723e-01 3.74595344e-01 -3.86261821e-01
-2.16790959e-01 -3.04943174e-01 3.05529594e-01 9.71098900e-01
5.98915331e-02 -2.24204902e-02 -1.28821743e+00 2.19987601e-01
-2.08803487e+00 -7.76335239e-01 -5.74735880e-01 2.09410930e+00
8.44293177e-01 9.54991728e-02 4.47871417e-01 -3.13927203e-01
5.98709643e-01 3.03022414e-01 -6.76743984e-01 -5.66176295e-01
4.11944449e-01 3.24760735e-01 8.63359213e-01 7.62769461e-01
-1.08183789e+00 1.28383291e+00 6.66251135e+00 7.69131899e-01
-1.19750249e+00 1.01431459e-01 3.82438689e-01 -1.03644148e-01
-1.78599000e-01 4.25671965e-01 -7.43681192e-01 2.09677011e-01
1.01021266e+00 -5.03694832e-01 9.57670331e-01 1.08833551e+00
2.68558890e-01 -3.40606794e-02 -1.19177365e+00 8.62004995e-01
-3.88654470e-01 -1.85550702e+00 -3.51027280e-01 2.21016005e-01
3.69537622e-01 8.30750585e-01 -3.30246508e-01 3.55204016e-01
7.70228624e-01 -8.69632006e-01 1.54600516e-01 1.88078061e-01
7.02646494e-01 -4.47615951e-01 1.86267078e-01 3.07846159e-01
-1.45442784e+00 2.49705493e-01 -3.97520602e-01 -3.68139982e-01
-1.03702620e-01 6.13627732e-01 -1.26157165e+00 3.26265335e-01
6.04469240e-01 1.09282362e+00 -1.01626658e+00 1.13072860e+00
-2.82932848e-01 8.48837674e-01 -6.49635911e-01 -2.45525807e-01
1.86072037e-01 -4.12710875e-01 4.54044491e-01 1.56506622e+00
-1.13849133e-01 -4.29441959e-01 4.26940769e-01 7.72626281e-01
-2.77081311e-01 3.25794250e-01 -7.55322933e-01 -2.44256750e-01
5.24859071e-01 1.67899203e+00 -1.01341188e+00 -2.95241356e-01
-5.18976808e-01 7.13071406e-01 9.48693156e-01 2.46401519e-01
-6.98208630e-01 -3.73898625e-01 6.56722546e-01 3.16196322e-01
4.50729355e-02 -7.63289392e-01 -1.63569227e-01 -1.22967172e+00
-1.90584615e-01 -9.02515411e-01 6.54129624e-01 -5.80290377e-01
-1.36536574e+00 2.99458504e-01 -5.68819940e-02 -6.73027158e-01
-1.17869586e-01 -7.69820988e-01 -8.44352961e-01 6.51222467e-01
-1.43318582e+00 -1.07813203e+00 -5.81178188e-01 6.54573023e-01
1.17195010e-01 -3.19632404e-02 8.58981490e-01 2.11204186e-01
-7.01606333e-01 6.08055353e-01 4.39102538e-02 1.35369703e-01
7.25730777e-01 -1.44599342e+00 1.05625856e+00 7.82496989e-01
2.33746171e-01 8.24437022e-01 4.33569610e-01 -6.58479750e-01
-1.74633849e+00 -1.07565415e+00 2.83153713e-01 -4.04995501e-01
1.34007800e+00 -7.13036239e-01 -9.20432746e-01 1.08622956e+00
1.57184586e-01 8.18939954e-02 6.33778512e-01 7.51781285e-01
-5.59787929e-01 2.02845000e-02 -6.89185023e-01 7.55614340e-01
1.29156601e+00 -5.41624069e-01 9.27959755e-02 8.27949822e-01
3.68227690e-01 -5.95910609e-01 -8.22898507e-01 -1.07975543e-01
5.27119637e-01 -8.86355281e-01 8.96326363e-01 -6.35963857e-01
1.38963282e-01 -1.49380639e-01 3.08312923e-01 -1.22384942e+00
-4.66972053e-01 -9.46176410e-01 -4.84604418e-01 9.16172206e-01
5.98857403e-01 -1.15071547e+00 1.21363306e+00 3.97020131e-01
-2.68134605e-02 -8.18751752e-01 -4.61328983e-01 -7.02756584e-01
-7.49526965e-03 -3.77333254e-01 3.48905802e-01 1.20712626e+00
4.14437167e-02 4.94010329e-01 -1.55547298e-02 3.91847901e-02
7.14100063e-01 7.29600117e-02 1.22079313e+00 -1.13525641e+00
-7.74100244e-01 -6.72197223e-01 -4.96153623e-01 -9.36973393e-01
2.68374979e-01 -1.76306129e+00 -6.83998406e-01 -1.86653030e+00
1.87511936e-01 -6.96734726e-01 5.18375188e-02 1.12786210e+00
-9.00928527e-02 2.09682852e-01 2.88988855e-02 2.54799157e-01
-4.96424437e-01 3.16556931e-01 1.07918143e+00 8.95551741e-02
-3.29467922e-01 -5.78518659e-02 -5.70183873e-01 6.57459617e-01
1.05587685e+00 -5.57218492e-01 -4.84315276e-01 -5.24999321e-01
3.42604220e-01 -2.02085555e-01 4.63203251e-01 -9.30148661e-01
4.07476455e-01 6.91337558e-03 1.31800890e-01 -2.94135362e-01
1.81844801e-01 -2.88122952e-01 5.31518534e-02 3.25728953e-01
-1.72833964e-01 4.65769887e-01 4.29946363e-01 4.03013438e-01
2.86147982e-01 -1.99759193e-02 6.34054124e-01 -2.66776830e-01
-6.86968327e-01 6.51198328e-01 -7.96819404e-02 2.80855417e-01
8.05297673e-01 -2.13489428e-01 -7.71126509e-01 -4.12128270e-01
-6.52255476e-01 4.00315702e-01 8.91244173e-01 2.62130827e-01
6.61458895e-02 -1.00554621e+00 -5.59876263e-01 6.27300143e-02
5.50229810e-02 -2.94519126e-01 -9.19529721e-02 9.92067575e-01
-1.10241938e+00 8.33672658e-02 -8.12203586e-02 -6.12665534e-01
-1.40846646e+00 2.83319265e-01 3.30573797e-01 -4.38978761e-01
-8.67213488e-01 8.10181499e-01 -2.23633051e-01 -7.47445941e-01
1.18056566e-01 -1.80314988e-01 3.84443551e-01 -1.46614864e-01
3.41866463e-01 6.19922042e-01 3.34491074e-01 1.33596048e-01
-3.90856594e-01 1.57395303e-01 -2.46534407e-01 2.67319173e-01
1.52601087e+00 1.78204373e-01 -5.25431693e-01 2.01352924e-01
1.34913838e+00 1.61718633e-02 -1.17745149e+00 -2.00272426e-02
1.08934715e-01 -2.03346819e-01 4.03244764e-01 -5.71030080e-01
-9.05679047e-01 7.10338533e-01 3.58693689e-01 2.53002763e-01
7.45431721e-01 -1.79197732e-02 4.76314336e-01 5.96675396e-01
4.74821001e-01 -9.72027421e-01 -9.58566889e-02 5.57880700e-01
5.45572937e-01 -1.01684964e+00 4.96738344e-01 -5.20376444e-01
-1.89357027e-01 1.43128824e+00 3.77422005e-01 -3.35774928e-01
7.65475035e-01 6.10244155e-01 -6.56176060e-02 -4.27915841e-01
-1.10801768e+00 -1.58088468e-02 -5.50819971e-02 5.13686776e-01
7.99553156e-01 1.59312785e-01 -1.45706952e-01 -2.32184544e-01
-4.29353833e-01 1.21859899e-02 7.22364783e-01 9.26803648e-01
-1.10761911e-01 -1.59432888e+00 3.01331505e-02 8.40977073e-01
-1.80529431e-01 -4.14965034e-01 -2.15298370e-01 8.57704759e-01
-7.24960923e-01 6.07717454e-01 -2.95900181e-02 -5.14704138e-02
-4.96907569e-02 -6.07885867e-02 6.50338888e-01 -8.12600017e-01
-6.26172960e-01 -3.02454323e-01 5.07458389e-01 -9.41850424e-01
7.11626112e-02 -3.28894973e-01 -1.11357498e+00 -8.20446908e-01
-1.86001837e-01 -5.54511510e-02 8.57286274e-01 4.00862128e-01
7.83035100e-01 3.31004769e-01 1.24800496e-01 -1.15297437e+00
-3.87039423e-01 -6.19279802e-01 -4.42010999e-01 -7.81021640e-02
8.60207751e-02 -4.92732406e-01 -3.20643783e-01 -1.81069955e-01] | [6.976859092712402, 5.923220634460449] |
5cccc5bc-23b1-405d-abe3-5d086acb6b1b | semantically-aware-attentive-neural | 1812.03402 | null | https://arxiv.org/abs/1812.03402v2 | https://arxiv.org/pdf/1812.03402v2.pdf | Semantically-Aware Attentive Neural Embeddings for Image-based Visual Localization | We present an approach that combines appearance and semantic information for 2D image-based localization (2D-VL) across large perceptual changes and time lags. Compared to appearance features, the semantic layout of a scene is generally more invariant to appearance variations. We use this intuition and propose a novel end-to-end deep attention-based framework that utilizes multimodal cues to generate robust embeddings for 2D-VL. The proposed attention module predicts a shared channel attention and modality-specific spatial attentions to guide the embeddings to focus on more reliable image regions. We evaluate our model against state-of-the-art (SOTA) methods on three challenging localization datasets. We report an average (absolute) improvement of $19\%$ over current SOTA for 2D-VL. Furthermore, we present an extensive study demonstrating the contribution of each component of our model, showing $8$--$15\%$ and $4\%$ improvement from adding semantic information and our proposed attention module. We finally show the predicted attention maps to offer useful insights into our model. | ['Han-Pang Chiu', 'Rakesh Kumar', 'Karan Sikka', 'Supun Samarasekera', 'Zachary Seymour'] | 2018-12-08 | null | null | null | null | ['deep-attention', 'image-based-localization', 'deep-attention'] | ['computer-vision', 'computer-vision', 'natural-language-processing'] | [-1.96420223e-01 -1.48893610e-01 1.84759706e-01 -5.42881787e-01
-8.17888379e-01 -3.09575230e-01 4.65202689e-01 1.88511714e-01
-4.14414555e-01 3.18301678e-01 2.68942297e-01 1.67239960e-02
8.62384811e-02 -4.22323495e-01 -8.53275120e-01 -2.88872242e-01
-4.66976345e-01 -1.08196974e-01 3.58775444e-02 -5.46229109e-02
3.09950382e-01 6.38827443e-01 -1.36201286e+00 4.06187475e-02
5.82705200e-01 1.36765516e+00 4.60125864e-01 4.53477770e-01
2.16918881e-03 5.50066173e-01 -2.38560960e-01 -1.45390064e-01
4.49032784e-01 -1.89119473e-01 -4.69561875e-01 1.00513682e-01
1.04888415e+00 -2.80417532e-01 -6.35454595e-01 8.89163077e-01
7.65233159e-01 3.78457129e-01 7.58333385e-01 -1.21268868e+00
-1.22931099e+00 -1.03370287e-01 -1.00633848e+00 2.56399274e-01
3.69827271e-01 2.86164969e-01 9.80543554e-01 -1.44689834e+00
6.99667037e-01 1.32122576e+00 6.15473986e-01 2.75476664e-01
-1.32713544e+00 -4.03283179e-01 6.95791781e-01 2.03758135e-01
-1.57892919e+00 -5.38066804e-01 9.30309951e-01 -4.37061608e-01
1.17180264e+00 -2.01186746e-01 3.02005082e-01 9.79673386e-01
2.97122836e-01 9.83209431e-01 1.17464495e+00 -4.20694441e-01
1.67190209e-01 1.87001958e-01 -1.27261683e-01 9.07313406e-01
2.97024827e-02 -5.07857800e-02 -7.47993946e-01 1.19246632e-01
8.51115704e-01 1.28144100e-01 -9.17442888e-02 -7.39554048e-01
-1.16275859e+00 6.03017926e-01 1.18272364e+00 1.33049386e-02
-4.34220016e-01 6.45110548e-01 1.62316307e-01 -3.96353155e-02
6.82353377e-01 4.91724342e-01 -3.27572405e-01 8.26567188e-02
-8.45773399e-01 5.42775262e-03 2.93378960e-02 1.12989700e+00
8.96674454e-01 3.05094700e-02 -3.24961901e-01 1.00504434e+00
5.76946497e-01 7.86435425e-01 -1.37396771e-02 -9.39921975e-01
1.84135675e-01 4.60377872e-01 3.32391530e-01 -1.23337007e+00
-4.00644630e-01 -6.49950147e-01 -4.81549531e-01 1.96525380e-01
-1.16528414e-01 1.92571983e-01 -1.06037474e+00 2.03444624e+00
7.25515857e-02 1.26788616e-01 -3.30902159e-01 1.11946130e+00
7.53788292e-01 4.57511663e-01 2.38458812e-01 2.62102902e-01
1.05749023e+00 -1.26212454e+00 -5.47098815e-01 -4.85040694e-01
2.85149425e-01 -7.48678386e-01 1.14961529e+00 -8.29924345e-02
-1.01785457e+00 -9.01199579e-01 -9.07871008e-01 -3.53828818e-01
-4.47856188e-01 4.04588848e-01 6.32943809e-01 1.32396609e-01
-1.36547530e+00 1.25306383e-01 -6.80530906e-01 -8.40940893e-01
5.83608270e-01 2.40749672e-01 -5.88655353e-01 -3.01071584e-01
-6.93603396e-01 8.01448464e-01 -1.98562145e-01 1.83574095e-01
-1.13500392e+00 -6.32418275e-01 -1.16809392e+00 -1.98304519e-01
1.50155108e-02 -6.90633774e-01 8.20741296e-01 -6.21127665e-01
-1.14306223e+00 9.15556669e-01 -6.68866038e-01 -2.30844676e-01
1.73781455e-01 -3.69434863e-01 -1.64853662e-01 2.43113846e-01
3.64770889e-01 1.29218256e+00 7.86027610e-01 -1.59006536e+00
-7.10596979e-01 -3.03520560e-01 9.57228243e-02 2.81967700e-01
-1.50540903e-01 -1.94947362e-01 -8.67068589e-01 -6.65204763e-01
2.24527434e-01 -8.46416473e-01 -3.03975344e-01 5.94289005e-01
-3.23060125e-01 -1.14856042e-01 7.14584291e-01 -4.98548925e-01
8.98488522e-01 -2.20852160e+00 2.58582264e-01 -4.61413600e-02
3.45891148e-01 -2.17771471e-01 -6.52605593e-01 4.15552914e-01
7.49252513e-02 4.02368046e-03 -1.58161521e-02 -9.16519105e-01
2.72391826e-01 -1.16761751e-01 -2.30396748e-01 6.51570797e-01
4.52094644e-01 1.11980975e+00 -1.01398337e+00 -1.57659411e-01
6.98913276e-01 6.32793128e-01 -4.51229453e-01 4.33435649e-01
-6.84824735e-02 3.28140259e-01 -3.08121592e-01 9.67643619e-01
8.76801014e-01 -2.84059495e-01 -2.09469974e-01 -4.96031106e-01
-7.54709169e-02 5.46555594e-02 -7.29852557e-01 2.27520108e+00
-8.04687262e-01 1.02103102e+00 5.59957847e-02 -6.97231829e-01
9.31210935e-01 -3.19772035e-01 3.81405056e-01 -1.09243166e+00
3.57062370e-02 9.27005410e-02 -3.33121866e-01 -4.06092823e-01
6.25650764e-01 3.94631803e-01 -2.33149320e-01 2.70281583e-01
4.53330636e-01 -5.41560352e-02 -2.53889203e-01 3.50264937e-01
1.01296329e+00 3.91785502e-01 2.71420460e-02 -3.49519134e-01
3.55088323e-01 -2.84316063e-01 2.41390809e-01 8.76543462e-01
-5.71799159e-01 6.51756108e-01 2.19162330e-01 -4.06400979e-01
-9.62036848e-01 -1.33055305e+00 -3.12660411e-02 1.15670335e+00
6.56749785e-01 -1.90516844e-01 -3.46969038e-01 -8.17582369e-01
1.97732121e-01 5.93996525e-01 -1.01090431e+00 3.74224596e-02
-1.06287345e-01 -2.18405098e-01 2.46680483e-01 7.36902177e-01
5.98646998e-01 -8.08947265e-01 -4.27149653e-01 -1.20412856e-01
-8.55387002e-02 -1.17090762e+00 -5.67895353e-01 1.00905292e-01
-4.48604494e-01 -8.25885653e-01 -7.19016254e-01 -8.69046628e-01
9.78422821e-01 8.03197861e-01 9.31183100e-01 -2.64463216e-01
-3.93280864e-01 8.79859269e-01 -3.98343474e-01 -1.90028608e-01
3.73939842e-01 -4.84937504e-02 2.86833733e-01 1.24839731e-01
3.56372297e-01 -4.94768590e-01 -1.08800340e+00 2.14682877e-01
-5.56874156e-01 -9.75521356e-02 9.07272696e-01 8.25484335e-01
7.80991077e-01 -5.49823284e-01 3.42104912e-01 -2.44150966e-01
4.35424030e-01 -5.20315886e-01 -4.69585806e-01 1.63167208e-01
-4.10137504e-01 9.98940915e-02 3.57009679e-01 -1.17083147e-01
-7.65554786e-01 1.02511704e-01 -7.40753338e-02 -8.51675093e-01
-3.11334103e-01 2.06325933e-01 -8.76179561e-02 -5.21465361e-01
3.79851401e-01 2.70135820e-01 -2.53356606e-01 -4.98144686e-01
8.06004703e-01 4.79378521e-01 4.10432428e-01 -5.09596169e-01
7.39529252e-01 7.66649663e-01 -1.53113350e-01 -8.16744089e-01
-7.17304587e-01 -4.89278913e-01 -7.08262622e-01 -2.26212040e-01
7.74864972e-01 -1.23701298e+00 -5.52601516e-01 3.23886901e-01
-1.17733109e+00 -4.71281976e-01 -1.98001176e-01 4.07445580e-01
-5.57559073e-01 2.50368953e-01 -3.06410372e-01 -8.23513031e-01
-1.16501354e-01 -1.22892797e+00 1.54680169e+00 2.12152854e-01
2.18748837e-03 -1.04995370e+00 1.58763617e-01 1.00998595e-01
5.42762876e-01 1.80760771e-01 5.93095601e-01 -1.24883749e-01
-8.28320205e-01 -1.41435549e-01 -9.00457442e-01 1.31503105e-01
1.66302934e-01 -2.79813349e-01 -1.21197486e+00 -5.34787416e-01
-5.58825910e-01 -2.70462185e-01 1.08971334e+00 5.49933195e-01
1.20018125e+00 8.37361142e-02 -5.01887321e-01 8.45223963e-01
1.67760491e+00 -1.10745035e-01 5.37613809e-01 9.55630764e-02
7.21489489e-01 4.17865187e-01 7.63756156e-01 2.77843297e-01
6.54902220e-01 9.58304465e-01 5.79280376e-01 -4.92452860e-01
-4.69829828e-01 -5.53383648e-01 3.48847866e-01 5.27940989e-01
1.99125454e-01 -3.27286750e-01 -6.43014550e-01 1.05845940e+00
-1.78428411e+00 -5.67753315e-01 2.52427310e-01 2.01012325e+00
3.56410831e-01 -1.54299065e-01 -9.03235972e-02 -6.24561727e-01
4.52306747e-01 3.52500528e-01 -5.08016586e-01 -3.42930734e-01
-1.47579581e-01 2.86338240e-01 5.79447448e-01 6.51049495e-01
-1.20774937e+00 1.16591311e+00 6.58911562e+00 7.94642270e-01
-1.38877428e+00 2.78082967e-01 4.99538898e-01 -2.42294207e-01
-3.48287612e-01 -1.78584859e-01 -5.76443374e-01 3.14238667e-01
4.15106267e-01 2.62779802e-01 2.80688882e-01 8.81906331e-01
3.44716281e-01 -3.80961090e-01 -1.14182460e+00 1.20311773e+00
4.54390228e-01 -1.31090879e+00 3.41989473e-02 9.11356956e-02
9.22769368e-01 3.85231704e-01 5.05756974e-01 8.99958834e-02
7.83933699e-02 -9.81679678e-01 9.69163179e-01 4.77408826e-01
1.10409820e+00 -5.83089471e-01 6.48335636e-01 -2.79387057e-01
-1.41842639e+00 -1.15256168e-01 -4.41807717e-01 3.95059921e-02
1.36487588e-01 3.35210502e-01 -6.43600345e-01 4.95736748e-01
8.26795697e-01 9.92333353e-01 -9.66704071e-01 1.10681283e+00
-4.22199398e-01 -1.28508136e-02 -1.54940382e-01 -9.86299887e-02
4.32825863e-01 2.15373278e-01 4.73900616e-01 1.39541864e+00
3.64077061e-01 -2.57582158e-01 6.79093152e-02 1.12043083e+00
-1.19653031e-01 1.78515445e-02 -8.81480396e-01 2.06308544e-01
5.22243142e-01 1.23731589e+00 -6.17417872e-01 -6.55206069e-02
-4.49687451e-01 1.55178535e+00 6.18009388e-01 6.36298656e-01
-9.29043174e-01 -6.15638852e-01 1.05264425e+00 5.74916415e-02
6.03768826e-01 -6.46536052e-01 -7.74627477e-02 -1.12695634e+00
-5.97780831e-02 -1.84363931e-01 -4.65635806e-02 -1.10631180e+00
-1.49349582e+00 6.45902634e-01 -2.60079861e-01 -1.19026411e+00
3.61697644e-01 -7.13478744e-01 -2.59239674e-01 9.68045056e-01
-1.79245794e+00 -1.49187887e+00 -5.48477829e-01 4.26166564e-01
6.09432995e-01 7.81238154e-02 7.53852606e-01 2.79641360e-01
-3.99798185e-01 9.64767396e-01 1.28675446e-01 1.10348798e-01
1.00153017e+00 -1.16658401e+00 4.87540781e-01 1.00602651e+00
2.92265147e-01 8.06937993e-01 4.00574327e-01 -3.44758689e-01
-1.54514492e+00 -1.40056443e+00 7.70140111e-01 -6.90015137e-01
4.67060745e-01 -6.86951041e-01 -3.80966574e-01 6.56455815e-01
4.37118173e-01 4.48816687e-01 5.54649830e-01 2.50001818e-01
-6.13575816e-01 -4.13041413e-01 -1.12221909e+00 6.07669473e-01
1.34481430e+00 -8.32663596e-01 -2.19358265e-01 -5.85005470e-02
7.21678138e-01 -3.03761482e-01 -6.19861901e-01 1.70518890e-01
6.02239072e-01 -9.63501334e-01 1.05398870e+00 -2.01902688e-01
3.39901567e-01 -5.97021699e-01 -6.03240609e-01 -1.36765373e+00
-6.61823690e-01 -2.47591898e-01 -1.07842535e-02 7.69289851e-01
3.44049603e-01 -4.23993826e-01 4.52386826e-01 3.72589111e-01
-3.49364460e-01 -7.30146170e-01 -9.45354939e-01 -6.48286819e-01
-1.64422750e-01 -5.76503038e-01 1.70576841e-01 7.62255669e-01
-1.16335832e-01 1.70071051e-01 -4.77759331e-01 4.78111923e-01
8.38033199e-01 5.22920564e-02 8.36145401e-01 -6.12856925e-01
3.62533182e-02 -2.79013634e-01 -8.67822289e-01 -1.56018174e+00
6.78106770e-02 -6.26804113e-01 1.68162957e-01 -1.67779064e+00
3.33690822e-01 -4.52487230e-01 -7.74520636e-01 4.70687866e-01
-3.43063846e-02 7.73164392e-01 2.30963588e-01 -1.65349133e-02
-1.17572033e+00 8.23987782e-01 9.64712620e-01 -2.54755229e-01
-9.76971462e-02 -7.08610654e-01 -7.91400135e-01 3.52907509e-01
3.12688708e-01 -1.15381286e-01 -4.31469500e-01 -9.40328538e-01
8.65825117e-02 -4.98763412e-01 6.88019991e-01 -8.88898253e-01
1.54168293e-01 -1.28135666e-01 6.77217364e-01 -6.14723027e-01
5.88226736e-01 -8.07663441e-01 -3.32549155e-01 1.70033842e-01
-3.91573846e-01 -7.96904638e-02 5.45976639e-01 8.51593912e-01
-1.24877296e-01 4.61672157e-01 6.84739232e-01 7.72287399e-02
-1.29511464e+00 4.71451461e-01 -1.87763557e-01 -2.19584078e-01
1.07965136e+00 -2.02906638e-01 -1.76093817e-01 -5.23041546e-01
-7.14364529e-01 4.00408745e-01 7.51405120e-01 7.25832582e-01
1.02363801e+00 -1.66455269e+00 -4.67600018e-01 3.39295477e-01
6.50503993e-01 -2.87575394e-01 4.76798177e-01 9.49968398e-01
-5.06535709e-01 6.08077228e-01 -3.62060696e-01 -8.25336754e-01
-8.54639113e-01 6.28994823e-01 3.06533694e-01 2.75379837e-01
-3.51407558e-01 1.41166699e+00 5.31005383e-01 -3.36089730e-01
3.58415723e-01 -3.40721995e-01 3.01633447e-01 -2.36689806e-01
3.76733720e-01 4.33292091e-02 -2.79447615e-01 -7.31567383e-01
-8.06836009e-01 9.16953027e-01 -6.75704563e-03 -1.71001717e-01
1.39447999e+00 -6.58773065e-01 7.01219589e-03 3.03220153e-01
1.33180261e+00 2.80925725e-02 -1.65467286e+00 -3.78371179e-01
-4.93191063e-01 -8.62115860e-01 1.84929103e-01 -9.78337348e-01
-1.01691115e+00 1.26181352e+00 1.02522147e+00 -2.77687699e-01
1.21117353e+00 2.39773065e-01 5.32868087e-01 6.07617795e-02
6.15857244e-01 -8.96594524e-01 3.44472498e-01 3.63315403e-01
1.05096519e+00 -1.55903614e+00 -4.22234200e-02 -8.23114291e-02
-5.33965349e-01 7.93696940e-01 8.28498542e-01 -2.95784235e-01
6.34634495e-01 -1.75805390e-01 6.07935190e-02 -2.79695481e-01
-5.64304829e-01 -5.31082511e-01 5.13917327e-01 9.10497963e-01
3.28634471e-01 -1.67494372e-01 9.94402245e-02 3.74593198e-01
5.75254142e-01 -2.42540330e-01 -2.06230637e-02 6.70939386e-01
-3.00720692e-01 -7.77304590e-01 -1.13370948e-01 -7.08250180e-02
2.70300712e-02 -1.03046775e-01 -4.82065588e-01 6.58554554e-01
4.34485137e-01 8.74053001e-01 2.33825505e-01 -5.33723295e-01
2.06016257e-01 -1.94264084e-01 7.77682245e-01 -5.05441248e-01
7.22430795e-02 8.23124945e-02 -2.87000984e-01 -1.04569507e+00
-4.63540405e-01 -4.48268950e-01 -8.96698236e-01 -2.21397340e-01
3.50201353e-02 -2.97039717e-01 8.09339881e-01 5.14039397e-01
1.02540302e+00 4.97840762e-01 7.27520347e-01 -1.30335331e+00
1.79394796e-01 -9.06043768e-01 -4.39575046e-01 3.95177305e-01
5.13435662e-01 -9.88472104e-01 -2.31408671e-01 -1.67675331e-01] | [7.942440509796143, -2.087804079055786] |
f47c6013-dd0a-42ce-b4f6-0d785ad6c378 | generative-adversarial-nets-for-multiple-text | 1712.09127 | null | http://arxiv.org/abs/1712.09127v1 | http://arxiv.org/pdf/1712.09127v1.pdf | Generative Adversarial Nets for Multiple Text Corpora | Generative adversarial nets (GANs) have been successfully applied to the
artificial generation of image data. In terms of text data, much has been done
on the artificial generation of natural language from a single corpus. We
consider multiple text corpora as the input data, for which there can be two
applications of GANs: (1) the creation of consistent cross-corpus word
embeddings given different word embeddings per corpus; (2) the generation of
robust bag-of-words document embeddings for each corpora. We demonstrate our
GAN models on real-world text data sets from different corpora, and show that
embeddings from both models lead to improvements in supervised learning
problems. | ['Baiyang Wang', 'Diego Klabjan'] | 2017-12-25 | null | https://openreview.net/forum?id=BkexaxBKPB | https://openreview.net/pdf?id=BkexaxBKPB | null | ['cross-corpus'] | ['computer-vision'] | [ 5.84499538e-01 3.37786913e-01 1.47571340e-01 -2.00077340e-01
-1.01076424e+00 -5.11039913e-01 1.15712786e+00 -3.48850101e-01
-3.44144613e-01 7.79865444e-01 6.04695857e-01 -2.04202473e-01
5.05860627e-01 -9.95624840e-01 -5.35935879e-01 -6.51915491e-01
2.93666452e-01 6.79389000e-01 -1.46177739e-01 -4.47244525e-01
-1.69261515e-01 1.10587478e-01 -1.24984288e+00 2.59465039e-01
7.23112047e-01 6.07012391e-01 2.77372617e-02 8.69147182e-01
-5.61818004e-01 6.90974414e-01 -1.11115479e+00 -9.02742624e-01
5.08379161e-01 -8.86257529e-01 -5.69890320e-01 1.80918217e-01
4.06165391e-01 -1.86797500e-01 -4.07454818e-01 1.06960595e+00
5.12994289e-01 -7.52004161e-02 9.58398104e-01 -1.49660981e+00
-1.43637013e+00 6.95248127e-01 -4.26667809e-01 -2.22810432e-01
3.45615819e-02 2.89494991e-01 1.00373685e+00 -8.72706115e-01
7.02252924e-01 1.30863345e+00 2.27574855e-01 1.02039146e+00
-1.40491438e+00 -6.04332387e-01 -2.08356962e-01 -2.19491139e-01
-1.15947223e+00 -1.79870009e-01 7.50853240e-01 -5.29988408e-01
8.30617666e-01 -1.80250723e-02 7.28135049e-01 1.68121672e+00
3.87042165e-01 8.25770199e-01 7.83342361e-01 -7.05666661e-01
2.37595528e-01 1.15061857e-01 -4.29653257e-01 1.70462623e-01
2.18836039e-01 9.76968929e-02 -4.20521408e-01 -8.28749612e-02
8.27053785e-01 -2.25883707e-01 -1.05340168e-01 -1.06864095e-01
-1.32474124e+00 1.49843001e+00 2.91842699e-01 4.83677626e-01
-3.30086112e-01 5.37845731e-01 4.70350832e-01 3.54929179e-01
8.17939997e-01 7.42952228e-01 1.47509407e-02 -5.09692132e-02
-6.27735198e-01 5.74438632e-01 4.95589226e-01 1.24281228e+00
5.07642448e-01 5.69908082e-01 -2.44687125e-01 9.45300341e-01
1.34664863e-01 6.84466243e-01 1.07971489e+00 -3.61437500e-01
6.48291230e-01 3.45662832e-01 -3.80005166e-02 -8.03712308e-01
2.93804705e-01 1.10914215e-01 -1.01354551e+00 1.01979978e-01
9.38017145e-02 -4.36113566e-01 -1.35501146e+00 1.94719720e+00
-1.26916066e-01 9.54686031e-02 5.41011631e-01 3.81455123e-01
6.73851907e-01 9.99551415e-01 5.88056445e-02 1.87444851e-01
1.20563328e+00 -1.04432118e+00 -9.51513648e-01 -7.41044700e-01
4.76325393e-01 -9.18550432e-01 9.29148078e-01 1.59905300e-01
-9.78284538e-01 -6.75738931e-01 -1.06816840e+00 -1.48300260e-01
-8.98193002e-01 -1.68336973e-01 4.62997377e-01 6.78369880e-01
-1.15374303e+00 4.95823808e-02 -5.21155715e-01 -1.90607205e-01
5.16013920e-01 6.46162555e-02 -3.80801499e-01 -2.45936751e-01
-1.33525252e+00 9.30086851e-01 6.13432646e-01 -7.72864297e-02
-8.86486173e-01 -4.96066213e-01 -1.02412903e+00 -1.14584677e-01
-2.04365309e-02 -9.18206215e-01 1.13223529e+00 -1.59829795e+00
-1.29974318e+00 8.94460022e-01 9.33076069e-02 -5.82256854e-01
3.86028141e-01 -1.61130428e-01 -5.05237758e-01 -1.58537850e-01
1.00606039e-01 1.04843903e+00 1.20177603e+00 -1.30229676e+00
-4.06740248e-01 -1.09363496e-01 -1.55701742e-01 3.76774818e-02
-5.10521770e-01 -4.12050448e-02 -6.40955865e-02 -1.38795412e+00
-5.69358766e-01 -9.44125950e-01 -3.44948649e-01 -1.84079796e-01
-4.15370524e-01 -3.05350900e-01 7.37210155e-01 -6.35479689e-01
7.56334066e-01 -1.94784570e+00 4.42919612e-01 -2.02819601e-01
2.20409989e-01 2.82396615e-01 -7.27903068e-01 5.74281514e-01
-2.27787912e-01 4.90135938e-01 -3.31308067e-01 -5.75800180e-01
8.11080933e-02 5.11594296e-01 -7.80420959e-01 -1.45329803e-01
6.74462438e-01 1.38397145e+00 -1.01099896e+00 -4.78703648e-01
2.42119916e-02 6.62558973e-01 -3.54926676e-01 6.38263881e-01
-3.17985237e-01 5.76191247e-02 -3.01306635e-01 1.45124912e-01
3.78043771e-01 1.00224661e-02 1.28403548e-02 1.49438411e-01
4.44425523e-01 1.50070935e-01 -6.67537391e-01 1.76974869e+00
-5.98004937e-01 8.39668870e-01 -5.07863641e-01 -8.86997819e-01
9.66215432e-01 5.53047836e-01 2.45249733e-01 -4.87297684e-01
9.16228220e-02 5.64168356e-02 6.88095391e-02 -2.58517921e-01
7.21996725e-01 -5.16934454e-01 -4.10685360e-01 8.06598842e-01
6.31892025e-01 -8.22052598e-01 3.72550786e-01 2.13644296e-01
9.77188468e-01 -3.86588946e-02 1.54986680e-01 -4.81293984e-02
1.08397178e-01 7.49214292e-02 1.65664613e-01 4.97001350e-01
1.73308074e-01 1.04216707e+00 2.27618083e-01 -5.82221925e-01
-1.57923162e+00 -1.11894226e+00 1.51379004e-01 8.76512229e-01
-2.97878921e-01 -3.77427340e-01 -8.56251657e-01 -7.96322584e-01
-1.56265318e-01 8.54464412e-01 -8.44593108e-01 -2.24502504e-01
-5.26756763e-01 -8.29264104e-01 7.27539599e-01 7.20671058e-01
1.48750082e-01 -1.37864852e+00 -3.65997821e-01 2.20839396e-01
-1.42498910e-01 -1.19937229e+00 -7.43661284e-01 1.45940080e-01
-5.39638638e-01 -7.63905108e-01 -8.14667225e-01 -1.02291405e+00
8.19234073e-01 3.42329517e-02 1.50343430e+00 -7.98674859e-03
-3.25774401e-01 4.42750722e-01 -5.30803621e-01 -8.11826169e-01
-1.15272510e+00 1.05183817e-01 3.21554323e-03 4.02728058e-02
5.08610308e-01 -4.39197600e-01 -1.18350275e-01 6.87284693e-02
-1.61688101e+00 3.54952574e-01 4.15146410e-01 1.28157687e+00
3.35818976e-01 -7.62279183e-02 8.58876228e-01 -9.87866521e-01
1.20225382e+00 -4.92158115e-01 -5.23970723e-01 1.86378881e-01
-5.57263494e-01 1.23382539e-01 7.94681847e-01 -7.51812577e-01
-9.11314070e-01 -2.44276062e-01 -2.56548703e-01 -2.91304499e-01
-1.27155378e-01 4.58058923e-01 -5.20281196e-02 4.41440642e-01
7.40183055e-01 3.49914491e-01 5.49272001e-02 4.18893695e-02
9.78936553e-01 8.19400966e-01 6.46882832e-01 -6.96907103e-01
1.16296768e+00 4.17084917e-02 -4.05326366e-01 -8.13649118e-01
-3.59141082e-01 1.55916020e-01 -6.91482544e-01 1.71849906e-01
1.27599823e+00 -9.62747335e-01 4.01016593e-01 4.07244414e-01
-1.41946805e+00 -4.89530861e-01 -8.10089946e-01 1.74426854e-01
-6.56357586e-01 -4.81425747e-02 -4.29626286e-01 -5.61753452e-01
-5.51557899e-01 -1.19658554e+00 1.12411702e+00 1.04604654e-01
-4.98894602e-01 -1.48601842e+00 4.13179725e-01 4.05434854e-02
4.55413491e-01 5.20459592e-01 1.07075679e+00 -7.30721474e-01
-1.93930522e-01 -4.58204329e-01 1.72252461e-01 8.81697237e-01
5.35196364e-01 2.78880876e-02 -1.04503524e+00 -4.19694066e-01
-4.03616764e-02 -6.52868211e-01 4.99624372e-01 1.63205132e-01
9.19327617e-01 -4.64614064e-01 -1.39896244e-01 4.06381667e-01
1.41170442e+00 4.67158228e-01 9.93466973e-01 4.39241864e-02
7.88869739e-01 4.54925656e-01 7.23444251e-03 1.44010363e-02
1.88161880e-02 5.98259032e-01 3.11159641e-01 -3.55811030e-01
-2.91584164e-01 -6.59996450e-01 4.57134515e-01 1.17410719e+00
2.25164920e-01 -7.69611359e-01 -8.47060800e-01 1.13772941e+00
-1.64973783e+00 -1.00701892e+00 2.78662503e-01 1.89167392e+00
9.08792853e-01 -4.27891426e-02 -2.54205428e-02 -1.64701149e-01
7.96942770e-01 3.66637051e-01 -4.84533221e-01 -5.45622230e-01
-1.86361179e-01 5.92274845e-01 2.00105861e-01 2.80409604e-01
-9.20789897e-01 9.24087405e-01 7.39624405e+00 8.12876880e-01
-1.00404596e+00 2.34153628e-01 6.96293473e-01 -6.62581474e-02
-6.82497144e-01 -1.55141667e-01 -4.43874389e-01 6.87119246e-01
8.76927018e-01 -6.31668866e-01 4.16275471e-01 7.87311554e-01
-3.50465089e-01 4.18946773e-01 -1.24327838e+00 9.84202683e-01
3.42359155e-01 -1.51508462e+00 6.51388049e-01 2.31973007e-01
1.33494878e+00 -1.12043388e-01 1.46741897e-01 2.80471236e-01
9.87297595e-01 -1.49980175e+00 6.22330129e-01 8.24537948e-02
1.03817534e+00 -8.16535652e-01 8.22856247e-01 1.14795215e-01
-8.53508115e-01 2.88893640e-01 -4.51268673e-01 2.83074856e-01
3.41380268e-01 5.10119081e-01 -9.40450251e-01 4.71298695e-01
2.29545206e-01 6.10913634e-01 -4.81265128e-01 2.80230582e-01
-5.11993587e-01 5.19809186e-01 5.01307310e-04 -1.05072362e-02
2.57836759e-01 -3.08573157e-01 2.38843217e-01 1.05762029e+00
4.45946544e-01 -2.97841519e-01 -8.73916596e-02 1.25013995e+00
-4.89456862e-01 -3.30362134e-02 -1.11896980e+00 -6.23690009e-01
3.61840785e-01 1.15643191e+00 -5.13189971e-01 -4.18322563e-01
-4.87198979e-01 1.15435445e+00 2.66653836e-01 4.23713326e-01
-6.91871941e-01 -4.78522569e-01 9.18377399e-01 -7.27093890e-02
3.17725271e-01 -2.94014543e-01 -3.15846443e-01 -1.17063451e+00
-9.74593759e-02 -1.08274460e+00 1.71293452e-01 -9.76380527e-01
-1.76549673e+00 1.05673814e+00 -7.21742734e-02 -1.06095827e+00
-8.32822263e-01 -6.05744660e-01 -8.13084185e-01 1.08690584e+00
-1.15714860e+00 -1.36604583e+00 -1.92251310e-01 4.35269296e-01
9.31383848e-01 -7.65591025e-01 1.23972249e+00 1.10519836e-02
-1.75182566e-01 6.67442143e-01 2.26420090e-01 5.07315278e-01
6.38189256e-01 -1.45390272e+00 1.30008328e+00 1.09210467e+00
7.57971883e-01 4.62807596e-01 5.47966838e-01 -5.81362247e-01
-1.27286780e+00 -1.30026495e+00 7.05241263e-01 -6.51718915e-01
6.82899415e-01 -7.71975875e-01 -6.29664242e-01 1.06323004e+00
8.83011222e-01 -1.71290904e-01 8.35257113e-01 -3.21518183e-01
-3.01774561e-01 2.84398526e-01 -1.10414267e+00 8.99879754e-01
7.43738592e-01 -5.65765500e-01 -6.20865107e-01 3.09868425e-01
1.03022158e+00 -3.29150885e-01 -8.84686768e-01 -9.93143246e-02
4.65727568e-01 -4.04546410e-01 8.29172790e-01 -9.63855505e-01
1.16385436e+00 -6.01951107e-02 -1.43917859e-01 -2.02593899e+00
-5.29944487e-02 -5.96734047e-01 2.40964383e-01 1.41597474e+00
4.40769225e-01 -6.60522819e-01 4.34331656e-01 5.55474341e-01
1.28092300e-02 -5.63756585e-01 -9.14548755e-01 -7.41032243e-01
7.40098059e-01 -4.87309337e-01 9.61787403e-01 9.29958880e-01
-3.35451305e-01 7.18534768e-01 -5.62951446e-01 -3.86845589e-01
2.92333126e-01 -1.28472388e-01 1.16936457e+00 -9.61582482e-01
-2.06180677e-01 -4.77230281e-01 -5.28015018e-01 -1.00342917e+00
3.49180937e-01 -9.47317719e-01 2.78089643e-01 -1.66283441e+00
6.94065168e-03 -2.43569911e-01 1.59163252e-01 6.06805205e-01
-2.93081135e-01 4.35324192e-01 2.74293363e-01 -4.51821052e-02
9.67144370e-02 8.70528400e-01 1.17636037e+00 -5.24825811e-01
2.71930426e-01 -4.90140975e-01 -8.82768810e-01 3.19782615e-01
7.40491927e-01 -6.56320810e-01 -7.43433297e-01 -9.03448701e-01
2.46044189e-01 -3.37398201e-01 5.52620552e-02 -7.17812836e-01
-1.77494571e-01 -3.05863470e-01 2.91115373e-01 -6.90086279e-03
3.72893900e-01 -7.50858366e-01 1.91417828e-01 2.34297886e-01
-5.23058534e-01 4.76385236e-01 2.04922989e-01 4.76835400e-01
-3.82503778e-01 -4.29548383e-01 7.30645895e-01 -2.61432856e-01
-3.30212623e-01 1.93911746e-01 -3.54958624e-01 5.32591045e-01
1.01716733e+00 -6.34983033e-02 -5.03493667e-01 -6.58667028e-01
-3.04161191e-01 5.73865371e-03 5.74835479e-01 9.97277379e-01
7.23161161e-01 -2.00298572e+00 -1.14059818e+00 3.56667817e-01
2.73741305e-01 2.15934411e-01 -2.41214260e-01 -1.98754191e-01
-3.96035343e-01 1.59272358e-01 -2.92071760e-01 -2.74565250e-01
-9.28232193e-01 6.31159484e-01 3.11585903e-01 -4.01481271e-01
-2.90936321e-01 6.88922048e-01 4.65527654e-01 -2.70021468e-01
-2.17965245e-01 -1.28369406e-01 1.61988229e-01 7.00806326e-04
5.62773287e-01 -8.53899717e-02 -2.16493413e-01 -6.69669688e-01
2.22120240e-01 3.56925160e-01 -1.43863484e-01 -4.80243206e-01
1.55087197e+00 3.29697579e-01 1.38872135e-02 3.95092338e-01
1.22006667e+00 -1.48234501e-01 -1.03822315e+00 -1.09965444e-01
-2.70142674e-01 -4.81629521e-01 -2.15603650e-01 -5.57173312e-01
-1.30201566e+00 8.48398209e-01 4.78567123e-01 5.37716985e-01
9.35050786e-01 -1.63237318e-01 9.38499272e-01 -2.41061486e-02
3.32633585e-01 -9.56424356e-01 4.53849047e-01 3.52385998e-01
1.14865518e+00 -1.20730913e+00 -1.78114384e-01 5.09082191e-02
-8.46351087e-01 1.13636029e+00 5.82197189e-01 -2.90545464e-01
3.93969208e-01 3.30581665e-01 2.77994305e-01 -5.40498421e-02
-8.33613753e-01 -9.36884359e-02 2.16494858e-01 1.10123873e+00
5.71444571e-01 1.08532570e-01 -2.18770415e-01 4.06374723e-01
-5.55331230e-01 -2.09964603e-01 7.56506383e-01 7.35929370e-01
1.66519120e-01 -1.55432582e+00 -2.01795936e-01 5.01806498e-01
-3.23570639e-01 -3.07619810e-01 -5.52110493e-01 7.32974768e-01
1.52265131e-01 9.21185017e-01 3.13134849e-01 -3.63951594e-01
3.71028222e-02 4.16156560e-01 4.70909476e-01 -9.35810983e-01
-3.88332129e-01 -3.37171942e-01 -4.94687594e-02 4.37628180e-02
-3.51628572e-01 -3.57426763e-01 -8.66915762e-01 -1.01931334e-01
-1.49399042e-01 2.24539742e-01 7.07290292e-01 8.80553842e-01
3.09283197e-01 5.53764880e-01 5.88442206e-01 -8.84093106e-01
-4.26791966e-01 -1.28402591e+00 -4.42846209e-01 8.79416764e-01
1.27753243e-01 -3.15595627e-01 -4.36159760e-01 5.40113091e-01] | [11.901880264282227, 9.40964412689209] |
56ff792a-9328-4ebc-874f-8fe54323a5de | a-hierarchical-neural-network-for-information | null | null | https://aclanthology.org/W16-4403 | https://aclanthology.org/W16-4403.pdf | A Hierarchical Neural Network for Information Extraction of Product Attribute and Condition Sentences | This paper describes a hierarchical neural network we propose for sentence classification to extract product information from product documents. The network classifies each sentence in a document into attribute and condition classes on the basis of word sequences and sentence sequences in the document. Experimental results showed the method using the proposed network significantly outperformed baseline methods by taking semantic representation of word and sentence sequential data into account. We also evaluated the network with two different product domains (insurance and tourism domains) and found that it was effective for both the domains. | ['Hisako Asano', 'Ryuichiro Higashinaka', 'Kugatsu Sadamitsu', 'Yukinori Homma', 'Yoshihiro Matsuo', 'Kyosuke Nishida'] | 2016-12-01 | null | null | null | ws-2016-12 | ['product-recommendation'] | ['miscellaneous'] | [ 5.36095023e-01 -1.66137796e-02 -5.14897227e-01 -1.02147722e+00
-1.84548825e-01 -5.55609941e-01 1.20566092e-01 5.82509756e-01
-4.14207309e-01 6.91273987e-01 4.65222210e-01 -4.92249340e-01
-8.25856775e-02 -1.04442859e+00 -3.10520679e-01 -2.47524187e-01
-1.46245092e-01 3.02195668e-01 6.97309710e-03 -4.97931927e-01
8.22618604e-01 2.75601685e-01 -1.62169731e+00 9.03074980e-01
5.41335046e-01 1.20024669e+00 4.09848571e-01 5.98587751e-01
-5.78630030e-01 1.13046062e+00 -9.25438643e-01 -3.62462014e-01
2.20076501e-01 -3.86394799e-01 -1.22096205e+00 5.58646500e-01
4.33624268e-01 -1.60871878e-01 1.79955646e-01 1.08429611e+00
2.36069858e-01 4.72314954e-01 6.87100947e-01 -6.73782408e-01
-1.05550599e+00 8.95288825e-01 -2.35644370e-01 2.94373453e-01
5.49147725e-01 -6.58480287e-01 1.31788754e+00 -7.55324423e-01
6.59422517e-01 1.26243210e+00 6.84875906e-01 1.38499051e-01
-8.70510697e-01 -2.66983181e-01 7.15174228e-02 3.33231419e-01
-1.03944135e+00 1.68908648e-02 7.92237878e-01 -2.63401359e-01
1.90968680e+00 -3.41806784e-02 3.05683881e-01 8.52808714e-01
6.20026052e-01 9.52968299e-01 7.89171100e-01 -6.19066834e-01
2.43832812e-01 2.48296991e-01 1.02664971e+00 3.12509179e-01
1.59515869e-02 -2.75574654e-01 -2.84898251e-01 -3.58253019e-03
2.78399736e-01 -1.24010190e-01 4.43833977e-01 2.88473397e-01
-4.98956949e-01 1.32318199e+00 1.13294721e-02 5.77646613e-01
-7.82537878e-01 -6.87150359e-01 1.03267503e+00 4.59487706e-01
7.10396409e-01 3.34307969e-01 -9.46731329e-01 1.58193737e-01
-7.35917687e-01 1.21088661e-01 1.06080449e+00 9.23989356e-01
1.83737129e-01 2.43489891e-01 -2.31584325e-01 9.06802118e-01
1.93873227e-01 1.20277695e-01 1.01234305e+00 -5.16049743e-01
6.28779054e-01 8.19554090e-01 -6.90657869e-02 -1.43907404e+00
-4.25045639e-01 -2.68146008e-01 -7.27535367e-01 -3.15362036e-01
-3.81393731e-01 -1.26237959e-01 -1.30359817e+00 1.11154723e+00
7.74328690e-03 -4.90168869e-01 6.53332651e-01 5.03284514e-01
1.43195331e+00 9.76144850e-01 4.39004838e-01 -3.04529637e-01
1.39920473e+00 -8.95493627e-01 -1.02190435e+00 -1.01559080e-01
6.68213606e-01 -7.02799737e-01 7.25101769e-01 4.42959934e-01
-1.02271605e+00 -1.03157532e+00 -1.44454932e+00 6.52233288e-02
-8.76156092e-01 2.76115686e-01 7.58834720e-01 4.72722024e-01
-6.88514650e-01 6.54601336e-01 -2.00886697e-01 -3.71201664e-01
1.71386868e-01 5.97948670e-01 -1.76710770e-01 3.40433449e-01
-1.82798195e+00 8.68621767e-01 9.50211883e-01 1.07701778e-01
-5.95843196e-01 6.97528571e-02 -1.34265459e+00 3.24580699e-01
-1.30412966e-01 -4.49137986e-01 1.57691550e+00 -1.17348492e+00
-1.39222717e+00 5.23529351e-01 -3.02755088e-01 -8.47410917e-01
-5.36492348e-01 7.38946572e-02 -9.18914557e-01 3.06611001e-01
1.70340165e-01 5.09879768e-01 6.54667437e-01 -8.19014192e-01
-7.78408766e-01 -4.77216274e-01 3.40472572e-02 3.15059692e-01
-5.54186642e-01 2.81115711e-01 3.30282092e-01 -5.57942152e-01
7.32847378e-02 -3.71257782e-01 -1.59373522e-01 -1.26904225e+00
-5.43709636e-01 -8.68822217e-01 7.41291463e-01 -7.40302026e-01
1.34313500e+00 -1.93988144e+00 -3.98829967e-01 1.25098288e-01
-2.64144838e-01 8.15756023e-02 -1.69446364e-01 7.85643637e-01
-3.38479578e-01 1.68997526e-01 4.13364172e-02 -1.15446605e-01
-7.25298077e-02 3.89080197e-01 -2.76441783e-01 -1.99803919e-01
1.35504812e-01 6.99702263e-01 -5.24793923e-01 -6.65258825e-01
-2.57453360e-02 2.22035006e-01 -1.37526793e-02 3.03471787e-03
-3.19568425e-01 -2.95834959e-01 -5.55332541e-01 5.66621721e-01
4.98439461e-01 -1.37887876e-02 2.52693325e-01 -1.82255805e-01
3.00073355e-01 6.67092741e-01 -9.51479554e-01 1.37151015e+00
-3.85142118e-01 5.12515306e-01 -3.21696490e-01 -1.33702302e+00
1.14157212e+00 5.61901569e-01 1.88637555e-01 -8.29614639e-01
3.26665521e-01 -1.27488524e-01 -3.18597183e-02 -1.09279048e+00
9.66397345e-01 -3.93613458e-01 -4.72083211e-01 1.16768353e-01
2.90873975e-01 1.26752377e-01 7.01152921e-01 5.47800399e-02
7.55629897e-01 -1.62887499e-01 4.93964136e-01 -2.06253693e-01
9.35446203e-01 2.63485909e-01 4.30266023e-01 7.15495527e-01
-1.00292698e-01 2.53819823e-01 4.11563426e-01 -8.57805550e-01
-8.61020923e-01 -7.02914000e-01 -6.55636936e-02 1.36595237e+00
-1.05728000e-01 -3.98993760e-01 -6.30252242e-01 -1.11515689e+00
-5.23825437e-02 9.15209234e-01 -7.02151179e-01 -8.50375593e-02
-2.07055539e-01 -5.14503419e-01 3.12166274e-01 7.48934209e-01
7.49652028e-01 -1.63487995e+00 -2.44276822e-01 3.09731632e-01
2.51275431e-02 -1.09718215e+00 -2.77822495e-01 4.10851300e-01
-9.97267425e-01 -5.82192004e-01 -1.23230346e-01 -1.58062768e+00
5.41836381e-01 1.25654321e-02 1.26353514e+00 -3.80418926e-01
-5.05780242e-02 -1.86271518e-01 -7.69879401e-01 -6.16374016e-01
-4.92895067e-01 4.24409807e-01 3.31923366e-02 -1.60092890e-01
1.23565805e+00 -1.24933630e-01 -1.69621244e-01 -6.50214916e-03
-1.10398042e+00 -4.13409084e-01 5.33259273e-01 7.51450837e-01
1.40137941e-01 1.01748276e+00 8.79253387e-01 -9.78848457e-01
1.51230407e+00 -4.37154144e-01 -2.15089783e-01 1.68521404e-01
-6.91994011e-01 -5.07597327e-02 4.88014281e-01 -1.88994020e-01
-1.23394966e+00 2.89389640e-01 -4.68604684e-01 6.10614061e-01
-6.98196352e-01 1.11984468e+00 2.61248299e-03 6.00804389e-01
4.63737339e-01 3.99378628e-01 6.17007986e-02 -3.73380125e-01
3.59212495e-02 1.32184577e+00 1.83299765e-01 1.03192262e-01
-1.55826017e-01 -7.57415444e-02 -1.78672493e-01 -9.85404909e-01
-1.02608883e+00 -6.67368650e-01 -8.25407147e-01 3.73887755e-02
9.90285575e-01 -5.52786231e-01 -9.26529050e-01 1.57638445e-01
-1.30749750e+00 6.02238059e-01 -2.16785297e-01 5.37773848e-01
-4.00114730e-02 4.57809389e-01 -1.13094151e+00 -9.08289313e-01
-6.90642416e-01 -8.53641570e-01 9.25720334e-01 9.40434933e-02
-6.22823179e-01 -1.15295136e+00 -1.66147262e-01 3.79399240e-01
2.13331252e-01 -2.03282312e-02 1.16324925e+00 -1.17055309e+00
4.90656048e-01 -6.95754707e-01 1.99014351e-01 9.41002846e-01
3.77445996e-01 -2.72438318e-01 -6.08886182e-01 1.08379617e-01
5.90744495e-01 -5.01821518e-01 1.01964271e+00 5.83691001e-01
8.27446699e-01 -5.02424002e-01 7.67119462e-03 -3.26401860e-01
1.54014266e+00 8.47659349e-01 5.10670424e-01 3.89534920e-01
2.86728203e-01 1.17552400e+00 6.73126817e-01 2.24565238e-01
1.47791296e-01 1.21227399e-01 1.63407877e-01 -8.69210660e-02
3.99791777e-01 -3.45705333e-03 4.29586351e-01 9.37408745e-01
4.66804504e-01 -4.10521001e-01 -6.68308377e-01 6.09899461e-01
-1.81099820e+00 -9.90694463e-01 -1.99464470e-01 1.46156120e+00
8.27476203e-01 6.53869212e-01 2.59142280e-01 1.90938443e-01
5.97786784e-01 1.16891868e-03 -3.50415885e-01 -1.29479516e+00
-2.16174752e-01 2.20843822e-01 4.86472934e-01 5.84937036e-01
-1.39664519e+00 9.57564473e-01 7.48213291e+00 7.25985169e-01
-7.42674768e-01 -2.35001773e-01 6.46609306e-01 2.07501918e-01
1.59694254e-01 -6.19835913e-01 -1.15633869e+00 3.54220718e-01
1.42850792e+00 2.01642811e-02 -4.34098244e-01 1.10511780e+00
2.02210531e-01 -2.15944767e-01 -7.61084795e-01 3.13539326e-01
4.44240928e-01 -1.32109761e+00 4.41697180e-01 -2.70967573e-01
4.58381802e-01 -1.35460272e-01 -1.23982914e-01 5.11839509e-01
4.13493991e-01 -9.89325762e-01 2.02856362e-01 -3.08484472e-02
-1.24372020e-01 -1.04533029e+00 1.58331120e+00 2.86602765e-01
-9.68816757e-01 -1.18189268e-01 -4.51490432e-01 -5.92038512e-01
1.69402748e-01 2.96983063e-01 -1.12968469e+00 6.59701467e-01
6.19205475e-01 1.01372087e+00 -5.63229144e-01 3.75890046e-01
-2.93436144e-02 7.46640563e-01 1.42685935e-01 -6.04651809e-01
6.46762252e-01 -2.79518127e-01 -2.34815888e-02 1.50760436e+00
3.66786271e-02 -1.85800325e-02 3.05326253e-01 3.71078044e-01
1.57210201e-01 4.29650754e-01 -8.45825195e-01 -2.80396104e-01
1.30521774e-01 1.07157028e+00 -1.01184857e+00 -5.48416257e-01
-4.50572520e-01 9.20513511e-01 -9.69081745e-02 3.43866855e-01
-3.11704963e-01 -1.08774841e+00 2.94967949e-01 -3.49283874e-01
8.11493635e-01 -7.47203082e-02 -3.87242019e-01 -6.47776544e-01
2.20447674e-01 -4.68346059e-01 5.71575105e-01 -6.17209435e-01
-1.54707539e+00 1.14213395e+00 1.04316950e-01 -1.16671455e+00
-5.54863870e-01 -9.73322988e-01 -1.56139627e-01 7.10592389e-01
-1.11471975e+00 -8.82806420e-01 4.12828237e-01 2.86793470e-01
1.11476505e+00 -5.36572516e-01 1.03279829e+00 5.66487312e-02
-3.93073142e-01 1.71563119e-01 1.77207415e-03 4.03180182e-01
1.41309276e-01 -1.25701606e+00 5.36143720e-01 3.57024014e-01
1.05752230e-01 9.22729254e-01 7.10285306e-01 -8.21319997e-01
-8.22664976e-01 -7.82797754e-01 1.63754952e+00 -1.38595760e-01
5.84516585e-01 -4.71008241e-01 -6.24577582e-01 5.29556632e-01
1.10829139e+00 -8.57963562e-01 1.17973518e+00 2.46455893e-01
-1.57017723e-01 -1.40796632e-01 -1.27296615e+00 -5.38617112e-02
3.52606028e-01 -4.44624573e-01 -1.10358739e+00 7.44797409e-01
1.09763157e+00 -8.81061703e-02 -8.57051551e-01 3.69668335e-01
2.80104756e-01 -7.19585538e-01 8.06650579e-01 -1.10286164e+00
8.74186158e-01 -5.55166565e-02 -2.24071681e-01 -1.14797270e+00
-4.36163008e-01 2.46194914e-01 -8.30361322e-02 1.30942488e+00
9.89625037e-01 -4.48339105e-01 9.08997774e-01 3.37016881e-01
-8.05471744e-03 -8.62385392e-01 -4.10878986e-01 -9.11733270e-01
-9.13715288e-02 -4.44598705e-01 5.75631559e-01 6.22954428e-01
2.16659918e-01 1.34632742e+00 -1.16430961e-01 -1.24050327e-01
8.06034133e-02 3.64975035e-01 -2.96988308e-01 -1.17344451e+00
-1.59664735e-01 -2.08330646e-01 -6.15000486e-01 -9.92394209e-01
5.89140892e-01 -9.38043892e-01 1.76986918e-01 -1.87671959e+00
1.05304413e-01 7.14597642e-01 -3.98334831e-01 3.52199197e-01
2.59059131e-01 9.84499976e-02 -1.98867321e-01 -8.41551349e-02
-5.92067719e-01 2.09581703e-01 8.43873322e-01 -4.69792873e-01
-2.48621315e-01 2.38772243e-01 -8.18208694e-01 4.39879537e-01
1.10783315e+00 -3.64682943e-01 -6.90486789e-01 -1.36115044e-01
3.85249645e-01 -1.97176784e-01 -4.08931881e-01 -4.86873686e-01
2.54565269e-01 1.47655115e-01 6.05127573e-01 -1.24020505e+00
4.48389706e-04 -9.75592852e-01 -4.23109233e-01 4.27551419e-01
-9.57247138e-01 3.21359754e-01 3.24204504e-01 3.45443696e-01
-7.79908180e-01 -7.79806852e-01 1.63390338e-01 -2.85338581e-01
-9.88624156e-01 -3.60309929e-01 -8.50872636e-01 -5.57710707e-01
8.54018748e-01 -4.39100683e-01 2.34629642e-02 -4.34292316e-01
-9.50249135e-01 2.29332551e-01 -4.14305389e-01 8.90484095e-01
8.44699442e-01 -1.35089552e+00 -6.51056290e-01 4.13131833e-01
5.03066443e-02 -5.28990507e-01 1.30533233e-01 1.45755947e-01
-4.87991214e-01 1.05892634e+00 -1.43776715e-01 -2.92046785e-01
-1.62981677e+00 8.28096688e-01 -5.10345250e-02 -4.27161813e-01
-1.05132103e-01 7.63633132e-01 -3.88759315e-01 -5.81074834e-01
5.41863084e-01 -4.59470093e-01 -1.12246203e+00 4.17976022e-01
4.88968104e-01 -9.22896527e-03 4.91352558e-01 -8.17654073e-01
-1.42789781e-01 4.22047466e-01 -7.40774155e-01 1.46152182e-02
1.36281753e+00 -2.57367432e-01 -4.30331588e-01 6.40642047e-01
1.62854254e+00 -8.48490596e-01 -3.83649170e-01 -1.60837859e-01
6.00879669e-01 5.16748168e-02 2.16516092e-01 -8.97218108e-01
-8.42361987e-01 3.45256597e-01 4.68942463e-01 1.02284491e+00
1.18768299e+00 -2.48953491e-01 1.06040585e+00 6.38494134e-01
1.06397614e-01 -1.46817315e+00 -1.68460846e-01 9.85362470e-01
6.59841239e-01 -1.35098124e+00 2.63723116e-02 -4.65800971e-01
-9.05345917e-01 1.36603332e+00 5.20939052e-01 -2.30632603e-01
8.82962167e-01 1.26153082e-01 1.81288034e-01 -5.24091840e-01
-5.93370736e-01 3.48458141e-02 3.34601551e-01 3.06211978e-01
9.06066179e-01 -1.55844614e-01 -7.76488245e-01 5.43204963e-01
-3.66381854e-01 7.15304986e-02 3.74705464e-01 1.44522464e+00
-5.84784269e-01 -1.17104602e+00 -1.16520114e-02 5.84705472e-01
-7.84262896e-01 -2.76184767e-01 -7.98797488e-01 6.81319118e-01
1.19959556e-01 1.30566525e+00 2.34033450e-01 -5.79061866e-01
5.39589226e-01 2.33160049e-01 1.86252370e-01 -1.06771827e+00
-1.10158324e+00 -5.40683009e-02 6.14166677e-01 -2.18803614e-01
-7.69049048e-01 -5.32453716e-01 -1.27180159e+00 1.97660193e-01
-2.02460513e-01 5.28599441e-01 7.80810654e-01 8.63842010e-01
4.99691695e-01 6.28506303e-01 8.45367968e-01 -6.20292664e-01
-6.58530176e-01 -1.59603906e+00 -8.82763922e-01 5.16771793e-01
4.55592811e-01 -1.75650254e-01 -2.69579023e-01 4.32439655e-01] | [11.125703811645508, 7.394969463348389] |
11cdfe38-6ad2-410b-b1b5-e25cf60c4934 | certified-graph-unlearning | 2206.09140 | null | https://arxiv.org/abs/2206.09140v2 | https://arxiv.org/pdf/2206.09140v2.pdf | Certified Graph Unlearning | Graph-structured data is ubiquitous in practice and often processed using graph neural networks (GNNs). With the adoption of recent laws ensuring the ``right to be forgotten'', the problem of graph data removal has become of significant importance. To address the problem, we introduce the first known framework for \emph{certified graph unlearning} of GNNs. In contrast to standard machine unlearning, new analytical and heuristic unlearning challenges arise when dealing with complex graph data. First, three different types of unlearning requests need to be considered, including node feature, edge and node unlearning. Second, to establish provable performance guarantees, one needs to address challenges associated with feature mixing during propagation. The underlying analysis is illustrated on the example of simple graph convolutions (SGC) and their generalized PageRank (GPR) extensions, thereby laying the theoretical foundation for certified unlearning of GNNs. Our empirical studies on six benchmark datasets demonstrate excellent performance-complexity trade-offs when compared to complete retraining methods and approaches that do not leverage graph information. For example, when unlearning $20\%$ of the nodes on the Cora dataset, our approach suffers only a $0.1\%$ loss in test accuracy while offering a $4$-fold speed-up compared to complete retraining. Our scheme also outperforms unlearning methods that do not leverage graph information with a $12\%$ increase in test accuracy for a comparable time complexity. | ['Olgica Milenkovic', 'Chao Pan', 'Eli Chien'] | 2022-06-18 | null | null | null | null | ['gpr', 'gpr'] | ['computer-vision', 'miscellaneous'] | [ 6.48180962e-01 5.79894960e-01 -2.50410765e-01 3.80511805e-02
-4.14175242e-01 -4.70556825e-01 2.12799534e-01 5.15026867e-01
-5.53266943e-01 8.39622617e-01 -3.29578578e-01 -7.26948321e-01
-6.27764344e-01 -1.15464032e+00 -9.73257959e-01 -5.77364981e-01
-7.01279879e-01 3.57661813e-01 1.90962017e-01 -1.23428114e-01
3.25888135e-02 5.21820486e-01 -1.28826189e+00 -2.58100510e-01
9.56136525e-01 9.87934530e-01 -2.29871348e-01 7.82393336e-01
-2.08383158e-01 7.37060964e-01 -4.63490933e-01 -7.94461489e-01
5.64962864e-01 -1.04532957e-01 -8.05539548e-01 6.53222855e-03
8.28176618e-01 -2.44920343e-01 -5.25345087e-01 1.41078866e+00
4.21695948e-01 6.26283586e-02 4.09443080e-01 -1.43958259e+00
-5.80530345e-01 9.90226865e-01 -7.08079815e-01 1.58494860e-01
2.03043199e-03 -6.57890439e-02 1.42045236e+00 -5.25037885e-01
6.03456855e-01 7.98793375e-01 8.98645759e-01 4.71544981e-01
-1.28534865e+00 -7.01967716e-01 3.69312286e-01 1.31916910e-01
-1.15026438e+00 -5.04184328e-02 6.71449661e-01 -2.90838256e-02
7.55224109e-01 4.04773533e-01 5.12537181e-01 8.34848762e-01
6.81580917e-04 7.71096945e-01 9.63420272e-01 -3.74134868e-01
1.81753382e-01 -1.37073025e-01 5.06726682e-01 9.76817369e-01
1.05005574e+00 4.13911380e-02 -4.16757941e-01 -8.48555043e-02
6.88038588e-01 -9.26361308e-02 -2.51668125e-01 -4.94998723e-01
-6.49868250e-01 7.19358802e-01 6.39316380e-01 1.28635451e-01
-1.26798853e-01 5.01377463e-01 4.15000081e-01 6.05014026e-01
4.96280611e-01 4.01361376e-01 -4.32620734e-01 9.03661251e-02
-6.63781464e-01 -9.35167968e-02 1.14854562e+00 9.13811743e-01
1.07853150e+00 2.64552355e-01 -4.40766402e-02 5.62397659e-01
-1.80698052e-01 5.19576967e-01 -1.54924482e-01 -5.25920451e-01
5.27787864e-01 9.41576779e-01 -3.35657567e-01 -1.25136566e+00
-4.18335825e-01 -1.05454409e+00 -1.33282959e+00 1.19595423e-01
5.43508053e-01 -1.03239141e-01 -1.05363762e+00 1.70992076e+00
2.32648000e-01 2.97843009e-01 -1.52873188e-01 3.89168710e-01
5.67942858e-01 4.46115583e-02 -1.72727332e-01 -2.90672034e-01
1.01992524e+00 -7.68426418e-01 -3.27120602e-01 -3.64733189e-01
6.09245539e-01 -3.16108525e-01 9.61229920e-01 5.33250213e-01
-8.12050819e-01 -1.27383381e-01 -1.22834992e+00 2.64243454e-01
-5.50552011e-01 -5.36507547e-01 9.94990885e-01 8.75934362e-01
-1.44254434e+00 9.60337162e-01 -5.04154205e-01 -3.11015606e-01
4.61875230e-01 6.17931485e-01 -5.11477590e-01 -4.41839188e-01
-1.02611470e+00 3.60366374e-01 4.39205140e-01 1.60999849e-01
-7.25162923e-01 -6.75178826e-01 -8.26583922e-01 3.54933321e-01
8.63313615e-01 -4.73430037e-01 8.67571652e-01 -1.05746150e+00
-8.39662075e-01 4.79820222e-01 3.04326802e-01 -7.28697717e-01
6.12547338e-01 -1.97408736e-01 -4.93139625e-01 2.57948279e-01
-8.77469853e-02 8.25815424e-02 9.15933490e-01 -1.07672763e+00
-5.20344555e-01 -2.31386706e-01 2.65437335e-01 -7.95015991e-02
-7.29010403e-01 -4.41279620e-01 -5.60458481e-01 -7.33778417e-01
2.67286468e-02 -8.16722453e-01 -4.64666843e-01 -1.74635991e-01
-5.93648434e-01 -4.59771045e-02 6.74947441e-01 -3.78739446e-01
1.27297115e+00 -1.88816702e+00 -1.41569018e-01 8.05328667e-01
9.70051706e-01 3.92616689e-01 -5.15664518e-01 5.17743766e-01
-5.90490550e-02 3.41098785e-01 -2.32896462e-01 -1.48433736e-02
1.30319610e-01 2.70888656e-01 -4.09600139e-02 4.50047284e-01
1.03520975e-01 9.30251837e-01 -1.12832367e+00 -1.85918733e-01
-1.78645954e-01 3.15222204e-01 -6.98215783e-01 -2.16326416e-01
-2.16261342e-01 -1.60514176e-01 -3.71218681e-01 7.03669786e-01
7.09202588e-01 -8.45367074e-01 6.23683572e-01 1.94602892e-01
4.72852826e-01 -1.57309249e-01 -1.23623848e+00 1.16902113e+00
-1.53440550e-01 4.26872998e-01 4.15330082e-01 -1.15850151e+00
6.21113896e-01 -1.11732028e-01 4.65485126e-01 -6.88042045e-01
4.47086580e-02 2.44502932e-01 5.17648794e-02 -2.91828644e-02
5.60021460e-01 2.38831475e-01 -1.40473023e-02 6.67220712e-01
1.13258086e-01 3.73323828e-01 5.07981896e-01 6.77091777e-01
2.01831698e+00 -4.05686855e-01 1.28520727e-01 -3.58868062e-01
3.17901343e-01 -2.20802173e-01 3.75237554e-01 1.19850564e+00
-7.46146142e-02 3.15691121e-02 9.10564840e-01 -4.42658991e-01
-5.82081854e-01 -1.04712486e+00 3.26133609e-01 1.15785098e+00
3.41412872e-02 -5.37085116e-01 -8.09961855e-01 -1.21755159e+00
1.70061797e-01 3.09649378e-01 -7.72455573e-01 -2.27684885e-01
-4.55927402e-01 -7.53558815e-01 5.07212758e-01 3.07007641e-01
5.07925749e-01 -8.69257092e-01 -8.17368627e-02 2.46406525e-01
2.89999336e-01 -1.08294451e+00 -3.91450405e-01 2.87457317e-01
-9.48096216e-01 -1.65727711e+00 -4.13205683e-01 -6.65841937e-01
1.14940691e+00 4.28706169e-01 1.27535570e+00 7.07940578e-01
-3.33151132e-01 6.78754687e-01 -3.64637196e-01 -7.69048035e-02
-3.23122203e-01 5.28328240e-01 -8.22098646e-03 -1.08493790e-01
1.69112831e-01 -1.10607040e+00 -6.87949002e-01 4.03242419e-03
-1.21621335e+00 -2.23354205e-01 1.09959054e+00 8.89560759e-01
3.01031709e-01 2.89426267e-01 5.88789403e-01 -1.56420290e+00
7.77458608e-01 -4.17910337e-01 -6.89598203e-01 3.37373435e-01
-1.33305967e+00 2.66216546e-01 7.64869213e-01 -3.01228315e-01
-3.68195474e-01 -1.42472118e-01 -1.12567814e-02 -3.18640053e-01
1.68314233e-01 6.88871443e-01 -9.14244447e-03 -5.17478347e-01
6.49875283e-01 1.49855629e-01 5.87653853e-02 -2.85861135e-01
4.26064521e-01 5.22481017e-02 4.09979165e-01 -5.91328800e-01
1.29696727e+00 4.20952678e-01 3.94446433e-01 -7.39476681e-01
-8.02761316e-01 -2.93504238e-01 -2.88858205e-01 -3.85700524e-01
3.33584309e-01 -3.41724128e-01 -9.31139767e-01 3.11128527e-01
-6.87215447e-01 -4.41655368e-01 -2.91997224e-01 4.06311154e-02
-5.08123711e-02 7.90665329e-01 -7.13223875e-01 -6.27053380e-01
-6.88309789e-01 -7.70265818e-01 3.95545483e-01 -2.92978454e-02
1.95877030e-01 -1.08833015e+00 -3.32169294e-01 6.69132844e-02
6.15560889e-01 4.15448010e-01 1.23989809e+00 -9.26149070e-01
-8.48593593e-01 -4.75543141e-01 -6.02603257e-01 5.12997091e-01
9.63351205e-02 -3.05323839e-01 -7.45904684e-01 -8.43396246e-01
-5.28578460e-01 -8.11024606e-02 9.76360142e-01 6.18416145e-02
1.05057812e+00 -7.18583524e-01 -2.32511550e-01 5.55700362e-01
1.60077488e+00 -3.23959500e-01 4.17693794e-01 -7.73102557e-03
8.85075450e-01 4.02251393e-01 7.54785389e-02 2.79925376e-01
8.77296254e-02 -9.21094716e-02 8.30839217e-01 -1.08385943e-01
-1.66977182e-01 -5.11829913e-01 1.63353607e-01 9.56659853e-01
-8.33511725e-02 -4.44364220e-01 -8.39445174e-01 3.75044525e-01
-1.74791837e+00 -5.48156202e-01 -1.98187664e-01 2.39109111e+00
4.99325365e-01 6.01948380e-01 -4.57796976e-02 4.55428302e-01
7.91906834e-01 2.79592484e-01 -6.38050497e-01 -1.09593734e-01
-3.17784003e-03 5.79074442e-01 1.00514770e+00 5.33246398e-01
-9.43392038e-01 7.41238952e-01 5.93938160e+00 9.10249531e-01
-9.04002964e-01 -1.14740945e-01 5.40903568e-01 1.86957121e-01
-4.25356179e-01 1.50781140e-01 -4.84275490e-01 8.57355893e-02
9.20777261e-01 -2.40485966e-01 8.25260222e-01 8.63623619e-01
-3.64322722e-01 -1.35490429e-02 -9.63590503e-01 7.75428832e-01
7.78955072e-02 -1.25909996e+00 8.13264847e-02 3.22960496e-01
6.44721508e-01 9.43427607e-02 -5.15760556e-02 5.59156477e-01
8.60969603e-01 -9.24556911e-01 1.58109069e-01 2.10792080e-01
1.04039896e+00 -8.73471975e-01 6.12923682e-01 2.20949799e-01
-1.18195212e+00 -1.32091314e-01 -1.81167468e-01 -1.55043930e-01
-3.14416796e-01 1.00835633e+00 -9.00460601e-01 8.21592331e-01
6.41040564e-01 5.18505275e-01 -8.18960667e-01 9.17241395e-01
-3.66292238e-01 7.75215745e-01 -3.40595931e-01 -1.83674037e-01
1.82620957e-01 -1.13459796e-01 6.30265892e-01 9.28980649e-01
1.08197421e-01 -2.38180399e-01 1.21211991e-01 4.03759062e-01
-6.09592140e-01 1.78778306e-01 -6.56894505e-01 -2.99632221e-01
3.32073867e-01 1.38966358e+00 -1.01994121e+00 -1.77561581e-01
-3.02816868e-01 9.69029725e-01 7.41827369e-01 4.96437848e-01
-5.83078980e-01 -6.63452327e-01 2.98650950e-01 3.82345706e-01
2.97720581e-01 -1.54726490e-01 8.28870833e-02 -9.51817334e-01
1.03397876e-01 -8.76827598e-01 7.43551791e-01 -1.23353727e-01
-1.55282140e+00 5.92726469e-01 -2.44340286e-01 -9.68011320e-01
5.60234534e-03 -7.67221153e-01 -4.59415972e-01 3.72983694e-01
-1.59905219e+00 -9.63808179e-01 -3.70581985e-01 4.90500003e-01
-1.14471793e-01 5.68069890e-02 6.07244849e-01 4.22489166e-01
-3.65574211e-01 8.98715079e-01 2.19366401e-02 4.25810814e-01
4.41763937e-01 -1.44563508e+00 4.70494568e-01 1.23034739e+00
1.49997100e-01 6.67353153e-01 5.84688723e-01 -7.28908777e-01
-1.82990599e+00 -1.18603337e+00 7.61135042e-01 -1.29459143e-01
9.23062265e-01 -5.76586783e-01 -7.50913560e-01 7.09898472e-01
-1.11944862e-02 3.22972715e-01 5.07103920e-01 4.87973481e-01
-6.46183074e-01 -4.54516023e-01 -1.05971217e+00 6.31399214e-01
1.51362181e+00 -3.67830634e-01 4.91082519e-02 2.34873593e-01
8.39066923e-01 -1.28338009e-01 -7.86591887e-01 5.27164221e-01
3.59305710e-01 -8.74534428e-01 7.22793639e-01 -6.66539192e-01
-3.63159105e-02 3.05679794e-02 5.31036071e-02 -1.12489212e+00
-2.34891996e-01 -1.03639627e+00 -3.22881520e-01 8.68142188e-01
5.52369773e-01 -9.46194351e-01 1.28224957e+00 4.20396358e-01
4.73091938e-02 -6.37827218e-01 -9.36459541e-01 -9.14699495e-01
-3.88132595e-02 -5.88603556e-01 4.22788411e-01 1.02644634e+00
-2.03539938e-01 3.70539010e-01 -3.75473738e-01 2.30817169e-01
9.96457100e-01 -1.09956510e-01 8.65413308e-01 -1.51184547e+00
-5.05739927e-01 -5.44346690e-01 -3.99523705e-01 -8.56182694e-01
1.46059945e-01 -1.28866434e+00 -2.61115223e-01 -1.57812476e+00
1.62915468e-01 -5.17886877e-01 -6.38897181e-01 7.33241260e-01
-1.74377009e-01 2.53479958e-01 5.61585091e-02 -2.11363748e-01
-8.22893739e-01 8.78385827e-02 1.15846741e+00 -2.99204767e-01
-5.13833715e-03 1.47552744e-01 -8.88149500e-01 4.27233219e-01
7.61529386e-01 -4.97447163e-01 -5.64902306e-01 -4.15430427e-01
8.52484941e-01 -6.96598291e-02 4.45193380e-01 -9.43503022e-01
4.24522728e-01 9.02422667e-02 2.46384460e-02 -3.81171644e-01
-1.65880039e-01 -8.65923584e-01 6.15964495e-02 7.28510201e-01
-8.23388323e-02 2.53475696e-01 1.02288134e-01 1.08513010e+00
6.48395121e-02 3.08889570e-03 4.70018774e-01 -9.54436734e-02
-5.44806600e-01 6.02236509e-01 -5.20855188e-02 2.10390866e-01
6.76904321e-01 -2.70029455e-01 -6.28940105e-01 -5.20563781e-01
-6.92884803e-01 2.65144885e-01 2.60572791e-01 1.62340641e-01
6.00074112e-01 -9.58747447e-01 -5.37229300e-01 2.61026680e-01
1.15180518e-02 -3.01855188e-02 3.73297185e-01 8.44799101e-01
-4.66762602e-01 7.58898854e-02 1.63205609e-01 -2.28236452e-01
-9.67277765e-01 6.24693274e-01 1.46596804e-01 -8.15291405e-01
-5.46225309e-01 8.07585299e-01 -1.46515071e-01 -4.64864850e-01
4.04395074e-01 -1.26092806e-01 2.71795094e-01 -8.97351429e-02
2.37371758e-01 5.10934591e-01 2.91195869e-01 2.48027965e-02
-1.51631996e-01 6.16048416e-03 -3.25210392e-01 3.23848754e-01
1.40378070e+00 2.38812208e-01 -3.76541138e-01 -5.41839525e-02
9.56417263e-01 3.21769714e-01 -1.10124063e+00 -4.58370626e-01
4.01167035e-01 -2.73132414e-01 -1.51201144e-01 -8.17424357e-01
-1.38438606e+00 5.35246313e-01 2.72509694e-01 7.43448257e-01
1.08485091e+00 -1.67465553e-01 6.04406238e-01 5.80740094e-01
5.07034957e-01 -1.05729437e+00 9.63376686e-02 5.26494205e-01
5.14797151e-01 -9.82735872e-01 1.26311943e-01 -5.76259792e-01
-4.57739923e-04 9.61193621e-01 6.14841163e-01 -2.73855120e-01
8.22125971e-01 3.04410458e-01 -2.62977123e-01 -4.11992013e-01
-6.53546810e-01 -2.50337690e-01 1.51441067e-01 5.95912755e-01
-1.11657053e-01 1.16966143e-01 -2.39060864e-01 4.73265171e-01
-8.33131671e-02 -2.35909805e-01 5.50157130e-01 8.98531377e-01
-4.15021092e-01 -1.17742693e+00 1.04354866e-01 1.05631375e+00
-4.91173118e-01 -4.35906649e-01 -4.76819813e-01 1.16355371e+00
-2.12890491e-01 7.56117225e-01 -1.76267669e-01 -5.92903316e-01
1.67357087e-01 -1.41690940e-01 2.59064883e-01 -5.29396057e-01
-4.76870477e-01 -1.97950289e-01 3.12688261e-01 -6.53099418e-01
-1.20599464e-01 -1.42981172e-01 -1.05732858e+00 -6.22880578e-01
-3.97149414e-01 2.74495304e-01 3.21505249e-01 5.86039245e-01
4.31173801e-01 5.66088676e-01 4.55039203e-01 -3.39522302e-01
-6.64183557e-01 -7.71463990e-01 -9.20128584e-01 2.79365748e-01
3.27076137e-01 -3.34926009e-01 -6.92439020e-01 -5.27527869e-01] | [7.018658638000488, 6.074635982513428] |
e9c76248-79bc-488d-b863-1cb55fd79648 | localeyenet-deep-attention-framework-for | 2303.12728 | null | https://arxiv.org/abs/2303.12728v1 | https://arxiv.org/pdf/2303.12728v1.pdf | LocalEyenet: Deep Attention framework for Localization of Eyes | Development of human machine interface has become a necessity for modern day machines to catalyze more autonomy and more efficiency. Gaze driven human intervention is an effective and convenient option for creating an interface to alleviate human errors. Facial landmark detection is very crucial for designing a robust gaze detection system. Regression based methods capacitate good spatial localization of the landmarks corresponding to different parts of the faces. But there are still scope of improvements which have been addressed by incorporating attention. In this paper, we have proposed a deep coarse-to-fine architecture called LocalEyenet for localization of only the eye regions that can be trained end-to-end. The model architecture, build on stacked hourglass backbone, learns the self-attention in feature maps which aids in preserving global as well as local spatial dependencies in face image. We have incorporated deep layer aggregation in each hourglass to minimize the loss of attention over the depth of architecture. Our model shows good generalization ability in cross-dataset evaluation and in real-time localization of eyes. | ['Akshansh Gupta', 'Somsukla Maiti'] | 2023-03-13 | null | null | null | null | ['deep-attention', 'facial-landmark-detection', 'deep-attention'] | ['computer-vision', 'computer-vision', 'natural-language-processing'] | [-3.01642984e-01 2.08457828e-01 2.78630495e-01 -6.83475912e-01
-1.47340655e-01 -8.95025730e-02 3.53611499e-01 -3.91554952e-01
-3.63891900e-01 4.23808128e-01 1.28338769e-01 1.91820338e-01
-9.28740650e-02 -3.73896599e-01 -5.85087061e-01 -5.41941702e-01
7.37578794e-02 1.35984123e-01 9.55407247e-02 -3.58566970e-01
5.53961456e-01 7.21457779e-01 -1.80250251e+00 1.05617426e-01
6.33792996e-01 1.04328525e+00 2.22288400e-01 4.19316322e-01
9.13313776e-02 4.40301239e-01 -3.46007109e-01 -3.21989149e-01
1.63591430e-01 -4.33346301e-01 -5.63234031e-01 2.18179915e-02
6.82318747e-01 -1.64745390e-01 1.27928406e-01 1.03554368e+00
6.73723817e-01 8.79245922e-02 7.53031015e-01 -1.45862114e+00
-7.06326485e-01 -2.57123888e-01 -1.09531689e+00 3.35996270e-01
3.46117258e-01 1.49185225e-01 6.39374852e-01 -1.02162015e+00
3.42558503e-01 1.30579996e+00 4.85560834e-01 8.75358105e-01
-8.62272084e-01 -7.11737931e-01 4.34068851e-02 2.69106209e-01
-1.61529577e+00 -8.15894306e-01 7.42253959e-01 -3.56072724e-01
1.09838796e+00 2.03065053e-02 4.48201001e-01 6.05896235e-01
4.74081546e-01 4.84947234e-01 9.74835157e-01 -7.00677872e-01
-3.57522815e-01 3.79896581e-01 -6.84121028e-02 1.22920322e+00
8.42915997e-02 -3.63718979e-02 -7.77383447e-01 3.45709741e-01
1.06634068e+00 2.70102859e-01 -2.06657112e-01 -4.60964113e-01
-8.22144985e-01 6.14000738e-01 9.94754732e-01 2.34404996e-01
-4.39260930e-01 -6.64428517e-04 -7.16860080e-03 3.96715738e-02
3.47444445e-01 3.87622625e-01 -3.54027510e-01 2.60254681e-01
-7.93827832e-01 -9.46116447e-02 2.13805333e-01 8.93270135e-01
8.11459541e-01 -2.33985230e-01 -1.22489795e-01 7.26665199e-01
6.98859811e-01 4.87216324e-01 5.07349312e-01 -6.97336257e-01
3.46064031e-01 1.12745655e+00 1.00271046e-01 -9.97912049e-01
-7.06549525e-01 -2.22875059e-01 -6.92055106e-01 6.94193065e-01
3.06988358e-01 -1.39450714e-01 -1.10607064e+00 1.77770686e+00
4.73981708e-01 -1.21415839e-01 -3.37377965e-01 1.03887928e+00
5.94698906e-01 2.85941690e-01 1.62868857e-01 1.45424619e-01
1.63285875e+00 -1.06791806e+00 -7.02325702e-01 -2.78624356e-01
3.30173522e-01 -7.24301994e-01 1.25493729e+00 1.49139777e-01
-1.03749204e+00 -6.34593368e-01 -1.01099670e+00 -5.63154876e-01
-6.51919484e-01 3.07073712e-01 3.87832344e-01 6.48956895e-01
-1.40608668e+00 2.23422930e-01 -6.37657106e-01 -7.76473820e-01
6.40465796e-01 9.49488461e-01 -8.88147831e-01 3.79504710e-01
-7.83051968e-01 1.04161811e+00 -8.26134160e-02 6.53037846e-01
-6.24246657e-01 -2.81523407e-01 -8.12841117e-01 1.27216548e-01
-9.08801099e-04 -7.21426666e-01 9.50378239e-01 -1.21253252e+00
-1.51551616e+00 1.13709402e+00 -6.07930541e-01 -7.64260516e-02
1.45457387e-01 -2.42338583e-01 -3.66280705e-01 1.53091028e-02
-9.09131393e-03 1.20411956e+00 1.15065563e+00 -8.96543324e-01
-9.45363760e-01 -7.61604667e-01 -2.47816533e-01 3.24816912e-01
-4.04892355e-01 3.55686098e-01 -4.31636542e-01 -1.97319835e-01
7.12370202e-02 -8.58064413e-01 1.94174439e-01 1.69020697e-01
-2.66904742e-01 -5.76208949e-01 9.32975352e-01 -5.29182851e-01
1.24044585e+00 -1.98310757e+00 1.37484036e-02 1.44676834e-01
3.35442573e-01 2.30925515e-01 -1.48265753e-02 2.98853479e-02
-2.00734705e-01 -7.36659169e-02 2.80407578e-01 -6.13909662e-01
-5.02886139e-02 -2.82473415e-01 -2.72199064e-02 6.47667170e-01
4.16701823e-01 9.33254123e-01 -3.18880230e-01 -4.56189990e-01
1.64318681e-01 7.75891662e-01 -4.30160701e-01 3.61757219e-01
1.55430168e-01 4.32231843e-01 -3.03230852e-01 7.01916099e-01
5.67021608e-01 -2.45295838e-01 -5.19407451e-01 -1.36654764e-01
-1.26350701e-01 -1.15055829e-01 -7.43730366e-01 1.66314006e+00
-3.13581973e-01 7.70212173e-01 6.98191226e-02 -3.23912591e-01
8.49362373e-01 2.32803792e-01 6.43879399e-02 -1.03785527e+00
3.62835437e-01 -1.45720512e-01 1.89709440e-01 -6.26478374e-01
2.57575214e-01 6.95571080e-02 2.72722453e-01 3.32445920e-01
1.11726262e-01 5.03136158e-01 -1.54404044e-01 -2.40826756e-01
3.70802790e-01 1.92946300e-01 3.09745759e-01 -4.03945535e-01
8.20380747e-01 -3.94030422e-01 1.97839752e-01 7.07200095e-02
-5.79082549e-01 7.78119445e-01 4.72877026e-01 -6.62640333e-01
-9.11340892e-01 -4.88191426e-01 6.38491195e-03 1.39650261e+00
1.61865968e-02 -5.41215437e-03 -1.10006928e+00 -6.49877012e-01
-1.71392992e-01 1.46568328e-01 -1.13980722e+00 -4.93224636e-02
-4.14257556e-01 -3.07134032e-01 2.80041784e-01 4.06242698e-01
7.50284255e-01 -1.30310905e+00 -9.04865384e-01 -2.98167676e-01
3.73875648e-01 -7.12706268e-01 -7.30297506e-01 -4.13269214e-02
-5.86423695e-01 -1.14035630e+00 -8.37803602e-01 -1.00393510e+00
1.01908410e+00 3.80943924e-01 7.07310915e-01 8.57190266e-02
-4.37088102e-01 6.45392155e-03 4.87599224e-02 -6.91315889e-01
4.52591181e-01 3.58060777e-01 8.30513909e-02 2.57180154e-01
8.48807514e-01 -1.34379625e-01 -1.03429687e+00 5.54083645e-01
-2.71686316e-01 -2.07266957e-01 5.14431655e-01 6.52192414e-01
2.57616162e-01 -3.34724307e-01 3.81795049e-01 -6.76323950e-01
6.57895863e-01 -2.20685393e-01 -7.00523019e-01 4.01783764e-01
-5.82919121e-01 2.05612302e-01 2.09300548e-01 1.04143381e-01
-1.07346869e+00 2.68613160e-01 -3.57076456e-03 -2.61139303e-01
-2.76026368e-01 -4.66500409e-02 -2.55580187e-01 -4.32616353e-01
6.71653867e-01 -3.30486894e-01 3.68279517e-01 -2.91353136e-01
1.53893292e-01 8.24233711e-01 1.85798720e-01 -7.69928917e-02
3.74433994e-01 4.37528104e-01 1.94945484e-01 -8.73428941e-01
-7.90785491e-01 -4.94392663e-01 -9.31978881e-01 -1.51683733e-01
1.08118546e+00 -8.65974009e-01 -1.21035147e+00 3.90914172e-01
-1.05331993e+00 -7.65837803e-02 3.89985293e-01 5.69744222e-02
-2.95335501e-01 -3.22597861e-01 2.17457749e-02 -7.33829319e-01
-5.99993229e-01 -1.21477127e+00 1.32740128e+00 8.78969014e-01
-2.94611633e-01 -9.67517853e-01 -1.20707229e-02 7.57591873e-02
4.36361074e-01 3.64786759e-02 4.09940004e-01 -2.11588889e-01
-5.55616021e-01 -2.93525398e-01 -4.13278878e-01 7.28540942e-02
2.63460517e-01 9.35432017e-02 -1.43867099e+00 -3.10189158e-01
-1.02763057e-01 -1.53015405e-01 6.83394730e-01 5.61632454e-01
8.85953426e-01 -2.11194590e-01 -5.01565576e-01 8.90574336e-01
1.16490853e+00 9.58823413e-02 7.92020380e-01 3.42458755e-01
8.71718705e-01 9.21568632e-01 4.95741457e-01 6.73844144e-02
4.37248498e-01 5.53520620e-01 3.59826833e-01 -4.01677012e-01
-1.37575552e-01 -1.72657639e-01 7.66488910e-02 5.66689409e-02
-1.89290151e-01 -7.64474273e-02 -8.91435146e-01 5.68828046e-01
-1.63266242e+00 -8.12063456e-01 7.00681731e-02 2.20584941e+00
5.12836576e-01 -9.90741327e-02 3.32149893e-01 -2.18233079e-01
7.38830268e-01 -1.38437137e-01 -4.59367692e-01 -5.71551979e-01
1.44180030e-01 1.76027969e-01 3.11473101e-01 5.42938411e-01
-1.18193543e+00 1.16716707e+00 5.93112659e+00 2.79375345e-01
-1.55958533e+00 6.66604489e-02 6.18483901e-01 -1.66992843e-01
3.66335690e-01 -4.00734484e-01 -1.21450329e+00 4.40409392e-01
8.56710136e-01 3.27107430e-01 3.38817120e-01 8.58309269e-01
1.91091597e-01 -2.22914085e-01 -1.05419886e+00 1.20340240e+00
3.37464362e-01 -9.87513065e-01 -2.59984970e-01 1.78166673e-01
6.09255254e-01 -9.67271924e-02 4.40974355e-01 -3.06048207e-02
-3.18685293e-01 -1.42886293e+00 3.94015163e-01 8.31599474e-01
9.57811475e-01 -9.24547076e-01 6.35014713e-01 9.70725268e-02
-9.31670845e-01 -2.03418225e-01 -2.49843583e-01 -1.35220945e-01
-2.56328851e-01 -4.56765085e-01 -1.02448356e+00 -2.06089213e-01
8.27291429e-01 4.29447800e-01 -9.76805687e-01 1.06656766e+00
-3.20715457e-01 -1.73700720e-01 -2.24330559e-01 -1.14768624e-01
1.07934564e-01 -2.59259269e-02 4.87664714e-02 7.81897485e-01
2.55772889e-01 -6.18450977e-02 -4.18338716e-01 5.43356240e-01
-1.19538279e-02 9.11146589e-03 -7.99860358e-01 2.99674094e-01
3.24728429e-01 1.34080553e+00 -6.03567958e-01 3.28015864e-01
-4.88175064e-01 1.04782331e+00 6.28415287e-01 3.85276318e-01
-6.03472769e-01 -7.29985416e-01 7.80271649e-01 6.08053148e-01
1.99521407e-01 -3.66418697e-02 -3.42485011e-01 -6.01243496e-01
-5.54513521e-02 -6.99998081e-01 2.96120167e-01 -8.73460770e-01
-8.38996649e-01 9.26144898e-01 -3.26819807e-01 -9.62984264e-01
-3.11081797e-01 -9.30205107e-01 -5.39530337e-01 1.31536353e+00
-1.45184553e+00 -1.56336367e+00 -6.42594934e-01 9.69138741e-01
3.02787155e-01 -3.73831064e-01 8.58284056e-01 1.81139722e-01
-9.20167506e-01 9.96452212e-01 -3.62151474e-01 1.20252028e-01
1.12688577e+00 -1.06114876e+00 1.65966123e-01 9.14386868e-01
-1.64029710e-02 1.21250153e+00 5.45843422e-01 -3.39322209e-01
-9.44801867e-01 -8.37913930e-01 1.06758726e+00 -7.71955192e-01
2.06622198e-01 -4.98552591e-01 -7.00428486e-01 7.98148155e-01
5.88102996e-01 6.30512387e-02 4.51674074e-01 3.60377580e-01
-1.35039791e-01 -4.99782085e-01 -1.20058870e+00 5.95330894e-01
7.53339767e-01 -6.21899366e-01 -3.55445743e-01 4.48735766e-02
1.65203676e-01 -3.72679085e-01 -3.33406150e-01 -1.32364286e-02
6.17980897e-01 -1.25378323e+00 6.27003908e-01 -5.50532997e-01
1.45334259e-01 -5.12400389e-01 3.91684234e-01 -9.19780433e-01
-3.73135269e-01 -9.25556481e-01 4.93825674e-02 1.16317081e+00
4.59998041e-01 -6.97318614e-01 8.59598398e-01 9.36402261e-01
8.48490223e-02 -7.43592978e-01 -7.18256295e-01 -3.78991663e-02
-2.83958703e-01 2.12722853e-01 6.61008179e-01 4.79065031e-01
1.61180571e-02 6.58777416e-01 -2.96545923e-01 2.27179259e-01
4.95247930e-01 -1.95233166e-01 9.09197271e-01 -1.33479261e+00
5.07632494e-01 -6.24458432e-01 -6.63913369e-01 -6.75745010e-01
1.12511873e-01 -3.45363975e-01 -4.23395783e-02 -1.21623468e+00
1.68774128e-01 -6.61886111e-02 -4.05232102e-01 7.21002281e-01
-1.82996064e-01 5.72887123e-01 -8.57848823e-02 8.23045745e-02
-5.56038380e-01 3.29459190e-01 1.36574852e+00 3.49831790e-01
-3.00952077e-01 -2.80249231e-02 -8.77762973e-01 7.43193209e-01
7.22594738e-01 -2.62492746e-01 -4.95701343e-01 -6.63367450e-01
1.86609164e-01 -4.36979473e-01 4.13291067e-01 -9.77770269e-01
7.39989161e-01 1.20440796e-01 8.71793091e-01 -4.69237447e-01
4.19792444e-01 -1.05702078e+00 -2.52781272e-01 -2.13948488e-02
-1.46335989e-01 3.46296400e-01 1.76285237e-01 2.89656103e-01
-1.87571749e-01 6.49265498e-02 1.11060429e+00 7.33775869e-02
-8.41061890e-01 4.03973967e-01 3.51551235e-01 -3.15502495e-01
1.28344131e+00 -5.27276099e-01 -2.70005703e-01 -1.31559789e-01
-6.33881211e-01 3.90265256e-01 6.88758492e-01 5.31871080e-01
5.94602942e-01 -9.66982543e-01 -2.70770937e-01 8.59884501e-01
1.61205277e-01 -1.87440991e-01 2.05212563e-01 8.41052651e-01
-6.86251998e-01 8.54211688e-01 -7.86569595e-01 -5.41060448e-01
-1.44741595e+00 5.56939125e-01 6.76065981e-01 3.60893935e-01
-2.55688936e-01 1.34897792e+00 4.91522431e-01 1.16261169e-02
5.58112860e-01 -2.47965425e-01 -6.54849291e-01 8.42914209e-02
8.13008428e-01 2.44450182e-01 5.53622805e-02 -1.05167484e+00
-6.05353594e-01 1.00079453e+00 -3.06539625e-01 9.60097685e-02
1.26829493e+00 -4.39867139e-01 -2.50564486e-01 5.75872958e-02
1.21081364e+00 -8.41536298e-02 -1.59969783e+00 1.74740911e-01
3.83634295e-04 -5.31822920e-01 1.75028950e-01 -8.64346743e-01
-1.11760545e+00 1.27817428e+00 1.08898461e+00 1.49459494e-02
1.25892448e+00 -2.14926451e-01 3.33778381e-01 1.49408609e-01
2.47428626e-01 -8.85467827e-01 -1.21898986e-01 3.55163485e-01
1.06413710e+00 -1.72653782e+00 -1.61567196e-01 -4.01591929e-03
-8.03543329e-01 1.03846693e+00 1.06349170e+00 -1.81312859e-01
7.66617060e-01 -1.11830488e-01 2.92113096e-01 -5.46756864e-01
-5.53400576e-01 -5.51799655e-01 7.79915035e-01 6.79881036e-01
7.08979726e-01 -3.12505424e-01 1.73327670e-01 2.52512991e-01
-2.18253825e-02 -1.80536255e-01 -2.17676073e-01 6.68089449e-01
-5.56495547e-01 -8.77383471e-01 -3.65254134e-01 1.44447088e-01
-4.78500754e-01 -5.73355518e-03 -4.26763415e-01 9.90496695e-01
2.30334848e-01 8.51580977e-01 2.68706799e-01 -1.25501409e-01
4.24380511e-01 1.24222629e-01 6.10853612e-01 -6.64533794e-01
-4.66927886e-01 9.93035808e-02 -3.71857733e-01 -6.28058255e-01
-2.15168908e-01 -4.19268668e-01 -1.14600492e+00 -2.47240558e-01
-1.73106909e-01 -1.16663672e-01 7.79324293e-01 6.79630995e-01
7.94754207e-01 3.38701248e-01 4.36992109e-01 -9.70422924e-01
-1.37150362e-01 -1.19470739e+00 -5.61748922e-01 8.64535645e-02
7.39832282e-01 -8.68340850e-01 -1.81925241e-02 8.14815387e-02] | [14.116055488586426, 0.047562167048454285] |
b784ce38-38d1-436c-84f7-46002e6db8fe | audio-driven-facial-animation-by-joint-end-to | null | null | https://research.nvidia.com/publication/2017-07_audio-driven-facial-animation-joint-end-end-learning-pose-and-emotion | https://dl.acm.org/doi/pdf/10.1145/3072959.3073658 | Audio-Driven Facial Animation by Joint End-to-End Learning of Pose and Emotion | We present a machine learning technique for driving 3D facial animation by audio input in real time and with low latency. Our deep neural network learns a mapping from input waveforms to the 3D vertex coordinates of a face model, and simultaneously discovers a compact, latent code that disambiguates the variations in facial expression that cannot be explained by the audio alone. During inference, the latent code can be used as an intuitive control for the emotional state of the face puppet. We train our network with 3–5 minutes of high-quality animation data obtained using traditional, vision-based performance capture methods. Even though our primary goal is to model the speaking style of a single actor, our model yields reasonable results even when driven with audio from other speakers with different gender, accent, or language, as we demonstrate with a user study. The results are applicable to in-game dialogue, low-cost localization, virtual reality avatars, and telepresence | ['Jaakko Lehtinen', 'Antti Herva', 'Samuli Laine', 'Timo Aila', 'Tero Karras'] | 2017-07-30 | null | null | null | siggraph-2017-7 | ['face-model'] | ['computer-vision'] | [-7.41087124e-02 3.09978932e-01 9.66230035e-03 -2.96930224e-01
-6.24961853e-01 -7.99292088e-01 4.82728750e-01 -4.92757529e-01
7.38795241e-03 1.50718451e-01 1.30172849e-01 1.61367714e-01
4.93505806e-01 -2.87258416e-01 -5.75247526e-01 -4.19822127e-01
-3.69080782e-01 5.25564194e-01 -2.69508958e-01 -1.35246873e-01
-2.08862185e-01 7.08110631e-01 -1.81832254e+00 5.63968830e-02
1.16810948e-01 9.07762825e-01 -1.33928433e-01 1.08076620e+00
3.62381697e-01 7.07605541e-01 -8.22381496e-01 -6.35624751e-02
1.95006013e-01 -1.89143598e-01 -3.37612540e-01 6.34016767e-02
5.39979219e-01 -6.39222682e-01 -4.05536801e-01 9.41407204e-01
5.64578354e-01 -1.17966972e-01 2.62111276e-01 -1.53335476e+00
-8.75062123e-02 2.28478625e-01 -4.02045876e-01 -2.45510697e-01
8.16638768e-01 2.64009565e-01 7.95560360e-01 -6.74131155e-01
7.77672768e-01 1.50378180e+00 6.94528401e-01 7.69671261e-01
-1.45365584e+00 -9.45631981e-01 7.26686791e-02 -1.84993133e-01
-1.65801489e+00 -7.97232807e-01 9.04738724e-01 -4.40326065e-01
6.07333958e-01 2.79710084e-01 1.01917732e+00 1.56489313e+00
-7.71994963e-02 4.82592791e-01 6.50909007e-01 -1.34312496e-01
1.66804448e-01 1.68578133e-01 -5.98667264e-01 1.04766035e+00
-6.14943445e-01 1.72301397e-01 -8.98687363e-01 -5.64568281e-01
9.71689403e-01 -5.16514301e-01 -4.31427568e-01 -3.44703287e-01
-9.19990838e-01 8.51782143e-01 -9.87630188e-02 -1.83451474e-01
-4.16975141e-01 5.00481665e-01 2.24457562e-01 3.24426562e-01
4.73465204e-01 1.92764953e-01 -4.49450761e-01 -9.34436977e-01
-7.56056249e-01 3.05306315e-01 1.02635431e+00 7.02651024e-01
4.24798936e-01 6.11349881e-01 3.86634499e-01 4.70956951e-01
4.69254732e-01 6.52480781e-01 1.09959304e-01 -1.60908115e+00
-2.15932548e-01 -8.52951314e-03 1.47736564e-01 -1.20891953e+00
-3.49188447e-01 3.90128940e-02 -2.26340055e-01 5.80106676e-01
3.07109594e-01 -6.39453113e-01 -5.19366324e-01 2.16595483e+00
3.96998554e-01 5.64390004e-01 -2.43499354e-01 1.08102202e+00
7.06081867e-01 5.64341486e-01 -3.34887743e-01 -2.62684017e-01
1.24180329e+00 -3.71078014e-01 -6.12618744e-01 -1.10961728e-01
1.71565369e-01 -6.45239830e-01 1.05949020e+00 5.57853937e-01
-1.11531818e+00 -4.89641666e-01 -8.97769392e-01 2.28326440e-01
2.17058659e-01 -3.62158865e-02 6.84699059e-01 6.07298076e-01
-1.30851901e+00 4.49752808e-01 -1.08279049e+00 -2.79381305e-01
-1.14187472e-01 6.00098073e-01 -4.50654387e-01 6.39862835e-01
-1.12133551e+00 6.17627561e-01 -5.20333052e-01 1.55072480e-01
-1.16290176e+00 -7.10742414e-01 -8.74357939e-01 4.07022536e-02
1.29985943e-01 -2.48870537e-01 1.44449961e+00 -1.39001119e+00
-2.19187427e+00 9.31282938e-01 -1.07037097e-01 -3.17342840e-02
3.83061647e-01 -1.67782173e-01 -2.17900991e-01 3.80272239e-01
-2.23289892e-01 1.08085382e+00 1.28990030e+00 -9.99797344e-01
5.11213951e-02 -1.81299344e-01 2.35891983e-01 1.67609468e-01
-9.44457278e-02 3.01317930e-01 -6.27860725e-01 -3.16679209e-01
-1.15487784e-01 -1.32563424e+00 2.05092832e-01 5.41150272e-01
-2.08047226e-01 2.05259874e-01 1.16093457e+00 -4.37328666e-01
5.44985890e-01 -2.27458811e+00 3.88976783e-01 2.22358659e-01
2.68635660e-01 -7.14375377e-02 -1.82205234e-02 4.97063668e-03
-1.36085942e-01 -1.81104913e-01 3.56703371e-01 -4.96318489e-01
4.44329269e-02 3.19839157e-02 -4.79475111e-01 5.75638711e-01
4.11188342e-02 6.83404267e-01 -6.90869331e-01 -4.11379963e-01
-2.67109531e-03 9.33125317e-01 -6.28486931e-01 4.32163864e-01
-2.14872375e-01 4.99814153e-01 -8.18787068e-02 6.17271006e-01
1.56266600e-01 -6.74109980e-02 2.96403289e-01 -9.07143354e-02
7.65375271e-02 1.34125233e-01 -1.08606827e+00 1.90988576e+00
-4.94249612e-01 1.24942625e+00 8.30582619e-01 -3.76779228e-01
1.04657519e+00 5.59429228e-01 6.30088687e-01 -2.52820998e-01
3.98016542e-01 -4.08106416e-01 -1.81705758e-01 -5.80483735e-01
3.13020200e-01 -9.11315903e-02 -2.47748926e-01 7.72937000e-01
7.96220005e-02 -2.87913948e-01 -4.78465766e-01 1.69417143e-01
9.76159632e-01 2.50714391e-01 -2.52175808e-01 1.79439828e-01
-1.84765577e-01 -4.13849205e-01 4.49179322e-01 2.14787558e-01
-1.97523534e-01 6.14974558e-01 9.32715893e-01 -3.17487121e-01
-8.10022295e-01 -9.59490001e-01 1.88742518e-01 1.16605318e+00
-5.98696582e-02 -4.92062330e-01 -7.09586799e-01 -1.58555523e-01
-6.02043830e-02 3.15975279e-01 -6.61511481e-01 -2.33333632e-01
-3.60086203e-01 -1.34113774e-01 9.15319622e-01 3.75360340e-01
-1.25017939e-02 -8.28139007e-01 -8.44670057e-01 2.18218621e-02
-3.16346973e-01 -1.30790842e+00 -5.55722177e-01 -1.37066573e-01
-4.26466554e-01 -8.05432856e-01 -1.67395115e-01 -4.49790806e-01
4.03584331e-01 -1.35156587e-01 1.03036463e+00 3.76180969e-02
-4.02782589e-01 6.65420115e-01 3.95576134e-02 -3.70731592e-01
-5.63016593e-01 -4.30561453e-01 4.69418764e-01 1.75605729e-01
1.43892184e-01 -9.01954353e-01 -2.79263258e-01 2.42178321e-01
-2.98297375e-01 1.92186117e-01 -2.18291000e-01 5.48056781e-01
1.62734792e-01 -3.98655027e-01 1.25214204e-01 -3.53823066e-01
6.54465437e-01 -3.32514793e-01 -8.46232712e-01 -2.64319539e-01
-7.87559971e-02 -8.65421593e-02 2.48042390e-01 -1.02432573e+00
-7.11320877e-01 3.63704324e-01 -8.31747949e-02 -1.09465599e+00
-3.09560061e-01 1.63780481e-01 -1.47201359e-01 -1.60787746e-01
6.77339256e-01 -2.47778282e-01 5.21349967e-01 -1.14101350e-01
3.63675117e-01 5.98945916e-01 7.67401397e-01 -5.19852698e-01
6.81810558e-01 5.50844729e-01 -4.83355112e-02 -9.97661591e-01
-2.71521121e-01 2.25215375e-01 -5.51761329e-01 -6.01806998e-01
8.51242065e-01 -1.17150784e+00 -1.47753942e+00 3.99274349e-01
-1.37041092e+00 -5.36479294e-01 2.60708164e-02 5.33118904e-01
-7.21109390e-01 -1.53747881e-02 -7.35151768e-01 -9.32255566e-01
-1.66933835e-01 -1.28763652e+00 1.43288326e+00 2.01481748e-02
-9.09070969e-01 -8.43492568e-01 1.18723065e-01 3.04193255e-02
1.62341520e-01 5.61235011e-01 7.55022049e-01 -1.28244877e-01
-4.36227471e-01 -3.08868796e-01 2.24728152e-01 -7.62741417e-02
1.80417538e-01 6.36579990e-01 -1.30039108e+00 -1.79031253e-01
-3.19871843e-01 -7.24460602e-01 3.84363197e-02 3.90739769e-01
9.19534445e-01 -3.41578007e-01 -5.55418134e-02 8.63963544e-01
7.12692738e-01 6.36460707e-02 8.24956968e-02 -3.39676291e-01
6.35926366e-01 5.16363144e-01 3.41589808e-01 7.19342053e-01
2.45362327e-01 1.01972532e+00 7.54415512e-01 -1.11256026e-01
2.50400752e-01 -3.56027454e-01 6.99009240e-01 6.13160849e-01
-1.05373763e-01 1.24920398e-01 -6.23159468e-01 1.39277101e-01
-1.52527130e+00 -1.20038080e+00 3.87370020e-01 2.01103997e+00
6.34914696e-01 1.44939765e-01 2.86454171e-01 -2.12273970e-01
6.78571463e-01 1.92483827e-01 -6.89710081e-01 -7.62486637e-01
2.35342816e-01 2.43196666e-01 -2.13797763e-02 7.09768474e-01
-8.58120680e-01 9.32242334e-01 7.30698252e+00 1.46872953e-01
-1.88271666e+00 -1.00639351e-01 3.13234448e-01 -7.55123615e-01
-4.53807354e-01 -1.71615764e-01 -2.91143626e-01 3.15847307e-01
1.05155623e+00 -2.44193543e-02 7.62632191e-01 8.94141972e-01
6.46859527e-01 2.33752370e-01 -1.29736662e+00 1.30072951e+00
1.88237980e-01 -1.22811067e+00 -5.31638205e-01 2.34052822e-01
1.81606472e-01 -8.86712864e-04 3.68963718e-01 1.82554722e-01
3.31731588e-01 -1.18281734e+00 9.46821690e-01 3.57459933e-01
1.32285106e+00 -7.46029377e-01 4.09514597e-03 4.26324576e-01
-9.32679117e-01 7.23672882e-02 1.74039066e-01 -2.64203548e-01
1.15608819e-01 -2.09404632e-01 -9.80619431e-01 -5.16463578e-01
5.68648279e-01 2.42180347e-01 -6.80419104e-03 1.02119707e-01
-2.73400813e-01 7.09725082e-01 -5.75940788e-01 4.00064923e-02
-9.32452083e-02 6.29599169e-02 7.57965863e-01 8.21131051e-01
1.86271086e-01 2.44677827e-01 7.92176425e-02 9.71929371e-01
-8.06647018e-02 -3.28616023e-01 -9.20099437e-01 -2.39629671e-01
5.46728194e-01 1.38517666e+00 -4.38492894e-01 1.63075328e-01
-1.70061633e-01 1.08543384e+00 1.75934553e-01 4.02228445e-01
-1.04248059e+00 -2.55392224e-01 1.26280153e+00 1.36266842e-01
2.35130563e-01 -5.21132886e-01 2.50137240e-01 -1.22177172e+00
-2.94557601e-01 -1.00111461e+00 -1.27059445e-01 -1.27766693e+00
-4.86335844e-01 7.82004058e-01 -4.36101854e-02 -1.15488303e+00
-1.04442656e+00 -2.21032307e-01 -6.50377929e-01 6.06564999e-01
-7.02601194e-01 -8.49308908e-01 -4.60213453e-01 6.33868694e-01
2.92936802e-01 -2.01883242e-01 1.36479890e+00 7.90693089e-02
-3.35264355e-01 6.93470001e-01 -4.22005653e-01 2.87772834e-01
7.01970577e-01 -8.77315283e-01 5.24306536e-01 3.69840533e-01
5.93499482e-01 4.29948956e-01 1.11635649e+00 -2.18172505e-01
-1.69772756e+00 -5.74549019e-01 3.00674677e-01 -4.67380762e-01
6.38095558e-01 -8.91207218e-01 -6.21665299e-01 7.32737839e-01
7.95895085e-02 1.75632283e-01 6.94445908e-01 3.31052214e-01
-5.03541052e-01 -1.61661863e-01 -9.85718250e-01 6.85036004e-01
8.05534840e-01 -9.81558263e-01 -1.32488087e-01 4.86083664e-02
6.20898426e-01 -8.35817814e-01 -5.70553482e-01 -6.77991137e-02
1.01655591e+00 -7.87564933e-01 6.97924614e-01 -6.23114526e-01
1.83157608e-01 -6.85449317e-02 -7.68022612e-02 -1.27362764e+00
3.19427364e-02 -1.24916637e+00 -6.76051676e-02 9.06242430e-01
1.56079113e-01 -7.05422461e-02 1.09478450e+00 8.76208782e-01
3.59596401e-01 -4.13521767e-01 -9.59915459e-01 -1.63798109e-01
-3.41767579e-01 -6.41009569e-01 5.71207762e-01 9.72169161e-01
2.29115888e-01 4.13115472e-01 -6.93195283e-01 4.41400617e-01
5.86915910e-01 2.39363328e-01 1.06041563e+00 -1.34575593e+00
-5.85982084e-01 -1.71895847e-01 -6.60055101e-01 -1.04730523e+00
7.77425110e-01 -4.30475116e-01 2.23413661e-01 -5.70890069e-01
-2.10699484e-01 -1.15284503e-01 3.35948557e-01 5.93914211e-01
4.30583328e-01 3.05671245e-01 1.77746817e-01 -1.83534220e-01
-3.40660065e-01 5.42919576e-01 6.98160350e-01 -3.76315042e-02
-2.66299486e-01 1.24709895e-02 -4.35815215e-01 1.00764024e+00
5.37552178e-01 -4.41848755e-01 -5.74647248e-01 -5.70530713e-01
2.03138992e-01 7.10538208e-01 6.03873372e-01 -7.07746983e-01
1.48215845e-01 -2.10321069e-01 3.20472091e-01 -6.89749345e-02
1.10256195e+00 -8.68779600e-01 3.22789609e-01 1.48005962e-01
-4.44622517e-01 4.09743451e-02 3.43201488e-01 2.36377433e-01
-6.14580698e-02 3.33238155e-01 6.36885345e-01 -6.47736117e-02
-3.43405634e-01 3.52364182e-01 -7.74717271e-01 2.88301837e-02
7.02911377e-01 -2.77260154e-01 6.94700330e-02 -1.40478480e+00
-9.22402561e-01 -1.26658559e-01 5.52073061e-01 6.29675329e-01
6.37949407e-01 -1.43406713e+00 -5.15351892e-01 6.04251444e-01
-1.48765132e-01 -3.49454671e-01 5.47119193e-02 4.06286955e-01
-4.55635518e-01 5.05838916e-03 -2.95119375e-01 -1.00892794e+00
-1.85128593e+00 2.59988829e-02 5.15060246e-01 5.55636108e-01
-4.99260038e-01 9.40829933e-01 1.94216184e-02 -1.72493339e-01
4.73900080e-01 4.24528122e-02 8.15170482e-02 -2.16869321e-02
4.52970386e-01 -3.62189338e-02 -1.29447266e-01 -9.11217153e-01
-2.76053220e-01 4.79855776e-01 4.43018168e-01 -6.30619466e-01
1.18068552e+00 1.32713569e-02 2.46607378e-01 7.30349302e-01
1.42572260e+00 3.47193211e-01 -1.72233045e+00 1.53042600e-01
-7.19275713e-01 -4.37527984e-01 7.17323646e-02 -1.54845402e-01
-1.09067452e+00 1.02078795e+00 6.33400202e-01 -8.90116096e-02
9.50006962e-01 4.38436456e-02 4.30306345e-01 3.23910862e-01
4.99936253e-01 -8.53804588e-01 2.96217084e-01 3.79878193e-01
7.87811756e-01 -1.00999939e+00 -2.92823762e-01 -3.21212150e-02
-7.31219530e-01 1.16298902e+00 6.91405773e-01 5.61113991e-02
7.85557568e-01 8.24737489e-01 5.96370876e-01 -2.46974155e-01
-1.27487171e+00 3.02998036e-01 -9.60995853e-02 6.04138076e-01
1.98804960e-01 -3.38775553e-02 8.11582148e-01 2.39795521e-01
-6.18408501e-01 -2.31207773e-01 6.88254058e-01 6.37434542e-01
-2.55244583e-01 -6.56199276e-01 -3.23019117e-01 -1.94213390e-01
-4.30109590e-01 2.88642436e-01 -5.50825357e-01 6.83897674e-01
-3.84159498e-02 7.85214186e-01 5.78829348e-01 -6.97004557e-01
2.09706306e-01 1.83428347e-01 5.85248828e-01 -5.31220853e-01
-3.70523125e-01 3.36824030e-01 1.62456751e-01 -9.40765142e-01
-2.40365073e-01 -7.36974418e-01 -1.38542080e+00 -5.19382894e-01
-6.68403357e-02 1.62277147e-01 8.32442760e-01 6.68379247e-01
4.72644657e-01 -5.60616096e-03 1.01702428e+00 -1.06512916e+00
-1.34249672e-01 -7.57939756e-01 -5.12543440e-01 1.73884153e-01
6.65608227e-01 -6.83513641e-01 -4.63005215e-01 2.28006914e-01] | [13.176382064819336, -0.4039621353149414] |
0491b97d-809d-449a-b9d7-ec8ec8cdadc9 | ais-a-nonlinear-activation-function-for | 2111.13861 | null | https://arxiv.org/abs/2111.13861v1 | https://arxiv.org/pdf/2111.13861v1.pdf | AIS: A nonlinear activation function for industrial safety engineering | In the task of Chinese named entity recognition based on deep learning, activation function plays an irreplaceable role, it introduces nonlinear characteristics into neural network, so that the fitted model can be applied to various tasks. However, the information density of industrial safety analysis text is relatively high, and the correlation and similarity between the information are large, which is easy to cause the problem of high deviation and high standard deviation of the model, no specific activation function has been designed in previous studies, and the traditional activation function has the problems of gradient vanishing and negative region, which also lead to the recognition accuracy of the model can not be further improved. To solve these problems, a novel activation function AIS is proposed in this paper. AIS is an activation function applied in industrial safety engineering, which is composed of two piecewise nonlinear functions. In the positive region, the structure combining exponential function and quadratic function is used to alleviate the problem of deviation and standard deviation, and the linear function is added to modify it, which makes the whole activation function smoother and overcomes the problem of gradient vanishing. In the negative region, the cubic function structure is used to solve the negative region problem and accelerate the convergence of the model. Based on the deep learning model of BERT-BiLSTM-CRF, the performance of AIS is evaluated. The results show that, compared with other activation functions, AIS overcomes the problems of gradient vanishing and negative region, reduces the deviation of the model, speeds up the model fitting, and improves the extraction ability of the model for industrial entities. | ['Dong Gao', 'Beike Zhang', 'Zhenhua Wang'] | 2021-11-27 | null | null | null | null | ['chinese-named-entity-recognition'] | ['natural-language-processing'] | [-4.05768305e-01 -2.55216777e-01 1.97565332e-02 -3.64152521e-01
1.16928540e-01 -8.94779712e-03 1.03345998e-01 -1.56547680e-01
-6.64221168e-01 5.57006061e-01 4.08279151e-03 -3.35328430e-01
2.72271801e-02 -8.80037367e-01 -4.93855327e-01 -9.44553077e-01
3.57511759e-01 1.17431834e-01 5.00923514e-01 -2.80594081e-01
3.04589216e-02 3.25447083e-01 -7.79767215e-01 7.36459941e-02
1.28182447e+00 1.05552876e+00 4.79684204e-01 -2.51349628e-01
-8.16852391e-01 6.88926041e-01 -9.37170804e-01 -6.43538963e-03
-3.75711508e-02 -2.61613131e-01 -4.02158022e-01 -2.99008429e-01
-4.42136943e-01 -2.05835536e-01 -3.51244152e-01 1.27875662e+00
5.99834979e-01 1.27724707e-01 5.75786471e-01 -1.03886104e+00
-9.50319469e-01 6.68716669e-01 -5.84580243e-01 -1.53319761e-02
-2.63555318e-01 -2.71830469e-01 3.43147606e-01 -7.11005509e-01
2.14135244e-01 1.35393071e+00 8.84713471e-01 3.80430788e-01
-7.17549562e-01 -1.03576910e+00 4.00820374e-01 4.92222123e-02
-1.58466244e+00 1.65601060e-01 6.86445892e-01 -3.12283069e-01
7.76914954e-01 -1.56728104e-01 6.75210238e-01 7.15989828e-01
3.09210449e-01 9.00486529e-01 6.50159597e-01 -9.70823318e-02
-1.99820831e-01 1.96358785e-01 2.24103302e-01 5.37910044e-01
1.39393643e-01 3.60391997e-02 3.20356995e-01 2.24739388e-01
1.05632520e+00 4.21938360e-01 -2.10418209e-01 1.35890529e-01
-9.83582437e-01 6.30087256e-01 9.91658151e-01 8.41246009e-01
-1.59863919e-01 -2.62997031e-01 4.99172002e-01 6.58897236e-02
4.34322476e-01 2.01512709e-01 -8.07358861e-01 -4.86052297e-02
-4.69856113e-01 5.60959727e-02 5.16885698e-01 9.74944293e-01
7.33890474e-01 4.81678963e-01 -1.57059848e-01 1.03757286e+00
3.78162354e-01 4.50121433e-01 8.58304143e-01 -2.52454728e-01
6.03309691e-01 1.18798137e+00 -1.14231698e-01 -1.13361382e+00
-8.12703669e-01 -4.43171263e-01 -1.27752888e+00 -1.48454458e-01
3.44012052e-01 -4.91468132e-01 -1.02262855e+00 1.57774937e+00
2.06565470e-01 3.02519836e-03 -6.32555708e-02 9.74134207e-01
1.10386610e+00 1.16561258e+00 3.32261771e-01 -3.80369633e-01
1.26087928e+00 -8.93024445e-01 -1.16331053e+00 -9.42855477e-02
5.46986461e-01 -6.99860632e-01 1.10828996e+00 2.39581227e-01
-6.43779039e-01 -9.03235078e-01 -1.01659119e+00 -1.67769849e-01
-6.38132572e-01 3.98500830e-01 5.23435235e-01 1.98011801e-01
-5.15753984e-01 4.58277881e-01 -7.28421390e-01 2.07226332e-02
4.63916302e-01 4.03918028e-01 -1.64793357e-01 7.78527111e-02
-1.66500151e+00 1.05193019e+00 1.11246085e+00 7.35449612e-01
-1.39107071e-02 -4.74746317e-01 -7.20756769e-01 2.71394283e-01
3.11589241e-01 -3.21735889e-01 1.00292122e+00 -1.08440626e+00
-1.52821326e+00 -1.13760792e-01 -5.31702153e-02 4.85870987e-02
4.95892227e-01 -2.92917550e-01 -9.66003597e-01 -5.21937311e-01
-1.67588189e-01 4.10980105e-01 3.94419432e-01 -9.00518000e-01
-5.34730136e-01 -3.19005251e-01 -3.21499795e-01 1.35605603e-01
-4.10833448e-01 5.33587188e-02 -5.01914442e-01 -6.56704605e-01
1.78929955e-01 -6.15292966e-01 -4.28010076e-01 -6.02520823e-01
-2.30819061e-01 -4.44919914e-01 1.01128876e+00 -9.96375620e-01
1.71605456e+00 -2.49085212e+00 -3.18460554e-01 1.86451852e-01
-1.74860917e-02 6.28535807e-01 1.02263778e-01 7.68322945e-02
-6.15880154e-02 2.92806834e-01 -4.68963861e-01 3.87941331e-01
-3.16886961e-01 3.22900951e-01 7.27262497e-02 7.23917335e-02
4.68468547e-01 8.55395138e-01 -7.97798753e-01 -5.75633824e-01
2.08896980e-01 5.54782629e-01 -1.09786004e-01 2.14477465e-01
-1.07614934e-01 3.91488850e-01 -9.18861508e-01 2.40402475e-01
1.08348668e+00 1.16008322e-03 -9.32953730e-02 -3.46375942e-01
-3.93972635e-01 3.00430924e-01 -1.37675166e+00 1.29306734e+00
-2.75213599e-01 6.04962558e-02 4.21654545e-02 -1.04492939e+00
1.53008199e+00 3.30587953e-01 2.88054138e-01 -8.74101102e-01
4.09682602e-01 2.16154084e-01 1.22764930e-01 -7.80461729e-01
2.60565251e-01 -2.59167612e-01 4.85683680e-02 -1.75301269e-01
-3.16377729e-01 1.98460445e-01 -1.27726600e-01 -1.58053011e-01
6.38762712e-01 2.21285746e-01 7.23306881e-03 -2.64830023e-01
7.75111079e-01 -1.37576491e-01 1.08848214e+00 4.97277416e-02
1.06433429e-01 3.26758862e-01 4.24135715e-01 -6.80837333e-01
-9.09616590e-01 -6.58776224e-01 -3.98047119e-01 6.53427899e-01
5.01916409e-01 -1.36880711e-01 -6.46120429e-01 -7.89299548e-01
-8.66639391e-02 5.30679524e-01 -2.66965270e-01 -5.37092924e-01
-9.50320363e-01 -8.62287581e-01 3.82183373e-01 7.58051395e-01
1.12355828e+00 -1.50063384e+00 9.47154164e-02 4.57326591e-01
-2.21031934e-01 -8.96657646e-01 -4.88727510e-01 3.96339186e-02
-1.11394298e+00 -7.30494499e-01 -7.84707963e-01 -1.14273441e+00
8.70625079e-01 -9.34414417e-02 7.79346883e-01 2.73787588e-01
2.98361599e-01 -6.63081229e-01 -3.31092894e-01 -7.08832860e-01
-2.60264486e-01 1.82825983e-01 -2.53059685e-01 -1.11018471e-01
5.41378319e-01 -1.13447078e-01 -3.11378241e-01 3.41333359e-01
-1.06625473e+00 -6.21667132e-02 8.21607649e-01 1.07470953e+00
2.17312202e-01 5.25291085e-01 8.57406914e-01 -9.36864913e-01
6.91448212e-01 -3.02584052e-01 -6.69999421e-01 1.76460490e-01
-8.21879208e-01 2.21427158e-02 1.07226622e+00 -5.64021051e-01
-1.37559021e+00 -4.36826386e-02 -4.11460936e-01 -2.52612144e-01
-5.43177687e-02 6.82840586e-01 -5.72573125e-01 2.00478196e-01
1.85868368e-01 2.26374984e-01 1.74031104e-03 -6.59775198e-01
1.04285680e-01 8.98560345e-01 2.98582792e-01 -1.97011992e-01
4.50596064e-01 -3.11833024e-01 -2.89896846e-01 -6.05729640e-01
-5.65711975e-01 -2.72063702e-01 -4.94780123e-01 8.52974318e-03
9.49776530e-01 -8.25564742e-01 -8.21761370e-01 8.34018052e-01
-1.55478454e+00 9.54781380e-03 1.37344864e-03 8.27436864e-01
1.89218834e-01 3.95731181e-01 -1.01712382e+00 -9.20820653e-01
-4.32400823e-01 -1.07107568e+00 6.03933632e-01 7.75976181e-01
2.82692254e-01 -9.38439369e-01 -4.56152439e-01 -5.47687300e-02
4.62687522e-01 5.11941463e-02 1.16761732e+00 -6.31510317e-01
-4.27242219e-01 -2.40554124e-01 -3.04147363e-01 7.97536910e-01
6.00770563e-02 1.81977242e-01 -6.81157649e-01 -4.24367981e-03
3.39480996e-01 1.74481615e-01 6.79351389e-01 4.25600201e-01
1.31340349e+00 -3.33407253e-01 -3.32681239e-01 4.76543248e-01
1.26242399e+00 9.06986177e-01 9.76726234e-01 4.87616152e-01
8.53716254e-01 4.49902147e-01 7.59146333e-01 -1.05369829e-01
3.82482827e-01 2.81455606e-01 1.66754857e-01 -6.22711599e-01
2.84327269e-01 -2.80636728e-01 2.48552144e-01 1.29233515e+00
-2.82325357e-01 8.49456713e-02 -6.98500812e-01 2.99874306e-01
-1.98004365e+00 -7.09522486e-01 -5.55831730e-01 2.06667042e+00
8.96493018e-01 2.96148956e-01 -8.27392936e-02 -1.33231670e-01
9.85514581e-01 -5.92101514e-02 -4.44890171e-01 -2.92500705e-01
-9.23052877e-02 -1.11798272e-01 5.48309505e-01 8.88745710e-02
-9.46387231e-01 9.68423545e-01 5.80769777e+00 1.19480383e+00
-1.38727760e+00 -2.35182736e-02 5.81293821e-01 5.88459492e-01
-9.45215300e-02 9.51551367e-03 -9.79399443e-01 9.12134171e-01
6.49772882e-01 1.60712734e-01 2.13632330e-01 9.22015190e-01
1.94605350e-01 1.73554301e-01 -4.94922131e-01 8.75498891e-01
-3.56435865e-01 -8.37563396e-01 -1.13492362e-01 -1.05441459e-01
5.00868678e-01 2.38619782e-02 -3.46101671e-01 7.55055487e-01
6.14755824e-02 -8.43838811e-01 5.62291205e-01 6.39269233e-01
4.26718473e-01 -9.38068330e-01 1.30804300e+00 6.84147418e-01
-1.35856712e+00 -2.97722936e-01 -8.72635067e-01 -3.83690223e-02
3.05392951e-01 8.83460462e-01 -6.92089796e-01 9.14105296e-01
6.39638960e-01 6.68829322e-01 -2.92415977e-01 1.06789851e+00
-3.45370770e-01 4.61686909e-01 -4.21108723e-01 -2.99575090e-01
2.64359534e-01 -7.71282256e-01 1.55995399e-01 1.21642911e+00
4.50532079e-01 3.16412672e-02 4.47970003e-01 1.07174027e+00
3.95769700e-02 3.71353418e-01 -4.39128846e-01 -4.49331291e-02
4.67416674e-01 1.14359963e+00 -5.10309994e-01 -4.34566677e-01
-4.11682665e-01 5.34264028e-01 3.00249934e-01 5.39455295e-01
-9.52630520e-01 -8.15004826e-01 -1.64461881e-01 -3.08516435e-02
2.87642449e-01 -2.45448500e-01 -5.89057803e-01 -9.68071818e-01
2.57958293e-01 -5.83384871e-01 3.13165128e-01 -6.47590220e-01
-1.29261649e+00 6.92113638e-01 -1.84672967e-01 -1.14801931e+00
1.45079404e-01 -6.12315238e-01 -9.04557168e-01 1.21671009e+00
-1.33708560e+00 -9.35193658e-01 -2.53381938e-01 4.88034308e-01
2.90396452e-01 -9.50959846e-02 5.91156244e-01 7.00127900e-01
-9.07635510e-01 3.47730756e-01 2.62365669e-01 4.17571634e-01
4.96569932e-01 -8.49609196e-01 2.23168135e-01 6.78709447e-01
-4.44084108e-01 8.02156925e-01 1.52057528e-01 -7.98565269e-01
-7.93376923e-01 -1.05948126e+00 9.91057277e-01 2.87078489e-02
4.39814329e-01 -2.25574374e-01 -1.48054194e+00 4.62823540e-01
-1.65405005e-01 -7.72753954e-02 2.38772929e-01 1.44586995e-01
-4.86417450e-02 -5.35886526e-01 -1.02637136e+00 3.55316997e-01
5.30322731e-01 -1.35964841e-01 -6.55360043e-01 2.09261760e-01
1.07420397e+00 -3.84342968e-01 -1.05579686e+00 7.35302687e-01
3.80136073e-01 -5.63641250e-01 7.36711264e-01 -4.37034994e-01
2.12386444e-01 -6.80149496e-01 5.67399323e-01 -1.16979063e+00
-6.55122042e-01 1.32759400e-02 1.21482089e-01 1.72214043e+00
7.03429759e-01 -9.55258965e-01 4.11924988e-01 6.44574523e-01
-4.45278585e-01 -8.69643688e-01 -7.52851129e-01 -8.78887951e-01
1.67117968e-01 2.17297580e-02 8.16363573e-01 1.04001391e+00
-9.05152857e-02 6.84739828e-01 -8.86322781e-02 -1.42181695e-01
-2.49608248e-01 -1.48021713e-01 3.73839021e-01 -1.22762930e+00
2.00108677e-01 -4.61625785e-01 -6.64061978e-02 -1.37669861e+00
-1.98067054e-02 -6.00279212e-01 1.91744059e-01 -1.75652409e+00
-1.72478184e-01 -8.23348880e-01 -5.91140330e-01 5.70940197e-01
-4.66451645e-01 -4.22199756e-01 2.30420411e-01 3.34332079e-01
-1.73170522e-01 8.18130016e-01 1.68316615e+00 -1.07510030e-01
-4.75338489e-01 1.74464270e-01 -5.31399965e-01 7.50116825e-01
7.23256826e-01 -3.88946623e-01 -1.70876086e-01 -4.95202690e-01
4.34015989e-02 -1.00850344e-01 -1.31950095e-01 -7.11511016e-01
3.22607249e-01 2.72286939e-03 8.43110681e-01 -7.18020558e-01
-4.36156914e-02 -1.10505867e+00 3.01925931e-02 6.10035241e-01
1.18232854e-01 2.34950110e-01 2.51552165e-01 3.18277031e-01
-4.63681340e-01 -3.75169069e-01 6.68773592e-01 -1.30241811e-01
-5.93595743e-01 4.34623033e-01 -1.11237234e-02 -5.99200577e-02
7.58104682e-01 -1.33374706e-01 -1.34701908e-01 3.19569409e-02
-3.27605277e-01 4.97297227e-01 5.76686487e-02 5.55542350e-01
4.14229512e-01 -1.65502512e+00 -5.75733483e-01 4.17432874e-01
-3.68134111e-01 7.45785832e-01 4.17824537e-01 8.44882905e-01
-5.98119795e-01 1.23559296e-01 -1.38492649e-02 -4.29768801e-01
-6.63036942e-01 7.22635984e-01 3.59776884e-01 -4.67132419e-01
-5.88259459e-01 5.90231180e-01 4.98512655e-01 -5.99757016e-01
3.14917825e-02 -5.08974314e-01 -7.68996060e-01 -1.38444275e-01
2.78154224e-01 3.70141298e-01 6.37597144e-02 -6.41084075e-01
-2.35993668e-01 6.53019488e-01 -1.64781973e-01 5.00952959e-01
1.21752787e+00 7.54772648e-02 -3.03273827e-01 6.38049603e-01
1.17200470e+00 -5.89461289e-02 -9.58597302e-01 -1.19150974e-01
-7.75396526e-02 -1.43882483e-01 -5.44080734e-02 -8.30739856e-01
-1.25822043e+00 9.25905585e-01 5.20883799e-01 5.25265098e-01
1.10444176e+00 -4.98477757e-01 1.27730393e+00 2.12195709e-01
9.18068066e-02 -1.27555990e+00 -4.00200635e-01 1.00908065e+00
5.01679540e-01 -1.07697189e+00 -1.76288873e-01 -5.07508576e-01
-7.69834757e-01 1.25815535e+00 8.79028320e-01 -2.10181519e-01
7.48192966e-01 4.48446155e-01 1.44089013e-01 1.40602263e-02
-4.31535691e-02 2.95157552e-01 3.34893465e-01 3.08142096e-01
5.63359201e-01 -7.75226951e-02 -8.19366574e-01 1.15875614e+00
-1.75655961e-01 -8.49997159e-03 -8.07985365e-02 6.36658192e-01
-4.20337737e-01 -1.04768026e+00 -2.53750056e-01 3.20796788e-01
-6.86262667e-01 -2.30236843e-01 -1.57061554e-02 9.02040064e-01
5.44899106e-01 9.09111559e-01 8.54955688e-02 -5.25189996e-01
7.72326350e-01 -6.53315336e-04 -1.43272325e-01 -2.04439208e-01
-7.73522556e-01 3.39595169e-01 -1.76919237e-01 -5.90914711e-02
-2.39676908e-01 -4.62568402e-02 -1.99406111e+00 -1.70256227e-01
-1.07341886e+00 4.76263046e-01 6.74297512e-01 1.12366855e+00
2.70245522e-01 8.63698483e-01 6.44589007e-01 -2.27925062e-01
-5.20649672e-01 -1.37640941e+00 -7.29587495e-01 5.11444211e-01
-4.16727029e-02 -5.59368312e-01 -1.50768831e-01 -2.02016100e-01] | [9.769743919372559, 9.692291259765625] |
fd69658e-2532-4722-83e0-c02bbd9dcfcb | claster-clustering-with-reinforcement | 2101.07042 | null | https://arxiv.org/abs/2101.07042v3 | https://arxiv.org/pdf/2101.07042v3.pdf | CLASTER: Clustering with Reinforcement Learning for Zero-Shot Action Recognition | Zero-shot action recognition is the task of recognizingaction classes without visual examples, only with a seman-tic embedding which relates unseen to seen classes. Theproblem can be seen as learning a function which general-izes well to instances of unseen classes without losing dis-crimination between classes. Neural networks can modelthe complex boundaries between visual classes, which ex-plains their success as supervised models. However, inzero-shot learning, these highly specialized class bound-aries may not transfer well from seen to unseen classes.In this paper we propose a centroid-based representation,which clusters visual and semantic representation, consid-ers all training samples at once, and in this way generaliz-ing well to instances from unseen classes. We optimize theclustering using Reinforcement Learning which we show iscritical for our approach to work. We call the proposedmethod CLASTER and observe that it consistently outper-forms the state-of-the-art in all standard datasets, includ-ing UCF101, HMDB51 and Olympic Sports; both in thestandard zero-shot evaluation and the generalized zero-shotlearning. Further, we show that our model performs com-petitively in the image domain as well, outperforming thestate-of-the-art in many settings. | ['Marcus Rohrbach', 'Frank Keller', 'Laura Sevilla-Lara', 'Shreyank N Gowda'] | 2021-01-18 | null | null | null | null | ['zero-shot-action-recognition'] | ['computer-vision'] | [ 7.05184415e-02 9.28983390e-02 -4.53508884e-01 -2.34480813e-01
-8.14161181e-01 -4.07564074e-01 6.78796113e-01 1.41192049e-01
-5.59587479e-01 6.57461703e-01 4.43062223e-02 8.20055008e-02
-1.67690516e-01 -8.17540407e-01 -7.75272846e-01 -7.34915316e-01
-2.04188917e-02 7.43480206e-01 4.89752948e-01 -2.28482902e-01
9.26036504e-04 4.23343420e-01 -1.99022138e+00 7.34995842e-01
5.48621535e-01 8.50899756e-01 -1.49121776e-01 7.16472447e-01
6.44196663e-03 9.85456049e-01 -6.10939026e-01 -3.12359840e-01
9.40693766e-02 -5.40472567e-01 -8.47240984e-01 6.74654663e-01
6.75874352e-01 -2.06886604e-02 -4.74280387e-01 8.00953686e-01
1.41844824e-01 6.12109601e-01 9.99179482e-01 -1.45674872e+00
-8.05128753e-01 2.73897555e-02 -4.60295379e-01 3.17462742e-01
2.32787460e-01 -5.80762215e-02 1.05505693e+00 -9.46797729e-01
7.98163950e-01 1.15291595e+00 3.74093771e-01 8.41737926e-01
-1.33346593e+00 -3.83551687e-01 1.14023373e-01 6.87134385e-01
-1.10183275e+00 -4.26885366e-01 5.56334198e-01 -6.36471689e-01
1.25218880e+00 1.75158471e-01 4.88456815e-01 1.13829541e+00
-2.52716430e-02 1.20211577e+00 1.01599741e+00 -4.77291435e-01
4.37819600e-01 5.99920452e-02 4.82692987e-01 4.81667399e-01
1.69451773e-01 3.07730049e-01 -4.00848985e-01 3.13701957e-01
3.23158771e-01 4.06762183e-01 -2.16699108e-01 -8.15542877e-01
-8.95019650e-01 7.53610969e-01 6.23114765e-01 4.76875573e-01
-1.98364183e-01 3.09448063e-01 4.25023496e-01 4.00904208e-01
2.75165409e-01 2.63527364e-01 -1.29260629e-01 -1.86168745e-01
-6.89768374e-01 1.66115746e-01 5.43842852e-01 8.06838155e-01
7.83468485e-01 2.59788781e-01 -6.05141707e-02 8.87501597e-01
-8.00002441e-02 3.46821904e-01 8.81957471e-01 -9.13288295e-01
2.38422185e-01 4.28339243e-01 -3.68316332e-03 -5.84806859e-01
-3.46370012e-01 -1.57797888e-01 -7.35296786e-01 6.71971679e-01
4.46906775e-01 3.45645919e-02 -1.38418794e+00 1.60898781e+00
6.93736374e-02 4.62861717e-01 3.92453372e-01 7.77302802e-01
7.88293362e-01 6.32600248e-01 9.25141945e-02 -1.39528379e-01
1.10922635e+00 -9.86993313e-01 -4.93999243e-01 -3.56637061e-01
7.15766609e-01 -2.53509462e-01 1.12689483e+00 4.14170176e-01
-8.40067327e-01 -7.67178476e-01 -1.24419546e+00 1.33534759e-01
-7.62847960e-01 -3.66473198e-01 4.71918404e-01 3.22965652e-01
-9.08577204e-01 8.98088336e-01 -9.32701528e-01 -7.70904779e-01
7.18423307e-01 1.30455703e-01 -7.04773903e-01 -1.09608635e-01
-8.62481892e-01 9.78871763e-01 8.19538355e-01 -4.73808676e-01
-1.18454981e+00 -5.40464580e-01 -1.05036294e+00 1.98223114e-01
6.16818666e-01 -3.28616828e-01 1.26039517e+00 -1.18447220e+00
-1.17432892e+00 1.09959114e+00 1.54559761e-01 -5.58869839e-01
2.79063493e-01 -1.24864653e-02 -5.13799310e-01 4.02469009e-01
1.70932692e-02 7.97564328e-01 8.08493495e-01 -1.36583006e+00
-6.89149797e-01 -5.13552904e-01 2.77489215e-01 3.43193740e-01
-5.06254733e-01 -1.94125682e-01 -2.07080364e-01 -5.64204514e-01
-4.40242514e-03 -7.98745573e-01 -1.28533272e-02 1.84962422e-01
-3.66617918e-01 -3.38818222e-01 1.01488090e+00 -8.68580043e-02
8.13138485e-01 -2.38038015e+00 3.18905920e-01 -2.05211580e-01
1.31084397e-01 3.18879396e-01 -9.28936526e-02 3.98747236e-01
-5.50478041e-01 -3.46492797e-01 -3.16991985e-01 -1.45008951e-01
2.56713361e-01 7.73374498e-01 -1.39790878e-01 5.46820164e-01
1.71123266e-01 9.87054169e-01 -1.07081079e+00 -4.35800195e-01
5.22504687e-01 1.15095101e-01 -3.35082412e-01 3.83681566e-01
-8.25339481e-02 -2.06100598e-01 1.16178662e-01 5.52527547e-01
-3.78413908e-02 -3.61662269e-01 3.87657657e-02 2.39156231e-01
3.43699247e-01 -5.98640561e-01 -1.19085145e+00 1.52992213e+00
-4.01411146e-01 7.31449544e-01 -3.96592051e-01 -1.51408243e+00
6.18572891e-01 3.05133075e-01 2.87371904e-01 -5.30102491e-01
1.58273965e-01 -4.91992496e-02 2.51517575e-02 -7.76283503e-01
1.49924532e-01 -6.69191957e-01 -9.47206467e-02 2.10673958e-01
5.17665923e-01 2.37964705e-01 2.86577731e-01 8.61432031e-02
7.64626920e-01 4.52787988e-02 4.40595537e-01 -6.73766285e-02
1.14033446e-01 -5.61245270e-02 2.66887933e-01 8.81052196e-01
-3.48239958e-01 6.03771210e-01 4.20180887e-01 -3.49303335e-01
-8.69887769e-01 -1.38621807e+00 -4.34182063e-02 1.29478586e+00
1.72005922e-01 -2.89558500e-01 -6.88888073e-01 -9.26667929e-01
3.61673944e-02 9.12026823e-01 -1.03707099e+00 -4.92071539e-01
-1.99553847e-01 -4.96759295e-01 2.40252122e-01 9.24678266e-01
2.55080849e-01 -1.16791821e+00 -9.25500095e-01 5.88973314e-02
1.56894937e-01 -8.92937839e-01 -1.27902985e-01 5.61941743e-01
-7.52901673e-01 -1.37848151e+00 -8.36493552e-01 -8.54167640e-01
4.25449103e-01 2.54567921e-01 1.07901525e+00 -2.95461882e-02
-6.48487866e-01 5.81742644e-01 -5.51973820e-01 -3.43597054e-01
-3.45264465e-01 -2.57028461e-01 -9.22385156e-02 2.14460760e-01
7.13842928e-01 -4.89275426e-01 -2.86178648e-01 2.54034728e-01
-9.33491886e-01 -2.19533801e-01 4.40367430e-01 9.82532740e-01
5.92916965e-01 -8.77342820e-02 3.74473244e-01 -8.66378248e-01
1.72152773e-01 -5.39693177e-01 -1.81976229e-01 5.35108149e-01
-2.71005571e-01 2.28978489e-02 7.60610759e-01 -5.15228510e-01
-6.70687914e-01 3.86404395e-02 1.05678067e-01 -8.61777604e-01
-6.10703230e-01 7.48963356e-02 -2.66735226e-01 1.58247620e-01
9.79174554e-01 1.85328767e-01 -1.12494081e-01 -4.53639656e-01
5.99419713e-01 7.34274387e-01 6.88066781e-01 -4.45602059e-01
6.43398583e-01 6.03783250e-01 1.21435069e-03 -1.08441758e+00
-9.49077487e-01 -7.74242282e-01 -7.37674773e-01 -1.42851576e-01
1.35758018e+00 -5.37369072e-01 -4.61458236e-01 3.78061295e-01
-8.43817413e-01 -3.68408263e-01 -8.19916010e-01 3.55417877e-01
-1.07302999e+00 3.15718412e-01 -6.31200910e-01 -7.08278596e-01
2.61114240e-01 -8.66723895e-01 1.08588398e+00 1.54497042e-01
-1.16897978e-01 -9.90657270e-01 1.38612956e-01 4.53111157e-02
-5.22707552e-02 5.45080721e-01 9.46581781e-01 -1.00734329e+00
-1.91132709e-01 -3.88497055e-01 -1.24148563e-01 3.79890591e-01
-9.62134674e-02 -2.70032376e-01 -1.17083943e+00 -2.49133274e-01
-2.53143638e-01 -7.30521917e-01 1.21335220e+00 2.61959404e-01
1.36729622e+00 3.47390361e-02 -1.58588514e-01 4.72240597e-01
1.59790015e+00 2.70443410e-01 8.49533737e-01 3.10554028e-01
5.57848334e-01 4.74203229e-01 4.35731709e-01 3.78385454e-01
1.04073457e-01 8.35363507e-01 5.79190850e-01 -1.83541939e-01
-2.09095180e-01 -1.86122566e-01 1.79528564e-01 2.70283967e-01
-1.94503754e-01 -2.52573162e-01 -7.25095987e-01 5.46035588e-01
-2.04839158e+00 -1.46313274e+00 1.79001942e-01 2.30436492e+00
4.76523310e-01 1.95303604e-01 5.16276300e-01 2.62526155e-01
6.29532039e-01 2.67525196e-01 -4.98656183e-01 -4.40851182e-01
-1.96849518e-02 3.97495091e-01 1.53476387e-01 3.83060992e-01
-1.38175106e+00 9.19136107e-01 5.95576000e+00 8.33942533e-01
-7.28330016e-01 2.70267010e-01 5.22335768e-01 -1.11834526e-01
3.44899684e-01 -2.32893243e-01 -5.97277462e-01 2.56931633e-01
1.00903559e+00 -2.51004785e-01 3.61450315e-01 1.14135861e+00
-2.67968476e-01 1.36415079e-01 -1.44455266e+00 1.10373807e+00
6.40438855e-01 -1.20222318e+00 5.83853722e-02 -1.00098819e-01
5.96457243e-01 -5.92044070e-02 -1.51339844e-01 7.49770761e-01
3.18988055e-01 -1.10840070e+00 4.13966626e-01 4.19028491e-01
8.07448149e-01 -8.27735722e-01 5.53589463e-01 3.88530433e-01
-9.10071909e-01 -4.11851734e-01 -6.92720890e-01 1.12188466e-01
-4.29614633e-02 -9.94019136e-02 -4.89152938e-01 4.51501131e-01
6.27108335e-01 1.17029262e+00 -4.76627946e-01 1.18833077e+00
-2.60990381e-01 2.02123895e-01 2.13321865e-01 7.25664124e-02
4.43743676e-01 9.75940973e-02 2.53183186e-01 1.13081300e+00
2.64480114e-01 2.66885698e-01 6.90885067e-01 3.84403139e-01
-9.27680172e-03 -8.40044022e-02 -7.36660182e-01 -1.56820923e-01
-5.02200313e-02 9.59653497e-01 -7.45499611e-01 -8.62199783e-01
-4.72895712e-01 1.22611213e+00 5.06473482e-01 5.67324638e-01
-7.68202424e-01 -7.67801166e-01 5.65447867e-01 1.84423551e-01
6.29087985e-01 1.68250099e-01 2.16518581e-01 -1.36962187e+00
-1.43169418e-01 -7.87502468e-01 8.74663413e-01 -8.43233824e-01
-1.40711701e+00 6.44150436e-01 4.11128253e-01 -1.48300529e+00
-4.55794156e-01 -8.92375648e-01 -6.23258173e-01 2.09927201e-01
-1.44052815e+00 -9.54959750e-01 -1.48584798e-01 7.63774276e-01
8.81151497e-01 -3.21299851e-01 1.03426111e+00 4.96850535e-02
-4.44587380e-01 5.22897184e-01 2.64669061e-01 1.89932600e-01
5.77014446e-01 -1.59514332e+00 -9.50203016e-02 6.48290277e-01
4.74457860e-01 2.92105079e-01 7.58223176e-01 -3.59647214e-01
-1.00856137e+00 -1.01316571e+00 7.06234872e-01 -2.73281097e-01
8.39795530e-01 -4.04555917e-01 -1.14007902e+00 7.00258195e-01
2.62890249e-01 3.91835064e-01 8.29977334e-01 3.18101496e-01
-4.78198409e-01 -9.45182592e-02 -8.80124688e-01 4.02243763e-01
9.70498979e-01 -5.32801032e-01 -1.17784178e+00 7.75444448e-01
3.72223020e-01 -1.82423100e-01 -6.28119767e-01 3.20267975e-02
4.25919443e-01 -1.13547039e+00 1.06041515e+00 -1.38393390e+00
5.45406699e-01 -5.88156171e-02 -3.79335493e-01 -1.49452889e+00
-3.60021919e-01 -1.02099657e-01 -4.48648751e-01 6.25309646e-01
1.77176028e-01 -3.39495778e-01 1.09898722e+00 2.27523237e-01
-3.32815349e-01 -6.78676963e-01 -9.94131923e-01 -1.40254378e+00
2.52572149e-01 -3.04984510e-01 1.67429432e-01 1.03482914e+00
1.57064050e-01 3.87096077e-01 -3.98886591e-01 -1.86186746e-01
7.41127431e-01 7.17642754e-02 6.17109537e-01 -1.26372266e+00
-7.89461315e-01 -3.67054671e-01 -1.17361283e+00 -6.87847495e-01
4.19603795e-01 -9.68719840e-01 1.27276108e-01 -1.48801672e+00
3.20485085e-01 2.33775839e-01 -5.88914752e-01 8.24165881e-01
-1.06253445e-01 4.35471445e-01 4.23704505e-01 -5.10339066e-02
-1.00878704e+00 5.35564542e-01 1.16824126e+00 -3.71180445e-01
4.95988652e-02 -1.67197868e-01 -3.67678165e-01 7.22166121e-01
5.42658567e-01 -4.70436931e-01 -7.23181129e-01 -1.27809390e-01
-3.72022867e-01 5.28760776e-02 6.07317626e-01 -1.18267250e+00
2.62092613e-02 -3.27349365e-01 6.27199829e-01 -3.21422279e-01
6.73698425e-01 -8.56419683e-01 -2.43185043e-01 6.33116066e-01
-4.73697454e-01 -2.00651854e-01 1.98304793e-03 8.44395876e-01
-2.93515265e-01 -5.18129408e-01 1.06043231e+00 -4.04608727e-01
-1.26478910e+00 2.89893746e-01 -2.44522840e-01 4.82264429e-01
1.52617288e+00 -6.86182380e-01 -2.96381116e-01 -3.91451001e-01
-1.36484659e+00 1.34910494e-01 5.89433014e-01 5.15951753e-01
6.26935780e-01 -1.43660712e+00 -5.81250668e-01 2.58345813e-01
6.11051977e-01 -6.17264330e-01 4.60154355e-01 5.64241171e-01
-3.37709457e-01 1.57830149e-01 -3.57965350e-01 -7.35852420e-01
-1.21171796e+00 9.86549199e-01 4.81780499e-01 1.27550930e-01
-7.74768710e-01 7.20788777e-01 4.77510899e-01 -2.48020411e-01
5.11393607e-01 -7.45631456e-02 -3.30330253e-01 7.56459758e-02
9.17629063e-01 3.15133184e-01 -8.65516365e-02 -7.20579743e-01
-2.44988024e-01 4.67380196e-01 -1.37400697e-03 1.97776213e-01
1.42434883e+00 3.94767404e-01 3.99136305e-01 9.84153628e-01
1.52366149e+00 -9.39299643e-01 -1.20555973e+00 -1.10931825e-02
8.69739726e-02 -6.47223532e-01 -3.18773091e-01 -5.49055099e-01
-8.65167558e-01 1.33480549e+00 8.54946554e-01 2.93147177e-01
8.63339186e-01 2.14387864e-01 3.90644282e-01 5.97416937e-01
2.74519265e-01 -1.34406769e+00 5.59182405e-01 2.74487674e-01
6.81309462e-01 -1.50397742e+00 -2.04959035e-01 -1.33857220e-01
-9.27667081e-01 1.29800403e+00 7.00806379e-01 -5.51636457e-01
5.33680856e-01 -8.96690562e-02 -5.78620844e-02 -3.48651826e-01
-7.52452254e-01 -4.63962555e-01 4.51890171e-01 9.28389907e-01
1.16734408e-01 1.07074447e-01 4.73279133e-02 2.89005131e-01
2.52460659e-01 -9.93046816e-03 5.54725587e-01 1.13174856e+00
-9.25403714e-01 -9.24710989e-01 -2.13062823e-01 4.94678229e-01
-1.62295029e-01 2.06019893e-01 -4.14142221e-01 1.24337077e+00
1.91846445e-01 8.83509994e-01 3.39001089e-01 -2.21164197e-01
4.78442878e-01 1.35592684e-01 5.92750132e-01 -1.03499067e+00
-2.50466466e-01 3.98056023e-02 1.04630202e-01 -6.58174217e-01
-4.12063479e-01 -4.97515619e-01 -1.27051544e+00 7.56497905e-02
-2.02315286e-01 1.08568624e-01 2.89773941e-01 1.17966390e+00
-8.08765963e-02 4.99322444e-01 4.42835480e-01 -7.52809405e-01
-8.10094833e-01 -5.79388082e-01 -8.86685312e-01 8.13035369e-01
2.95455784e-01 -9.14514184e-01 -4.25541848e-01 1.41244475e-02] | [8.72701358795166, 1.1684004068374634] |
177af928-9c1a-449a-b9c2-4d10fd0ed08a | deep-instance-segmentation-and-visual | 2211.07977 | null | https://arxiv.org/abs/2211.07977v2 | https://arxiv.org/pdf/2211.07977v2.pdf | Deep Instance Segmentation and Visual Servoing to Play Jenga with a Cost-Effective Robotic System | The game of Jenga represents an inspiring benchmark for developing innovative manipulation solutions for complex tasks. Indeed, it encouraged the study of novel robotics methods to successfully extract blocks from the tower. A Jenga game round undoubtedly embeds many traits of complex industrial or surgical manipulation tasks, requiring a multi-step strategy, the combination of visual and tactile data, and the highly precise motion of the robotic arm to perform a single block extraction. In this work, we propose a novel, cost-effective architecture for playing Jenga with e.Do, a 6-DOF anthropomorphic manipulator manufactured by Comau, a standard depth camera, and an inexpensive monodirectional force sensor. Our solution focuses on a visual-based control strategy to accurately align the end-effector with the desired block, enabling block extraction by pushing. To this aim, we train an instance segmentation deep learning model on a synthetic custom dataset to segment each piece of the Jenga tower, allowing visual tracking of the desired block's pose during the motion of the manipulator. We integrate the visual-based strategy with a 1D force sensor to detect whether the block can be safely removed by identifying a force threshold value. Our experimentation shows that our low-cost solution allows e.DO to precisely reach removable blocks and perform up to 14 consecutive extractions in a row. | ['Marcello Chiaberge', 'Francesco Salvetti', 'Simone Angarano', 'Mauro Martini', 'Giulio Pugliese', 'Luca Marchionna'] | 2022-11-15 | null | null | null | null | ['visual-tracking'] | ['computer-vision'] | [-1.08948156e-01 3.37490797e-01 -1.03946142e-01 3.02994251e-01
-3.72602493e-01 -8.23726475e-01 9.66795906e-02 -4.40171473e-02
-5.05454361e-01 3.34926665e-01 -6.79013193e-01 -3.80809575e-01
-4.28956836e-01 -3.79063219e-01 -8.82788241e-01 -6.38664007e-01
-5.45353293e-02 7.56783187e-01 1.97402194e-01 -5.95525146e-01
4.20711160e-01 8.82087231e-01 -1.25013041e+00 -1.71268180e-01
5.93098342e-01 8.35024118e-01 1.02249134e+00 5.74494421e-01
5.05944431e-01 4.28945094e-01 -3.77584338e-01 -1.56906486e-01
6.07265592e-01 -8.11402947e-02 -7.10002840e-01 1.53195215e-02
-1.64884344e-01 -5.26654422e-01 -4.55910265e-02 7.74599075e-01
3.67428720e-01 -1.00904405e-01 5.75565696e-01 -1.17333519e+00
2.65012592e-01 6.49266422e-01 -6.00912631e-01 -5.76883197e-01
2.23403558e-01 5.19240618e-01 5.70603490e-01 -3.96123409e-01
1.27302361e+00 6.69857860e-01 4.99021232e-01 6.70180798e-01
-9.97094393e-01 -4.64813858e-01 -3.18003893e-01 -7.30774477e-02
-9.91996229e-01 -4.13125707e-03 9.22522306e-01 -8.34113479e-01
5.25441587e-01 3.13194215e-01 1.00893450e+00 1.03497291e+00
7.94867396e-01 7.31056392e-01 8.35487843e-01 -2.67244935e-01
2.02951133e-01 -1.41137928e-01 -4.51498479e-01 9.61621821e-01
8.59070476e-03 2.70719379e-01 2.48701703e-02 2.89732188e-01
1.48668504e+00 2.75167584e-01 -2.13184848e-01 -1.08687007e+00
-1.41792476e+00 3.43690604e-01 8.77926171e-01 2.67644882e-01
-5.26994407e-01 4.69575673e-01 2.19409212e-01 2.49035403e-01
-3.89384955e-01 9.79674399e-01 -5.31871498e-01 -3.87738526e-01
-4.65517104e-01 2.29422361e-01 6.24752223e-01 1.15973055e+00
4.33672190e-01 -3.52180749e-01 1.23217531e-01 2.63487250e-01
8.13792795e-02 2.14673728e-01 7.40782097e-02 -1.07186580e+00
3.50870818e-01 7.47072458e-01 2.19531804e-01 -7.57038772e-01
-9.22195196e-01 -2.66155541e-01 -7.50068069e-01 8.63472164e-01
5.50489545e-01 -1.51791915e-01 -1.01670408e+00 1.21930313e+00
4.79535758e-01 -7.16880023e-01 -4.08632755e-01 1.31214714e+00
3.21561426e-01 2.08930477e-01 -4.11940634e-01 9.53484327e-02
1.43647325e+00 -7.86386132e-01 -3.16233456e-01 -2.54529178e-01
5.72392941e-01 -5.36169469e-01 9.17030513e-01 9.12283421e-01
-1.20252705e+00 -5.39094746e-01 -1.17951012e+00 -2.03580782e-01
-1.53918624e-01 6.58403516e-01 6.75295889e-01 1.53480247e-01
-3.71319652e-01 9.21109855e-01 -1.17339563e+00 5.77942356e-02
3.70099604e-01 7.37701416e-01 -8.16806734e-01 1.72497407e-01
-5.99545836e-01 1.24641597e+00 4.34442282e-01 5.36941469e-01
-9.95861173e-01 -4.89040554e-01 -7.60693073e-01 4.91135614e-03
6.73327088e-01 -8.38204622e-01 1.07089186e+00 -4.40346420e-01
-2.00588560e+00 1.04810441e+00 6.80257916e-01 -2.99456060e-01
1.00952125e+00 -3.38680565e-01 6.36011779e-01 3.71696025e-01
-2.11233214e-01 8.15896749e-01 1.06901693e+00 -1.14853346e+00
-3.03961426e-01 -5.67476094e-01 4.43034023e-01 6.84948489e-02
5.76318018e-02 -5.71244180e-01 -4.96138662e-01 -6.01273715e-01
3.07320654e-01 -1.20878839e+00 -4.54532295e-01 4.87016708e-01
-6.28602386e-01 -7.22244382e-03 5.01020372e-01 -6.68195486e-01
6.46971345e-01 -2.24609613e+00 9.81426716e-01 2.77846158e-01
4.63882625e-01 7.80458003e-02 1.31146491e-01 5.11658967e-01
3.16657662e-01 -2.34352127e-01 -2.03384552e-02 3.68158966e-02
-2.09659293e-01 -7.56095871e-02 2.30431989e-01 5.82393289e-01
-1.09102271e-01 8.63658428e-01 -9.35666502e-01 -2.76679695e-01
3.56893092e-01 5.36438599e-02 -5.80590725e-01 2.87161291e-01
-3.74018461e-01 7.84733415e-01 -5.18703401e-01 6.68025792e-01
5.71053505e-01 1.34969980e-01 3.61057729e-01 -2.76708156e-01
-1.28144965e-01 -1.86830088e-01 -8.01221788e-01 2.32055902e+00
-6.80487394e-01 3.69282216e-01 7.53068209e-01 -6.04002178e-01
1.11043215e+00 9.41477418e-02 6.04162276e-01 -3.16431165e-01
7.09577322e-01 5.02576530e-01 6.13638014e-02 -6.06135070e-01
1.79607123e-01 -2.79610530e-02 -2.59061903e-01 9.52102467e-02
-5.09801246e-02 -9.72407341e-01 7.90942311e-02 -2.05638647e-01
1.09951448e+00 4.48847950e-01 1.32784605e-01 -1.80372819e-01
1.92647353e-01 5.49120188e-01 1.09386444e-01 5.39908588e-01
8.34339485e-02 4.83658075e-01 6.20912611e-01 -3.77551585e-01
-1.06213737e+00 -1.10772824e+00 3.81440669e-01 5.24148107e-01
5.04979491e-01 -4.86954749e-02 -7.64756441e-01 -5.04054666e-01
2.18024760e-01 2.13261232e-01 -5.26558936e-01 -3.83676082e-01
-8.45441878e-01 2.38220930e-01 2.06690375e-02 3.76857311e-01
6.29246905e-02 -1.12030649e+00 -1.40428138e+00 2.47203052e-01
1.82800919e-01 -8.52704108e-01 1.19822480e-01 5.53249538e-01
-9.58370149e-01 -1.30920947e+00 -7.42793739e-01 -1.05859494e+00
7.26287723e-01 -2.07254514e-02 6.10375404e-01 2.94909589e-02
-6.65468156e-01 7.13454734e-04 -3.29765856e-01 -3.50373000e-01
-4.89910543e-01 4.40368950e-01 -3.33212167e-01 -6.99954450e-01
-4.28618103e-01 -3.90348136e-01 -7.53188372e-01 5.28246284e-01
-5.66277027e-01 5.09942293e-01 1.03760719e+00 6.74145222e-01
3.44321400e-01 -1.96848676e-01 2.25569069e-01 -2.55017191e-01
4.42637742e-01 -8.23615640e-02 -8.24280381e-01 -2.03988761e-01
2.01347008e-01 -2.05540478e-01 7.59674847e-01 -5.52516878e-01
-5.39297879e-01 7.48336732e-01 -2.29305789e-01 -7.23748446e-01
1.97311305e-02 3.80168349e-01 2.89038867e-02 9.27412137e-03
6.09025538e-01 -1.63183078e-01 5.76399148e-01 -2.87582308e-01
4.66307133e-01 6.02113605e-01 7.80780911e-01 -3.92288059e-01
7.28851795e-01 6.23759329e-01 2.29694769e-01 -5.91242075e-01
-9.16137472e-02 -3.97183925e-01 -1.07791948e+00 -6.05883837e-01
7.59276688e-01 -4.46849883e-01 -1.39756036e+00 4.95790720e-01
-1.21911192e+00 -6.37991011e-01 -1.83579162e-01 6.50461733e-01
-1.09600282e+00 -6.12159912e-03 -7.59357870e-01 -5.00941575e-01
-4.64150637e-01 -1.51593792e+00 1.26286256e+00 7.09169209e-02
-3.77340794e-01 -2.91880429e-01 -2.81104892e-01 3.41581732e-01
6.71924427e-02 7.08162189e-01 7.37095237e-01 1.15257323e-01
-7.18435645e-01 -5.86793482e-01 2.63537198e-01 4.26991470e-02
1.30940750e-01 7.18816593e-02 -3.70678544e-01 -5.06143212e-01
1.00418687e-01 -3.20807278e-01 5.04839659e-01 3.46570045e-01
1.08656371e+00 -3.68878022e-02 -6.40349925e-01 5.70830047e-01
1.26114511e+00 2.60266513e-01 5.96086681e-01 4.59552705e-01
7.77930796e-01 6.12894714e-01 1.05675256e+00 2.61281222e-01
-2.36853480e-01 1.01053298e+00 1.01912129e+00 8.11357275e-02
2.10463762e-01 1.26997933e-01 2.57334769e-01 5.12814879e-01
-3.57157469e-01 2.26298556e-01 -8.49776685e-01 3.84947598e-01
-1.61476791e+00 -4.23455954e-01 -1.33104533e-01 2.13706088e+00
7.53960967e-01 3.83152455e-01 9.49921608e-02 1.85411155e-01
3.90176535e-01 -4.22894090e-01 -5.19666493e-01 -3.25918943e-01
7.84319937e-01 6.28436431e-02 6.24919593e-01 2.67083257e-01
-9.12048221e-01 8.86291027e-01 4.83339262e+00 7.75680840e-01
-1.53495824e+00 -4.61829424e-01 -1.45332590e-01 -2.79341131e-01
2.80202836e-01 -3.52242172e-01 -3.82298023e-01 2.30041116e-01
1.45566925e-01 2.69594193e-01 3.36706340e-01 9.26693857e-01
1.83606207e-01 -2.66013384e-01 -1.23899937e+00 9.03334856e-01
-1.71518415e-01 -1.44441462e+00 -4.05884981e-01 1.25835268e-02
3.35818619e-01 -2.69992530e-01 -3.89891900e-02 4.36658114e-02
-2.72706617e-03 -7.78946877e-01 9.75326121e-01 5.86402059e-01
7.64258921e-01 -6.06224179e-01 4.70724583e-01 9.04624224e-01
-8.42320085e-01 -5.12787104e-01 -2.53942996e-01 -8.51920173e-02
2.20187500e-01 4.18707013e-01 -1.20297635e+00 7.84626126e-01
5.16664386e-01 3.10937703e-01 -9.43998992e-02 1.04417622e+00
-3.82102758e-01 -4.39160556e-01 -2.76671469e-01 -1.82898894e-01
3.30731034e-01 -2.66950995e-01 7.79225290e-01 5.34282506e-01
2.41193354e-01 -2.64156699e-01 -4.47509624e-02 8.61157358e-01
9.95469838e-02 -1.43532082e-01 -6.07434332e-01 1.08891711e-01
-4.27882969e-02 1.51664269e+00 -1.08694351e+00 1.14047870e-01
5.07053733e-01 1.00749540e+00 3.10730875e-01 -2.24183068e-01
-8.91524076e-01 -6.35306656e-01 4.29614335e-01 2.52445757e-01
4.15131301e-01 -6.96370602e-01 -3.46064031e-01 -6.76968813e-01
3.95933539e-01 -6.55930817e-01 -3.28288376e-01 -1.11939859e+00
-3.60329419e-01 3.94586384e-01 -2.17518687e-01 -1.53720105e+00
-4.19450998e-01 -9.72095132e-01 -2.12471470e-01 4.50145006e-01
-9.07258689e-01 -1.12349844e+00 -5.09232998e-01 2.57169127e-01
6.30607426e-01 2.91788965e-01 5.90053797e-01 -2.11824417e-01
-2.11479753e-01 -3.89118567e-02 -1.89721212e-01 1.41661614e-01
5.76061845e-01 -1.11958897e+00 1.64448127e-01 3.18949699e-01
-2.85169572e-01 5.23494899e-01 6.43051088e-01 -6.47988856e-01
-1.94759667e+00 -5.50797403e-01 4.50107781e-03 -4.10524905e-01
4.29568827e-01 -7.42806554e-01 -4.89717215e-01 8.05667520e-01
-2.71844417e-01 -3.96379143e-01 -1.22177832e-01 -4.00163740e-01
3.11957240e-01 -1.63928598e-01 -1.23927796e+00 7.97977269e-01
1.07306027e+00 -2.91202217e-02 -6.87611043e-01 4.28265125e-01
3.48955512e-01 -1.16036117e+00 -9.26034808e-01 5.59363723e-01
1.17022216e+00 -7.94391215e-01 9.86128926e-01 -3.74943614e-01
8.66068304e-01 -4.67043638e-01 7.36187696e-01 -1.38329554e+00
-1.40822515e-01 -8.22197855e-01 3.39463465e-02 4.07819659e-01
1.22755937e-01 -3.19822073e-01 1.02191687e+00 1.96401685e-01
-3.91144812e-01 -1.13214743e+00 -9.35388267e-01 -8.16302896e-01
-1.38245761e-01 -6.02982119e-02 2.33865753e-01 5.51403999e-01
5.50043941e-01 -4.33083922e-02 -1.94400311e-01 1.50750339e-01
1.35563001e-01 3.49490881e-01 1.26469815e+00 -1.33604777e+00
-2.92605489e-01 -6.96071088e-01 -6.80815458e-01 -1.21023679e+00
-1.81916773e-01 -6.41215026e-01 3.77213031e-01 -1.61656439e+00
-3.72831792e-01 -4.11521077e-01 3.58132303e-01 4.47142243e-01
3.29820722e-01 -9.25199501e-03 4.39482182e-01 2.48430729e-01
2.26234142e-02 7.54541457e-02 1.77798545e+00 -1.52932465e-01
-4.51294661e-01 2.23047584e-01 -1.15901850e-01 7.69133389e-01
7.79771447e-01 -4.01347011e-01 -1.42570511e-01 -3.10613692e-01
5.45432746e-01 4.91160631e-01 3.84558856e-01 -9.63897824e-01
4.91363764e-01 -1.16554666e-02 3.17807347e-01 -5.23982406e-01
4.24188048e-01 -1.47494853e+00 3.46097916e-01 1.28071487e+00
-5.49631603e-02 -1.25878111e-01 2.95829475e-01 2.15200081e-01
1.32509559e-01 -3.40795368e-01 6.41951084e-01 -3.47163141e-01
-4.14762259e-01 -1.44066140e-01 -4.20268655e-01 -7.16008306e-01
1.52665281e+00 -3.25066328e-01 -1.15158528e-01 2.33676389e-01
-1.18886709e+00 3.75885755e-01 7.68209338e-01 4.45745587e-01
6.26512885e-01 -7.28714049e-01 -4.80996966e-01 4.74155426e-01
-1.38611808e-01 5.21672845e-01 4.26641524e-01 1.14760041e+00
-1.33420897e+00 5.64856678e-02 -7.69023538e-01 -1.01559460e+00
-1.22298837e+00 7.55662918e-01 5.49181402e-01 -2.92210579e-01
-8.25799108e-01 5.44654846e-01 7.03655649e-03 -5.08453250e-01
-1.44615900e-02 -8.44603837e-01 -3.71621363e-02 7.26091629e-03
8.55047777e-02 3.54420304e-01 2.47695014e-01 1.50218129e-01
-1.88970745e-01 6.91660047e-01 7.73739666e-02 1.10504851e-01
1.47896945e+00 2.22834930e-01 -4.51692520e-03 1.76919445e-01
8.46394241e-01 5.12122326e-02 -1.44556379e+00 6.98824883e-01
-3.89661282e-01 -4.12316948e-01 -3.04190427e-01 -8.47967565e-01
-1.12307322e+00 6.36776924e-01 1.47276610e-01 1.17133260e-01
8.87065768e-01 3.57456692e-02 6.48345172e-01 4.55267012e-01
8.57746720e-01 -9.13790882e-01 1.13791555e-01 2.41796046e-01
1.35067379e+00 -9.52094674e-01 -2.53587984e-03 -7.81198919e-01
-3.02254379e-01 1.36171186e+00 5.41274965e-01 -3.66188705e-01
3.85936856e-01 3.78916800e-01 3.18553112e-02 -6.40196860e-01
-1.52794734e-01 1.14881158e-01 7.32864439e-02 4.70972836e-01
-3.23862910e-01 1.60665601e-01 -2.64615208e-01 2.65727103e-01
-4.47808027e-01 3.10036182e-01 5.21823704e-01 1.35820556e+00
-4.90665704e-01 -5.69798648e-01 -2.53629327e-01 2.84045517e-01
-2.08433747e-01 5.80986142e-01 -3.10334146e-01 1.09396553e+00
-9.57482234e-02 4.86884922e-01 1.10317655e-02 -2.36893743e-01
7.32884884e-01 -3.70945126e-01 9.09623682e-01 -5.38949311e-01
-9.16559398e-01 1.03251167e-01 -2.27433443e-01 -8.30456674e-01
1.39218364e-02 -1.74554616e-01 -1.29618585e+00 1.24137504e-02
-3.34185630e-01 -3.10134888e-01 1.17926574e+00 7.62935698e-01
1.58417463e-01 6.79407775e-01 5.40642262e-01 -1.61882389e+00
-5.75119615e-01 -7.73921072e-01 -6.50561392e-01 2.94976205e-01
3.25092763e-01 -1.02465928e+00 -1.39182061e-01 -2.19623238e-01] | [5.944459915161133, -0.7116377949714661] |
6c88180b-f65e-431e-917c-9a462fb8d741 | semattack-natural-textual-attacks-via-1 | 2205.01287 | null | https://arxiv.org/abs/2205.01287v3 | https://arxiv.org/pdf/2205.01287v3.pdf | SemAttack: Natural Textual Attacks via Different Semantic Spaces | Recent studies show that pre-trained language models (LMs) are vulnerable to textual adversarial attacks. However, existing attack methods either suffer from low attack success rates or fail to search efficiently in the exponentially large perturbation space. We propose an efficient and effective framework SemAttack to generate natural adversarial text by constructing different semantic perturbation functions. In particular, SemAttack optimizes the generated perturbations constrained on generic semantic spaces, including typo space, knowledge space (e.g., WordNet), contextualized semantic space (e.g., the embedding space of BERT clusterings), or the combination of these spaces. Thus, the generated adversarial texts are more semantically close to the original inputs. Extensive experiments reveal that state-of-the-art (SOTA) large-scale LMs (e.g., DeBERTa-v2) and defense strategies (e.g., FreeLB) are still vulnerable to SemAttack. We further demonstrate that SemAttack is general and able to generate natural adversarial texts for different languages (e.g., English and Chinese) with high attack success rates. Human evaluations also confirm that our generated adversarial texts are natural and barely affect human performance. Our code is publicly available at https://github.com/AI-secure/SemAttack. | ['Bo Li', 'Yu Cheng', 'Xiangyu Liu', 'Chejian Xu', 'Boxin Wang'] | 2022-05-03 | null | https://aclanthology.org/2022.findings-naacl.14 | https://aclanthology.org/2022.findings-naacl.14.pdf | findings-naacl-2022-7 | ['adversarial-text'] | ['adversarial'] | [-9.13253650e-02 -1.03481352e-01 9.22998637e-02 6.55692667e-02
-9.29911911e-01 -1.25722945e+00 7.56288826e-01 -2.02374622e-01
-4.54639554e-01 6.91934347e-01 1.85554862e-01 -5.49777746e-01
3.04835200e-01 -1.03713071e+00 -8.63009810e-01 -5.85826993e-01
3.96297842e-01 4.88310814e-01 2.04716429e-01 -5.88887632e-01
4.25997563e-02 5.21341324e-01 -8.05154562e-01 1.06955923e-01
1.00163388e+00 3.11100513e-01 -2.71088809e-01 7.44595289e-01
-6.40549809e-02 3.25483620e-01 -1.10106015e+00 -9.59642649e-01
3.72609079e-01 -3.11466217e-01 -7.45637417e-01 -5.25044143e-01
2.30641216e-01 -2.20105007e-01 -7.53277004e-01 1.42820823e+00
7.66214728e-01 1.92614183e-01 5.90025902e-01 -1.61594057e+00
-1.01883578e+00 1.03106081e+00 -1.19831540e-01 2.38661543e-02
3.39566767e-01 6.78358257e-01 7.48123586e-01 -8.35930705e-01
4.15001452e-01 1.68006480e+00 5.35911024e-01 1.02507210e+00
-1.12849367e+00 -1.07756948e+00 1.07559711e-01 7.46380836e-02
-1.55218589e+00 -2.08552495e-01 6.47652149e-01 -9.46472958e-02
6.12047672e-01 8.07818949e-01 -6.56073764e-02 2.07520390e+00
2.02116132e-01 7.42279530e-01 9.41961646e-01 -3.04988235e-01
4.28566843e-01 2.66287923e-01 -2.65105098e-01 3.35053951e-01
3.28673095e-01 2.67452002e-01 -2.08108991e-01 -5.29118955e-01
2.34624684e-01 -1.83000460e-01 -2.73563236e-01 1.02616869e-01
-9.81679022e-01 1.08202374e+00 3.00027311e-01 2.61856019e-01
7.22160414e-02 1.46807954e-01 6.83328629e-01 1.91061169e-01
2.65562415e-01 8.46537650e-01 -3.50209355e-01 4.52573262e-02
-4.60826010e-01 4.93811488e-01 8.36927295e-01 9.95853961e-01
2.71129102e-01 4.48565274e-01 -3.12721789e-01 6.59217298e-01
1.95830271e-01 1.07159221e+00 7.32403338e-01 -4.64546800e-01
7.98103452e-01 -1.17451057e-03 -4.01411876e-02 -1.22495377e+00
-1.10985160e-01 -2.07318500e-01 -1.05691075e+00 -8.31020996e-02
3.63861978e-01 -6.21970713e-01 -7.58084297e-01 2.10909081e+00
3.18425357e-01 5.13751149e-01 4.52188730e-01 6.66983008e-01
6.32268608e-01 6.72400355e-01 3.23750824e-01 2.93585032e-01
1.14305687e+00 -8.33964884e-01 -5.05845249e-01 -5.28609633e-01
6.26205504e-01 -7.76562393e-01 1.67863441e+00 1.19801790e-01
-9.87149775e-01 -2.67958820e-01 -9.27096128e-01 4.77176011e-01
-7.10333049e-01 -3.86369795e-01 2.19845787e-01 1.09260905e+00
-6.70226038e-01 4.13247466e-01 -7.25877643e-01 -1.54431984e-01
2.82482773e-01 1.12902768e-01 -3.12260866e-01 6.76893517e-02
-2.03548074e+00 7.59438932e-01 6.95450127e-01 -3.04858506e-01
-1.05641627e+00 -5.59193015e-01 -9.30575490e-01 3.78496312e-02
4.73924547e-01 -6.21856809e-01 1.15342999e+00 -7.00775564e-01
-1.58225632e+00 6.00621819e-01 3.61067653e-01 -6.15011930e-01
6.77201927e-01 -3.39941740e-01 -5.59141457e-01 1.35858476e-01
-1.73611417e-01 3.62723649e-01 9.67651606e-01 -1.30716598e+00
-3.99951339e-02 -5.71795553e-02 2.24474564e-01 2.07031503e-01
-9.61802840e-01 3.33008587e-01 -1.86916947e-01 -1.29922605e+00
-5.76244533e-01 -1.11748552e+00 -5.13656199e-01 -3.69562447e-01
-1.01775992e+00 -2.28059161e-02 8.18827808e-01 -5.37996709e-01
1.50543427e+00 -2.17734456e+00 2.17053786e-01 4.66212928e-01
-1.98384896e-02 7.69182086e-01 -4.22196865e-01 5.40163159e-01
-1.26825556e-01 8.95051360e-01 -3.29951674e-01 -1.60558403e-01
4.89746809e-01 9.44296420e-02 -6.90683365e-01 3.54197741e-01
-1.91737935e-01 1.09456992e+00 -9.45830107e-01 -2.99698591e-01
1.70030832e-01 2.22119346e-01 -6.73165083e-01 1.81899473e-01
-3.19520414e-01 2.14332610e-01 -6.13529265e-01 3.00974220e-01
5.51250219e-01 1.25716463e-01 -7.70994872e-02 9.26078856e-02
6.10005558e-01 1.87982231e-01 -1.04565763e+00 1.26627016e+00
-5.46156585e-01 4.35112804e-01 -1.63056359e-01 -5.18883944e-01
6.84359670e-01 2.70334154e-01 -2.56715715e-02 -1.86243236e-01
2.53868341e-01 1.48287132e-01 -9.02055278e-02 -1.04669936e-01
3.54634136e-01 -4.37881537e-02 -5.43490052e-01 5.38355708e-01
-1.94164380e-01 -3.44375819e-01 -9.54073295e-02 5.40532589e-01
1.14523757e+00 -5.01848936e-01 2.36696810e-01 -6.36743605e-02
6.75511837e-01 -2.81609923e-01 3.98872972e-01 9.08452272e-01
-3.00115645e-01 4.53853369e-01 3.62403333e-01 4.93953675e-02
-1.05320823e+00 -1.38962853e+00 1.82732761e-01 9.15479064e-01
2.27892116e-01 -7.91982651e-01 -1.29409444e+00 -1.14568686e+00
2.07935367e-02 1.29077923e+00 -4.01011735e-01 -8.48635972e-01
-6.20335698e-01 -5.66574454e-01 1.45537400e+00 3.69805783e-01
3.24286401e-01 -1.34644198e+00 1.12792119e-01 2.63421386e-02
-2.39840254e-01 -1.22522128e+00 -9.74088669e-01 -4.02168304e-01
-2.92044997e-01 -8.62697661e-01 -3.19653600e-01 -5.32171607e-01
6.26378417e-01 -5.65816015e-02 1.00700414e+00 8.41904338e-03
-1.11728281e-01 2.86601663e-01 -5.86497903e-01 -4.48761940e-01
-1.03331959e+00 2.02193528e-01 4.76371288e-01 -2.19172671e-01
2.87769258e-01 -4.40925121e-01 -3.92955124e-01 4.83861208e-01
-1.45908868e+00 -3.47334623e-01 1.92545027e-01 8.33359778e-01
2.69647390e-01 4.09107894e-01 5.82262516e-01 -1.07137096e+00
1.04032004e+00 -6.31167114e-01 -3.83149594e-01 3.09058338e-01
-2.86488384e-01 -4.73417677e-02 1.46607888e+00 -9.94220316e-01
-7.53290474e-01 -5.02351880e-01 -5.14422953e-01 -6.28585696e-01
-3.68501961e-01 2.41189852e-01 -6.72753751e-01 -2.37154099e-03
1.27932823e+00 2.96318799e-01 -3.47614765e-01 -9.66324285e-02
6.55798256e-01 7.32283056e-01 6.27314806e-01 -8.20217907e-01
1.75701308e+00 2.12888777e-01 -3.49043787e-01 -4.95459110e-01
-6.40233397e-01 1.37831673e-01 -1.09395429e-01 1.84996650e-01
6.36203229e-01 -6.62532628e-01 -4.37940627e-01 7.09101379e-01
-9.65427160e-01 -5.04094899e-01 -1.84463546e-01 2.10609451e-01
-3.72593313e-01 6.73300445e-01 -7.75458694e-01 -4.19801444e-01
-7.23645508e-01 -1.14259720e+00 7.59236217e-01 1.37909818e-02
-4.27279800e-01 -1.21582055e+00 -4.67924215e-02 3.48306715e-01
4.98564780e-01 4.95701581e-01 9.23651636e-01 -1.23248696e+00
-1.58433020e-01 -3.49990815e-01 3.25116992e-01 5.49197376e-01
1.36257634e-01 1.65660128e-01 -6.76458955e-01 -5.64824164e-01
-9.23422277e-02 -4.32888836e-01 3.17562044e-01 -6.85464665e-02
1.37684417e+00 -1.00024438e+00 -1.59354806e-01 7.94691980e-01
1.04114819e+00 5.20559140e-02 7.52990961e-01 4.00055796e-01
7.97811866e-01 3.43014032e-01 5.75590849e-01 3.77733767e-01
-2.20275838e-02 6.93375587e-01 4.76520389e-01 9.53212678e-02
1.87314257e-01 -5.49088240e-01 8.43123794e-01 6.72563136e-01
6.80455446e-01 -8.15367460e-01 -1.07845592e+00 3.21881324e-01
-1.45338202e+00 -1.02137363e+00 2.42278919e-01 1.96103668e+00
1.21804976e+00 5.47511280e-01 -1.09616015e-03 2.73861401e-02
8.40150595e-01 2.04535559e-01 -6.34786248e-01 -5.62767029e-01
-3.63407314e-01 4.41083938e-01 6.12466156e-01 6.58306301e-01
-1.17103720e+00 1.71043992e+00 5.35444784e+00 1.52963781e+00
-9.60642219e-01 1.89771786e-01 4.67659414e-01 -2.42055133e-01
-6.39440119e-01 -2.22838014e-01 -7.66963542e-01 7.82182634e-01
9.93267715e-01 -8.08961451e-01 6.78458393e-01 9.18836534e-01
-2.64885109e-02 8.08567643e-01 -6.86844170e-01 6.57471895e-01
6.44914061e-02 -1.11098194e+00 4.66083258e-01 -1.79679036e-01
7.59751201e-01 -1.36105210e-01 5.23277938e-01 5.38213134e-01
9.60372210e-01 -1.32383144e+00 7.21210063e-01 -6.58083037e-02
1.09160113e+00 -1.07983029e+00 5.89256942e-01 5.49241483e-01
-8.23443770e-01 -3.33088962e-03 -3.53155524e-01 5.72259247e-01
8.39420184e-02 1.47050858e-01 -6.64869249e-01 4.41444308e-01
4.94297415e-01 1.92523554e-01 -6.59422874e-01 2.98526734e-01
-6.51980281e-01 9.14249778e-01 -3.14352304e-01 -1.25293776e-01
4.48956728e-01 1.07354708e-01 9.58384871e-01 1.19372714e+00
2.64932036e-01 8.63183737e-02 2.37356037e-01 8.38660121e-01
-4.11100656e-01 2.53268629e-01 -8.36787105e-01 -2.76804268e-01
9.44102108e-01 1.03323126e+00 -2.69570500e-01 -1.72731802e-01
-3.48202176e-02 1.10402739e+00 1.47058278e-01 6.27344429e-01
-1.21337211e+00 -6.59274399e-01 1.08561563e+00 -6.11809678e-02
-1.34436890e-01 -7.54444525e-02 -1.86791703e-01 -1.34681833e+00
-1.52823210e-01 -1.68203855e+00 4.27318126e-01 -5.31965137e-01
-1.65014863e+00 8.65716755e-01 -7.35426992e-02 -9.72923338e-01
-3.06613088e-01 -4.90669280e-01 -8.65890443e-01 9.56341207e-01
-7.13290632e-01 -9.55674946e-01 6.08470142e-02 9.75856721e-01
4.59033459e-01 -6.86056972e-01 1.01396215e+00 5.01144156e-02
-8.51228714e-01 1.54170525e+00 2.50567138e-01 4.91537422e-01
8.97291958e-01 -1.17206705e+00 1.09249473e+00 1.20819581e+00
2.06783339e-01 7.69315541e-01 7.84176350e-01 -8.08886170e-01
-1.06945169e+00 -1.50732434e+00 4.39352483e-01 -7.31485188e-01
9.94730532e-01 -7.77292669e-01 -9.09787476e-01 6.52268052e-01
1.09788589e-03 -2.23498568e-01 7.20719039e-01 -3.56397748e-01
-6.58749342e-01 2.30239674e-01 -1.43015075e+00 1.46994340e+00
9.63724911e-01 -7.13398635e-01 -3.21342319e-01 5.43907940e-01
1.21314359e+00 -6.78261697e-01 -6.14084423e-01 3.57975751e-01
-4.47562039e-02 -6.24139667e-01 1.24594295e+00 -1.10259998e+00
3.10989052e-01 -1.07422836e-01 -2.59712368e-01 -1.60594499e+00
-2.00681567e-01 -1.08948481e+00 -3.71623456e-01 1.22654343e+00
4.07397956e-01 -8.95378470e-01 6.96528554e-01 5.02993286e-01
5.72329313e-02 -6.71936154e-01 -8.37699294e-01 -1.02148724e+00
7.31477022e-01 -4.52323645e-01 9.03256595e-01 1.12325776e+00
-2.24210545e-01 1.67685524e-01 -3.47430378e-01 5.10335088e-01
5.71653366e-01 -4.96197999e-01 9.63507414e-01 -5.85838974e-01
-4.05689538e-01 -5.38234353e-01 -2.59472430e-01 -5.39027214e-01
8.33183408e-01 -1.03371465e+00 -1.32088363e-01 -8.82415533e-01
-1.85988620e-01 -4.57000285e-01 -2.17560753e-01 5.17163336e-01
-7.80621648e-01 1.31424919e-01 3.76051247e-01 8.47634394e-03
-3.13774526e-01 6.79529428e-01 9.09657896e-01 -2.21465379e-01
-2.76524257e-02 -3.93118449e-02 -9.35659945e-01 8.65072548e-01
1.40583873e+00 -6.42109513e-01 -5.69142163e-01 -3.16311598e-01
1.05909675e-01 -3.85357976e-01 3.15079749e-01 -8.26991439e-01
1.48494057e-02 -5.16577899e-01 -1.37876580e-02 -6.63878163e-03
-2.17543524e-02 -6.14709258e-01 -2.53953282e-02 5.90857804e-01
-5.41102707e-01 1.76475912e-01 4.23481494e-01 5.11577785e-01
1.75307151e-02 -2.80065984e-01 9.79408741e-01 5.19796833e-02
-1.50655344e-01 5.84709883e-01 -2.80274600e-01 6.85939431e-01
1.13368344e+00 2.51160264e-01 -5.77761650e-01 -5.27071178e-01
-4.98576939e-01 2.64598936e-01 5.60599029e-01 6.71391368e-01
6.44965410e-01 -1.44878244e+00 -1.01642179e+00 6.11710697e-02
-9.67946425e-02 -1.87407836e-01 2.40719542e-01 4.60158400e-02
-5.76422036e-01 2.95537710e-01 5.17444871e-02 -6.92952937e-03
-1.35453892e+00 8.82348180e-01 3.40577543e-01 -5.89793563e-01
-5.23113012e-01 9.75721776e-01 3.77668828e-01 -7.09640026e-01
2.57733047e-01 2.64771163e-01 1.69573784e-01 -5.46536684e-01
5.11482239e-01 2.89697021e-01 -3.70758504e-01 -6.64178312e-01
-4.07741815e-01 1.33263499e-01 -1.95096314e-01 -2.54832834e-01
7.02904940e-01 2.38706023e-01 9.65035707e-02 -2.00750157e-01
1.13600743e+00 5.13376415e-01 -7.93735027e-01 -3.96740824e-01
-1.75963655e-01 -5.38695335e-01 -5.16993284e-01 -6.72449231e-01
-9.53126311e-01 6.53450370e-01 1.64397478e-01 1.99869692e-01
1.03353512e+00 -2.04337329e-01 1.21464503e+00 5.01706958e-01
2.81877160e-01 -8.79438162e-01 3.07850361e-01 6.20525599e-01
1.01717079e+00 -7.69579411e-01 -2.67622739e-01 -3.75201344e-01
-8.90909970e-01 5.58157384e-01 9.66531634e-01 -9.50802639e-02
4.52442855e-01 3.60554129e-01 2.24147752e-01 3.26303929e-01
-6.86693728e-01 2.70851880e-01 2.43886441e-01 6.39258862e-01
-8.02697241e-02 2.74398088e-01 7.75295601e-04 8.77244234e-01
-8.20983231e-01 -9.16850805e-01 5.37326634e-01 5.45414090e-01
-2.73117814e-02 -1.29538155e+00 -6.93364084e-01 1.02521233e-01
-8.67622316e-01 -4.26181436e-01 -6.99300647e-01 6.69575393e-01
-1.34588674e-01 1.19048381e+00 -5.58853447e-01 -7.44455993e-01
4.36876178e-01 2.46118545e-03 2.30335654e-03 -7.23144293e-01
-8.38953733e-01 -4.40083951e-01 -8.72291997e-02 -5.12238085e-01
5.08864939e-01 -3.49414557e-01 -1.18250895e+00 -8.16858709e-01
-1.38800204e-01 3.30226809e-01 4.43360776e-01 6.41881526e-01
1.85454413e-01 3.51786613e-01 9.64031398e-01 -3.24242175e-01
-1.17142284e+00 -8.64675999e-01 -2.61053294e-01 9.34683084e-01
-9.61721092e-02 -1.80401355e-01 -7.24661171e-01 -2.50932932e-01] | [6.005535125732422, 8.089773178100586] |
8032c8a0-8d71-433d-9dfd-9ef9078a2f06 | multilinear-wavelets-a-statistical-shape | 1401.2818 | null | http://arxiv.org/abs/1401.2818v2 | http://arxiv.org/pdf/1401.2818v2.pdf | Multilinear Wavelets: A Statistical Shape Space for Human Faces | We present a statistical model for $3$D human faces in varying expression,
which decomposes the surface of the face using a wavelet transform, and learns
many localized, decorrelated multilinear models on the resulting coefficients.
Using this model we are able to reconstruct faces from noisy and occluded $3$D
face scans, and facial motion sequences. Accurate reconstruction of face shape
is important for applications such as tele-presence and gaming. The localized
and multi-scale nature of our model allows for recovery of fine-scale detail
while retaining robustness to severe noise and occlusion, and is
computationally efficient and scalable. We validate these properties
experimentally on challenging data in the form of static scans and motion
sequences. We show that in comparison to a global multilinear model, our model
better preserves fine detail and is computationally faster, while in comparison
to a localized PCA model, our model better handles variation in expression, is
faster, and allows us to fix identity parameters for a given subject. | ['Timo Bolkart', 'Stefanie Wuhrer', 'Alan Brunton'] | 2014-01-13 | null | null | null | null | ['3d-face-modeling'] | ['computer-vision'] | [-1.51613001e-02 -9.05505046e-02 -1.52327148e-02 -5.78240275e-01
-8.22141230e-01 -5.74110985e-01 3.08087915e-01 -5.72583139e-01
-4.94802147e-02 3.95529807e-01 2.63904065e-01 3.34376961e-01
-2.84960240e-01 -4.45300132e-01 -5.62091947e-01 -7.25157619e-01
-3.23541850e-01 3.96945029e-01 -2.23487809e-01 7.16507658e-02
-1.53963596e-01 8.59300375e-01 -1.58528113e+00 2.71411031e-01
-7.71864736e-03 8.94913971e-01 -1.62185967e-01 5.00190318e-01
2.08279893e-01 4.37784940e-01 -1.68501705e-01 -3.58990490e-01
3.95560890e-01 -3.44743460e-01 -4.23850268e-01 4.21229243e-01
1.10044420e+00 -3.25026542e-01 -3.04817557e-01 8.72820437e-01
5.36698461e-01 -1.42018110e-01 5.39298475e-01 -1.03290629e+00
-3.78769308e-01 -2.62021661e-01 -1.10597813e+00 -1.23328380e-01
5.55620313e-01 -2.94886231e-01 8.41338456e-01 -1.10008919e+00
1.00896156e+00 1.79666805e+00 1.22158575e+00 6.63239300e-01
-1.84235275e+00 -7.50639379e-01 -2.73332605e-03 -2.74447262e-01
-1.63083124e+00 -1.13406658e+00 9.03048813e-01 -4.04735595e-01
4.14805472e-01 2.44500414e-01 6.52666867e-01 9.87633348e-01
2.47287586e-01 4.98887688e-01 1.21225989e+00 -1.80999741e-01
4.95242560e-03 -2.58173317e-01 -1.76164731e-01 1.29390001e+00
3.54658850e-02 -7.29124025e-02 -9.04234767e-01 -6.44338131e-01
1.10902345e+00 -1.14721708e-01 -2.50929266e-01 -7.67082810e-01
-8.13656628e-01 5.92012465e-01 -7.81085938e-02 1.98038697e-01
-4.32290584e-01 3.27062100e-01 -4.24075797e-02 2.04319999e-01
6.15908563e-01 7.60902476e-04 -4.65039521e-01 -2.27522314e-01
-1.20829213e+00 2.95793772e-01 9.04695570e-01 6.35854840e-01
8.21383595e-01 2.12804615e-01 3.20126981e-01 1.10051095e+00
3.32892120e-01 9.29038465e-01 1.95595652e-01 -1.69757688e+00
-1.42401725e-01 5.09346612e-02 -1.14225954e-01 -1.34649253e+00
-6.02205694e-01 -5.66451484e-03 -7.77203500e-01 3.83846432e-01
3.37043732e-01 3.39642689e-02 -7.22412407e-01 2.17648244e+00
3.63691211e-01 6.05570853e-01 -3.63948971e-01 7.11137950e-01
7.20149994e-01 1.67922914e-01 -1.72043592e-01 -6.81213140e-01
1.41653991e+00 -7.53303915e-02 -7.57815480e-01 -7.23066553e-02
-4.05141823e-02 -8.86589766e-01 7.55618393e-01 4.80051041e-01
-1.38327837e+00 -4.67839479e-01 -7.00357318e-01 -2.25187957e-01
3.36764067e-01 7.21600577e-02 6.47313833e-01 7.25881696e-01
-1.45582688e+00 6.44593477e-01 -1.05423641e+00 -3.59788835e-01
5.22413194e-01 6.19795799e-01 -9.32625175e-01 -4.03049558e-01
-4.16557193e-01 6.71205640e-01 -7.88117766e-01 1.07164904e-01
-6.42246366e-01 -8.55846226e-01 -1.01067257e+00 -1.91677928e-01
-2.43206412e-01 -6.23488843e-01 9.88942623e-01 -8.73514652e-01
-1.46600449e+00 1.14186490e+00 -9.39655542e-01 1.70264572e-01
2.50648499e-01 3.26284133e-02 -2.69729346e-01 3.54582012e-01
-1.90654695e-02 6.49482906e-01 1.37191355e+00 -1.37039578e+00
1.80989638e-01 -9.58774567e-01 -4.59692419e-01 -4.79569472e-02
-1.71253189e-01 1.73421532e-01 -6.09814048e-01 -6.82446778e-01
6.19281948e-01 -1.10757399e+00 -1.86493881e-02 6.09262884e-01
2.39600703e-01 3.09692562e-01 1.05195379e+00 -9.88207042e-01
8.19029212e-01 -2.24768496e+00 3.85034412e-01 4.52713430e-01
3.74192655e-01 -2.85810471e-01 -4.23880398e-01 -1.50735423e-01
-1.90355882e-01 3.37885432e-02 -2.21861169e-01 -9.18049634e-01
-1.63789302e-01 3.80955577e-01 5.56528233e-02 8.68805110e-01
2.01680407e-01 7.65862048e-01 -4.77267146e-01 -4.40694183e-01
-1.45320103e-01 1.12357700e+00 -8.63304257e-01 -6.02387413e-02
1.29696727e-01 5.24131835e-01 -3.39305609e-01 1.05571616e+00
8.22682023e-01 -2.45978702e-02 3.96026224e-01 -3.76084089e-01
2.94443607e-01 -4.26137418e-01 -1.36860371e+00 1.71790564e+00
-5.65593004e-01 7.25666404e-01 1.10915422e+00 -7.99242437e-01
8.69342506e-01 4.99798864e-01 9.56527889e-01 -5.38293719e-01
1.39731154e-01 9.98339355e-02 -4.55599666e-01 -4.13498700e-01
4.35415283e-02 -4.95362997e-01 2.63641268e-01 4.82528269e-01
1.50261611e-01 -3.22947800e-01 -3.82478446e-01 -1.40226707e-01
1.20783389e+00 2.77715564e-01 7.42041841e-02 -4.45155650e-01
3.20252061e-01 -7.26264834e-01 6.33848906e-01 -5.04145026e-02
-1.43779516e-01 7.85630226e-01 4.89289045e-01 -5.50209939e-01
-8.28736305e-01 -1.05499184e+00 -3.89725208e-01 8.16820145e-01
-2.64323771e-01 -3.54222685e-01 -7.47850478e-01 -2.05885649e-01
3.99015188e-01 -2.87949685e-02 -6.94022775e-01 2.19424024e-01
-6.90684378e-01 -7.17756510e-01 3.98643702e-01 2.10683376e-01
2.69511968e-01 -6.82719171e-01 -1.45628616e-01 7.58443028e-02
-3.20490986e-01 -1.31767595e+00 -5.80798447e-01 -3.17968816e-01
-1.00752938e+00 -1.14882767e+00 -4.85109538e-01 -7.71228671e-01
8.15737844e-01 2.84509242e-01 1.23204148e+00 1.44785980e-03
-7.04382002e-01 9.49423134e-01 1.39571354e-01 -1.51398778e-02
-6.42602891e-02 -7.05376267e-01 4.12922800e-01 4.45805937e-01
1.85096532e-01 -9.98000443e-01 -3.84355605e-01 4.72516179e-01
-6.89871669e-01 -5.67060828e-01 2.80292958e-01 8.63165200e-01
8.35974514e-01 -9.39264800e-03 1.04875602e-01 -7.21810043e-01
4.10097301e-01 -2.85247594e-01 -4.90920395e-01 4.66125160e-02
-2.29302272e-01 1.76142186e-01 9.01808441e-02 -4.29627717e-01
-9.43318069e-01 3.92124355e-01 -1.47700563e-01 -7.68722773e-01
3.83656844e-02 8.02583620e-02 -1.90677077e-01 -7.69073248e-01
7.14044809e-01 -2.61332951e-02 5.63708901e-01 -7.76349127e-01
2.63927758e-01 1.30068302e-01 5.38869739e-01 -8.41929913e-01
8.42839241e-01 9.56602931e-01 4.46437657e-01 -1.50393689e+00
-2.07705811e-01 -2.39504203e-01 -8.97034764e-01 -1.69860661e-01
6.61538363e-01 -9.50394988e-01 -9.34971750e-01 2.86203116e-01
-1.04047096e+00 -1.53307766e-01 -1.51921570e-01 2.30872363e-01
-7.62629807e-01 5.04505157e-01 -5.85362017e-01 -8.85125756e-01
6.73798025e-02 -9.69338655e-01 1.49055588e+00 -2.55391359e-01
-4.92073953e-01 -9.52529490e-01 6.97124153e-02 3.17808330e-01
1.57614216e-01 5.26606619e-01 7.92151451e-01 4.16391551e-01
-3.89281750e-01 -2.27760583e-01 8.46352503e-02 2.12929949e-01
2.92527378e-01 1.46889329e-01 -1.10967326e+00 -4.98022318e-01
2.87761897e-01 -3.30554008e-01 5.86810589e-01 6.41687572e-01
1.14382589e+00 -3.34425867e-01 -1.95198789e-01 9.35990214e-01
1.09336913e+00 -2.00453535e-01 5.20165145e-01 -3.34200829e-01
5.16764224e-01 8.87152433e-01 1.40729606e-01 3.99733990e-01
3.51669312e-01 7.99081802e-01 7.66870752e-02 -7.77449459e-02
-2.72167534e-01 -5.88721968e-02 4.58524376e-01 6.06137931e-01
-4.35687184e-01 4.02221680e-01 -6.32530749e-01 4.21396255e-01
-1.50128913e+00 -1.18075943e+00 2.37371042e-01 2.06614113e+00
6.64392531e-01 -5.59848011e-01 2.01918989e-01 9.87398997e-02
2.92482346e-01 2.23407760e-01 -4.46369648e-01 -2.28979260e-01
-3.07033539e-01 8.25285673e-01 2.27790967e-01 7.76947320e-01
-8.89641106e-01 7.03401744e-01 7.89042330e+00 5.53867280e-01
-9.82311904e-01 2.64674306e-01 6.30467474e-01 -4.76684183e-01
-4.57688153e-01 -4.12943333e-01 -5.46162844e-01 7.11899772e-02
7.67455041e-01 2.16452658e-01 7.58427322e-01 6.27376318e-01
3.56365800e-01 1.99119091e-01 -1.02660477e+00 1.33807778e+00
3.82915705e-01 -1.01716053e+00 -2.75517672e-01 2.25631833e-01
6.36653543e-01 -2.09408373e-01 3.34782928e-01 -2.70865887e-01
2.03470588e-01 -1.27920032e+00 5.23358047e-01 5.49983621e-01
9.65641141e-01 -6.98446572e-01 1.34630233e-01 1.17686935e-01
-1.30547965e+00 8.40601772e-02 -2.43972957e-01 2.65625089e-01
1.20410766e-03 4.95424271e-01 -2.05735520e-01 8.75374600e-02
8.37290764e-01 6.40040874e-01 -2.25978836e-01 3.39922041e-01
8.33082348e-02 2.41558805e-01 -5.23010314e-01 5.59175134e-01
-4.27383006e-01 -4.26541209e-01 3.73790294e-01 1.04895139e+00
5.76748788e-01 6.22655451e-01 5.88752143e-02 4.51857895e-01
-1.00172520e-01 9.94309485e-02 -6.93203270e-01 1.45535067e-01
1.32712632e-01 1.28500807e+00 -3.46321315e-01 1.08781658e-01
-5.01327991e-01 1.07296646e+00 2.94941425e-01 5.11349976e-01
-3.60316753e-01 3.33400398e-01 1.29169118e+00 4.22291070e-01
3.92428815e-01 -6.48269236e-01 -1.67319417e-01 -1.24226797e+00
1.73305064e-01 -9.95066941e-01 1.74945697e-01 -6.67453110e-01
-1.27097690e+00 6.15936458e-01 -1.49920434e-01 -9.06658471e-01
-3.22543442e-01 -6.40571296e-01 -2.15253666e-01 7.66201079e-01
-1.29544604e+00 -1.26586628e+00 -2.68182456e-01 9.42549646e-01
1.94534168e-01 3.75102460e-02 1.13942492e+00 3.30471396e-01
-3.64118904e-01 6.22271836e-01 -1.16968162e-01 -1.69801176e-01
7.74475276e-01 -7.57835209e-01 1.47204831e-01 5.53647041e-01
4.14141834e-01 9.28852737e-01 6.37612045e-01 -4.83546734e-01
-1.94445443e+00 -6.63367033e-01 6.85635149e-01 -6.34726286e-01
3.87712449e-01 -3.99205208e-01 -7.79850423e-01 7.64760792e-01
-2.59384185e-01 5.55003226e-01 9.74853992e-01 3.08725506e-01
-7.11386085e-01 -2.76554316e-01 -1.58786500e+00 2.62290150e-01
1.28556335e+00 -7.65695572e-01 -1.72925651e-01 1.93276465e-01
4.48994013e-03 -3.68433505e-01 -1.07125998e+00 2.38545150e-01
1.26376891e+00 -1.03520119e+00 1.31680155e+00 -5.28913617e-01
9.60959569e-02 -1.26496524e-01 -4.88008946e-01 -1.11592543e+00
-4.56433862e-01 -7.93415010e-01 -2.78200656e-01 8.38549316e-01
1.84072986e-01 -4.26749766e-01 1.05423582e+00 8.67156208e-01
4.07621950e-01 -5.61744750e-01 -1.32154405e+00 -6.37542069e-01
-1.83633506e-01 -3.61090153e-01 3.38306159e-01 1.08787048e+00
-2.79808164e-01 -1.44580305e-01 -6.28791630e-01 2.83148736e-01
9.80628908e-01 1.70085117e-01 8.41497183e-01 -1.46570361e+00
-3.42677474e-01 -2.33032107e-01 -6.10044181e-01 -9.17699277e-01
7.18194664e-01 -7.32884407e-01 -1.30470321e-01 -8.42724741e-01
1.37224063e-01 -2.48449087e-01 1.26952291e-01 6.05310738e-01
2.54011393e-01 8.81215572e-01 4.08591367e-02 1.18905582e-01
-2.51157265e-02 4.32358205e-01 1.06584704e+00 -2.11096499e-02
-8.05957466e-02 -1.52166605e-01 -6.97416961e-01 1.10608721e+00
1.65545151e-01 -2.31239527e-01 -2.60213763e-01 -4.36389774e-01
-1.50475398e-01 3.53031188e-01 3.66775900e-01 -6.25654817e-01
-3.46892662e-02 -6.52797818e-02 8.81288588e-01 -1.46267354e-03
9.81353939e-01 -9.86272752e-01 5.75720251e-01 1.72932908e-01
1.31941140e-01 7.49818161e-02 3.27175677e-01 5.10134876e-01
-1.98891386e-01 3.59826654e-01 1.10077846e+00 -1.34638771e-01
-3.32304806e-01 6.53170466e-01 2.47009639e-02 -1.89990669e-01
7.20804453e-01 -2.70364285e-01 2.07326487e-01 -7.36139774e-01
-1.09363639e+00 -1.76050082e-01 6.58890069e-01 2.78307021e-01
6.40523493e-01 -1.54705501e+00 -6.81652367e-01 8.33500862e-01
-2.19161943e-01 -3.76925498e-01 3.18729341e-01 6.91966653e-01
-4.55511719e-01 3.24984491e-02 -2.29121923e-01 -6.68380678e-01
-1.89281309e+00 9.36964005e-02 3.27595830e-01 2.68386602e-01
-3.54717046e-01 9.53306377e-01 1.41468436e-01 -3.26305687e-01
-4.02380750e-02 -8.69419202e-02 9.83885750e-02 1.53271362e-01
6.86689973e-01 3.16449732e-01 1.50243808e-02 -1.23215365e+00
-5.29006541e-01 1.51739037e+00 4.23238367e-01 -3.17792058e-01
1.55964017e+00 -2.67253313e-02 -5.16287208e-01 1.97339937e-01
1.53432441e+00 4.67086911e-01 -1.45769811e+00 -2.24218234e-01
-3.81954670e-01 -8.34074259e-01 3.85244228e-02 -2.00103655e-01
-1.40359163e+00 6.71909094e-01 4.87480491e-01 -2.42261752e-01
1.23692000e+00 3.81011516e-02 4.02354598e-01 1.68635175e-01
5.19505918e-01 -5.86556971e-01 1.46373957e-01 2.32948259e-01
1.11989665e+00 -9.02593613e-01 2.69491762e-01 -7.59868145e-01
-3.11864525e-01 1.11071432e+00 1.97361261e-01 -1.09548658e-01
1.10278952e+00 6.13936901e-01 1.57391727e-01 -4.18414921e-01
-5.08891940e-01 1.52639389e-01 3.72396350e-01 8.14503670e-01
4.21445161e-01 -7.72379786e-02 1.18541852e-01 3.17352146e-01
-3.56145144e-01 -1.91421136e-01 3.63215320e-02 8.02772880e-01
-1.52466461e-01 -1.21882892e+00 -6.48892999e-01 2.81290650e-01
-4.92128551e-01 2.92170316e-01 -3.05642128e-01 6.53469980e-01
1.02525190e-01 7.83926070e-01 8.42624307e-02 -2.24622369e-01
4.58197922e-01 3.72705966e-01 1.00859129e+00 -4.50042844e-01
-6.26976565e-02 5.46464920e-01 -6.71648234e-03 -1.21326649e+00
-5.52267611e-01 -1.15925670e+00 -9.33871627e-01 -5.25098264e-01
1.99273109e-01 -1.90019891e-01 8.00542355e-01 5.08060038e-01
5.22948384e-01 -4.75481525e-02 7.95692801e-01 -1.16102123e+00
-1.98810712e-01 -5.88969946e-01 -9.14594769e-01 5.28201520e-01
6.79069221e-01 -9.29120839e-01 -4.99560714e-01 3.66313279e-01] | [13.163264274597168, -0.015258114784955978] |
a174bc10-6df9-4443-a19e-315cd43b54ec | an-ensemble-approach-for-facial-expression | 2203.12891 | null | https://arxiv.org/abs/2203.12891v1 | https://arxiv.org/pdf/2203.12891v1.pdf | An Ensemble Approach for Facial Expression Analysis in Video | Human emotions recognization contributes to the development of human-computer interaction. The machines understanding human emotions in the real world will significantly contribute to life in the future. This paper will introduce the Affective Behavior Analysis in-the-wild (ABAW3) 2022 challenge. The paper focuses on solving the problem of the valence-arousal estimation and action unit detection. For valence-arousal estimation, we conducted two stages: creating new features from multimodel and temporal learning to predict valence-arousal. First, we make new features; the Gated Recurrent Unit (GRU) and Transformer are combined using a Regular Networks (RegNet) feature, which is extracted from the image. The next step is the GRU combined with Local Attention to predict valence-arousal. The Concordance Correlation Coefficient (CCC) was used to evaluate the model. | ['Soo-Hyung Kim', 'Van-Thong Huynh', 'Hong-Hai Nguyen'] | 2022-03-24 | null | null | null | null | ['action-unit-detection'] | ['computer-vision'] | [ 2.22160682e-01 -8.19485486e-02 8.12132508e-02 -7.05113530e-01
-3.94346446e-01 -1.98384225e-01 4.20894653e-01 1.17287494e-01
-5.35710931e-01 7.77097225e-01 2.82786995e-01 5.01570761e-01
3.23626906e-01 -3.48266959e-01 3.76460738e-02 -7.69449949e-01
-3.84010732e-01 -2.07843900e-01 -2.55590796e-01 -2.99950600e-01
1.56876579e-01 2.99275458e-01 -1.82113898e+00 8.27108264e-01
3.69955391e-01 1.64832306e+00 -1.93483040e-01 8.79372120e-01
9.23289824e-03 9.70846832e-01 -3.42661887e-01 -5.70630394e-02
5.37745580e-02 -8.54062855e-01 -7.15750158e-01 -2.55658418e-01
-4.79819924e-01 1.82605013e-01 4.61507469e-01 8.15012693e-01
6.15352988e-01 5.99844396e-01 5.65717041e-01 -1.43548095e+00
-5.37591800e-02 9.69302207e-02 -5.41048050e-01 2.84350634e-01
4.17000234e-01 3.85929085e-02 7.84380496e-01 -9.25409555e-01
4.08390254e-01 1.12957489e+00 5.76568305e-01 8.63016903e-01
-7.45858312e-01 -4.89417285e-01 2.49183625e-02 6.19458199e-01
-1.08424568e+00 -1.60018608e-01 8.86593759e-01 -2.48872057e-01
1.36092222e+00 3.99539798e-01 1.13209760e+00 1.13178992e+00
3.88831049e-01 7.18918204e-01 1.34558380e+00 -3.50856632e-01
2.12523997e-01 3.12709630e-01 4.44817185e-01 5.42042971e-01
-4.80577797e-01 8.12146463e-04 -6.20240629e-01 2.55318452e-02
1.77679300e-01 -1.88175365e-01 2.51952887e-01 3.45789522e-01
-5.88907838e-01 7.80911267e-01 3.27664405e-01 5.04589736e-01
-9.76640701e-01 8.08359459e-02 9.79839563e-01 2.68409967e-01
5.36398530e-01 5.49619496e-01 -4.02022213e-01 -5.95733523e-01
-6.35376453e-01 -9.71838683e-02 4.86714929e-01 1.83057383e-01
7.02745199e-01 7.89936706e-02 -2.23896936e-01 9.70619857e-01
1.26806304e-01 1.18657321e-01 9.69184399e-01 -6.84997022e-01
-3.56507987e-01 8.11727762e-01 -1.09455876e-01 -9.88447249e-01
-8.55388701e-01 1.70984969e-01 -9.20822442e-01 9.94703695e-02
-2.68652827e-01 -5.20640492e-01 -7.44869649e-01 1.72385085e+00
3.49419385e-01 9.20882300e-02 1.32749453e-01 9.15431023e-01
9.65014040e-01 1.18747520e+00 4.83962566e-01 -8.84617448e-01
1.39583874e+00 -7.75327206e-01 -1.22102916e+00 -1.55239716e-01
6.16080523e-01 -5.12829065e-01 8.32861543e-01 6.69460177e-01
-9.30885077e-01 -6.58027470e-01 -1.09957683e+00 1.73041880e-01
-7.39848971e-01 4.28374223e-02 8.58369946e-01 5.69819391e-01
-9.26875889e-01 3.68385643e-01 -7.07946479e-01 -6.23214185e-01
1.80326954e-01 4.71185476e-01 -2.86258459e-01 8.04590642e-01
-1.64336956e+00 1.15757990e+00 4.71953750e-01 4.53122169e-01
-7.64271796e-01 -7.59832114e-02 -7.96422362e-01 -3.94541137e-02
-2.24055767e-01 -1.40756473e-01 1.02506936e+00 -1.85794985e+00
-1.78616571e+00 1.02598548e+00 -1.54446587e-01 -3.73030245e-01
-9.52545032e-02 -2.27953106e-01 -5.93197465e-01 -1.07296675e-01
-4.71340895e-01 6.07169449e-01 5.20213246e-01 -9.21330631e-01
-5.69318950e-01 -4.97045189e-01 -4.92327571e-01 4.58236963e-01
-4.69583213e-01 3.37814182e-01 -2.35976860e-01 -2.68920004e-01
-2.04125732e-01 -7.78257072e-01 -3.52778673e-01 -7.60381341e-01
2.10506395e-02 -6.08301103e-01 5.03502369e-01 -7.89840817e-01
1.52360523e+00 -2.17445564e+00 1.21205814e-01 6.33068562e-01
-9.84174982e-02 2.71019071e-01 -2.03711569e-01 1.96582019e-01
-4.69158411e-01 -2.49145687e-01 2.22079962e-01 -2.42114499e-01
-1.76575199e-01 -1.32526442e-01 -1.00726314e-01 2.04666957e-01
2.85498142e-01 9.17555928e-01 -6.06953084e-01 -5.45098484e-01
2.13496342e-01 4.22707438e-01 -3.67105991e-01 6.83299720e-01
6.22428060e-02 2.17463255e-01 -2.70800710e-01 4.47222948e-01
2.36494973e-01 3.05310726e-01 2.23714933e-01 -3.46576631e-01
-2.70401597e-01 -1.08297735e-01 -9.36550915e-01 1.24883318e+00
-5.44781327e-01 6.62949860e-01 -1.87276810e-01 -8.24008226e-01
1.37885654e+00 4.83653873e-01 7.21360922e-01 -1.17018008e+00
7.75617063e-01 -8.77508000e-02 -1.04223922e-01 -8.20657849e-01
4.27937120e-01 -2.95637578e-01 -2.90869623e-01 3.15555632e-01
1.24514692e-01 1.69047460e-01 8.14630687e-02 -1.11876063e-01
8.61983716e-01 1.64773628e-01 7.00323343e-01 -3.94224338e-02
6.79212391e-01 -3.06471437e-01 8.86383235e-01 9.66746137e-02
-6.11670315e-01 1.36192709e-01 6.97199583e-01 -6.43171728e-01
-6.19309902e-01 -7.49411106e-01 2.93066621e-01 1.25643241e+00
-1.29826907e-02 -4.67338830e-01 -7.83421338e-01 -4.06339437e-01
-6.07996464e-01 6.32773638e-01 -1.09559965e+00 -6.73901379e-01
-1.64940462e-01 -1.10133898e+00 2.78865099e-01 6.59623265e-01
4.56338286e-01 -1.72262275e+00 -1.10299003e+00 7.19788671e-02
-2.37571537e-01 -7.80534923e-01 4.78134863e-02 5.57840466e-01
-5.08718729e-01 -6.23037934e-01 -1.52546957e-01 -6.50305212e-01
2.99577564e-01 -5.18835127e-01 8.53508413e-01 -2.80367076e-01
-6.57976449e-01 3.53561759e-01 -3.79590094e-01 -6.20326400e-01
1.97059810e-02 -3.28589648e-01 1.44605696e-01 3.61412227e-01
7.77241349e-01 -5.20351231e-01 -4.72806275e-01 -4.84898724e-02
-4.11530465e-01 4.31648120e-02 3.91110092e-01 6.63857281e-01
6.83825612e-01 -2.41571203e-01 8.65058780e-01 -7.12311089e-01
1.03529608e+00 -5.42000115e-01 -1.62362173e-01 3.13074201e-01
-6.95423484e-01 -3.03558201e-01 5.66043079e-01 -5.45827746e-01
-1.37455297e+00 2.84806520e-01 -2.68207759e-01 -2.22811893e-01
-2.15318706e-02 5.16695321e-01 -6.96674660e-02 3.81468415e-01
4.95593905e-01 -8.32583159e-02 -3.41399521e-01 9.80029106e-02
2.27977037e-01 8.42860341e-01 3.68769348e-01 -1.41506314e-01
-2.24954069e-01 -7.15935305e-02 -1.03548527e-01 -8.16095471e-01
-7.48487115e-01 -5.05195320e-01 -4.47183937e-01 -9.28355753e-01
1.34803963e+00 -7.54550755e-01 -1.28381586e+00 3.66030604e-01
-1.23455119e+00 -2.47067019e-01 -1.27935693e-01 6.38384700e-01
-7.41212785e-01 -1.08966328e-01 -6.64703369e-01 -1.52746713e+00
-9.07636523e-01 -6.74974620e-01 5.71428478e-01 6.67335272e-01
-6.81044340e-01 -6.59413218e-01 6.10494316e-01 8.46739784e-02
2.73669034e-01 5.40005147e-01 6.76670790e-01 -7.37743974e-01
5.33550799e-01 -2.61344552e-01 4.55744751e-02 7.37656057e-01
-2.25194871e-01 7.88075849e-02 -1.19758534e+00 3.20754260e-01
2.02553570e-01 -7.77152777e-01 8.22578669e-01 4.25598592e-01
1.29186571e+00 2.31587954e-04 2.35514835e-01 4.23898578e-01
1.21124196e+00 7.68736064e-01 9.66127396e-01 3.96259241e-02
4.27380055e-01 7.91237831e-01 9.98132408e-01 8.57798755e-01
1.31859690e-01 1.98617861e-01 3.96454543e-01 -1.59759521e-01
4.42722082e-01 4.70493287e-02 7.27905273e-01 1.04733276e+00
-4.84084576e-01 -1.94821209e-02 -8.28840673e-01 3.94467294e-01
-2.07651472e+00 -1.10014737e+00 -5.08866347e-02 1.95083249e+00
6.92063451e-01 1.05264513e-02 2.04970419e-01 -3.69658018e-03
6.95798397e-01 6.39295056e-02 -3.72981787e-01 -1.39621413e+00
-7.92882219e-02 3.82680595e-01 -2.16484070e-01 4.87522155e-01
-1.00938940e+00 9.52600896e-01 6.13820553e+00 6.02764308e-01
-1.22536695e+00 9.28085670e-03 1.08379328e+00 -3.12146038e-01
1.10810399e-01 -4.29940313e-01 -4.30736870e-01 2.35880986e-01
1.36784995e+00 -1.29715562e-01 6.45094991e-01 8.58329892e-01
4.75788057e-01 -3.86425406e-01 -7.49784648e-01 1.39377415e+00
4.62460488e-01 -4.78176266e-01 -2.82479525e-01 -4.71592754e-01
5.61706662e-01 -3.12909514e-01 -7.23421350e-02 7.00939059e-01
-1.43672466e-01 -9.72057223e-01 6.02699965e-02 1.05030024e+00
5.28601646e-01 -1.32684600e+00 9.53365862e-01 5.40015958e-02
-1.12294424e+00 -1.63925201e-01 -2.81378090e-01 -2.85825431e-01
-3.82463150e-02 3.58722359e-01 -6.28565729e-01 4.69965115e-02
8.23308885e-01 6.80100143e-01 -4.64607239e-01 4.16208148e-01
-9.05179605e-02 4.26063031e-01 -5.23742288e-02 -6.68572962e-01
1.37320712e-01 -3.66144478e-01 1.72548845e-01 1.47447896e+00
1.24661267e-01 4.27561313e-01 -1.06752805e-01 5.86652517e-01
2.31893566e-02 5.82635224e-01 -3.94726127e-01 -3.07978958e-01
-5.71783371e-02 1.98186457e+00 -7.68422544e-01 -3.76411885e-01
-8.57784674e-02 1.22389877e+00 2.55330294e-01 2.12376826e-02
-1.10894525e+00 -5.22550225e-01 4.18802172e-01 -4.35834229e-01
-3.04917246e-01 4.28892761e-01 -2.89271414e-01 -8.46552849e-01
-2.81179696e-01 -6.71720505e-01 5.65367222e-01 -9.77513731e-01
-1.16014695e+00 9.13868248e-01 -2.25488216e-01 -9.26652551e-01
-4.22417641e-01 -5.43671191e-01 -9.74980950e-01 6.67133152e-01
-8.92081261e-01 -9.49012876e-01 -5.01114428e-01 6.45294666e-01
5.89457631e-01 1.11702327e-02 1.15025151e+00 2.13830724e-01
-8.23388278e-01 3.28935534e-01 -4.23134178e-01 -1.12151325e-01
8.35603833e-01 -1.09033930e+00 -3.92790109e-01 5.86511850e-01
-3.90173435e-01 2.90417045e-01 6.98123753e-01 -4.42445546e-01
-9.60346162e-01 -8.97751927e-01 1.00855613e+00 -2.06098109e-01
4.45894808e-01 -4.38853920e-01 -5.57403743e-01 5.18267214e-01
4.88141805e-01 -9.04463157e-02 1.11804152e+00 3.17584544e-01
7.33283982e-02 -2.41891399e-01 -1.06576085e+00 4.09881443e-01
4.13831979e-01 -4.09838021e-01 -4.60716486e-01 -2.06139520e-01
2.41377339e-01 1.06148317e-01 -8.66528809e-01 4.81977075e-01
9.87425506e-01 -9.95536685e-01 4.30912316e-01 -8.88372719e-01
5.52307069e-01 6.45596907e-02 -2.01141134e-01 -1.33048129e+00
-2.33103067e-01 -4.54198331e-01 -1.10265352e-01 1.26730287e+00
3.42655152e-01 -1.48278192e-01 4.69074816e-01 7.59992898e-01
3.49100493e-02 -1.11334097e+00 -6.82306826e-01 -1.52141973e-01
-3.94505560e-01 -5.13473690e-01 1.82277799e-01 9.19407666e-01
8.09395075e-01 7.46417761e-01 -6.68580770e-01 -4.34857249e-01
1.10071152e-02 -1.78000219e-02 2.58566350e-01 -1.15296960e+00
-1.11914299e-01 -3.27977985e-01 -5.73230803e-01 -1.08156271e-01
1.77282602e-01 -6.12020373e-01 3.17534447e-01 -1.25839603e+00
4.92794573e-01 4.42799419e-01 -9.85692918e-01 5.46552837e-01
-9.62484553e-02 3.20933044e-01 1.06205031e-01 -5.25754750e-01
-1.11413288e+00 8.41098785e-01 7.67567873e-01 1.87416285e-01
-5.94488025e-01 -2.28488952e-01 -4.51037794e-01 8.63294244e-01
1.12412572e+00 -1.79972500e-01 -3.04031909e-01 5.29149830e-01
4.72784728e-01 8.18498209e-02 -8.41938406e-02 -9.97372925e-01
3.66433933e-02 -2.00220346e-01 8.19735706e-01 -6.23263955e-01
5.78850925e-01 -6.35109425e-01 -1.89844191e-01 3.37060690e-01
-5.46495974e-01 3.60324383e-01 3.07216346e-01 8.92909691e-02
-3.11308682e-01 -5.42163849e-04 1.00289214e+00 -8.31237957e-02
-1.04582918e+00 3.43055315e-02 -6.83683574e-01 -2.04792157e-01
1.41269004e+00 -7.32454211e-02 -2.78921649e-02 -4.00956035e-01
-1.15939736e+00 2.93246746e-01 -2.86252290e-01 4.63972896e-01
7.21083939e-01 -1.49854374e+00 -4.18541104e-01 2.54812837e-01
1.41590402e-01 -9.31007504e-01 6.55739963e-01 1.04642725e+00
-4.98687029e-02 1.62815731e-02 -7.45510280e-01 -1.39414400e-01
-1.71260846e+00 4.90650475e-01 4.62183684e-01 -5.05937994e-01
3.33968997e-01 7.88556337e-01 5.70632741e-02 -7.15652034e-02
2.20327005e-01 1.99519083e-01 -9.31339264e-01 6.52819693e-01
6.76167190e-01 4.85445768e-01 -5.36559224e-02 -6.79911137e-01
-4.84705716e-01 2.30034679e-01 8.25051367e-02 -2.59560257e-01
1.58797610e+00 -7.09067956e-02 -4.98564482e-01 1.06498694e+00
1.33212960e+00 -5.26928544e-01 -7.09872305e-01 2.58880705e-01
-4.48328741e-02 1.38369948e-01 1.28923923e-01 -1.00856197e+00
-9.69088078e-01 1.12021387e+00 1.26617301e+00 1.82792440e-01
1.55346572e+00 -4.15356845e-01 6.76783264e-01 4.49063957e-01
-2.32571587e-01 -1.95179760e+00 3.07393134e-01 6.81553900e-01
7.23471940e-01 -1.20585155e+00 -1.93048716e-01 1.52235150e-01
-1.33979774e+00 1.25186133e+00 1.03366220e+00 -1.21268436e-01
7.81588614e-01 1.88095525e-01 2.40872324e-01 -2.91461200e-01
-1.17747009e+00 -4.29758489e-01 3.35587144e-01 2.88468570e-01
8.91420305e-01 1.57573283e-01 -7.83211887e-01 1.22094119e+00
6.41629919e-02 3.11556412e-03 2.22020775e-01 6.52536213e-01
-5.62386036e-01 -7.93625891e-01 -1.07845537e-01 5.43258190e-01
-4.59928036e-01 9.91018340e-02 -7.81190455e-01 1.94634214e-01
3.38029534e-01 1.00807238e+00 2.06027225e-01 -8.57070804e-01
3.11292499e-01 5.45580328e-01 2.78828681e-01 -4.11153048e-01
-1.07633317e+00 2.09266782e-01 1.48120224e-01 -9.46233511e-01
-5.33362627e-01 -5.82192242e-01 -1.59384942e+00 1.19997039e-01
-4.59565818e-02 2.42122129e-01 8.21560800e-01 7.80019641e-01
2.61820912e-01 6.94456160e-01 1.05867946e+00 -8.28581929e-01
1.65572166e-01 -1.08030760e+00 -7.02126145e-01 6.28700852e-01
-7.09137693e-02 -3.54887217e-01 -4.53266263e-01 3.41597274e-02] | [13.490983963012695, 2.6867470741271973] |
99ca63ee-3dd3-47b3-969f-72273d9c1868 | short-segment-heart-sound-classification | 1810.11573 | null | http://arxiv.org/abs/1810.11573v1 | http://arxiv.org/pdf/1810.11573v1.pdf | Short-segment heart sound classification using an ensemble of deep convolutional neural networks | This paper proposes a framework based on deep convolutional neural networks
(CNNs) for automatic heart sound classification using short-segments of
individual heart beats. We design a 1D-CNN that directly learns features from
raw heart-sound signals, and a 2D-CNN that takes inputs of two- dimensional
time-frequency feature maps based on Mel-frequency cepstral coefficients
(MFCC). We further develop a time-frequency CNN ensemble (TF-ECNN) combining
the 1D-CNN and 2D-CNN based on score-level fusion of the class probabilities.
On the large PhysioNet CinC challenge 2016 database, the proposed CNN models
outperformed traditional classifiers based on support vector machine and hidden
Markov models with various hand-crafted time- and frequency-domain features.
Best classification scores with 89.22% accuracy and 89.94% sensitivity were
achieved by the ECNN, and 91.55% specificity and 88.82% modified accuracy by
the 2D-CNN alone on the test set. | ['Sh-Hussain Salleh', 'Chee-Ming Ting', 'Fuad Noman', 'Hernando Ombao'] | 2018-10-27 | null | null | null | null | ['sound-classification'] | ['audio'] | [ 4.42893542e-02 -1.30097777e-01 -6.11678734e-02 -4.62668955e-01
-6.91542923e-01 -1.93379924e-01 -9.73230824e-02 1.61149770e-01
-5.87839603e-01 6.18165612e-01 4.97781783e-02 -2.99339116e-01
-1.15985431e-01 -7.19407499e-01 -8.37721080e-02 -6.31888688e-01
-6.20836794e-01 -8.54675565e-03 -6.15469329e-02 2.52573252e-01
-1.00399517e-01 2.95429021e-01 -1.21789789e+00 4.61638600e-01
4.52470750e-01 1.47109258e+00 -3.43505204e-01 1.38561213e+00
2.57323742e-01 5.65291703e-01 -9.53123510e-01 -1.04855977e-01
-5.48647298e-03 -6.54080391e-01 -4.10474390e-01 -7.90883541e-01
1.63000464e-01 -3.28588217e-01 -4.64256763e-01 4.89958376e-01
1.08958852e+00 -1.40604660e-01 3.79736245e-01 -9.23356056e-01
-4.41793472e-01 5.09061337e-01 9.35396478e-02 7.39233732e-01
1.32899076e-01 2.81016324e-02 8.03353250e-01 -7.86532760e-01
-4.23474051e-02 6.62476420e-01 1.44348288e+00 4.54496831e-01
-9.65302110e-01 -8.28007221e-01 -6.63131952e-01 1.04802027e-01
-1.45725429e+00 1.03581443e-01 7.71438360e-01 -5.20954788e-01
1.30967200e+00 1.37064964e-01 9.72592711e-01 9.13579702e-01
5.77897727e-01 3.03375214e-01 8.02638412e-01 -4.19448733e-01
5.79879573e-03 -2.58176267e-01 7.27283880e-02 7.39699185e-01
1.30055815e-01 4.61443335e-01 -4.56392974e-01 -5.47610879e-01
9.90509927e-01 2.18722045e-01 -2.28859499e-01 4.29942548e-01
-1.32990551e+00 8.21186423e-01 2.57142663e-01 6.18576586e-01
-5.29260457e-01 1.59581870e-01 6.67635977e-01 1.27331316e-01
2.94651479e-01 5.32673895e-01 -9.71792459e-01 -4.74466622e-01
-1.10410357e+00 2.55164742e-01 8.46662045e-01 2.41439611e-01
2.98939109e-01 3.81290168e-01 -3.91356230e-01 7.91666985e-01
6.24852292e-02 5.79958618e-01 7.38845170e-01 -8.62670064e-01
9.93340611e-02 9.55919251e-02 -2.22238719e-01 -1.09100616e+00
-9.79265809e-01 -9.12878931e-01 -1.24206221e+00 -3.86502743e-01
2.72692472e-01 -4.52393264e-01 -8.03556561e-01 1.71168256e+00
2.25596633e-02 6.39804542e-01 2.61134595e-01 8.17385912e-01
1.22617233e+00 4.21983629e-01 2.29223832e-01 -3.71203154e-01
1.35485530e+00 -5.32110512e-01 -9.54153180e-01 5.53898990e-01
5.55177569e-01 -3.32493424e-01 4.82952982e-01 4.49905872e-01
-8.77528071e-01 -1.10273826e+00 -1.10456884e+00 1.33012176e-01
-2.15404689e-01 4.76552784e-01 2.90272355e-01 8.81303012e-01
-7.35886753e-01 1.11457837e+00 -7.86681235e-01 1.59791470e-01
6.22870862e-01 4.27894264e-01 -1.29253924e-01 4.35452074e-01
-1.84569311e+00 5.27762055e-01 1.17627680e-01 2.00911045e-01
-9.08504903e-01 -9.38372493e-01 -8.15389752e-01 4.01030868e-01
-6.32599890e-01 -7.84624636e-01 1.32561195e+00 -4.90994960e-01
-1.62370515e+00 5.31440139e-01 2.77115673e-01 -8.23348463e-01
3.13092321e-01 -3.41644585e-01 -6.90409124e-01 5.39902627e-01
-6.99199438e-02 3.23286533e-01 7.25368381e-01 -3.33724678e-01
-6.19554758e-01 -1.10693149e-01 -4.74286944e-01 -2.35717177e-01
-2.65224636e-01 -2.25126907e-01 1.35690957e-01 -8.79473984e-01
2.29287408e-02 -8.76074612e-01 -1.29176617e-01 -3.15668285e-01
-3.46584678e-01 -2.72061467e-01 6.26055002e-01 -8.59003127e-01
1.49556482e+00 -2.27295518e+00 -3.61317664e-01 -3.88250761e-02
5.22907078e-01 7.89628208e-01 1.25910714e-01 1.83819860e-01
-2.96654075e-01 2.32017651e-01 -7.68463761e-02 -7.84685165e-02
-3.12903851e-01 2.14286242e-02 -8.02735984e-02 3.02274019e-01
3.57488155e-01 8.73032629e-01 -9.24155414e-01 -4.54103082e-01
3.24219316e-01 9.73815799e-01 -4.95755464e-01 2.38046944e-01
3.93434078e-01 2.86192626e-01 -2.01102585e-01 4.04005736e-01
3.56054127e-01 -9.27703306e-02 1.66068614e-01 -3.99728596e-01
2.02204715e-02 3.66147757e-01 -6.95259869e-01 1.66446376e+00
-3.88298810e-01 6.40679061e-01 -6.90266848e-01 -1.09322798e+00
1.26709557e+00 9.34395194e-01 7.36952066e-01 -4.38493431e-01
2.75695920e-01 2.96324164e-01 2.89118260e-01 -7.34190702e-01
-3.78883541e-01 -4.98155087e-01 -2.11472943e-01 1.39951989e-01
5.50887406e-01 2.55674217e-02 -2.44292870e-01 -3.78835410e-01
1.25474334e+00 -1.83414623e-01 3.20935667e-01 -2.24380627e-01
3.71997625e-01 -5.09083271e-01 8.58666539e-01 9.97435510e-01
-5.20379841e-01 1.02744734e+00 4.99300480e-01 -1.22365606e+00
-5.90426385e-01 -8.95784855e-01 -4.91580367e-01 5.26462197e-01
-5.83351016e-01 -5.92473030e-01 -5.41525006e-01 -7.09968150e-01
9.24180970e-02 2.86340535e-01 -8.32545042e-01 -3.92482251e-01
-5.43531597e-01 -7.94215083e-01 1.35539937e+00 8.80279005e-01
4.60696608e-01 -1.24051440e+00 -1.05831456e+00 5.98107278e-01
-2.60511994e-01 -9.32880044e-01 -1.75598666e-01 6.93626881e-01
-9.94533956e-01 -9.51465547e-01 -9.29971099e-01 -7.34103262e-01
-2.23597154e-01 -5.11538863e-01 1.05034506e+00 8.91496316e-02
-7.84242868e-01 4.93167900e-02 -2.69828528e-01 -8.92335892e-01
-1.11321941e-01 1.83677912e-01 2.32509956e-01 4.09306176e-02
2.75703847e-01 -9.38733995e-01 -8.52781594e-01 -1.86972246e-01
-2.69322306e-01 -1.42688826e-01 5.41438758e-01 9.71236050e-01
4.97548431e-01 -1.00590467e-01 1.18465960e+00 -4.95420069e-01
5.77881277e-01 -2.98030794e-01 3.68045121e-02 -1.37603521e-01
-6.21191919e-01 -4.90342766e-01 6.48929477e-01 -5.02582967e-01
-3.27329040e-01 2.08958089e-01 -4.96160090e-01 -7.91478932e-01
-2.43387669e-01 4.33871031e-01 3.57307911e-01 4.27605152e-01
8.20283175e-01 3.77238542e-01 6.60603940e-02 -4.16098416e-01
-2.30901197e-01 7.59455502e-01 8.74900281e-01 -2.68215865e-01
2.57872611e-01 1.77349329e-01 -6.82773143e-02 -6.17792249e-01
-9.04887915e-01 -3.42502534e-01 -8.32457304e-01 -1.72253922e-01
1.38212204e+00 -9.31923807e-01 -7.82264173e-01 7.39554167e-01
-1.10172677e+00 -7.88821056e-02 -3.46127421e-01 9.12832856e-01
-5.61586916e-01 2.52244979e-01 -8.98244798e-01 -8.16503644e-01
-9.80965316e-01 -6.04764462e-01 8.50439787e-01 3.79151583e-01
-5.29411256e-01 -8.96713138e-01 5.64255238e-01 -1.06310032e-01
6.72121584e-01 7.46680737e-01 8.50189328e-01 -9.67377603e-01
3.31283718e-01 -6.09215617e-01 9.30192769e-02 7.99443543e-01
2.37214237e-01 1.58448517e-01 -1.26257384e+00 -9.68075618e-02
-2.25207642e-01 -1.01377107e-01 9.16257799e-01 9.03897464e-01
1.43548453e+00 -1.37703195e-01 -8.77415016e-02 7.55046904e-01
1.07203400e+00 5.45293033e-01 4.97828394e-01 -2.03653291e-01
4.27787513e-01 -9.40899700e-02 -6.99085370e-02 7.20450878e-01
2.81670183e-01 2.14108899e-01 6.93177879e-02 -4.47500229e-01
-1.33315131e-01 -1.84006676e-01 -1.96964577e-01 1.03244710e+00
-4.87279147e-01 1.88987061e-01 -1.15050733e+00 5.79662621e-01
-1.58959222e+00 -1.00602937e+00 -1.75902471e-01 1.93059921e+00
8.77480745e-01 2.66490906e-01 2.81114519e-01 7.46818185e-01
7.77568996e-01 -1.29374087e-01 -3.64324898e-01 -3.91552776e-01
5.98300174e-02 8.81136000e-01 -1.54805630e-01 2.67192312e-02
-1.57258439e+00 1.50790080e-01 6.60872173e+00 3.27710241e-01
-1.46995187e+00 3.30063403e-02 7.59586275e-01 1.01744853e-01
5.38509846e-01 -5.31580091e-01 -4.48509067e-01 4.07495022e-01
1.54013455e+00 -7.31330924e-03 1.22577520e-02 8.37188661e-01
-2.76232120e-02 3.95521313e-01 -8.20893645e-01 1.28374863e+00
-1.28732726e-01 -1.53839493e+00 -4.22318697e-01 -1.75848648e-01
4.54373062e-01 1.24326169e-01 -2.89987847e-02 6.15495145e-01
-3.55895340e-01 -1.15356183e+00 3.16162378e-01 7.62968063e-01
1.35931695e+00 -8.04019511e-01 1.16374695e+00 2.31199518e-01
-1.24636948e+00 -2.36856803e-01 -1.08967267e-01 -2.91412234e-01
-1.70717925e-01 8.56690168e-01 -9.25709844e-01 4.13686782e-01
1.13718212e+00 8.55333090e-01 -2.13118941e-01 9.93035793e-01
5.84898330e-02 1.07215428e+00 -3.25865597e-01 9.76337306e-03
-1.97087124e-01 6.08944893e-01 1.61121696e-01 1.63347101e+00
3.01175773e-01 3.14526618e-01 2.69371849e-02 5.56205630e-01
1.24406770e-01 -7.00919256e-02 -2.35819936e-01 -5.07570840e-02
3.80886108e-01 1.42677081e+00 -4.81399536e-01 -6.70808077e-01
-2.42403328e-01 3.62405956e-01 -4.70166691e-02 1.48696095e-01
-9.76290286e-01 -1.27649474e+00 5.25050521e-01 -1.07425176e-01
3.36555362e-01 1.59215525e-01 -4.44865584e-01 -1.00334656e+00
-4.75170583e-01 -5.60586154e-01 6.97314143e-01 -4.47339773e-01
-1.30204785e+00 9.64072406e-01 -2.36246929e-01 -1.30271518e+00
-2.43506059e-01 -3.63341659e-01 -8.32672715e-01 9.62619483e-01
-1.39935732e+00 -6.13776386e-01 -3.13304156e-01 5.63525259e-01
2.43152961e-01 -1.95486888e-01 1.50910163e+00 4.55739051e-01
-5.01182735e-01 7.80227005e-01 -2.70053446e-01 7.86791682e-01
5.60931683e-01 -1.28121817e+00 2.60897934e-01 1.94126740e-01
8.75335708e-02 5.59143305e-01 1.57526687e-01 -3.14132243e-01
-6.73951507e-01 -1.35016286e+00 1.43749511e+00 -2.68561900e-01
-1.11809880e-01 -8.70860741e-02 -7.83549428e-01 2.55449444e-01
-1.13281436e-01 7.70810902e-01 1.20651472e+00 4.49579358e-02
-3.51748407e-01 -2.19605371e-01 -1.13704085e+00 -3.39456856e-01
5.41325569e-01 -7.93270767e-01 -7.39840209e-01 2.20638476e-02
5.34220397e-01 -5.19074738e-01 -1.47449207e+00 9.84243333e-01
1.09122097e+00 -9.43119586e-01 1.03610659e+00 -7.67686129e-01
4.62612063e-01 -1.20250367e-01 -1.83290392e-02 -1.13557351e+00
-7.04009652e-01 -5.21459341e-01 -3.13027352e-01 6.67497575e-01
3.05769801e-01 -4.91173416e-01 3.70895952e-01 -8.03071558e-02
-3.82915139e-01 -1.07150197e+00 -1.17288578e+00 -5.29940367e-01
3.85388657e-02 -6.56429887e-01 4.62590843e-01 9.37424242e-01
9.50723886e-02 2.83954203e-01 -3.73793662e-01 1.73515245e-01
3.30130279e-01 1.45002782e-01 2.36884221e-01 -1.49129128e+00
-2.78899699e-01 -1.89850003e-01 -6.53076708e-01 -3.15096527e-01
-2.34249964e-01 -8.29855800e-01 -7.21208081e-02 -1.20771456e+00
1.14257686e-01 -1.77060708e-01 -1.14517844e+00 7.79676974e-01
-2.44377315e-01 5.90117514e-01 1.01355694e-01 -1.14854142e-01
-3.97194177e-01 2.16891959e-01 9.94605660e-01 1.58835560e-01
-3.59299392e-01 3.87526721e-01 -2.31013343e-01 5.99814773e-01
1.08104241e+00 -6.43492520e-01 -2.07679734e-01 1.73569173e-01
-2.38110304e-01 7.72089839e-01 4.58511293e-01 -1.55153072e+00
-1.65240005e-01 3.16498309e-01 1.03860056e+00 -6.30262613e-01
2.31589749e-01 -1.99023992e-01 1.96675390e-01 1.11325443e+00
-6.26144052e-01 -4.58331369e-02 2.21970931e-01 4.82077599e-01
-2.99788833e-01 3.48912001e-01 8.49295795e-01 -6.73826560e-02
1.07589737e-01 4.21552479e-01 -6.71968639e-01 1.03698879e-01
6.71681106e-01 1.65425576e-02 -9.28057879e-02 -2.36643732e-01
-1.25268793e+00 -4.50973839e-01 -6.78306103e-01 5.02185762e-01
7.23097622e-01 -1.54152405e+00 -7.88445354e-01 5.91586530e-01
-1.15109524e-02 -2.28299767e-01 6.79541349e-01 9.55279827e-01
-5.46334803e-01 6.47619247e-01 -5.31903744e-01 -8.48814070e-01
-1.13711572e+00 1.17882751e-01 8.58129501e-01 -2.73952216e-01
-7.41959572e-01 1.02638817e+00 -9.74534526e-02 -2.55277514e-01
3.95031214e-01 -8.66579652e-01 -3.80278945e-01 -4.34738733e-02
5.89785516e-01 2.75739729e-01 2.81198204e-01 -2.29086891e-01
-5.99770725e-01 5.46877742e-01 2.98067540e-01 2.32051224e-01
1.52429628e+00 5.31261981e-01 3.09834212e-01 5.52505493e-01
1.42504525e+00 -6.38780653e-01 -8.56098711e-01 -1.69682816e-01
-4.08116132e-01 -6.11889847e-02 2.75580436e-01 -1.16275322e+00
-1.27244270e+00 1.43723965e+00 1.08829856e+00 3.44911724e-01
1.19302189e+00 -4.12637472e-01 1.14921534e+00 1.36127114e-01
-1.33873925e-01 -8.32929373e-01 1.16415694e-01 4.04474020e-01
5.96484661e-01 -7.19303370e-01 -2.69433916e-01 2.46509477e-01
-5.09388030e-01 1.81613207e+00 3.77552360e-01 -3.17115098e-01
1.37677431e+00 2.03401700e-01 3.88362348e-01 -1.72335118e-01
-9.15119410e-01 8.87265336e-03 3.38515013e-01 6.07156515e-01
5.59797883e-01 2.77192056e-01 -1.62694320e-01 1.52856886e+00
-3.31837893e-01 5.07978976e-01 1.25256345e-01 8.20830882e-01
-1.85080022e-01 -4.51977193e-01 -3.84361036e-02 6.31691694e-01
-9.93031800e-01 -6.74439743e-02 1.30362049e-01 4.53839064e-01
6.37327254e-01 8.15000355e-01 2.66689628e-01 -9.57159340e-01
3.35380495e-01 5.79666436e-01 1.36095822e-01 -4.88024294e-01
-9.76237774e-01 1.99895263e-01 -1.56557381e-01 -2.91832358e-01
-2.15953365e-01 -5.53020179e-01 -1.31889164e+00 2.22851202e-01
-2.92048365e-01 2.16184750e-01 6.23716295e-01 5.63607752e-01
5.71981251e-01 8.55033934e-01 8.13688099e-01 -7.36504734e-01
-6.96156502e-01 -1.18518507e+00 -6.03443384e-01 1.34979263e-01
9.87231135e-01 -3.27573210e-01 -4.38239574e-01 2.48960391e-01] | [14.283003807067871, 3.3641231060028076] |
f36202c8-96ea-4f70-a4e4-a7dc9fe2da92 | me-d2n-multi-expert-domain-decompositional | 2210.05280 | null | https://arxiv.org/abs/2210.05280v1 | https://arxiv.org/pdf/2210.05280v1.pdf | ME-D2N: Multi-Expert Domain Decompositional Network for Cross-Domain Few-Shot Learning | Recently, Cross-Domain Few-Shot Learning (CD-FSL) which aims at addressing the Few-Shot Learning (FSL) problem across different domains has attracted rising attention. The core challenge of CD-FSL lies in the domain gap between the source and novel target datasets. Though many attempts have been made for CD-FSL without any target data during model training, the huge domain gap makes it still hard for existing CD-FSL methods to achieve very satisfactory results. Alternatively, learning CD-FSL models with few labeled target domain data which is more realistic and promising is advocated in previous work~\cite{fu2021meta}. Thus, in this paper, we stick to this setting and technically contribute a novel Multi-Expert Domain Decompositional Network (ME-D2N). Concretely, to solve the data imbalance problem between the source data with sufficient examples and the auxiliary target data with limited examples, we build our model under the umbrella of multi-expert learning. Two teacher models which can be considered to be experts in their corresponding domain are first trained on the source and the auxiliary target sets, respectively. Then, the knowledge distillation technique is introduced to transfer the knowledge from two teachers to a unified student model. Taking a step further, to help our student model learn knowledge from different domain teachers simultaneously, we further present a novel domain decomposition module that learns to decompose the student model into two domain-related sub parts. This is achieved by a novel domain-specific gate that learns to assign each filter to only one specific domain in a learnable way. Extensive experiments demonstrate the effectiveness of our method. Codes and models are available at https://github.com/lovelyqian/ME-D2N_for_CDFSL. | ['Yu-Gang Jiang', 'Jingjing Chen', 'Yanwei Fu', 'Yu Xie', 'Yuqian Fu'] | 2022-10-11 | null | null | null | null | ['cross-domain-few-shot', 'cross-domain-few-shot-learning'] | ['computer-vision', 'computer-vision'] | [ 2.58264124e-01 2.47370720e-01 -4.52391207e-01 -4.96575147e-01
-8.96898687e-01 -5.92221200e-01 3.69882137e-01 2.22786721e-02
-2.25589156e-01 8.78567219e-01 -1.35102257e-01 -1.73294678e-01
-2.67194092e-01 -8.65833700e-01 -6.44622326e-01 -5.35304844e-01
6.14257872e-01 6.34184897e-01 6.59845054e-01 -4.37428176e-01
-3.18491191e-01 1.14417680e-01 -1.62234759e+00 1.74804300e-01
1.34540486e+00 7.91617274e-01 3.23919207e-01 2.52226859e-01
-2.43517876e-01 8.82281363e-01 -5.42832017e-01 -4.30831522e-01
1.28773406e-01 -5.89158654e-01 -9.34561133e-01 1.83603913e-01
3.65566432e-01 -1.67331934e-01 -1.49150848e-01 1.12679684e+00
5.84111631e-01 6.48246586e-01 7.49397695e-01 -1.41454744e+00
-6.94363952e-01 5.95407128e-01 -6.58354163e-01 1.60997927e-01
1.81420207e-01 5.81834465e-02 8.78434956e-01 -9.98297989e-01
5.58479488e-01 1.21537900e+00 3.44573110e-01 1.00449550e+00
-1.20192206e+00 -8.98531556e-01 2.83628911e-01 5.17222881e-01
-1.08214927e+00 -2.51696169e-01 9.17976856e-01 -4.13310885e-01
4.33371574e-01 -2.03495085e-01 4.31670815e-01 1.22818911e+00
-5.72306752e-01 1.13780177e+00 1.11887074e+00 -5.27307868e-01
3.89214277e-01 5.57687938e-01 4.69829232e-01 3.38553131e-01
1.01002797e-01 -1.31866083e-01 -4.92269397e-01 1.26538053e-01
4.21469390e-01 2.56913360e-02 -4.46626246e-01 -6.01028740e-01
-5.18098891e-01 8.50288749e-01 1.42600074e-01 3.82441849e-01
4.21687886e-02 -4.80110407e-01 2.65440166e-01 6.33574545e-01
6.76525593e-01 2.32656106e-01 -6.23974919e-01 -3.67948711e-02
-7.43600607e-01 1.39592022e-01 7.41885126e-01 1.12724304e+00
7.98025191e-01 5.20822476e-04 -2.86507756e-02 1.18188441e+00
1.82908103e-01 1.81319609e-01 5.64400852e-01 -6.65998399e-01
5.33217967e-01 7.33825505e-01 5.24183596e-03 -6.35258794e-01
-1.65350139e-02 -4.05175865e-01 -6.90486193e-01 1.71929047e-01
3.55991602e-01 -2.70748347e-01 -8.56500387e-01 1.92732036e+00
7.87060797e-01 8.64867687e-01 2.74586052e-01 6.93256021e-01
1.12844121e+00 6.93650126e-01 1.07028477e-01 -3.55425537e-01
1.23946202e+00 -1.05183458e+00 -5.38083434e-01 -2.94162244e-01
6.00565612e-01 -4.78055239e-01 1.15862286e+00 5.30500829e-01
-1.09262860e+00 -8.31039906e-01 -9.89025712e-01 -1.52808443e-01
-5.09697855e-01 -1.35104969e-01 1.22815311e-01 4.39371020e-01
-5.74933827e-01 4.63307798e-01 -4.36874568e-01 -2.21577495e-01
6.27054751e-01 1.20013423e-01 -1.77658945e-01 -5.35379887e-01
-1.54456031e+00 9.41284597e-01 6.35378242e-01 -5.66119194e-01
-1.07508695e+00 -9.70624030e-01 -8.96691382e-01 2.60434031e-01
7.41625726e-01 -3.59573871e-01 1.55643237e+00 -9.37045455e-01
-1.48439622e+00 8.59602630e-01 6.92685246e-02 -2.59051144e-01
4.09792185e-01 -1.07817508e-01 -4.09970194e-01 1.45623609e-02
2.13854790e-01 4.22331721e-01 8.12106073e-01 -1.22712111e+00
-9.25351501e-01 -3.10895145e-01 1.58189893e-01 3.01820248e-01
-7.13822484e-01 -1.21266820e-01 -4.76024836e-01 -5.67945957e-01
-2.72842944e-01 -5.21094561e-01 -1.46786645e-01 -1.96025044e-01
1.21154552e-02 -8.68834972e-01 7.59730160e-01 -3.86751086e-01
1.54721498e+00 -2.09100127e+00 2.29949385e-01 -5.47279902e-02
2.48339504e-01 9.02441084e-01 -3.23620498e-01 2.10743815e-01
-3.54292154e-01 -2.24342257e-01 -3.37139994e-01 -2.90550590e-01
-1.01133185e-02 3.02796602e-01 -4.07315433e-01 1.42096713e-01
3.72927964e-01 6.44808829e-01 -1.27110946e+00 -5.34505665e-01
1.89323515e-01 2.68058270e-01 -5.03145874e-01 4.89307076e-01
-4.61188823e-01 4.72566515e-01 -5.42448282e-01 5.09710312e-01
8.36931646e-01 -2.24297434e-01 2.27036729e-01 1.13572828e-01
9.53737348e-02 1.59318641e-01 -1.48789978e+00 1.75350988e+00
-5.16677439e-01 2.60521472e-01 1.99536055e-01 -1.60444248e+00
9.74478185e-01 6.33348286e-01 5.28044105e-01 -5.36305785e-01
1.64431348e-01 2.08491653e-01 -5.10671176e-02 -5.06474376e-01
7.09205270e-02 -5.61608493e-01 -8.86440575e-02 4.10594434e-01
6.89227462e-01 -2.40407988e-01 2.65563190e-01 1.36289656e-01
9.46752310e-01 1.05167791e-01 5.05554676e-01 -3.86139983e-03
6.33550823e-01 2.53945775e-02 9.97121811e-01 4.76472527e-01
-5.75334072e-01 3.87950063e-01 3.83969635e-01 -1.65714979e-01
-5.95769644e-01 -1.05844796e+00 5.85877113e-02 1.57048368e+00
1.56268895e-01 -1.99694201e-01 -5.66416264e-01 -1.01831782e+00
-7.01017380e-02 8.47416818e-01 -5.00956357e-01 -4.81741875e-01
-3.69856387e-01 -3.65360498e-01 3.16935718e-01 4.51158762e-01
4.56924945e-01 -9.08232987e-01 -4.65132684e-01 2.95473605e-01
-9.97328311e-02 -9.09050643e-01 -2.28647694e-01 3.85118157e-01
-6.59448862e-01 -1.26140916e+00 -8.83927226e-01 -1.06684268e+00
5.43202162e-01 3.95637691e-01 1.21452940e+00 -3.01354736e-01
-1.57200336e-01 3.29241961e-01 -5.60595810e-01 -6.90347850e-01
-2.70630985e-01 4.78449911e-02 1.97247818e-01 -2.21993346e-02
9.16644216e-01 -8.43244135e-01 -1.59781873e-01 2.32815519e-01
-9.33998108e-01 -1.26096783e-02 3.88734400e-01 9.94354308e-01
4.12966341e-01 3.71190816e-01 8.91327560e-01 -1.16697073e+00
6.64024532e-01 -8.84185433e-01 -5.01991451e-01 4.88802403e-01
-4.80694056e-01 -6.30694255e-02 8.82512450e-01 -7.66315937e-01
-1.35984707e+00 -1.06516378e-02 -1.76472276e-01 -9.71700907e-01
-4.95565593e-01 3.94895017e-01 -5.37066340e-01 1.94706023e-01
8.05452406e-01 1.70250878e-01 -2.18413889e-01 -7.49431968e-01
3.98152560e-01 8.27113926e-01 4.04317081e-01 -8.23646069e-01
1.01822591e+00 8.72203335e-02 -5.14941335e-01 -6.40326321e-01
-1.44928443e+00 -7.09813297e-01 -7.71022320e-01 -6.52942955e-02
6.04343474e-01 -1.11772716e+00 -4.14533794e-01 3.94253850e-01
-1.03309560e+00 -4.03097272e-01 -6.35346293e-01 3.66286725e-01
-4.25044775e-01 1.30039364e-01 -1.30314440e-01 -6.75984263e-01
1.09486291e-02 -1.07383168e+00 4.25513685e-01 5.67166209e-01
1.89008638e-02 -1.19246864e+00 3.72599930e-01 5.11103988e-01
1.49133593e-01 -4.29951847e-02 1.01196659e+00 -1.32711089e+00
-2.92042438e-02 8.01026002e-02 -1.37840793e-01 6.23400569e-01
2.08578095e-01 -4.13010210e-01 -1.09941316e+00 -3.83904487e-01
2.36308098e-01 -1.00698400e+00 8.02415073e-01 1.26532346e-01
1.24992526e+00 -2.02471130e-02 -2.99062759e-01 3.18920046e-01
1.45723426e+00 2.72401839e-01 9.72752571e-02 1.26700148e-01
6.73223257e-01 7.35233784e-01 8.87702465e-01 4.51319814e-01
4.79719251e-01 5.16487598e-01 8.52015465e-02 1.29458517e-01
-2.29230464e-01 -3.30709338e-01 3.91540766e-01 9.93045211e-01
1.67552650e-01 -2.27298304e-01 -1.04693592e+00 1.00515151e+00
-1.81589007e+00 -8.25911641e-01 3.22035998e-01 1.95007277e+00
1.15027094e+00 9.40366015e-02 1.84004098e-01 6.57949829e-03
7.56318808e-01 1.09247625e-01 -8.97391796e-01 -3.35982293e-01
3.09332252e-01 5.57189882e-01 -6.07434362e-02 4.31946725e-01
-1.09986389e+00 1.02041960e+00 4.75276470e+00 1.34613109e+00
-9.36216414e-01 3.96876991e-01 3.61800313e-01 -1.73943892e-01
-3.07731718e-01 -5.73108457e-02 -1.16176832e+00 4.27249163e-01
9.00661945e-01 -4.50915575e-01 1.21624276e-01 1.10894859e+00
-6.31439835e-02 1.92442626e-01 -1.06044924e+00 7.86603570e-01
-6.05560653e-03 -9.61899459e-01 -5.86420763e-03 -2.29439989e-01
9.93380666e-01 -1.66400194e-01 6.15079654e-03 1.06702006e+00
7.37174571e-01 -7.69014299e-01 1.07654467e-01 1.64112315e-01
8.19588840e-01 -8.94192815e-01 3.13712358e-01 9.00600433e-01
-1.18457806e+00 -3.82977784e-01 -6.25869751e-01 -8.61855298e-02
-1.42995313e-01 4.97853726e-01 -8.04153800e-01 7.14368165e-01
6.30260170e-01 9.47636783e-01 -3.16127896e-01 8.40308964e-01
-3.80273789e-01 7.45177567e-01 -1.18125975e-02 2.74566770e-01
1.72310099e-01 -1.79047540e-01 4.59017217e-01 8.10811758e-01
2.37753272e-01 5.15457392e-01 4.41209257e-01 8.52899969e-01
-3.08094949e-01 -2.13826112e-02 -7.20085502e-01 -4.52894233e-02
6.92567885e-01 1.17371595e+00 -1.54318139e-01 -5.42801380e-01
-8.84117126e-01 6.74768925e-01 6.77958012e-01 3.40269476e-01
-7.45050371e-01 -4.57373738e-01 7.99555719e-01 -9.14959237e-03
3.45149815e-01 4.07724947e-01 5.37196882e-02 -1.46350062e+00
-9.32614505e-02 -1.10753632e+00 6.18312776e-01 -4.51770604e-01
-1.69961286e+00 3.02748173e-01 1.69698671e-01 -1.52763534e+00
-2.17277482e-01 -4.81808126e-01 -9.16184247e-01 8.70041251e-01
-1.80726755e+00 -1.07986987e+00 -1.57695130e-01 9.55826938e-01
8.84911358e-01 -3.61021131e-01 7.40098774e-01 4.58230048e-01
-8.33940506e-01 7.26847589e-01 2.89815962e-01 1.65297136e-01
1.03197050e+00 -1.38361526e+00 1.18314371e-01 8.21877956e-01
-2.68754512e-02 3.92964751e-01 5.57526648e-01 -5.85553050e-01
-9.77019846e-01 -1.22230136e+00 8.29168081e-01 -3.72872531e-01
6.59833133e-01 -1.92369878e-01 -1.30350947e+00 6.37793362e-01
1.99826226e-01 1.29716247e-01 9.24740911e-01 7.23694712e-02
-3.59965533e-01 -6.82393387e-02 -1.16389346e+00 2.25829616e-01
8.37031484e-01 -3.93233180e-01 -1.12175131e+00 1.77525982e-01
8.67320538e-01 -2.36110568e-01 -9.58781183e-01 4.55419749e-01
5.68172447e-02 -7.26627350e-01 8.97927761e-01 -1.03340232e+00
6.14225984e-01 -1.06312484e-01 3.04998476e-02 -1.69862235e+00
-1.41921461e-01 -2.07419544e-01 -3.50011259e-01 1.54259014e+00
4.73939851e-02 -3.15785229e-01 8.05373013e-01 4.19810832e-01
-1.70903906e-01 -9.07373488e-01 -7.96951473e-01 -9.01154578e-01
5.34884810e-01 -5.22240758e-01 4.88644391e-01 1.46752071e+00
1.29765347e-01 6.07755721e-01 -2.63476759e-01 6.72768950e-02
6.19991243e-01 3.41646075e-01 6.82969272e-01 -1.69933116e+00
-4.03774261e-01 -2.45483562e-01 1.96035475e-01 -1.15400136e+00
4.48117226e-01 -9.05233324e-01 8.21504593e-02 -1.48271608e+00
1.70504406e-01 -5.53865790e-01 -4.92791623e-01 6.24081552e-01
-4.24187273e-01 -1.41579047e-01 2.31410280e-01 -2.23997846e-01
-8.28397274e-01 7.12031186e-01 1.44107354e+00 -1.07984833e-01
-2.50153631e-01 3.48656356e-01 -8.44761252e-01 8.10773075e-01
6.10558152e-01 -4.61281151e-01 -9.36713696e-01 -2.72449523e-01
-1.33777976e-01 3.15374099e-02 2.79472657e-02 -9.92741764e-01
4.53152567e-01 -4.19715971e-01 2.32646734e-01 -4.13912982e-01
3.16319704e-01 -9.66825545e-01 -3.91831815e-01 2.51492202e-01
-2.52799749e-01 -6.58120215e-01 1.16776511e-01 5.65138042e-01
-4.68888730e-01 -5.27650893e-01 1.11417556e+00 -3.37570727e-01
-1.08172369e+00 4.36925799e-01 -8.16582516e-02 5.66957533e-01
1.28532577e+00 4.29215766e-02 -2.67213285e-01 -1.71068877e-01
-9.55587149e-01 6.75202429e-01 2.24918649e-01 5.02350211e-01
5.82259238e-01 -1.34199023e+00 -6.89117610e-01 3.07548612e-01
3.02668869e-01 5.55428982e-01 5.38958251e-01 6.17961645e-01
2.29357824e-01 2.76666194e-01 -2.86594957e-01 -2.40368724e-01
-1.16829777e+00 7.87619948e-01 1.49252295e-01 -6.08349383e-01
-2.16080949e-01 1.19923770e+00 2.41804317e-01 -9.19607103e-01
4.38171893e-01 -4.41620611e-02 -4.79731679e-01 5.21382451e-01
6.68410599e-01 4.92410749e-01 -1.42171234e-01 -4.15980309e-01
-1.71347782e-01 4.80205357e-01 -2.61567086e-01 1.62634969e-01
1.49571073e+00 -6.11664765e-02 3.18154514e-01 4.87120599e-01
1.00591326e+00 -3.67050827e-01 -1.28027666e+00 -7.79732227e-01
2.50860974e-02 -1.31861612e-01 -2.41034716e-01 -7.83030987e-01
-8.88621449e-01 1.35621381e+00 4.95276421e-01 4.70773950e-02
1.35665393e+00 1.34498656e-01 7.37366557e-01 8.93807411e-02
1.88567370e-01 -1.29430568e+00 3.21406126e-01 7.53627121e-01
3.36813897e-01 -1.60140240e+00 -2.26730332e-01 -3.41564059e-01
-6.02033079e-01 8.99059474e-01 1.23984265e+00 1.29514053e-01
6.80043042e-01 -7.23473281e-02 -2.96914931e-02 -6.04239553e-02
-9.84326243e-01 -5.12283146e-01 3.94226789e-01 7.48661458e-01
3.27721238e-01 -2.91192997e-02 -9.97201651e-02 1.21454751e+00
1.65337726e-01 2.26630837e-01 3.73917043e-01 8.76657248e-01
-8.06894243e-01 -1.44151139e+00 -2.23084033e-01 2.43766636e-01
-1.11901186e-01 1.20394766e-01 -1.90950185e-01 6.36190772e-01
5.30602515e-01 9.81698811e-01 -1.75216109e-01 -3.50216150e-01
5.18660724e-01 4.59797144e-01 2.93889403e-01 -1.16430485e+00
-5.61844885e-01 -9.53330249e-02 -2.85502464e-01 -3.05138558e-01
-3.96922559e-01 -2.97480822e-01 -1.11291099e+00 -6.14224449e-02
-1.88327581e-01 3.81516337e-01 5.43850623e-02 1.12894309e+00
1.08682051e-01 6.64816380e-01 6.47229731e-01 -4.16548014e-01
-7.28819072e-01 -9.35260296e-01 -5.91015339e-01 3.83533925e-01
2.59647608e-01 -8.82852972e-01 -3.56750727e-01 -2.37977188e-02] | [10.128108024597168, 2.9794766902923584] |
143eb561-9a7c-4219-8d64-fdd418c9841c | morphological-analysis-and-disambiguation-for-1 | null | null | https://aclanthology.org/2020.lrec-1.480 | https://aclanthology.org/2020.lrec-1.480.pdf | Morphological Analysis and Disambiguation for Gulf Arabic: The Interplay between Resources and Methods | In this paper we present the first full morphological analysis and disambiguation system for Gulf Arabic. We use an existing state-of-the-art morphological disambiguation system to investigate the effects of different data sizes and different combinations of morphological analyzers for Modern Standard Arabic, Egyptian Arabic, and Gulf Arabic. We find that in very low settings, morphological analyzers help boost the performance of the full morphological disambiguation task. However, as the size of resources increase, the value of the morphological analyzers decreases. | ['Nizar Habash', 'Salam Khalifa', 'Nasser Zalmout'] | 2020-05-01 | null | null | null | lrec-2020-5 | ['morphological-disambiguation', 'morphological-tagging'] | ['natural-language-processing', 'natural-language-processing'] | [-4.20987725e-01 -2.80362576e-01 2.76031464e-01 -1.53640494e-01
-4.65590209e-01 -1.11918652e+00 4.24855441e-01 7.74981618e-01
-7.65849948e-01 6.10840738e-01 -1.87771954e-02 -9.05512214e-01
-3.98841426e-02 -7.93789148e-01 1.25301391e-01 -4.51191247e-01
-8.64948891e-03 7.88643897e-01 4.07456934e-01 -1.04188323e+00
4.15332645e-01 7.45950997e-01 -8.93516123e-01 1.35864979e-02
1.02799439e+00 1.79284260e-01 -6.85195066e-03 4.72868323e-01
-1.11657582e-01 -1.51236475e-01 -9.97725785e-01 -7.07318366e-01
4.63712007e-01 -2.06820861e-01 -8.15605879e-01 3.75088602e-02
1.77303687e-01 -8.05483982e-02 6.17388226e-02 1.22198164e+00
6.28019035e-01 3.02416861e-01 5.49587667e-01 -5.90739191e-01
-7.04330325e-01 1.13653636e+00 -5.72815835e-01 4.98221725e-01
5.41126311e-01 -2.17274744e-02 7.41439223e-01 -1.00150001e+00
7.89593160e-01 1.40940475e+00 4.05496299e-01 -5.89513220e-02
-6.30909562e-01 -4.46776837e-01 -1.17032170e-01 -2.46928018e-02
-1.52437413e+00 -5.00427663e-01 2.62354285e-01 4.21892479e-02
1.27335024e+00 -1.12666644e-01 4.42179173e-01 9.40815778e-04
6.47965372e-02 9.95346755e-02 1.05101776e+00 -1.02272213e+00
9.22242031e-02 -1.47286326e-01 3.52422088e-01 8.97314370e-01
6.10672891e-01 -4.93181735e-01 -2.55226552e-01 -2.34245881e-01
5.27197957e-01 -7.37434149e-01 1.40909418e-01 4.05299127e-01
-1.08555174e+00 9.56375718e-01 -9.88630131e-02 7.08562791e-01
-1.89804524e-01 -7.49902427e-01 4.23743725e-01 5.71021140e-01
1.79024130e-01 9.78676736e-01 -6.59255683e-01 -1.05730072e-01
-9.86811221e-01 1.01076096e-01 9.85694230e-01 7.17020631e-01
5.48862576e-01 2.79983312e-01 4.34503287e-01 8.31057847e-01
2.36439198e-01 6.82613671e-01 7.03693688e-01 -6.76284194e-01
6.09793425e-01 7.96286166e-01 1.91303045e-01 -8.43138278e-01
-6.50275528e-01 -4.09584679e-02 -1.37562249e-02 1.28915077e-02
9.24152672e-01 -5.27003646e-01 -1.09255266e+00 1.12247646e+00
3.01433921e-01 -7.94305384e-01 2.26541653e-01 6.31089568e-01
5.26087046e-01 5.32237709e-01 7.94923678e-02 -3.94333363e-01
1.65694284e+00 -6.59924150e-01 -9.31099832e-01 -3.67362320e-01
7.45092392e-01 -1.43486273e+00 1.11594748e+00 5.32163978e-01
-1.31834233e+00 -2.84133554e-01 -1.45132053e+00 -4.09052009e-03
-8.15198004e-01 1.32766396e-01 6.89415395e-01 1.25968695e+00
-9.39171672e-01 3.42317730e-01 -1.05986702e+00 -4.70210671e-01
-2.77909219e-01 5.71120381e-01 -5.76582551e-01 -2.12384298e-01
-1.20182407e+00 1.50160551e+00 7.31429935e-01 1.58058733e-01
1.27639741e-01 7.99362138e-02 -9.56054151e-01 -1.92850813e-01
3.46871048e-01 1.22622490e-01 9.22344089e-01 -7.46298373e-01
-1.20668900e+00 8.47914457e-01 4.71783802e-02 -1.31935820e-01
-1.07774936e-01 -1.57684639e-01 -8.27728271e-01 1.66671827e-01
2.44646948e-02 3.98084432e-01 5.17289102e-01 -8.89479399e-01
-7.49448121e-01 -6.04216993e-01 2.32084678e-03 3.19503725e-01
-2.27081880e-01 8.16305459e-01 -4.89437789e-01 -8.53787780e-01
2.69704551e-01 -1.06738067e+00 -2.42368549e-01 -8.97169232e-01
2.48612583e-01 -1.10244438e-01 5.38176298e-01 -1.20337570e+00
1.51973200e+00 -2.30276537e+00 9.62979421e-02 4.97859895e-01
-3.79989743e-01 6.95162177e-01 -3.17247123e-01 4.17535335e-01
1.26585662e-01 4.54722464e-01 -2.54970372e-01 3.83323021e-02
-3.63899395e-02 5.74784815e-01 2.58798122e-01 4.76561189e-01
4.20672774e-01 6.16568804e-01 -7.66871512e-01 -6.49146020e-01
2.47476026e-02 1.82101488e-01 -1.90191120e-01 -4.19750959e-01
3.36514972e-02 -2.13150829e-01 4.99993116e-02 1.16595352e+00
6.77417159e-01 6.28643453e-01 8.54837418e-01 1.62818924e-01
-2.80700177e-01 5.84348619e-01 -1.53370881e+00 1.44550645e+00
-2.34594345e-01 3.33421558e-01 1.50138274e-01 -5.37224054e-01
8.51333916e-01 2.44297192e-01 -1.02512851e-01 -7.25156188e-01
4.04746681e-01 6.29566252e-01 7.46450007e-01 -4.56654578e-02
1.19294465e+00 -4.89226207e-02 -4.48951393e-01 5.72008371e-01
1.83704942e-01 -1.90490320e-01 1.16549444e+00 7.29921982e-02
7.12598741e-01 -1.90012395e-01 7.07785368e-01 -4.33527648e-01
5.35538793e-01 3.66643965e-01 2.51785308e-01 3.25263917e-01
-5.92420362e-02 9.40650403e-02 2.39285618e-01 -1.65587276e-01
-7.84505129e-01 -8.96462798e-01 -1.71768665e-01 1.53931022e+00
-2.47484416e-01 -8.08849335e-01 -8.19236040e-01 -6.88731790e-01
-1.36857063e-01 5.88090837e-01 -2.02311292e-01 1.77127540e-01
-1.22066450e+00 -1.30807900e+00 8.97737920e-01 3.91894996e-01
2.18338028e-01 -1.24208391e+00 -5.84679246e-01 4.58313286e-01
-7.12747127e-02 -1.04504454e+00 -1.54793918e-01 2.34595761e-01
-8.32539022e-01 -9.13245738e-01 -1.86223328e-01 -1.00235510e+00
4.44650710e-01 9.62310359e-02 1.00682473e+00 4.04393077e-01
2.02783003e-01 -1.77534491e-01 -8.92186999e-01 -5.43916762e-01
-7.87580907e-01 3.56744885e-01 5.61559126e-02 -6.71932220e-01
4.37987447e-01 -1.14009574e-01 2.02605546e-01 1.16886668e-01
-1.23759782e+00 -9.00699317e-01 5.83250403e-01 3.93033177e-01
2.23532066e-01 3.82045746e-01 2.45128781e-01 -8.68921578e-01
1.03747571e+00 -1.70172706e-01 -5.76448321e-01 4.54941154e-01
-6.81685507e-01 3.34933549e-01 4.43965912e-01 -4.74667996e-01
-9.55144465e-01 -8.33638608e-02 -2.20154941e-01 5.26851535e-01
5.62711135e-02 1.19245708e+00 -3.32464159e-01 -2.79829800e-01
6.59537613e-01 -3.11299682e-01 -1.01544291e-01 -5.44140577e-01
5.57099223e-01 7.56749868e-01 4.21261728e-01 -4.34340775e-01
6.08747721e-01 -1.31715443e-02 -1.60874538e-02 -7.77331412e-01
-1.47604674e-01 -4.05472130e-01 -1.04381430e+00 1.74431950e-01
2.22149745e-01 -6.22484326e-01 -1.71967745e-01 5.89989007e-01
-8.57850611e-01 -3.06961298e-01 8.14809725e-02 2.05056563e-01
1.50307685e-01 8.31764340e-01 -7.58063674e-01 -1.89452872e-01
-3.48023623e-01 -1.16453528e+00 6.33822978e-01 1.73149914e-01
-4.87736046e-01 -1.09766650e+00 -4.77789566e-02 4.95963059e-02
1.57930076e-01 -4.13753577e-02 1.20032918e+00 -9.29581404e-01
2.99679369e-01 -1.16942659e-01 -3.16221341e-02 4.98900637e-02
4.48657870e-01 3.26573730e-01 -1.59877151e-01 -2.39798918e-01
-1.07386991e-01 7.22295493e-02 3.42824996e-01 -8.15220326e-02
-1.47044793e-01 -2.35044688e-01 2.30616972e-01 2.39129871e-01
1.30736983e+00 4.03864115e-01 5.40408194e-01 7.58220613e-01
3.55959088e-01 5.88262200e-01 1.12624848e+00 3.10795486e-01
3.95755261e-01 4.33393687e-01 -1.40624180e-01 5.08212820e-02
5.11121675e-02 4.74653512e-01 6.59358084e-01 1.10845566e+00
-5.61387166e-02 -1.29557267e-01 -1.48638213e+00 7.33151853e-01
-1.43399572e+00 -8.52763504e-02 -7.17539266e-02 1.84608853e+00
1.20466435e+00 1.88798495e-02 3.14738750e-01 5.08333564e-01
4.79998708e-01 -1.41851783e-01 2.97769785e-01 -9.50434804e-01
-4.46349591e-01 8.61589372e-01 6.17674172e-01 8.04329991e-01
-1.11157370e+00 1.53000760e+00 7.12775135e+00 7.88783073e-01
-1.16230345e+00 2.87462696e-02 -1.47326976e-01 2.00651109e-01
1.02973185e-01 1.38047710e-01 -8.12229395e-01 5.28984889e-02
1.24578917e+00 2.84184534e-02 4.22237754e-01 4.62517798e-01
-1.28220558e-01 -3.94345611e-01 -4.12378341e-01 6.13396406e-01
2.14262515e-01 -8.53546143e-01 3.02413017e-01 -1.18325345e-01
4.27727818e-01 4.66910936e-03 -2.89295912e-01 -7.69010885e-03
5.00436425e-01 -9.23738480e-01 7.49296784e-01 -1.30843356e-01
5.23522079e-01 -1.11670983e+00 1.12738526e+00 -1.99751675e-01
-8.83815646e-01 4.08198833e-02 -2.93981999e-01 -1.86926365e-01
1.99703246e-01 1.53042540e-01 -8.55724871e-01 5.53267777e-01
2.81078756e-01 5.53309843e-02 -1.36631095e+00 7.58213460e-01
-6.32236898e-01 7.48050690e-01 -5.32344878e-01 5.19224592e-02
4.05736864e-01 -4.19984818e-01 3.57082069e-01 1.69404387e+00
9.21006575e-02 2.06422329e-01 3.84775817e-01 -2.69183815e-01
2.13157594e-01 6.56014860e-01 5.30312061e-02 -3.14974993e-01
8.13333511e-01 1.11648190e+00 -1.44953609e+00 -4.26309377e-01
-1.72653779e-01 5.03725290e-01 2.52763808e-01 1.09044030e-01
-3.34557980e-01 -7.49731064e-01 2.28752345e-01 4.35834341e-02
3.74637097e-01 -1.00934160e+00 -6.52210116e-01 -9.20330644e-01
-2.42025346e-01 -1.57837856e+00 8.85072470e-01 -4.29879606e-01
-9.19026017e-01 5.81879437e-01 1.47498667e-01 -5.76076269e-01
-4.52244282e-01 -9.03097749e-01 -1.27749592e-01 8.95702124e-01
-1.14570332e+00 -9.66834426e-01 3.74887168e-01 5.01545370e-01
3.04994732e-01 -2.72817910e-01 1.08727121e+00 1.15407497e-01
-3.28228176e-01 7.47222900e-01 1.99511480e-02 5.60836554e-01
1.11346114e+00 -1.14750469e+00 4.80377853e-01 1.43019569e+00
4.43094224e-01 1.05938065e+00 6.63315177e-01 -8.94867301e-01
-1.24594557e+00 -4.88836497e-01 1.08289838e+00 -3.94955128e-01
1.00172770e+00 -2.37569902e-02 -9.37600374e-01 5.10274053e-01
5.01321435e-01 -5.60635448e-01 7.84109414e-01 2.86540806e-01
-3.48781794e-01 2.22971454e-01 -1.11848056e+00 7.62720704e-01
4.21143383e-01 -3.23462278e-01 -9.78504181e-01 -1.17841270e-02
3.42395335e-01 -4.69623238e-01 -1.17839754e+00 1.75073698e-01
6.20169520e-01 -1.98892191e-01 6.88038766e-01 -5.82729459e-01
-1.53342277e-01 -5.09386301e-01 -1.76580459e-01 -1.44961631e+00
-2.87597865e-01 -7.66684592e-01 2.28720769e-01 1.12286520e+00
8.58855009e-01 -5.46676934e-01 1.15943544e-01 3.47283185e-01
-1.26631167e-02 -7.35386759e-02 -7.20650911e-01 -6.16942346e-01
2.35973358e-01 -9.93707106e-02 7.23716021e-01 1.15962744e+00
5.56470096e-01 6.58921003e-01 5.08587360e-01 1.81223318e-01
-8.13245028e-02 5.63409850e-02 1.82296455e-01 -9.55492616e-01
-5.78980222e-02 -4.50959474e-01 -4.64473724e-01 3.14991847e-02
4.48316485e-02 -4.20723915e-01 -7.57858157e-02 -1.27985251e+00
-5.61241746e-01 -5.87575257e-01 -1.12341419e-01 9.61194694e-01
-5.95202684e-01 3.99760395e-01 5.35230398e-01 7.55855292e-02
-2.56693721e-01 -2.91999906e-01 6.26839042e-01 5.76470681e-02
-6.00641608e-01 -4.39390451e-01 -7.36305833e-01 7.74503827e-01
1.10422003e+00 -5.69884658e-01 7.36382753e-02 -6.01199806e-01
4.44342613e-01 -2.25642905e-01 -5.85888803e-01 -6.39682055e-01
1.51986927e-01 -2.99071789e-01 2.61353165e-01 -2.73006469e-01
-1.66244611e-01 -2.94541061e-01 -2.91771144e-01 4.24679279e-01
4.22585815e-01 1.14952648e+00 7.19959974e-01 -4.06773210e-01
-2.74709642e-01 -6.34168983e-01 7.71657228e-01 -2.68613547e-01
-8.19691956e-01 -3.01263064e-01 -8.78917992e-01 1.01132885e-01
5.09660721e-01 -2.36226007e-01 -3.07686776e-01 1.48513064e-01
-5.92528284e-01 -2.28753865e-01 9.37011778e-01 2.98773408e-01
2.70657271e-01 -6.74424827e-01 -7.61886477e-01 2.07415879e-01
-2.40825117e-01 -3.77510726e-01 -6.99881136e-01 5.62799513e-01
-1.09823406e+00 3.12214375e-01 -8.53094101e-01 2.05494031e-01
-1.54499292e+00 5.47419786e-01 -5.52924648e-02 -4.34518605e-01
7.34317303e-02 5.52504599e-01 -8.83815765e-01 -3.86898071e-01
-2.09846646e-01 -1.03779405e-01 -5.47076583e-01 5.95530808e-01
3.93541336e-01 5.02019644e-01 7.64968276e-01 -1.00205779e+00
-5.80754817e-01 2.08986938e-01 -1.28238842e-01 -8.04044425e-01
1.11835742e+00 -6.91596121e-02 -6.06207967e-01 1.29015446e-01
1.30386919e-01 7.90384710e-01 -1.39191851e-01 1.71404362e-01
4.10356164e-01 -3.09318900e-01 -1.65691748e-01 -9.05969977e-01
-5.37421644e-01 4.38536853e-01 2.89294690e-01 1.85448572e-01
1.24159741e+00 -2.97238737e-01 5.95392704e-01 6.73146248e-01
3.40764701e-01 -1.45472360e+00 -5.47366142e-01 1.28045833e+00
3.67804199e-01 -7.80239046e-01 4.01636928e-01 -5.89250088e-01
-6.00486875e-01 1.17857313e+00 3.20156276e-01 4.78533609e-03
5.87233961e-01 5.29150009e-01 5.01454115e-01 -6.81395382e-02
-1.12920299e-01 -5.79012930e-01 1.07653745e-01 7.32508123e-01
6.04149759e-01 2.36924678e-01 -1.18713367e+00 6.54106021e-01
-8.03106010e-01 -7.50376642e-01 8.80554616e-01 1.31663179e+00
-4.35767770e-01 -1.74215102e+00 -8.13150942e-01 -1.09574804e-02
-9.30836439e-01 -6.80494070e-01 -9.00619447e-01 1.31853592e+00
2.12794378e-01 1.27814484e+00 -8.01053494e-02 -9.72200334e-02
1.39303997e-01 3.37826133e-01 8.90049100e-01 -7.74871945e-01
-1.04516077e+00 2.78119564e-01 5.38839877e-01 3.64128910e-02
-6.55221343e-01 -9.42561984e-01 -1.46611786e+00 -5.00727713e-01
-6.71380460e-01 3.15732896e-01 6.76594675e-01 1.17034912e+00
3.77106317e-03 -1.31451339e-01 -6.87692836e-02 -3.81998122e-01
-4.97966111e-01 -1.46604335e+00 -6.27591014e-01 1.33912593e-01
-2.21783713e-01 -5.02557755e-01 1.34972394e-01 2.99274355e-01] | [10.35727310180664, 10.38680648803711] |
9644caf4-46c2-4f55-8f4f-32ffa7a9d5fd | unsupervised-semantic-segmentation-with-self | 2207.05027 | null | https://arxiv.org/abs/2207.05027v2 | https://arxiv.org/pdf/2207.05027v2.pdf | Unsupervised Semantic Segmentation with Self-supervised Object-centric Representations | In this paper, we show that recent advances in self-supervised feature learning enable unsupervised object discovery and semantic segmentation with a performance that matches the state of the field on supervised semantic segmentation 10 years ago. We propose a methodology based on unsupervised saliency masks and self-supervised feature clustering to kickstart object discovery followed by training a semantic segmentation network on pseudo-labels to bootstrap the system on images with multiple objects. We present results on PASCAL VOC that go far beyond the current state of the art (50.0 mIoU), and we report for the first time results on MS COCO for the whole set of 81 classes: our method discovers 34 categories with more than $20\%$ IoU, while obtaining an average IoU of 19.6 for all 81 categories. | ['Thomas Brox', 'Francesco Locatello', 'Yi Zhu', 'Matthaeus Kleindessner', 'Andrii Zadaianchuk'] | 2022-07-11 | null | null | null | null | ['unsupervised-semantic-segmentation'] | ['computer-vision'] | [ 5.39069712e-01 5.23977220e-01 -1.81636244e-01 -5.61208844e-01
-5.13681591e-01 -5.52875638e-01 2.94338167e-01 1.59508169e-01
-5.57775438e-01 7.54358590e-01 -3.82074326e-01 -5.75529039e-02
-4.41593900e-02 -5.07625341e-01 -7.56508708e-01 -4.55619335e-01
-1.19769417e-01 8.55799377e-01 1.04449427e+00 -3.81097756e-02
5.26824653e-01 1.26790062e-01 -2.25404835e+00 3.16320211e-01
1.05629671e+00 1.14710879e+00 4.77856010e-01 6.99953318e-01
-4.01820868e-01 6.44008875e-01 -8.86657059e-01 -9.09671932e-02
1.26454249e-01 -3.49626124e-01 -1.42405856e+00 4.92563218e-01
6.41477585e-01 2.00207576e-01 2.63489962e-01 1.17315185e+00
1.12030454e-01 1.42312348e-01 7.22642004e-01 -1.13399374e+00
-3.26829165e-01 8.37704480e-01 -5.18200636e-01 4.13256049e-01
1.47292271e-01 -1.80293992e-02 1.07332587e+00 -7.19239712e-01
8.12849462e-01 8.88211668e-01 4.92301136e-01 6.53426766e-01
-9.92671430e-01 -5.27615786e-01 2.07803305e-02 2.52068609e-01
-1.32308352e+00 -3.12645018e-01 6.64153695e-01 -4.27822441e-01
9.91928637e-01 2.67261237e-01 7.96139181e-01 4.30771053e-01
-4.50253606e-01 1.35419929e+00 1.36531579e+00 -5.24349213e-01
4.47271645e-01 3.27869415e-01 6.37118042e-01 6.38397753e-01
2.45272815e-01 -1.54474467e-01 -4.43336725e-01 4.87200081e-01
6.45462513e-01 -3.99601400e-01 2.95582145e-01 -5.20746768e-01
-1.20968604e+00 7.76376307e-01 6.08687758e-01 7.73519993e-01
-1.06744356e-02 8.26227218e-02 1.10602602e-01 -1.47097245e-01
3.38999629e-01 9.12471354e-01 -7.12950826e-01 1.67698972e-02
-1.47896779e+00 4.45495881e-02 6.25044048e-01 1.28850639e+00
1.15119314e+00 1.31734028e-01 -2.45817937e-02 6.95209861e-01
2.06027720e-02 3.63629997e-01 6.16766930e-01 -1.02197587e+00
-1.60558939e-01 8.55667949e-01 6.16955943e-02 -4.99264002e-01
-6.20236218e-01 -5.81265509e-01 -1.29474327e-02 3.14038306e-01
5.75971305e-01 -1.46670192e-02 -1.66745913e+00 9.78889585e-01
2.21744582e-01 -2.77738664e-02 1.37065807e-02 8.32247674e-01
1.07994080e+00 1.92197844e-01 1.12351120e-01 7.76511803e-02
1.09471679e+00 -1.22232711e+00 -5.58257103e-01 -3.58639687e-01
4.40165192e-01 -7.80932188e-01 9.20121551e-01 5.01408935e-01
-8.23675454e-01 -8.79103601e-01 -1.04627562e+00 1.94119886e-01
-7.40466475e-01 1.64851964e-01 1.02030241e+00 6.44884646e-01
-1.19483519e+00 6.30397797e-01 -6.67421520e-01 -6.16263866e-01
9.00925934e-01 5.93442440e-01 1.48744240e-01 3.70297313e-01
-8.68502378e-01 6.25703514e-01 1.09517264e+00 -3.36169600e-01
-9.34690297e-01 -3.89547169e-01 -6.22059524e-01 -1.84145465e-01
7.13805676e-01 -1.76159859e-01 1.25757146e+00 -1.57356334e+00
-1.25673771e+00 1.24730301e+00 -8.02146420e-02 -8.70624959e-01
2.80635774e-01 -2.47988403e-01 -4.11328167e-01 6.23823464e-01
5.79501390e-01 1.58129990e+00 8.96235704e-01 -1.52572250e+00
-1.23615348e+00 -6.49457499e-02 -8.78319368e-02 1.47597358e-01
-6.96232393e-02 3.08833886e-02 -4.10962373e-01 -5.01640618e-01
4.06409591e-01 -7.30766773e-01 -2.92972386e-01 -5.55900455e-01
-5.76003015e-01 -4.37057495e-01 8.86702061e-01 -2.98282623e-01
8.74289453e-01 -1.90510237e+00 4.83123213e-02 9.08439755e-02
2.96813697e-01 3.79358172e-01 1.71161965e-01 -3.90235066e-01
-1.85503647e-01 6.24082573e-02 -7.01245904e-01 -2.43150830e-01
-2.13266328e-01 1.83526784e-01 6.15635049e-03 1.15400232e-01
2.89608568e-01 1.16170609e+00 -1.15377116e+00 -8.39564383e-01
2.87256420e-01 -2.14649558e-01 -3.26393038e-01 9.13178325e-02
-5.58493555e-01 4.52069312e-01 -3.43622386e-01 1.01652133e+00
4.68681604e-01 -2.49938309e-01 -1.94008082e-01 1.99727938e-01
-2.81740248e-01 7.05203135e-03 -1.10541964e+00 1.89038646e+00
4.38432306e-01 6.44330561e-01 -1.32680193e-01 -1.19665730e+00
8.89314353e-01 -1.26572475e-01 6.03570580e-01 -4.80679065e-01
3.57721627e-01 3.73914540e-01 -5.16071394e-02 -3.01758647e-01
6.54760838e-01 2.32033834e-01 -1.47626370e-01 7.54374191e-02
8.12328935e-01 -4.86207187e-01 5.28339028e-01 1.65162310e-01
7.50988901e-01 2.77467608e-01 -2.05317289e-02 -8.32476020e-01
2.82669336e-01 7.11039603e-01 2.81180799e-01 1.02245080e+00
-5.02664387e-01 9.24832284e-01 2.60736316e-01 -1.97022974e-01
-9.31546450e-01 -1.16942263e+00 -2.83396125e-01 1.21924520e+00
4.96587962e-01 -2.42859498e-03 -1.48217010e+00 -9.58375931e-01
-2.45450109e-01 7.35377491e-01 -5.62887847e-01 3.41904193e-01
-2.46849954e-01 -7.20742226e-01 3.59064609e-01 4.91155833e-01
7.52235651e-01 -1.58832765e+00 -9.22325015e-01 1.70740187e-01
4.90886979e-02 -1.19762743e+00 3.27826440e-02 5.68856895e-01
-9.81456697e-01 -1.19979382e+00 -7.20171809e-01 -1.26117110e+00
8.02652717e-01 1.36440739e-01 1.27289915e+00 -1.34979382e-01
-6.91537797e-01 3.05788428e-01 -5.62147677e-01 -6.66769505e-01
-1.79856941e-01 3.92641604e-01 -1.81237414e-01 -1.04674421e-01
6.14771128e-01 -2.40798876e-01 -5.28012216e-01 3.73101920e-01
-8.26174021e-01 3.37960482e-01 4.92244959e-01 4.26679999e-01
7.56834984e-01 5.85418865e-02 7.70816028e-01 -1.16621256e+00
-3.03542644e-01 -2.23507613e-01 -6.66620493e-01 -4.68150713e-03
-5.98797560e-01 -4.53173183e-02 1.49548993e-01 -1.77496895e-01
-1.02957046e+00 6.74237788e-01 -9.63011459e-02 -2.62223691e-01
-8.63993883e-01 -1.09408058e-01 -3.50430347e-02 -8.28817040e-02
8.09627891e-01 1.90824300e-01 -2.05725029e-01 -4.96947587e-01
6.53283000e-01 7.49963403e-01 7.90962100e-01 -1.41218111e-01
7.00991511e-01 6.80798590e-01 -4.56748724e-01 -8.16606998e-01
-1.27264118e+00 -9.77302670e-01 -1.05804694e+00 -2.69904494e-01
1.29173899e+00 -7.59659171e-01 -1.19068146e-01 6.24598384e-01
-6.83871806e-01 -4.11607534e-01 -6.93069160e-01 1.69787869e-01
-8.86321366e-01 1.23001724e-01 -2.46915981e-01 -5.45220435e-01
-1.79265335e-01 -9.50172246e-01 1.08778393e+00 6.87828243e-01
-1.94783270e-01 -7.11593270e-01 -3.88586819e-01 7.11023510e-01
2.51301438e-01 1.82856873e-01 2.28232533e-01 -9.52568352e-01
-7.18958139e-01 1.74780823e-02 -4.56827432e-01 3.54604125e-01
4.89411876e-02 -4.03400660e-01 -1.18775773e+00 1.00875616e-01
-1.75439253e-01 -3.47999722e-01 1.15691042e+00 4.89847064e-01
1.23661363e+00 3.35579604e-01 -4.58462089e-01 3.38724494e-01
1.35818684e+00 2.95842677e-01 5.02684832e-01 3.73281717e-01
9.63057697e-01 5.52379489e-01 1.00100636e+00 -3.82550433e-02
1.36464909e-01 2.39405274e-01 5.32054484e-01 -3.30465198e-01
-5.07073224e-01 -3.04797497e-02 -2.91361690e-01 4.65061039e-01
-9.12958570e-03 3.25965345e-01 -1.12013805e+00 1.16189957e+00
-1.77541685e+00 -4.97558922e-01 -1.66568190e-01 1.89350820e+00
8.88985157e-01 7.05714405e-01 3.50586116e-01 2.97615826e-01
9.11231875e-01 -6.81072995e-02 -4.90663230e-01 -2.73642123e-01
-1.53278157e-01 6.40002787e-01 8.45377445e-01 4.30035263e-01
-1.67388725e+00 1.80863166e+00 7.24079657e+00 1.02785015e+00
-7.87047327e-01 2.13608474e-01 7.55190372e-01 1.54540002e-01
4.97639440e-02 1.37837693e-01 -9.84314084e-01 2.32254446e-01
6.97433352e-01 2.67955780e-01 1.09335840e-01 1.13318706e+00
-3.01809877e-01 -7.35968173e-01 -7.58476377e-01 7.37429798e-01
3.34250510e-01 -1.24333644e+00 -2.39297330e-01 -3.80808592e-01
1.26960814e+00 1.68383062e-01 -9.78999436e-02 6.77535087e-02
3.42372149e-01 -1.03928661e+00 8.78136337e-01 2.65679598e-01
7.14824438e-01 -6.03329182e-01 6.15319073e-01 1.92534134e-01
-9.32283103e-01 5.67788258e-03 -1.25402406e-01 1.94679957e-03
-1.25470608e-01 6.06263340e-01 -1.01057196e+00 3.37389380e-01
9.72223759e-01 8.50596130e-01 -1.03714633e+00 1.28341603e+00
-3.57941538e-01 7.99021244e-01 -4.82395411e-01 -2.18863666e-01
4.98254389e-01 1.75411046e-01 6.21760309e-01 1.36532021e+00
-1.43879578e-01 -1.83457896e-01 2.82173336e-01 7.73873687e-01
2.02313691e-01 1.05206735e-01 -1.64183021e-01 1.61978956e-02
6.00877628e-02 1.18188572e+00 -1.76565516e+00 -7.24054337e-01
-8.16315711e-02 1.23616099e+00 1.60243958e-01 2.80329823e-01
-8.21318448e-01 -6.27432644e-01 8.06233883e-02 -3.21839470e-03
7.83624053e-01 -1.29457593e-01 -6.29416943e-01 -1.00601387e+00
-3.70301455e-01 -3.26521277e-01 3.39278042e-01 -5.89788198e-01
-9.07203138e-01 5.43920517e-01 1.23416677e-01 -8.96280944e-01
-1.79764599e-01 -5.92603207e-01 -3.03107381e-01 2.69741654e-01
-1.54330838e+00 -1.06654501e+00 -8.43755156e-02 2.70664841e-01
8.54055822e-01 -1.77345514e-01 6.74395442e-01 4.61907350e-02
-2.16853574e-01 2.55121678e-01 -1.00801654e-01 2.20898971e-01
2.67990917e-01 -1.71983886e+00 3.05560470e-01 8.41601253e-01
5.46561897e-01 3.01225223e-02 7.85794079e-01 -6.58791959e-01
-6.64492786e-01 -1.05597472e+00 7.17056036e-01 -6.45919621e-01
6.07661426e-01 -4.16425318e-01 -6.28993392e-01 4.48340178e-01
1.37803987e-01 3.63658369e-02 3.62495512e-01 -8.30884874e-02
4.09103669e-02 1.49971321e-01 -1.24100971e+00 2.64991283e-01
1.24946582e+00 -1.34605810e-01 -9.57238078e-01 3.85910600e-01
8.56993139e-01 -4.66716290e-01 -5.18437684e-01 6.11747801e-01
1.59122214e-01 -1.08062363e+00 8.49121869e-01 -2.91605145e-01
1.54860556e-01 -6.29419744e-01 1.18403867e-01 -7.47219920e-01
9.90414340e-03 -4.79568630e-01 1.47190705e-01 1.11126149e+00
5.77990055e-01 -2.44896337e-01 1.20144808e+00 -4.76606265e-02
-3.59474480e-01 -5.20554245e-01 -9.85188782e-01 -6.95743442e-01
-1.73534423e-01 -4.48777765e-01 1.34664699e-01 7.87168324e-01
-3.01651627e-01 2.66547173e-01 1.95187479e-01 -1.13721162e-01
8.98024499e-01 4.94639635e-01 3.42999786e-01 -1.33274841e+00
-1.26057610e-01 -4.75645751e-01 -7.33798265e-01 -7.56197631e-01
2.08609894e-01 -7.66707480e-01 4.86255318e-01 -1.56830931e+00
2.94128031e-01 -4.55519229e-01 -4.68710214e-01 7.07991421e-01
-1.63485602e-01 8.41825426e-01 8.58362392e-02 1.15155093e-01
-1.29574513e+00 -9.74204578e-03 9.96948421e-01 -1.79188251e-01
-3.95243406e-01 -5.97165786e-02 -7.76778698e-01 1.09399223e+00
8.19796503e-01 -4.89413410e-01 -2.09129855e-01 -1.06662750e-01
-3.99509370e-01 -7.04881072e-01 1.66467220e-01 -1.62834489e+00
5.18270545e-02 3.49303298e-02 5.02064466e-01 -9.81415629e-01
-4.00447622e-02 -6.84540629e-01 -2.40909487e-01 4.37136382e-01
-1.74304992e-01 -8.08512747e-01 2.30602577e-01 2.67284840e-01
-3.63665104e-01 -5.25129557e-01 9.88243639e-01 -4.02566522e-01
-1.69253469e+00 -5.00932969e-02 -4.06937480e-01 1.72886699e-01
1.19920683e+00 -5.57177424e-01 -1.82159524e-03 1.54581875e-01
-1.27085292e+00 3.39639395e-01 5.20626664e-01 5.20656705e-01
4.17749614e-01 -7.88501084e-01 -2.94714332e-01 8.51643980e-02
3.07992429e-01 2.26000682e-01 -7.86535069e-02 3.33370388e-01
-6.08940005e-01 4.70857978e-01 -3.59921128e-01 -1.02664316e+00
-1.10110986e+00 4.77668434e-01 9.56607535e-02 8.75703245e-02
-3.18014324e-01 9.91852462e-01 -5.13313152e-03 -3.08985621e-01
2.48026535e-01 -6.56635538e-02 -6.47326887e-01 2.64294505e-01
1.36574104e-01 2.85310477e-01 -1.44394845e-01 -7.95261919e-01
-4.63126868e-01 6.77823722e-01 4.31046169e-03 -9.74811614e-02
1.05168045e+00 9.88168735e-03 -6.54838011e-02 6.81861460e-01
1.00308681e+00 -3.02786708e-01 -1.37598979e+00 -6.07961826e-02
4.66264933e-01 -1.92793190e-01 -1.43227875e-02 -1.01752400e+00
-9.50156868e-01 5.41018426e-01 1.02912116e+00 2.04897970e-01
1.09238589e+00 6.70752227e-01 6.45297945e-01 2.64184564e-01
5.33631325e-01 -1.69881094e+00 2.72568077e-01 5.76775312e-01
2.85952538e-01 -1.50395048e+00 -2.84304097e-02 -9.23803091e-01
-6.22763276e-01 7.44153917e-01 6.20486259e-01 -6.03803217e-01
5.17733574e-01 2.18079351e-02 1.70353606e-01 -3.16840082e-01
1.05028391e-01 -1.17940187e+00 4.37433064e-01 7.30233669e-01
1.21204808e-01 2.88893640e-01 -4.78715628e-01 5.17416000e-01
-2.70020604e-01 3.52971107e-02 4.07086790e-01 1.17400408e+00
-1.20953596e+00 -8.43695581e-01 -2.99902856e-01 5.54903507e-01
-6.27420902e-01 -7.03311786e-02 -5.25849044e-01 6.11490011e-01
5.69630563e-01 1.07089627e+00 3.75963509e-01 -2.40632072e-01
2.50812396e-02 2.85964310e-01 4.64025557e-01 -1.04929745e+00
-5.32175303e-01 3.12780589e-01 1.10932440e-02 -5.40722132e-01
-9.24525857e-01 -6.89739645e-01 -1.74097824e+00 6.72395587e-01
-5.14081001e-01 1.83689609e-01 7.32343137e-01 1.03554106e+00
2.13324074e-02 5.45867920e-01 5.05024970e-01 -1.00318348e+00
2.53828973e-01 -1.01009488e+00 -6.17628992e-01 5.07657528e-01
4.59080338e-02 -7.49620199e-01 -3.25645953e-01 4.48490709e-01] | [9.532651901245117, 0.5669164657592773] |
2d8ad2c9-d739-491f-8a9e-7f29260617a7 | 3d-model-shapenet-core-classification-using | 2205.15869 | null | https://arxiv.org/abs/2205.15869v1 | https://arxiv.org/pdf/2205.15869v1.pdf | 3D-model ShapeNet Core Classification using Meta-Semantic Learning | Understanding 3D point cloud models for learning purposes has become an imperative challenge for real-world identification such as autonomous driving systems. A wide variety of solutions using deep learning have been proposed for point cloud segmentation, object detection, and classification. These methods, however, often require a considerable number of model parameters and are computationally expensive. We study a semantic dimension of given 3D data points and propose an efficient method called Meta-Semantic Learning (Meta-SeL). Meta-SeL is an integrated framework that leverages two input 3D local points (input 3D models and part-segmentation labels), providing a time and cost-efficient, and precise projection model for a number of 3D recognition tasks. The results indicate that Meta-SeL yields competitive performance in comparison with other complex state-of-the-art work. Moreover, being random shuffle invariant, Meta-SeL is resilient to translation as well as jittering noise. | ['Hamid R. Arabnia', 'Beshoy Morkos', 'M. Hadi Amini', 'Farzan Shenavarmasouleh', 'Cheng Chen', 'Farid Ghareh Mohammadi'] | 2022-05-28 | null | null | null | null | ['point-cloud-segmentation'] | ['computer-vision'] | [-3.19338180e-02 -2.69892395e-01 -1.65335432e-01 -5.16520023e-01
-8.71877432e-01 -6.52870595e-01 7.42186725e-01 5.07649891e-02
-2.81752378e-01 -1.90138206e-01 -5.90996563e-01 -3.64001900e-01
-5.78333996e-02 -8.06961060e-01 -1.13050807e+00 -4.49572265e-01
6.83103353e-02 1.04491389e+00 6.11229360e-01 -1.67606529e-02
4.84871596e-01 1.07814443e+00 -1.88894522e+00 1.07289813e-02
5.58998168e-01 1.17725801e+00 2.13045180e-01 3.06414038e-01
-5.42171061e-01 -7.62595749e-03 -4.55290288e-01 -3.79269421e-01
7.33646452e-01 2.52615839e-01 -6.71775877e-01 2.51548558e-01
6.08961403e-01 -1.61548764e-01 -9.72060189e-02 1.11425626e+00
3.43665600e-01 -2.15388015e-02 6.70458615e-01 -1.54602039e+00
-1.02365814e-01 -3.91750075e-02 -6.04157686e-01 -1.69705540e-01
-1.29814059e-01 1.71578258e-01 6.04904890e-01 -1.18926227e+00
4.32545662e-01 1.42562783e+00 1.00610018e+00 4.01002914e-01
-1.11981308e+00 -9.21851695e-01 1.99212357e-02 2.65286088e-01
-1.26651609e+00 -9.37346742e-02 9.19340730e-01 -5.89919746e-01
1.00703979e+00 1.26869500e-01 7.73325562e-01 8.56127918e-01
1.62478358e-01 9.22920287e-01 1.06349683e+00 -2.35376656e-01
4.74521250e-01 -6.86736032e-02 2.40114480e-01 5.50645351e-01
2.49338895e-01 1.45966381e-01 -4.62658525e-01 -6.00273199e-02
6.06653094e-01 2.02029675e-01 5.01542807e-01 -1.04660451e+00
-1.14574564e+00 6.62899375e-01 4.85005498e-01 -2.09282503e-01
-1.94904044e-01 2.66082138e-01 2.65293151e-01 1.39776185e-01
6.54740810e-01 1.93249270e-01 -5.34515738e-01 -2.01464564e-01
-9.26891744e-01 4.67436373e-01 6.59350812e-01 1.29501510e+00
1.13111353e+00 -2.86655992e-01 3.49708885e-01 7.80151546e-01
3.74991179e-01 8.96127641e-01 5.95393740e-02 -1.14605165e+00
3.84592235e-01 8.87352288e-01 -7.94284698e-03 -9.15839851e-01
-5.48052251e-01 -3.55375379e-01 -5.81341982e-01 6.76320791e-01
-8.41638446e-03 3.80449355e-01 -1.42165256e+00 1.31040311e+00
5.61554015e-01 3.81223083e-01 -6.42474294e-02 8.71471226e-01
8.62162054e-01 4.30594802e-01 -8.96274447e-02 4.98795241e-01
1.12734306e+00 -8.69145870e-01 2.79794987e-02 -5.07397294e-01
4.25865561e-01 -6.30918205e-01 8.30161512e-01 7.51111703e-03
-9.05486286e-01 -7.20100820e-01 -1.09071946e+00 -2.30442375e-01
-4.82861340e-01 -1.70027480e-01 6.66155875e-01 6.60886526e-01
-9.61843729e-01 5.77562332e-01 -1.27195227e+00 -4.47129905e-01
8.98123324e-01 6.38692975e-01 -4.82585460e-01 -4.37453836e-02
-4.30829704e-01 9.49412107e-01 1.80534959e-01 -1.25345096e-01
-8.85122836e-01 -7.72254407e-01 -7.43028283e-01 -2.59733170e-01
2.06510082e-01 -8.90603662e-01 1.35744345e+00 -3.21427673e-01
-1.47441196e+00 1.24797559e+00 -3.86544317e-01 -5.71034729e-01
6.71785951e-01 -4.42494690e-01 -1.15914340e-03 1.51504993e-01
2.27491319e-01 1.04915786e+00 1.01459050e+00 -1.49776006e+00
-7.19188154e-01 -7.83032060e-01 -2.16843575e-01 3.10512722e-01
2.89333194e-01 -2.49691337e-01 -7.80258834e-01 -1.15953103e-01
8.68614018e-01 -1.29201353e+00 -3.41408014e-01 8.69954750e-02
-4.35990989e-01 -3.13149035e-01 1.17528975e+00 -1.55869290e-01
1.29910111e-01 -2.16124392e+00 -1.23570912e-01 2.07205102e-01
3.62241775e-01 3.74900430e-01 -3.08915582e-02 2.05067500e-01
1.49059087e-01 1.50228813e-01 -4.15210128e-01 -6.79841220e-01
3.32956463e-01 2.27071702e-01 -3.25305700e-01 6.39025629e-01
2.79836267e-01 1.14867187e+00 -6.64614856e-01 -3.39782119e-01
7.44440913e-01 3.84615332e-01 -4.03142422e-01 3.01014390e-02
-2.95168251e-01 3.90482575e-01 -4.39249426e-01 9.37920034e-01
1.03634536e+00 -2.36242875e-01 -5.32539129e-01 6.76241592e-02
5.91548905e-02 3.24393004e-01 -9.77730334e-01 1.92552733e+00
-1.45526990e-01 5.67878485e-01 -1.51549414e-01 -9.41398382e-01
1.27357292e+00 -1.96794510e-01 6.97384775e-01 -6.31184518e-01
2.15261802e-01 3.92919928e-01 -3.40310514e-01 -1.64934397e-01
7.22341776e-01 -7.77608901e-02 -1.37404144e-01 3.50932956e-01
2.52387188e-02 -6.89480186e-01 -3.26174796e-01 -1.19279437e-01
1.02407205e+00 3.28842372e-01 -1.29065484e-01 -1.11340687e-01
2.21086487e-01 3.91805559e-01 4.99916911e-01 8.34882736e-01
-4.26753879e-01 7.47138560e-01 1.90649316e-01 -4.45298284e-01
-1.18733311e+00 -1.04317594e+00 -1.60321862e-01 4.69629377e-01
8.03835332e-01 -1.77534044e-01 -7.54447281e-01 -5.78461587e-01
6.37493193e-01 5.77404201e-01 -2.62205392e-01 -1.80115834e-01
-5.93024671e-01 -4.31325376e-01 3.67031127e-01 5.99126160e-01
5.61670184e-01 -7.36721933e-01 -8.16420376e-01 4.83056270e-02
2.62591422e-01 -1.39229321e+00 -6.34758472e-02 2.25158215e-01
-1.36572909e+00 -9.22801435e-01 -3.77369612e-01 -6.94984674e-01
6.01724088e-01 7.75618076e-01 1.21292949e+00 -2.32521623e-01
-1.66881427e-01 4.36733276e-01 -2.38849044e-01 -7.56529450e-01
-2.76328802e-01 3.26178998e-01 6.99662641e-02 -9.90995690e-02
1.00739968e+00 -6.19314909e-01 -4.67276722e-01 5.68592012e-01
-5.60327232e-01 8.91367123e-02 7.73863435e-01 5.38826168e-01
1.11646235e+00 -2.01332152e-01 9.84962285e-02 -7.35576570e-01
1.91820383e-01 -2.16075480e-01 -9.40505266e-01 -1.53412893e-01
-4.10016298e-01 -4.77770641e-02 9.91619676e-02 -8.70271958e-03
-5.36641479e-01 3.72091472e-01 -3.14479798e-01 -1.07046056e+00
-6.02496326e-01 -3.38708386e-02 -2.51652658e-01 -4.51285243e-01
3.74149978e-01 2.21030071e-01 2.76457816e-01 -7.40983486e-01
4.54384208e-01 6.25491917e-01 5.69859445e-01 -4.54903305e-01
1.08733869e+00 8.14127088e-01 3.73101741e-01 -9.05321956e-01
-6.17451251e-01 -8.31452191e-01 -1.14868903e+00 -1.05540074e-01
8.06098223e-01 -9.40105081e-01 -6.97972536e-01 8.64374340e-01
-1.23783553e+00 -1.82997048e-01 -3.12097102e-01 3.15577388e-01
-7.95149624e-01 2.40536585e-01 -1.69626012e-01 -6.11818314e-01
-2.29735687e-01 -1.47001410e+00 1.55672705e+00 -6.51055872e-02
8.63569081e-02 -6.41742408e-01 -1.38691083e-01 4.93188024e-01
1.05413489e-01 3.74719948e-01 9.33432996e-01 -7.02084363e-01
-1.19888484e+00 -4.44173515e-01 -3.48651767e-01 1.93942145e-01
-8.88681114e-02 -1.96546018e-01 -1.00855565e+00 -1.71662107e-01
-2.07195035e-03 -1.42084882e-01 8.14316332e-01 3.01850885e-01
1.18360436e+00 2.66944110e-01 -5.91717303e-01 9.59981918e-01
1.15489805e+00 9.06072110e-02 3.07696015e-01 5.39827824e-01
9.10534322e-01 5.01911700e-01 7.12069869e-01 7.77695999e-02
6.77802801e-01 6.11803412e-01 7.73283005e-01 -1.65662363e-01
-6.45417944e-02 -3.37215960e-01 -8.11699480e-02 8.35076332e-01
2.31036976e-01 1.15992315e-01 -1.31852829e+00 6.32399738e-01
-1.78824437e+00 -7.29264200e-01 -2.04719350e-01 2.08969593e+00
1.89560816e-01 4.46772158e-01 8.21349323e-02 1.50563255e-01
6.11582875e-01 1.25625744e-01 -1.12503517e+00 -4.93446849e-02
-9.57015455e-02 2.52737463e-01 9.35995638e-01 1.38479441e-01
-1.31465662e+00 1.18310142e+00 6.43135881e+00 7.06747770e-01
-1.20141923e+00 1.71316162e-01 3.89883757e-01 7.81868920e-02
-2.06365079e-01 1.60742383e-02 -9.72142577e-01 2.55123913e-01
6.75710559e-01 1.02995723e-01 7.50145167e-02 1.22852933e+00
-1.83380201e-01 7.11852759e-02 -1.17236829e+00 1.40353703e+00
-2.36165952e-02 -1.41412032e+00 1.01691455e-01 2.29720503e-01
6.18162036e-01 7.97239542e-01 7.48784319e-02 1.10495865e-01
1.96164504e-01 -8.39632869e-01 1.10346007e+00 3.27559471e-01
5.70209086e-01 -9.02166724e-01 6.00194454e-01 6.09062731e-01
-9.84458268e-01 8.55380446e-02 -5.03917515e-01 1.99018210e-01
1.41498536e-01 6.21941805e-01 -8.54005516e-01 4.88872260e-01
9.93969142e-01 8.53217661e-01 -6.20109320e-01 1.28629315e+00
1.68067552e-02 3.91251564e-01 -7.04909146e-01 1.02104209e-01
3.48845720e-01 -2.03147903e-01 8.56269121e-01 8.97429466e-01
3.43038052e-01 -1.88983962e-01 2.32809335e-01 1.03101408e+00
-1.43158376e-01 -2.43823260e-01 -7.37842143e-01 1.00852869e-01
6.53953731e-01 9.41241145e-01 -1.02652752e+00 -1.10512063e-01
-3.40760857e-01 7.50799000e-01 8.50961208e-02 4.15074155e-02
-5.26229322e-01 -2.41745442e-01 1.10545099e+00 -4.81337681e-02
5.28367400e-01 -7.31577635e-01 -8.41842175e-01 -7.26915061e-01
1.98417023e-01 -5.64136326e-01 -1.31207630e-01 -6.04481757e-01
-1.14419961e+00 5.23439407e-01 1.96656272e-01 -1.53885150e+00
4.95301746e-03 -8.31500113e-01 -1.61128402e-01 6.73899591e-01
-1.77463567e+00 -1.36042285e+00 -4.24302638e-01 4.35572326e-01
6.03328645e-01 -2.15059459e-01 5.07908463e-01 2.09105968e-01
-1.53619841e-01 3.63261253e-01 4.41345107e-03 -7.42786825e-02
2.52294898e-01 -1.15268600e+00 1.28586936e+00 7.97275126e-01
3.60402554e-01 3.63927573e-01 3.98223728e-01 -5.71872652e-01
-1.77676547e+00 -1.30248797e+00 5.93586028e-01 -8.44301224e-01
1.73677146e-01 -6.85060978e-01 -8.38717163e-01 6.04650497e-01
-4.45296168e-01 1.15363792e-01 4.33975607e-01 -6.99234679e-02
-3.10631126e-01 -2.22802460e-01 -1.13670194e+00 2.91564226e-01
1.40089357e+00 -6.47120178e-01 -7.31173694e-01 2.69773513e-01
9.35027003e-01 -1.00723112e+00 -5.56683123e-01 7.15571761e-01
3.25729519e-01 -1.14641643e+00 1.26523495e+00 -3.06556910e-01
8.80033821e-02 -4.99616444e-01 -2.59836137e-01 -9.70213711e-01
-2.82655507e-01 -4.33776200e-01 -1.46692172e-01 7.82616735e-01
2.14469433e-01 -6.62629247e-01 1.28504312e+00 5.79760849e-01
-6.24179304e-01 -5.90350389e-01 -1.12627888e+00 -1.01095271e+00
1.29500434e-01 -1.04716396e+00 1.12553465e+00 6.68985128e-01
-7.78639376e-01 5.25827892e-02 9.92654935e-02 4.95323986e-01
9.56627190e-01 5.56829691e-01 1.35541642e+00 -1.64156032e+00
2.72225559e-01 -6.37564421e-01 -9.75793004e-01 -1.33095169e+00
2.97926724e-01 -1.06199825e+00 -1.04458667e-02 -1.54668260e+00
-2.15258703e-01 -8.59112322e-01 -7.99626932e-02 5.16575933e-01
2.11144015e-01 3.70235920e-01 2.66164124e-01 6.32578492e-01
-5.16463578e-01 6.01009786e-01 1.04068053e+00 -2.41507053e-01
-2.50913084e-01 4.59066808e-01 -2.45981440e-01 8.68008673e-01
6.95286810e-01 -5.77888727e-01 -2.21165746e-01 -7.32179165e-01
-9.59136933e-02 -3.69724423e-01 6.31783903e-01 -1.20234549e+00
3.58790517e-01 -9.24501345e-02 2.61401802e-01 -1.50555015e+00
7.61317134e-01 -9.77079511e-01 1.57637432e-01 3.27267289e-01
1.81715310e-01 1.15135953e-01 4.28589135e-01 6.11921072e-01
-2.02940002e-01 5.93595486e-03 7.96455324e-01 -2.22163364e-01
-1.10038412e+00 6.24134064e-01 1.56910360e-01 -1.51492417e-01
1.21528625e+00 -7.79306352e-01 -3.89291309e-02 1.12350345e-01
-2.78123647e-01 1.82860047e-01 7.80328393e-01 7.68123507e-01
7.45855629e-01 -1.21403682e+00 -4.47241277e-01 6.29032075e-01
3.51641655e-01 6.67362273e-01 7.96333551e-02 5.92807949e-01
-7.97810614e-01 5.43033242e-01 -1.54844359e-01 -1.51596582e+00
-1.00271249e+00 3.83424789e-01 2.78815329e-01 3.27963740e-01
-1.10325003e+00 9.27951992e-01 -5.99020310e-02 -9.91329610e-01
2.32574940e-01 -5.09591758e-01 3.06149483e-01 -1.61124796e-01
-2.15927325e-02 3.86302024e-01 4.22501385e-01 -9.38246906e-01
-6.17189229e-01 1.08609664e+00 1.79687459e-02 3.15368362e-02
1.37964332e+00 1.13756647e-02 -3.38651496e-03 4.89391595e-01
1.14070785e+00 -6.05889201e-01 -1.58895600e+00 -3.54209423e-01
3.45121890e-01 -6.68367505e-01 8.90854001e-02 -4.32084829e-01
-9.41579700e-01 1.15392780e+00 7.57207751e-01 -3.79037410e-02
6.98679209e-01 1.41747057e-01 9.69506860e-01 5.47042310e-01
7.97974348e-01 -8.84025514e-01 -2.34096661e-01 7.03182399e-01
5.08847475e-01 -1.41287684e+00 -5.09391800e-02 -4.33744639e-01
-3.80802512e-01 8.51498663e-01 4.53968734e-01 -2.03119278e-01
7.94840038e-01 1.20589249e-01 2.83682793e-01 -3.78371865e-01
-3.62727672e-01 -2.93460488e-01 3.23781550e-01 8.48858297e-01
-4.57986772e-01 7.87345916e-02 4.17149842e-01 2.81727254e-01
-4.75090832e-01 -2.14789957e-01 -7.74069503e-02 1.04596126e+00
-4.28177893e-01 -1.06741309e+00 -3.89220387e-01 3.69300365e-01
4.01333198e-02 3.38106662e-01 -4.82662767e-01 8.04010332e-01
1.27230421e-01 6.15175962e-01 2.73519099e-01 -6.36798441e-01
4.74545211e-01 1.75649866e-01 3.09236825e-01 -5.30543983e-01
-3.51366937e-01 -1.54921323e-01 -4.32393759e-01 -7.97951698e-01
-6.06390119e-01 -7.97213912e-01 -1.24210608e+00 -2.70326704e-01
-2.75224417e-01 -2.70144850e-01 1.25303745e+00 1.05755687e+00
9.46968496e-01 1.09212473e-01 6.76119745e-01 -1.38774574e+00
-6.26657903e-01 -6.10524356e-01 -4.14420873e-01 3.69766891e-01
4.10802335e-01 -9.30316567e-01 -2.24340260e-01 -1.57369152e-01] | [7.976344585418701, -3.1036345958709717] |
0e4cd476-039c-4e21-af88-839ae7435fc0 | optimized-views-photogrammetry-precision | 2206.12216 | null | https://arxiv.org/abs/2206.12216v1 | https://arxiv.org/pdf/2206.12216v1.pdf | Optimized Views Photogrammetry: Precision Analysis and A Large-scale Case Study in Qingdao | UAVs have become one of the widely used remote sensing platforms and played a critical role in the construction of smart cities. However, due to the complex environment in urban scenes, secure and accurate data acquisition brings great challenges to 3D modeling and scene updating. Optimal trajectory planning of UAVs and accurate data collection of onboard cameras are non-trivial issues in urban modeling. This study presents the principle of optimized views photogrammetry and verifies its precision and potential in large-scale 3D modeling. Different from oblique photogrammetry, optimized views photogrammetry uses rough models to generate and optimize UAV trajectories, which is achieved through the consideration of model point reconstructability and view point redundancy. Based on the principle of optimized views photogrammetry, this study first conducts a precision analysis of 3D models by using UAV images of optimized views photogrammetry and then executes a large-scale case study in the urban region of Qingdao city, China, to verify its engineering potential. By using GCPs for image orientation precision analysis and TLS (terrestrial laser scanning) point clouds for model quality analysis, experimental results show that optimized views photogrammetry could construct stable image connection networks and could achieve comparable image orientation accuracy. Benefiting from the accurate image acquisition strategy, the quality of mesh models significantly improves, especially for urban areas with serious occlusions, in which 3 to 5 times of higher accuracy has been achieved. Besides, the case study in Qingdao city verifies that optimized views photogrammetry can be a reliable and powerful solution for the large-scale 3D modeling in complex urban scenes. | ['San Jiang', 'Wenshuai Yu', 'Qingquan Li'] | 2022-06-24 | null | null | null | null | ['trajectory-planning'] | ['robots'] | [-2.33776733e-01 -4.80394542e-01 9.54398140e-02 -1.73206069e-03
-6.93520755e-02 -2.86926001e-01 2.71399051e-01 -2.64884949e-01
-3.60809825e-02 4.94173318e-01 -5.54723918e-01 -5.97434223e-01
-4.13414717e-01 -1.57729328e+00 -4.23186511e-01 -6.45892620e-01
1.69615299e-01 7.07343936e-01 2.36706749e-01 -6.00011587e-01
9.68546048e-03 1.14348185e+00 -1.44257438e+00 -4.81233835e-01
1.09271264e+00 7.43377030e-01 6.19812667e-01 2.72752970e-01
1.03336409e-01 -2.61327177e-01 -1.74669355e-01 -1.80836618e-01
5.97330272e-01 2.73142129e-01 -3.84290457e-01 6.69168651e-01
9.74337533e-02 -7.28431702e-01 -3.41768950e-01 9.35773790e-01
3.97885382e-01 -1.45055473e-01 2.74820298e-01 -1.39766419e+00
-3.62486571e-01 -7.70696625e-02 -6.44560695e-01 -3.77374113e-01
5.68791032e-02 2.20933452e-01 6.14983201e-01 -8.98708761e-01
4.43667501e-01 8.96286845e-01 7.10895717e-01 -2.70195544e-01
-8.27467084e-01 -6.11282468e-01 -1.65805981e-01 2.19263539e-01
-1.91556871e+00 -1.79295197e-01 5.76665223e-01 -4.65664655e-01
7.39327490e-01 5.01246750e-01 1.34488142e+00 -4.92113233e-02
2.24364281e-01 -8.24887305e-02 8.34925473e-01 -3.23819041e-01
-6.21688552e-02 -8.32973421e-02 -1.69428766e-01 6.69056475e-01
6.29800141e-01 4.86024380e-01 3.14521551e-01 1.46601181e-02
1.16439068e+00 4.17737693e-01 -4.67345446e-01 -9.98078659e-02
-1.23839521e+00 6.03441119e-01 8.34371030e-01 2.07238853e-01
-5.17502487e-01 3.05553041e-02 -2.52388477e-01 4.38241586e-02
4.73878711e-01 -4.70116064e-02 -3.19278359e-01 1.40742883e-01
-8.23314607e-01 -6.71926886e-02 -6.70909211e-02 1.28148782e+00
8.72602463e-01 4.27105457e-01 7.95554221e-01 6.92614555e-01
5.64734519e-01 1.26639855e+00 -2.25164995e-01 -1.01073146e+00
6.77505314e-01 9.39082086e-01 1.07205525e-01 -1.34788179e+00
-3.92593890e-01 -2.59316772e-01 -1.05157590e+00 3.14053625e-01
-6.19416498e-02 2.57898476e-02 -5.30416310e-01 7.70016789e-01
5.40748775e-01 -6.78968281e-02 -7.21183792e-02 7.45851159e-01
5.96672177e-01 9.43915129e-01 -3.45474184e-01 -3.81953090e-01
1.23049247e+00 -3.62819850e-01 -2.66763180e-01 9.58056748e-02
4.49792504e-01 -8.73118520e-01 9.48665738e-01 1.61796480e-01
-6.92669690e-01 -4.63486344e-01 -8.30397308e-01 3.56972039e-01
1.62106585e-02 4.21357483e-01 5.22036970e-01 6.04136586e-01
-8.91306520e-01 1.80022940e-01 -7.31258273e-01 -6.07029915e-01
4.29056644e-01 1.50588319e-01 -3.61592829e-01 -4.31741327e-01
-9.80805874e-01 8.80290031e-01 1.76269755e-01 2.66268015e-01
-2.65841872e-01 -6.55962348e-01 -6.34368300e-01 6.34452775e-02
3.51803713e-02 -8.37754309e-01 6.63818002e-01 -3.22066963e-01
-1.24278378e+00 5.89527845e-01 -1.00271724e-01 -8.21363330e-02
3.02973181e-01 2.74855614e-01 -5.50335228e-01 2.72959858e-01
2.36322865e-01 2.62439847e-01 3.85304034e-01 -1.42960680e+00
-6.12466156e-01 -5.45286596e-01 5.83659932e-02 2.35838637e-01
-9.31187645e-02 -2.32025981e-01 -2.54354477e-01 -5.52607998e-02
7.64026821e-01 -1.07848358e+00 -5.41245639e-01 1.26215070e-01
-1.28344417e-01 2.44797751e-01 1.25954032e+00 -5.57573855e-01
1.01486933e+00 -1.91750109e+00 -5.12646973e-01 5.07180095e-01
1.70482155e-02 3.86233926e-01 1.43393382e-01 6.40461385e-01
1.29372239e-01 3.20087612e-01 -1.14808701e-01 2.09263176e-01
-5.30183792e-01 4.13433373e-01 -1.11204885e-01 5.22467315e-01
-4.11047757e-01 7.19910264e-01 -6.21577621e-01 -3.56594920e-01
7.86495686e-01 4.89128351e-01 -8.03247169e-02 -3.79751205e-01
8.83129314e-02 4.52518523e-01 -8.24728847e-01 8.92419875e-01
9.52893913e-01 -1.32065877e-01 -1.63527820e-02 -3.15109849e-01
-6.47664666e-01 -3.82414341e-01 -1.17243886e+00 1.13442278e+00
-7.15806484e-01 5.61095834e-01 1.52332425e-01 -5.62177181e-01
1.30936587e+00 4.10181075e-01 7.14088321e-01 -6.76617682e-01
8.49880800e-02 2.50124305e-01 -3.90447736e-01 -6.03738248e-01
7.20009446e-01 -7.35390708e-02 1.82134002e-01 1.61209777e-01
-8.77298892e-01 -1.12933373e+00 -1.95679232e-01 8.48232284e-02
2.15279832e-01 -5.87552823e-02 3.46159995e-01 -9.23543051e-02
4.34708595e-01 6.79984152e-01 3.97334546e-01 -2.71654222e-02
2.02093184e-01 4.12647873e-01 -3.55944157e-01 -7.94580400e-01
-1.25554025e+00 -8.10718238e-01 -4.00994807e-01 -3.60491842e-01
6.40075326e-01 -3.12691152e-01 -3.23160231e-01 -2.74424106e-02
-1.04613118e-01 6.47846699e-01 2.98895359e-01 4.49846745e-01
-3.65527689e-01 -8.56065810e-01 1.87754467e-01 1.24902181e-01
1.11627269e+00 -1.27563849e-01 -6.34016752e-01 1.28582492e-01
-2.98360169e-01 -1.18211150e+00 -1.47159230e-02 -9.60415661e-01
-1.25584328e+00 -1.40311265e+00 -3.40598822e-01 -3.19204330e-01
9.93207276e-01 1.39039910e+00 7.79901266e-01 6.99151039e-01
-5.93086630e-02 5.65876186e-01 -3.50728452e-01 -4.45594937e-01
-3.33187342e-01 -3.19268316e-01 1.21478125e-01 -2.82681465e-01
-2.07031399e-01 -8.84082079e-01 -4.50394064e-01 9.33050096e-01
-6.21505558e-01 3.04487705e-01 3.14655423e-01 3.30697298e-01
8.67793322e-01 8.71660590e-01 1.98867887e-01 -3.38761806e-01
7.13376747e-03 -2.19929382e-01 -1.39605653e+00 1.81517318e-01
-6.18239760e-01 -9.07820642e-01 5.01530170e-01 3.75830978e-01
-9.22676206e-01 2.12249801e-01 -9.84744877e-02 -3.54059726e-01
-1.30538866e-01 6.36393785e-01 -2.87330806e-01 -4.00811374e-01
3.95305693e-01 1.55537665e-01 1.79714352e-01 -1.86335102e-01
1.46408647e-01 8.21562588e-01 2.07563620e-02 -2.74338722e-01
1.31410444e+00 9.79018152e-01 5.27976453e-01 -1.56441939e+00
2.25097197e-03 -3.13273460e-01 -9.06315327e-01 -7.56309927e-01
7.09733129e-01 -1.06595588e+00 -6.37151301e-01 5.72099686e-01
-1.15042114e+00 -4.27852049e-02 -5.95229194e-02 8.03561091e-01
-2.47273326e-01 4.62616146e-01 -8.55239928e-02 -6.82553649e-01
-1.58413649e-01 -1.41373110e+00 9.17939723e-01 -9.94159002e-03
2.61726469e-01 -8.04386318e-01 -4.31888640e-01 6.45687521e-01
3.94610792e-01 3.89327943e-01 8.20763350e-01 5.15526652e-01
-1.40742886e+00 -2.96607494e-01 -2.21401095e-01 -1.91846918e-02
2.48259649e-01 6.50773346e-01 -5.05591452e-01 -1.29489571e-01
-7.83849061e-02 4.92011875e-01 1.92197904e-01 4.95274603e-01
6.09756052e-01 -7.58913830e-02 -4.73933071e-01 7.63285220e-01
1.83731496e+00 3.93348753e-01 9.69668210e-01 5.81892848e-01
6.50540113e-01 4.87661511e-01 1.04051507e+00 3.67294550e-01
6.67770267e-01 7.69476593e-01 1.06977963e+00 -1.05253577e-01
8.32681805e-02 -2.10953802e-01 3.17336060e-02 1.01945388e+00
-7.44693398e-01 1.10359617e-01 -1.01761484e+00 3.95621121e-01
-1.45879352e+00 -1.13626587e+00 -1.09497654e+00 2.44688535e+00
-6.85754716e-02 -5.07202744e-01 -2.42615640e-01 3.24257135e-01
8.27222168e-01 1.55214965e-01 -2.77930107e-02 1.34619221e-01
-2.75091380e-02 -1.89795867e-01 7.93837070e-01 7.16399550e-01
-5.82395196e-01 8.50803852e-01 4.86606884e+00 7.08831847e-01
-1.40548301e+00 6.01164065e-02 1.48073379e-02 2.67789274e-01
-5.87783217e-01 3.54951620e-01 -8.25044215e-01 3.01537335e-01
5.15960097e-01 -1.48387134e-01 2.96461731e-01 7.93966234e-01
9.81258512e-01 -4.02192622e-01 -5.78529127e-02 1.04885781e+00
-4.39786881e-01 -1.73585308e+00 2.22990498e-01 5.18048882e-01
9.43809927e-01 3.05690169e-01 -1.99493602e-01 -5.39931297e-01
1.90843567e-01 -5.92566729e-01 6.67095184e-01 4.02947456e-01
1.04851997e+00 -7.89558113e-01 4.80885327e-01 7.67117620e-01
-1.63319230e+00 1.55563995e-01 -6.92609191e-01 -2.15270191e-01
6.78516388e-01 9.75171864e-01 -8.77496183e-01 1.27170825e+00
5.90117693e-01 8.41013908e-01 -1.16278902e-01 1.09910154e+00
-5.32127142e-01 3.15200567e-01 -5.66542327e-01 1.90347195e-01
1.60027564e-01 -9.27998781e-01 6.40553057e-01 8.55288282e-02
1.04350364e+00 5.46989501e-01 2.04760149e-01 6.20027721e-01
3.19085151e-01 1.03980377e-01 -1.13085651e+00 4.50986773e-01
9.05308008e-01 1.22653210e+00 -5.86525023e-01 -1.21505193e-01
-3.81106347e-01 1.21947356e-01 -4.33932424e-01 3.24191302e-01
-8.82441163e-01 -5.28187305e-02 4.55938488e-01 6.24974668e-01
-1.15096301e-01 -9.80490923e-01 -4.86633450e-01 -1.04047823e+00
-2.23501876e-01 -2.13243350e-01 -2.14206144e-01 -1.25969815e+00
-5.23434043e-01 6.28893554e-01 2.81561673e-01 -1.86780775e+00
1.30039901e-01 -2.92179883e-01 -7.56859303e-01 8.73401999e-01
-1.68998516e+00 -1.60376275e+00 -6.13924325e-01 6.75183952e-01
6.08882189e-01 -2.68954393e-02 7.16573417e-01 1.64132550e-01
-5.22192717e-01 -4.47285801e-01 3.54744285e-01 -3.27916205e-01
-5.38422018e-02 -2.93667138e-01 3.33525985e-01 1.03822947e+00
-6.81530535e-02 2.61597842e-01 2.15490565e-01 -8.99269462e-01
-1.42891276e+00 -1.43516231e+00 8.93154383e-01 -8.65167081e-02
3.37593287e-01 3.23266178e-01 -3.75836521e-01 7.27530718e-01
-1.47329956e-01 -9.01688859e-02 1.43488422e-01 -5.29994488e-01
3.94665331e-01 -3.91702235e-01 -1.27490127e+00 5.31780362e-01
9.28742111e-01 -2.63022687e-02 -1.38964295e-01 6.51227415e-01
5.46732783e-01 -3.47089201e-01 -9.70928133e-01 4.62080121e-01
5.02810597e-01 -1.08287776e+00 1.12152815e+00 1.77257225e-01
3.43637727e-02 -6.80830419e-01 -3.33997786e-01 -1.35127378e+00
-2.85290092e-01 -2.50568062e-01 8.49452257e-01 1.05881786e+00
1.36507750e-01 -1.10525107e+00 7.01157928e-01 4.95797873e-01
-3.89908403e-01 -4.96439099e-01 -1.01564336e+00 -8.84078860e-01
-3.52255702e-01 -7.16646612e-01 1.13011146e+00 9.33376729e-01
-5.79287529e-01 6.93820789e-02 -6.24770001e-02 1.02066517e+00
6.97032213e-01 3.38767678e-01 1.17664850e+00 -1.64636874e+00
4.80780333e-01 -4.15099971e-02 -4.35278386e-01 -9.01348412e-01
1.07171396e-02 -6.10722125e-01 -5.86505175e-01 -2.04131103e+00
-4.72680181e-01 -1.10206771e+00 1.08795166e+00 8.99406672e-02
4.33024526e-01 2.14837849e-01 2.56591678e-01 7.86331296e-01
3.78913164e-01 6.06522202e-01 1.59250784e+00 -1.96979240e-01
-1.32772908e-01 5.13472497e-01 -1.01914510e-01 8.90615344e-01
7.82903850e-01 -8.27632248e-02 -6.75886452e-01 -9.46791828e-01
4.30810869e-01 2.94772387e-01 5.67960083e-01 -9.40304756e-01
2.85054803e-01 -5.68592131e-01 -2.78634578e-02 -1.09262919e+00
6.45814657e-01 -1.66074646e+00 8.86208475e-01 7.27821946e-01
1.04890645e+00 3.03101301e-01 8.93435776e-02 2.12586135e-01
-1.40925959e-01 -1.01511747e-01 7.27467060e-01 -3.60872775e-01
-9.27357733e-01 7.18783617e-01 -3.49211514e-01 -6.40021086e-01
1.46502423e+00 -7.82017231e-01 -2.77617782e-01 -4.20949042e-01
-5.40397286e-01 3.01913023e-01 9.47598636e-01 -3.09565634e-01
1.03177643e+00 -1.45741189e+00 -7.36116707e-01 4.94067460e-01
-8.85714144e-02 4.49505359e-01 5.43642879e-01 8.92005026e-01
-1.14129663e+00 4.38973993e-01 -1.07562974e-01 -9.66074765e-01
-1.49775863e+00 6.38922025e-03 2.99011409e-01 1.82221144e-01
-5.37009835e-01 2.16911688e-01 -2.63457477e-01 -5.67708015e-01
-7.70169556e-01 -4.50803995e-01 -4.58981134e-02 -1.45961538e-01
2.02890188e-01 7.09901869e-01 2.88696527e-01 -1.24477923e+00
-1.57866403e-01 1.35405731e+00 7.77656674e-01 6.37803078e-02
1.52732408e+00 -6.96893692e-01 -3.08240414e-01 -7.83630833e-02
5.93139291e-01 3.65204722e-01 -9.53161180e-01 1.38222009e-01
-6.88319206e-01 -1.00037873e+00 2.92570889e-01 -2.41264716e-01
-1.39481294e+00 1.01814318e+00 1.01312853e-01 2.38520697e-01
1.09757781e+00 -2.33096927e-01 9.87087548e-01 2.58337587e-01
1.28732586e+00 -5.95541716e-01 -5.77648997e-01 5.74963570e-01
8.99290681e-01 -8.96801710e-01 3.32178295e-01 -9.72056210e-01
-3.34447056e-01 1.42204225e+00 3.36982608e-01 1.36674166e-01
7.31634378e-01 -1.17738098e-01 1.09280562e-02 -3.23559433e-01
3.20201144e-02 -1.12739541e-01 -2.11325914e-01 8.62512708e-01
-2.92873979e-01 2.52801001e-01 -6.27672598e-02 -5.53717539e-02
-3.12991023e-01 -3.34074683e-02 6.86859906e-01 7.20742762e-01
-6.98041439e-01 -1.13730657e+00 -1.01384783e+00 3.42290580e-01
4.60570395e-01 -1.29197547e-02 2.83978015e-01 1.39051306e+00
-3.74485105e-02 9.85544622e-01 9.90180597e-02 -5.12789071e-01
7.16580391e-01 -5.76079369e-01 3.22067589e-01 -4.35185552e-01
-1.20227233e-01 -6.77059144e-02 1.04944348e-01 -3.90296876e-01
-6.21169567e-01 -5.45349121e-01 -1.34140086e+00 -8.78445625e-01
-7.12709308e-01 2.44696677e-01 8.93361628e-01 8.25598419e-01
4.50816005e-01 2.90034013e-03 1.09919214e+00 -9.81643498e-01
-1.55068085e-01 -4.71207708e-01 -9.62341487e-01 -1.21287040e-01
-2.56336331e-01 -6.26616418e-01 -9.80757400e-02 -5.95583282e-02] | [8.34218692779541, -2.669382095336914] |
f858a6c7-2a5b-4987-879b-9e526adc7038 | ptw-pivotal-tuning-watermarking-for-pre | 2304.07361 | null | https://arxiv.org/abs/2304.07361v2 | https://arxiv.org/pdf/2304.07361v2.pdf | PTW: Pivotal Tuning Watermarking for Pre-Trained Image Generators | Deepfakes refer to content synthesized using deep generators, which, when misused, have the potential to erode trust in digital media. Synthesizing high-quality deepfakes requires access to large and complex generators only a few entities can train and provide. The threat is malicious users that exploit access to the provided model and generate harmful deepfakes without risking detection. Watermarking makes deepfakes detectable by embedding an identifiable code into the generator that is later extractable from its generated images. We propose Pivotal Tuning Watermarking (PTW), a method for watermarking pre-trained generators (i) three orders of magnitude faster than watermarking from scratch and (ii) without the need for any training data. We improve existing watermarking methods and scale to generators $4 \times$ larger than related work. PTW can embed longer codes than existing methods while better preserving the generator's image quality. We propose rigorous, game-based definitions for robustness and undetectability and our study reveals that watermarking is not robust against an adaptive white-box attacker who has control over the generator's parameters. We propose an adaptive attack that can successfully remove any watermarking with access to only $200$ non-watermarked images. Our work challenges the trustworthiness of watermarking for deepfake detection when the parameters of a generator are available. Source code to reproduce our experiments is available at https://github.com/dnn-security/gan-watermark. | ['Florian Kerschbaum', 'Nils Lukas'] | 2023-04-14 | null | null | null | null | ['face-swapping'] | ['computer-vision'] | [ 3.85226727e-01 6.53257430e-01 -2.02978075e-01 4.94946390e-01
-1.08700633e+00 -1.43118441e+00 5.95278144e-01 -1.97472155e-01
-3.91843438e-01 5.75630903e-01 1.78229511e-02 -7.55645633e-01
4.59273309e-01 -9.33796287e-01 -1.08256495e+00 -6.02965593e-01
-3.73315632e-01 -3.17816615e-01 2.83301204e-01 -2.04853475e-01
6.58394620e-02 2.37765536e-01 -8.27673435e-01 1.16573460e-01
4.28413779e-01 7.63890922e-01 -9.76293758e-02 1.18684757e+00
7.38622963e-01 6.13500834e-01 -1.20292532e+00 -1.20233977e+00
9.65183675e-01 -4.44508582e-01 -3.92199367e-01 -3.46268594e-01
2.13679716e-01 -1.04052460e+00 -6.29573941e-01 1.38355160e+00
8.08292508e-01 -6.26580477e-01 3.16164643e-01 -1.78754723e+00
-9.88685548e-01 1.23885131e+00 -4.70062673e-01 -1.25596941e-01
7.19904229e-02 7.22092271e-01 9.39091802e-01 -3.73662144e-01
7.82092869e-01 8.74044240e-01 7.14404404e-01 1.17662323e+00
-1.05944490e+00 -1.22638929e+00 -3.75623912e-01 -3.54888856e-01
-1.46903265e+00 -8.31278741e-01 6.37714684e-01 -3.74863029e-01
7.43399799e-01 3.96616966e-01 4.03301805e-01 1.61141205e+00
1.53836042e-01 4.36322749e-01 7.91259944e-01 -4.48815614e-01
2.54418969e-01 2.38718405e-01 -7.17288435e-01 7.45622933e-01
8.15728307e-01 7.57698596e-01 -3.75556707e-01 -4.65713561e-01
8.27229202e-01 -7.71572828e-01 -5.09456873e-01 -7.40861893e-02
-1.35728860e+00 1.08598423e+00 9.63441730e-02 2.32236862e-01
-6.43535256e-02 1.12844980e+00 4.59347427e-01 7.29391515e-01
7.57932430e-03 6.25393450e-01 -3.67496312e-01 -3.38179439e-01
-9.91357028e-01 6.78722337e-02 8.11738193e-01 9.10739481e-01
2.98271656e-01 5.23110569e-01 3.50555569e-01 -5.73666207e-02
5.64148128e-01 6.18714929e-01 6.68885946e-01 -1.17220795e+00
6.04510486e-01 -9.07409340e-02 6.89346790e-02 -1.27558005e+00
4.11500156e-01 -2.13565126e-01 -6.46181345e-01 6.81441963e-01
5.33640802e-01 -7.86427975e-01 -5.91089368e-01 1.98801315e+00
1.47586698e-02 1.32427245e-01 2.73909956e-01 5.09461582e-01
4.05616015e-01 6.28979981e-01 -2.73997784e-01 3.15849215e-01
1.28810227e+00 -5.40972769e-01 -3.96904618e-01 -1.32984608e-01
5.64624012e-01 -6.16203308e-01 9.19651866e-01 4.44521308e-01
-1.29645669e+00 -1.88154608e-01 -1.87091887e+00 1.59908980e-01
-4.81544316e-01 -1.49548531e-01 1.22814693e-01 1.90191567e+00
-1.42622399e+00 2.93085635e-01 -6.60828292e-01 3.71707350e-01
7.36598670e-01 7.00700462e-01 -5.05160034e-01 3.80546480e-01
-1.64231002e+00 6.64077222e-01 3.78423661e-01 -3.41199934e-01
-1.57054448e+00 -4.41107810e-01 -9.61779296e-01 -1.78560868e-01
1.43560588e-01 -5.81299007e-01 1.35241425e+00 -1.16227019e+00
-1.58638775e+00 8.16915452e-01 6.25500798e-01 -9.17621076e-01
1.07781339e+00 1.93807423e-01 -3.70113790e-01 6.22552872e-01
-1.16670199e-01 9.09029186e-01 1.56708753e+00 -1.36348033e+00
-4.26862612e-02 2.80370086e-01 4.79145020e-01 -4.21338499e-01
-7.68045664e-01 9.42496303e-03 1.64116442e-01 -1.14453125e+00
-6.44961059e-01 -9.57250893e-01 -1.36436775e-01 4.16704983e-01
-6.37357831e-01 5.18372118e-01 1.02135706e+00 -8.00727427e-01
1.03277528e+00 -2.29302382e+00 -3.30892116e-01 3.52658033e-01
5.71374178e-01 5.17874479e-01 -4.38745528e-01 5.39764225e-01
1.09953113e-01 1.16718471e+00 -1.38235763e-01 -1.29820198e-01
6.02924526e-01 -8.71791840e-02 -6.12348735e-01 7.58477271e-01
1.41214281e-01 1.17107165e+00 -7.90691018e-01 -1.55539200e-01
-3.16832870e-01 5.57672620e-01 -7.40636468e-01 -3.47748958e-02
-3.49334806e-01 -1.23449162e-01 -1.69389829e-01 4.71842140e-01
7.47465611e-01 -1.97326764e-02 1.31720960e-01 1.55246094e-01
3.30468029e-01 -4.24083732e-02 -1.29378307e+00 1.42381287e+00
-2.45717809e-01 9.68216598e-01 3.03668737e-01 -4.64070827e-01
6.31423295e-01 7.59136319e-01 -1.87386572e-01 -1.84275120e-01
4.62973237e-01 5.36802888e-01 1.17851049e-01 -1.20683976e-01
3.98876518e-01 -3.27611752e-02 -3.58336270e-01 1.02956975e+00
-5.91975711e-02 -2.31275454e-01 -1.67507634e-01 5.09543419e-01
1.52524281e+00 -2.58789212e-01 -3.59714702e-02 1.88104451e-01
1.99008882e-01 -5.51972389e-01 2.30227783e-01 1.01995480e+00
-2.56733060e-01 3.52093458e-01 8.06370199e-01 1.52668327e-01
-1.39064753e+00 -8.80314708e-01 5.58900416e-01 7.30126441e-01
-6.40224665e-02 -5.91041148e-01 -1.12231600e+00 -8.11212778e-01
2.09189393e-02 4.10054445e-01 -6.92003071e-01 -3.45682889e-01
-2.74778068e-01 -1.60361841e-01 1.91151512e+00 3.30933124e-01
7.98781395e-01 -8.40414166e-01 -6.83177769e-01 7.46611133e-02
-5.35419323e-02 -1.12156904e+00 -8.47759008e-01 -2.10318163e-01
-4.18835700e-01 -9.06534076e-01 -9.27945018e-01 -7.16761410e-01
7.38340855e-01 -1.19521916e-01 6.49100840e-01 6.34422481e-01
-4.74113673e-02 3.70051354e-01 -2.46409193e-01 -4.96799111e-01
-1.36389184e+00 1.06808573e-01 -1.37966692e-01 -2.00451195e-01
-2.63253361e-01 -5.50117552e-01 -7.28965282e-01 2.54049957e-01
-1.36065352e+00 -1.98162898e-01 4.52681243e-01 5.51347613e-01
-1.67218387e-01 4.68049437e-01 8.07816327e-01 -7.08004177e-01
8.64994884e-01 -2.57421762e-01 -8.15089881e-01 -1.09092250e-01
-4.11788076e-01 -9.58005711e-02 4.36905742e-01 -8.16070378e-01
-3.42859417e-01 -2.30864644e-01 -7.06714317e-02 -2.41836935e-01
2.56095052e-01 2.23439574e-01 -4.61771876e-01 -6.81306779e-01
9.81253505e-01 4.37902194e-03 -3.13124061e-02 1.47978708e-01
8.94417465e-01 5.76555312e-01 4.95659351e-01 -5.89112639e-01
1.72467041e+00 4.77634430e-01 -3.27076167e-01 -3.32280427e-01
3.21155906e-01 5.75652897e-01 -4.29201275e-02 -2.40228757e-01
6.17997587e-01 -1.04570913e+00 -7.58557081e-01 9.40472543e-01
-1.23676348e+00 -7.55543828e-01 -3.87917519e-01 -2.93467473e-02
-4.31008160e-01 5.81049979e-01 -8.32613587e-01 -5.83302081e-01
-3.97914767e-01 -1.25220084e+00 7.09589064e-01 -1.28034309e-01
-2.13681519e-01 -9.17617202e-01 -1.01314075e-01 2.90299952e-01
5.15424311e-01 6.53046131e-01 6.49154484e-01 -5.71532190e-01
-8.79982948e-01 -6.58377051e-01 1.08523898e-01 7.66877770e-01
-9.55195054e-02 2.65210956e-01 -1.14252305e+00 -5.39587677e-01
-7.27196317e-03 -3.70258451e-01 5.37726641e-01 -6.17231987e-02
7.38562405e-01 -1.07506001e+00 1.38283312e-01 5.73626161e-01
1.36412144e+00 1.55818984e-01 9.47012484e-01 5.95639408e-01
3.49493444e-01 1.83880553e-01 -2.82652736e-01 4.29110855e-01
2.56400164e-02 1.77233413e-01 5.50733566e-01 9.22682956e-02
-1.20801896e-01 -6.91337466e-01 1.03821957e+00 5.05927742e-01
1.79029599e-01 -5.75663865e-01 -5.00522196e-01 4.79072750e-01
-1.16923106e+00 -1.03741205e+00 3.00834000e-01 1.89321661e+00
1.28026581e+00 6.68906748e-01 1.76403716e-01 5.46910048e-01
7.66287208e-01 2.52750307e-01 -3.94862354e-01 -7.06978619e-01
-1.71172500e-01 6.45299375e-01 1.16613626e+00 5.37918746e-01
-9.46002603e-01 9.32582438e-01 5.76809120e+00 9.62604582e-01
-1.05244720e+00 4.95268732e-01 6.02743566e-01 4.42753769e-02
-8.23843300e-01 6.50126338e-02 -3.29777062e-01 5.79762399e-01
1.19042099e+00 -5.13880432e-01 4.92472917e-01 6.48517251e-01
3.83799151e-02 3.63314718e-01 -9.81877327e-01 5.20491838e-01
-1.97407845e-02 -1.69453120e+00 -1.49991408e-01 4.57789004e-01
7.27777779e-01 -3.35376173e-01 7.49773920e-01 -6.79010004e-02
8.34186077e-01 -1.04971528e+00 1.32120657e+00 -1.37070730e-01
1.22009027e+00 -7.72636712e-01 4.41192061e-01 1.89380974e-01
-8.80577505e-01 -2.01762035e-01 -2.25882843e-01 2.65038699e-01
-2.00324897e-02 1.84591189e-01 -8.44630659e-01 1.40091866e-01
1.34464994e-01 1.11760631e-01 -6.20904863e-01 6.61732495e-01
-7.41284132e-01 8.20284426e-01 -2.44100124e-01 1.01684205e-01
2.64204890e-01 5.98288059e-01 5.81009686e-01 1.01565957e+00
7.87321329e-01 -1.55161262e-01 -3.60842377e-01 9.61801529e-01
-8.11025858e-01 -4.53242004e-01 -8.53046715e-01 -6.07255936e-01
7.37621665e-01 8.85192692e-01 -7.22859740e-01 -2.00541660e-01
-1.84886213e-02 1.01941788e+00 -6.08136296e-01 2.45115772e-01
-1.02936697e+00 -7.76820064e-01 5.70795894e-01 1.92528114e-01
4.73012388e-01 -5.59706032e-01 -2.38145784e-01 -1.04450190e+00
-7.21094608e-02 -1.60023737e+00 1.45688906e-01 -7.86831379e-01
-1.01698768e+00 4.16954637e-01 -3.37249041e-01 -1.23246527e+00
-2.61070639e-01 -3.87793809e-01 -4.51533020e-01 6.72378361e-01
-1.39580083e+00 -1.29741824e+00 3.46538484e-01 7.69817173e-01
-5.44296019e-02 -3.91014844e-01 9.32573974e-01 1.90682441e-01
-1.98583826e-01 1.30801487e+00 -1.45904198e-01 7.91160166e-01
4.96986419e-01 -1.00233686e+00 1.12598467e+00 1.41614449e+00
2.66163468e-01 5.34667611e-01 5.89092255e-01 -7.73842812e-01
-1.49272549e+00 -8.75581920e-01 6.00720644e-01 -3.14952344e-01
1.02852309e+00 -7.77809083e-01 -3.46196562e-01 6.67267680e-01
5.86831689e-01 -2.14367837e-01 7.08914459e-01 -1.07228923e+00
-9.22761440e-01 1.94771737e-01 -1.52313113e+00 7.53917515e-01
6.16551578e-01 -9.93087530e-01 -1.16511926e-01 7.98951369e-03
1.15409613e+00 -3.68993938e-01 -6.39758885e-01 -2.37696186e-01
6.75600052e-01 -7.61442006e-01 9.32003319e-01 -1.15109950e-01
3.16808105e-01 -3.93857539e-01 -1.74635127e-01 -9.73592162e-01
3.19749206e-01 -1.65956438e+00 -3.24090511e-01 1.24845970e+00
6.13484919e-01 -9.35281217e-01 8.13135386e-01 6.01319849e-01
4.31122929e-01 1.59236923e-01 -1.12664711e+00 -1.03230894e+00
5.36046147e-01 -6.30841851e-01 9.20380771e-01 9.77086186e-01
-1.00524046e-01 -5.32877743e-01 -5.19242704e-01 6.59142971e-01
7.48812139e-01 -7.82540381e-01 7.66161561e-01 -5.85487604e-01
-5.03968000e-01 -5.52443922e-01 -6.72259629e-01 -5.94743133e-01
1.19685404e-01 -8.83896232e-01 -1.17415540e-01 -9.08913612e-01
-4.19596970e-01 -2.19153449e-01 2.32326258e-02 7.03502774e-01
4.26874042e-01 8.34564447e-01 6.25742972e-01 -3.62444744e-02
6.14986494e-02 4.24555428e-02 9.56001401e-01 -4.63203549e-01
2.53835529e-01 -1.01318479e-01 -1.10952723e+00 5.38169444e-01
1.03207242e+00 -9.23555970e-01 -7.23461270e-01 -3.68669331e-01
7.84531772e-01 -7.15448037e-02 7.43032515e-01 -1.09261656e+00
1.65750951e-01 2.30809361e-01 1.95183903e-01 3.90565932e-01
3.35380286e-02 -7.77509272e-01 3.86567354e-01 9.56238091e-01
-5.22303760e-01 1.86956838e-01 1.59645885e-01 5.58872104e-01
3.11862290e-01 -6.48581505e-01 6.38032615e-01 -1.69604868e-01
-2.32809499e-01 2.42350012e-01 -8.29929233e-01 1.95968479e-01
1.01410115e+00 -2.17638850e-01 -7.68080533e-01 -9.23985243e-01
-6.13521874e-01 -3.83523464e-01 8.21899176e-01 1.89001024e-01
8.20702493e-01 -1.25566816e+00 -6.92557216e-01 2.71311551e-01
-3.44603837e-01 -6.28113627e-01 -1.31695226e-01 -1.17387287e-02
-8.82993937e-01 -3.68197531e-01 -9.72457528e-02 2.21170172e-01
-1.25139248e+00 5.26928186e-01 3.74953866e-01 2.20959866e-03
-3.85967702e-01 9.79381800e-01 -3.07877868e-01 4.95320968e-02
3.25200886e-01 -2.85290718e-01 4.21810091e-01 2.18759347e-02
6.80819213e-01 9.34829116e-02 -2.01976016e-01 -4.12942111e-01
-1.68503344e-01 1.80398658e-01 1.28059760e-01 -8.31181586e-01
9.56203640e-01 -8.88484903e-03 5.24932519e-02 -3.71816039e-01
1.23521864e+00 5.79185665e-01 -1.33173084e+00 2.34422892e-01
-2.11634383e-01 -6.04350448e-01 -4.43212241e-02 -7.58080244e-01
-1.44064164e+00 8.93414080e-01 5.10808885e-01 6.18553102e-01
8.86000633e-01 -2.93922514e-01 1.38195252e+00 1.89172074e-01
4.70424443e-01 -7.71280050e-01 4.21759784e-01 -1.86742291e-01
7.17379272e-01 -5.89828670e-01 -2.31118098e-01 -7.28837475e-02
-5.77843845e-01 1.06487584e+00 2.10393801e-01 -1.22776255e-01
7.41650224e-01 8.52278471e-01 2.23129869e-01 2.06086382e-01
-5.95369935e-01 4.57547367e-01 -2.72756159e-01 1.16035962e+00
-2.43988737e-01 1.46800235e-01 8.76557827e-03 7.38576472e-01
-5.48419237e-01 -1.39560122e-02 1.25091159e+00 9.59449708e-01
-1.46659061e-01 -1.48897028e+00 -4.03304964e-01 -9.54265818e-02
-9.84045744e-01 -3.32048386e-01 -4.42103952e-01 6.62382782e-01
3.37110817e-01 1.20741355e+00 -5.19460618e-01 -6.30376875e-01
-1.32527426e-01 -2.26071060e-01 2.37738863e-01 -2.11163074e-01
-7.79934466e-01 -6.39899820e-02 2.48740554e-01 -2.83262104e-01
-4.26617026e-01 -2.86709845e-01 -9.71753836e-01 -8.06216717e-01
-4.19396013e-01 4.26596031e-02 5.83774269e-01 2.00477645e-01
2.45575264e-01 1.65491272e-02 8.27086806e-01 -6.17654443e-01
-6.83507025e-01 -1.86396554e-01 -5.86571932e-01 -1.11065336e-01
5.82817912e-01 4.47127856e-02 -7.06151783e-01 5.15016794e-01] | [5.609605312347412, 7.776146411895752] |
74eb599b-4932-4e0a-bbc4-756049876ad2 | spatial-probabilistic-pulsatility-model-for | 1607.08129 | null | http://arxiv.org/abs/1607.08129v1 | http://arxiv.org/pdf/1607.08129v1.pdf | Spatial probabilistic pulsatility model for enhancing photoplethysmographic imaging systems | Photolethysmographic imaging (PPGI) is a widefield non-contact biophotonic
technology able to remotely monitor cardiovascular function over anatomical
areas. Though spatial context can provide increased physiological insight,
existing PPGI systems rely on coarse spatial averaging with no anatomical
priors for assessing arterial pulsatility. Here, we developed a continuous
probabilistic pulsatility model for importance-weighted blood pulse waveform
extraction. Using a data-driven approach, the model was constructed using a 23
participant sample with large demographic variation (11/12 female/male, age
11-60 years, BMI 16.4-35.1 kg$\cdot$m$^{-2}$). Using time-synchronized
ground-truth waveforms, spatial correlation priors were computed and projected
into a co-aligned importance-weighted Cartesian space. A modified
Parzen-Rosenblatt kernel density estimation method was used to compute the
continuous resolution-agnostic probabilistic pulsatility model. The model
identified locations that consistently exhibited pulsatility across the sample.
Blood pulse waveform signals extracted with the model exhibited significantly
stronger temporal correlation ($W=35,p<0.01$) and spectral SNR ($W=31,p<0.01$)
compared to uniform spatial averaging. Heart rate estimation was in strong
agreement with true heart rate ($r^2=0.9619$, error $(\mu,\sigma)=(0.52,1.69)$
bpm). | ['David A. Clausi', 'Alexander Wong', 'Robert Amelard'] | 2016-07-27 | null | null | null | null | ['heart-rate-estimation'] | ['medical'] | [ 8.44317600e-02 1.75843108e-02 1.28220618e-01 -1.98271230e-01
-7.40359843e-01 -4.55508828e-01 8.22429061e-02 -4.50217463e-02
-2.61987716e-01 1.13109779e+00 4.45344746e-01 4.61346805e-02
-4.06894326e-01 -5.68395495e-01 -1.90192387e-01 -8.16279590e-01
-9.38078582e-01 8.96350175e-05 -5.67486733e-02 4.76632208e-01
1.45557672e-01 2.57141471e-01 -6.29705131e-01 -2.53215015e-01
9.06703055e-01 1.20302582e+00 -3.52850050e-01 1.03189492e+00
3.25247377e-01 2.71534234e-01 -6.45893037e-01 -2.88657039e-01
2.04569325e-01 -8.28280926e-01 -1.41850278e-01 -5.96020341e-01
4.61985201e-01 -1.19298570e-01 -2.85166651e-01 8.02695513e-01
1.05899692e+00 9.70411673e-02 7.36406624e-01 -6.55887783e-01
-5.02892375e-01 5.38545787e-01 -7.84468830e-01 7.69724488e-01
1.52721286e-01 4.68394190e-01 3.42375070e-01 -7.40794539e-01
4.32962894e-01 4.23883587e-01 1.11463284e+00 9.89611968e-02
-1.77072299e+00 -6.90757930e-01 -6.85489118e-01 -2.30593443e-01
-1.73007011e+00 -5.01474142e-01 7.66750753e-01 -6.85038269e-01
6.21111095e-01 5.48021793e-01 1.16756248e+00 6.00109756e-01
8.72845173e-01 -3.14032257e-01 1.65042162e+00 -1.69892594e-01
1.63749740e-01 2.02710945e-02 -1.08798750e-01 5.10018587e-01
3.57325822e-01 4.03501183e-01 -6.18241727e-01 -4.57305133e-01
1.16871321e+00 -5.53099036e-01 -5.01992345e-01 -2.75747441e-02
-1.19405627e+00 3.45800281e-01 1.50022581e-01 2.89566815e-01
-6.68717682e-01 4.77532774e-01 2.04604238e-01 -1.46724790e-01
4.71616566e-01 5.16898811e-01 -1.30165592e-01 -6.89388692e-01
-1.10192227e+00 4.96651307e-02 8.43545318e-01 3.84747982e-01
1.43094167e-01 4.10322070e-01 -5.35195708e-01 6.45850122e-01
4.71390933e-01 1.02215409e+00 1.14675209e-01 -1.46032691e+00
-3.05143923e-01 -1.74884528e-01 2.56566882e-01 -1.24216807e+00
-6.90752804e-01 -7.07035303e-01 -1.06327224e+00 -1.20282480e-02
8.18492293e-01 -5.13946712e-01 -5.45694828e-01 1.59553313e+00
2.86085904e-01 5.35832226e-01 -4.07981396e-01 1.13176537e+00
6.60975277e-01 2.15913877e-01 5.80130279e-01 -6.68287992e-01
1.77790344e+00 1.56066567e-01 -6.23203635e-01 1.20485581e-01
-1.57370821e-01 -6.00953281e-01 6.19311273e-01 2.11397305e-01
-1.31040704e+00 -4.81848925e-01 -6.89647377e-01 5.72928727e-01
3.11538011e-01 -1.06186159e-01 3.15668523e-01 1.09306288e+00
-1.05604863e+00 8.55585217e-01 -7.90563464e-01 -6.07923791e-02
6.49900913e-01 -9.51202810e-02 -3.17345448e-02 4.99961199e-03
-1.22325504e+00 1.11703181e+00 -2.55778342e-01 1.96787372e-01
-4.78780091e-01 -1.69514835e+00 -6.30379438e-01 -9.16974247e-02
-3.74227613e-01 -7.66755879e-01 3.64061326e-01 7.03214705e-02
-1.63932633e+00 5.49544096e-01 -1.08485065e-01 -3.64739150e-01
3.61286700e-01 -6.30860627e-02 -7.16509521e-01 7.89523005e-01
8.73849913e-02 1.33434162e-01 8.64280462e-01 -9.95527267e-01
3.29615235e-01 -3.06373924e-01 -8.76737475e-01 -4.21512239e-02
3.70201141e-01 1.16054192e-01 2.83299744e-01 -5.17371237e-01
2.20818356e-01 -6.90499902e-01 -1.58260107e-01 3.18628848e-01
-8.77087489e-02 6.35697842e-01 -4.68307026e-02 -1.09047639e+00
1.34694707e+00 -1.90919149e+00 -4.69594479e-01 6.86918736e-01
6.36618197e-01 -1.90292159e-03 3.30870062e-01 4.96831387e-02
8.71807262e-02 3.03228825e-01 -6.23441875e-01 2.56875843e-01
-5.07133007e-02 -3.16707581e-01 3.72037552e-02 1.07611203e+00
-1.62957788e-01 1.04833269e+00 -1.07738602e+00 -5.43090045e-01
5.34727395e-01 8.40551555e-01 -1.70127422e-01 -8.85033160e-02
5.43850183e-01 8.08581948e-01 1.17336184e-01 6.09934747e-01
7.94604540e-01 -2.35913917e-02 2.26573646e-01 -6.41467273e-01
-5.44938684e-01 -4.04165149e-01 -9.76022601e-01 1.73693597e+00
-1.49836645e-01 6.71313643e-01 2.03276068e-01 -5.95650852e-01
1.18798792e+00 5.56595385e-01 1.11620557e+00 -7.78967083e-01
1.45072713e-01 1.00944065e-01 5.99601984e-01 -6.14310801e-01
-1.34522066e-01 -9.04634595e-01 8.73735920e-02 4.09170955e-01
1.40306458e-01 -3.59211236e-01 -2.19281703e-01 -2.12311566e-01
1.08460188e+00 7.94725046e-02 2.81067520e-01 -1.05245864e+00
9.93821099e-02 -8.29930305e-02 4.93699402e-01 9.81273413e-01
-1.04011011e+00 8.60719025e-01 5.45535982e-01 -3.03665251e-01
-7.86626041e-01 -1.76847470e+00 -1.00234520e+00 1.81844831e-01
1.66445240e-01 -9.28883553e-02 -2.00828090e-01 7.46822581e-02
2.58593082e-01 7.29469836e-01 -5.43762088e-01 -2.07020402e-01
-3.91878009e-01 -1.14463747e+00 9.54574227e-01 2.69425452e-01
5.08739173e-01 -8.61217558e-01 -1.15401542e+00 4.67078924e-01
-2.07582042e-01 -7.56015539e-01 -2.99967080e-01 -3.74986142e-01
-8.14730644e-01 -8.04075718e-01 -1.07264876e+00 7.47343823e-02
2.41438642e-01 -5.01579463e-01 1.34787226e+00 -5.88067889e-01
-1.15567732e+00 6.90850675e-01 -3.88450436e-02 -3.80689323e-01
2.53270626e-01 -7.27297425e-01 2.63214022e-01 -2.16949612e-01
2.32525274e-01 -8.60424817e-01 -1.47899592e+00 5.60308918e-02
-3.03602573e-02 -5.68676591e-01 3.02195370e-01 6.22745395e-01
6.09307349e-01 -4.29940313e-01 8.90803337e-01 -3.86508018e-01
6.60965502e-01 -5.84562302e-01 -3.36013287e-01 -1.98689520e-01
-5.31603098e-01 -5.32735050e-01 2.02540770e-01 -4.68264997e-01
-9.45837498e-01 -4.07753795e-01 8.76485035e-02 -3.73751193e-01
-1.54313639e-01 6.49724662e-01 7.74914980e-01 -8.18136856e-02
1.20266652e+00 3.77368927e-01 4.69950914e-01 7.03783482e-02
3.39874566e-01 2.33384483e-02 8.95383298e-01 -8.09399426e-01
2.97746807e-01 4.69093293e-01 3.27090681e-01 -1.07389247e+00
5.59902303e-02 -6.90706521e-02 -5.89474499e-01 -5.85443795e-01
9.36479747e-01 -5.84974945e-01 -9.67847109e-01 1.28768399e-01
-3.76967371e-01 -4.92710173e-01 -5.21586299e-01 1.18078446e+00
-5.46931326e-01 4.44663018e-01 -5.99120378e-01 -1.16732466e+00
-7.79301047e-01 -4.71568525e-01 2.12058231e-01 4.62809980e-01
-8.04814398e-01 -1.09118283e+00 4.96922016e-01 -5.98007515e-02
1.27920830e+00 1.13699734e+00 4.55244273e-01 1.61541939e-01
1.54929245e-02 -2.53634512e-01 -3.07720184e-01 2.73957718e-02
2.62873352e-01 2.47013886e-02 -9.72284317e-01 2.00099945e-01
7.45060146e-02 3.04281801e-01 3.14336240e-01 1.43929112e+00
8.92369688e-01 -1.09470062e-01 -1.35745585e-01 4.63007271e-01
1.39184308e+00 3.13471198e-01 1.00475442e+00 -6.46917582e-01
3.13880891e-01 4.32456434e-01 7.35483766e-02 9.40436304e-01
1.10144444e-01 2.07218841e-01 -1.47246018e-01 -7.31290877e-02
-2.65830457e-01 1.40530482e-01 -5.16586341e-02 2.51528263e-01
-3.64959449e-01 6.02524757e-01 -8.76063347e-01 6.00023806e-01
-1.01916277e+00 -1.02306461e+00 -4.15790349e-01 2.58729434e+00
9.63342786e-01 -3.34734283e-02 1.98202446e-01 -4.50593382e-01
6.51632428e-01 2.74177850e-03 -4.50364858e-01 -2.51780510e-01
1.43906206e-01 9.41668749e-01 5.44573247e-01 5.49338281e-01
-8.08727205e-01 9.30702463e-02 6.34026766e+00 1.15942731e-01
-1.09875894e+00 6.17300384e-02 6.60125554e-01 -2.17210725e-01
-6.46382421e-02 5.19042015e-02 -9.40064341e-03 9.14119065e-01
1.57029188e+00 -6.37300134e-01 3.05938810e-01 2.45150238e-01
4.92940634e-01 -7.96487451e-01 -4.47622448e-01 1.10447562e+00
-1.38994440e-01 -1.40073979e+00 -1.02616715e+00 -5.86050656e-03
2.56942332e-01 -1.00870036e-01 3.27162743e-02 -6.99406341e-02
-5.38854264e-02 -1.20789587e+00 1.94966972e-01 1.40253282e+00
1.67782903e+00 -3.87568325e-01 4.44107682e-01 -1.50076345e-01
-1.34525394e+00 5.30726016e-01 -2.82258838e-01 -1.21025732e-02
6.21985376e-01 1.39776134e+00 -4.55839694e-01 4.79883701e-01
6.53598011e-01 3.79237443e-01 2.12136991e-02 1.41637135e+00
8.38764757e-02 6.99547708e-01 -6.07988834e-01 2.50368267e-01
-5.93405426e-01 -3.84495467e-01 8.13435435e-01 1.19292891e+00
6.34270191e-01 6.23915553e-01 -2.62843698e-01 1.46247399e+00
5.02228558e-01 -1.45643532e-01 -4.29463565e-01 9.12594274e-02
7.98232019e-01 1.39219952e+00 -7.00944960e-01 -3.41645092e-01
-3.48852068e-01 2.56101191e-01 -5.03421426e-01 4.17341441e-01
-8.56775820e-01 -6.70822918e-01 6.38964117e-01 5.00903845e-01
-4.19628054e-01 -3.37992191e-01 -8.15728188e-01 -8.00721586e-01
-2.45697767e-01 -3.51822488e-02 3.18093598e-01 -6.74795806e-01
-1.46535301e+00 3.36034447e-01 2.92040020e-01 -9.22220826e-01
2.73029134e-03 -1.83897212e-01 -7.82410324e-01 1.76330447e+00
-1.14367187e+00 -3.76111895e-01 -2.98024148e-01 1.30026489e-01
-5.11070192e-01 4.29650098e-01 1.08818197e+00 2.18753338e-01
-2.76551783e-01 3.48445684e-01 -5.15858471e-01 4.42339815e-02
7.80293465e-01 -1.37280965e+00 -3.25191259e-01 6.63291156e-01
-5.31406701e-01 1.09448409e+00 6.52059615e-01 -8.95505786e-01
-1.30201721e+00 -7.96209037e-01 7.73555160e-01 -3.80295247e-01
4.81757939e-01 2.04419583e-01 -6.66204751e-01 -8.99189562e-02
2.90708840e-01 7.59486556e-01 1.16791463e+00 -1.55767977e-01
1.47502482e-01 -2.55330026e-01 -1.86116457e+00 3.73330355e-01
5.95132768e-01 -3.44496667e-01 -3.05275410e-01 -1.96628839e-01
-6.05389066e-02 -5.59658289e-01 -2.02971983e+00 6.06847525e-01
1.26983702e+00 -7.81031609e-01 1.10860884e+00 -3.06719518e-03
1.15909435e-01 -2.72653013e-01 1.51412070e-01 -1.08492339e+00
-5.09920478e-01 -8.26597273e-01 -2.80992746e-01 8.15604687e-01
-4.67671454e-02 -1.01560616e+00 5.07801354e-01 1.11093175e+00
-2.99971759e-01 -5.52137494e-01 -1.42346573e+00 -4.34665859e-01
2.24592835e-01 -4.22841161e-01 -8.60071182e-02 8.27853382e-01
6.19951129e-01 -3.68260324e-01 -2.52171099e-01 8.10931548e-02
1.18697250e+00 -5.79294711e-02 -1.04381954e-02 -1.27144945e+00
-2.70144403e-01 -4.22825962e-01 -2.58756548e-01 -2.30836034e-01
-6.16877496e-01 -5.66144049e-01 8.28242674e-03 -1.46829379e+00
-6.82653338e-02 -6.78467155e-01 -6.10746145e-01 1.63720936e-01
-1.49716303e-01 5.03471434e-01 -2.15033054e-01 -6.64601400e-02
4.28886086e-01 1.37779132e-01 1.11400712e+00 3.22708696e-01
-4.33480263e-01 -2.53199905e-01 -6.57560945e-01 2.41545476e-02
9.45360661e-01 -3.05535555e-01 -4.66300875e-01 4.57116574e-01
-1.42317459e-01 6.90586269e-01 6.90240145e-01 -1.39993966e+00
-6.86302111e-02 -6.03833161e-02 9.31727469e-01 -2.49220163e-01
3.43617916e-01 -2.95859486e-01 7.64715493e-01 7.64997900e-01
5.71537837e-02 -2.19660506e-01 2.96090215e-01 1.69858754e-01
3.46446842e-01 1.84799597e-01 1.14043307e+00 -4.67235744e-01
-7.65721351e-02 2.05971166e-01 -5.19434333e-01 4.69452471e-01
6.95231855e-01 -2.44125411e-01 -5.13729632e-01 -2.98360527e-01
-1.13713646e+00 -2.59831578e-01 1.18908351e-02 -3.94026816e-01
6.15304291e-01 -1.18604207e+00 -9.28041995e-01 1.28797457e-01
-1.16253473e-01 -5.80577135e-01 1.09996498e+00 1.79918885e+00
-6.06543303e-01 1.28944427e-01 -4.45911676e-01 -7.24787116e-01
-4.07171905e-01 -5.02650402e-02 7.76711583e-01 1.60399884e-01
-9.23125088e-01 9.42695260e-01 -3.43775213e-01 4.14037436e-01
-4.38944638e-01 -1.92086846e-01 1.78283229e-01 -3.66777591e-02
5.75233638e-01 6.42911911e-01 -3.46112698e-01 -4.82825249e-01
-5.55969954e-01 6.26764894e-01 7.81760573e-01 -4.24859464e-01
8.38612497e-01 -3.56012344e-01 -9.73752514e-02 4.89569873e-01
8.75986338e-01 1.83645487e-01 -1.42870414e+00 7.29211941e-02
-4.25531149e-01 -6.41956985e-01 -7.37015828e-02 -1.24576485e+00
-9.49989021e-01 7.11325526e-01 9.61522102e-01 2.86172181e-01
9.89491224e-01 -2.58717746e-01 3.25224042e-01 -9.34134364e-01
2.40613341e-01 -8.59674871e-01 -4.19974595e-01 -3.58718969e-02
8.14896584e-01 -4.78431821e-01 3.74314278e-01 -3.47810179e-01
-5.92254043e-01 8.87583673e-01 2.63202399e-01 -3.99653882e-01
1.18658602e+00 3.59097868e-01 9.40505564e-02 -3.31987470e-01
-9.51161310e-02 -2.48662569e-02 3.52693915e-01 8.56444061e-01
7.72910535e-01 3.95850599e-01 -8.39587688e-01 5.21354675e-01
-8.29221979e-02 3.14137936e-01 3.70219320e-01 7.80302942e-01
-4.42596942e-01 -4.65093136e-01 -2.18632728e-01 9.68759060e-01
-5.83324552e-01 -1.66018456e-01 5.37934601e-01 5.31866908e-01
-1.58713870e-02 8.16011548e-01 1.81020513e-01 4.72440710e-03
2.77738720e-01 4.79442269e-01 6.42931402e-01 -3.05048287e-01
-3.95704716e-01 3.28875601e-01 -1.97372064e-02 -5.58534682e-01
-2.99019873e-01 -7.96926439e-01 -1.21343791e+00 -9.87306610e-02
2.94684529e-01 2.09903494e-02 5.53457916e-01 4.85467166e-01
6.10021174e-01 5.59851885e-01 3.11204076e-01 -5.88942587e-01
8.53316952e-03 -1.14912283e+00 -1.04016483e+00 5.29888757e-02
9.00506452e-02 -7.99831510e-01 -3.25077742e-01 2.27086961e-01] | [14.026573181152344, 2.9612128734588623] |
280c530a-f497-438d-8dac-daf29b331d39 | divide-and-conquer-text-semantic-matching-1 | 2203.02898 | null | https://arxiv.org/abs/2203.02898v1 | https://arxiv.org/pdf/2203.02898v1.pdf | Divide and Conquer: Text Semantic Matching with Disentangled Keywords and Intents | Text semantic matching is a fundamental task that has been widely used in various scenarios, such as community question answering, information retrieval, and recommendation. Most state-of-the-art matching models, e.g., BERT, directly perform text comparison by processing each word uniformly. However, a query sentence generally comprises content that calls for different levels of matching granularity. Specifically, keywords represent factual information such as action, entity, and event that should be strictly matched, while intents convey abstract concepts and ideas that can be paraphrased into various expressions. In this work, we propose a simple yet effective training strategy for text semantic matching in a divide-and-conquer manner by disentangling keywords from intents. Our approach can be easily combined with pre-trained language models (PLM) without influencing their inference efficiency, achieving stable performance improvements against a wide range of PLMs on three benchmarks. | ['Daniel Wang', 'Haixiang Li', 'Meng Tang', 'Qi Zhang', 'Junzhe Wang', 'Tao Gui', 'Hongwei Liu', 'Yicheng Zou'] | 2022-03-06 | null | https://aclanthology.org/2022.findings-acl.287 | https://aclanthology.org/2022.findings-acl.287.pdf | findings-acl-2022-5 | ['community-question-answering', 'community-question-answering'] | ['miscellaneous', 'natural-language-processing'] | [ 4.78837222e-01 -1.46578044e-01 -5.27747273e-01 -5.76864898e-01
-6.99337363e-01 -4.28766608e-01 9.54972565e-01 9.70954835e-01
-5.94974041e-01 3.66011381e-01 5.34107566e-01 -4.43781793e-01
-4.08330001e-03 -1.05551004e+00 -4.86987203e-01 -1.12689503e-01
5.65580249e-01 4.67979997e-01 6.38939261e-01 -2.86574453e-01
4.65233266e-01 -2.50260308e-02 -1.62036693e+00 5.46322227e-01
8.72084320e-01 1.01738560e+00 9.26548764e-02 5.33171557e-02
-8.61284554e-01 7.68495023e-01 -5.46997130e-01 -8.01985383e-01
-2.18033984e-01 -3.80532801e-01 -9.94560003e-01 -3.49151045e-01
5.11710227e-01 -1.42320961e-01 -4.24012244e-01 1.15649319e+00
2.40101144e-01 3.93251568e-01 6.26332581e-01 -1.13452089e+00
-4.84385103e-01 7.77534425e-01 -3.28774780e-01 1.31923571e-01
8.62520039e-01 -2.67649293e-01 1.49529648e+00 -6.83389664e-01
3.21048707e-01 1.49936569e+00 4.13104147e-01 5.04803598e-01
-9.63962495e-01 -8.51240575e-01 3.42370361e-01 5.92808545e-01
-1.05141890e+00 -2.34576091e-01 6.63311183e-01 -1.43495709e-01
9.38420296e-01 5.67540586e-01 1.86541930e-01 8.58853102e-01
4.52782176e-02 9.94706571e-01 6.31344020e-01 -3.72089267e-01
3.62237126e-01 5.80078661e-02 3.62917364e-01 4.37975854e-01
3.05953681e-01 -5.56178510e-01 -5.85053742e-01 -5.47744274e-01
1.96955442e-01 4.99903619e-01 -1.55719176e-01 -7.72820413e-02
-1.16826510e+00 7.55563736e-01 3.30164194e-01 4.35541213e-01
-1.21243484e-01 1.88036889e-01 7.55665779e-01 2.93889970e-01
2.63979644e-01 4.08660382e-01 -3.69303048e-01 1.02391332e-01
-9.85264659e-01 5.67151368e-01 8.63133490e-01 1.02852547e+00
9.12551701e-01 -7.72240162e-01 -6.89126074e-01 9.01442826e-01
3.43271941e-01 3.42117190e-01 8.44214320e-01 -8.31720173e-01
7.79031396e-01 1.01215172e+00 1.19541124e-01 -1.21140981e+00
-8.17773640e-02 -1.75046057e-01 -6.71655416e-01 -6.52939975e-01
2.34266788e-01 1.91814706e-01 -5.46456993e-01 1.67340040e+00
2.73601055e-01 4.50154096e-01 5.69023415e-02 7.34568775e-01
1.03223312e+00 4.89782840e-01 4.50108856e-01 1.85943455e-01
1.90344620e+00 -9.95616138e-01 -5.80276966e-01 -6.42031372e-01
7.06997871e-01 -8.16818833e-01 1.12115407e+00 -2.40051150e-01
-9.44877386e-01 -3.71708602e-01 -8.00670326e-01 -3.99920732e-01
-5.89771032e-01 -2.63536602e-01 7.05087662e-01 4.31026995e-01
-6.28335178e-01 6.22072220e-01 -5.12907088e-01 -6.08893752e-01
2.16500238e-01 -1.36513272e-02 -2.19906330e-01 -2.61547834e-01
-1.56910002e+00 7.00183570e-01 6.27449691e-01 -3.91813874e-01
-2.54118383e-01 -7.47598767e-01 -8.91887188e-01 4.40796554e-01
6.63070619e-01 -1.06932294e+00 1.56954575e+00 -5.98766625e-01
-1.09774387e+00 1.11821699e+00 -6.10233486e-01 -5.23447037e-01
3.01966548e-01 -3.35915565e-01 -5.14076233e-01 1.69683576e-01
3.56204212e-01 4.78274614e-01 7.10293829e-01 -6.04047894e-01
-7.79056013e-01 -4.47461396e-01 2.82632887e-01 2.27572799e-01
-5.55917799e-01 4.45192695e-01 -6.19319141e-01 -8.22558641e-01
3.68919164e-01 -4.68369693e-01 -1.28691301e-01 3.18154335e-01
-2.14612037e-01 -6.57465756e-01 5.45747876e-01 -4.97654259e-01
1.51770413e+00 -1.86405075e+00 -4.90096547e-02 -1.02764510e-01
9.17785168e-02 2.08723173e-01 8.22287872e-02 7.41024077e-01
2.48758599e-01 1.18630745e-01 -2.98983216e-01 -2.75360674e-01
4.09286261e-01 8.45012367e-02 -7.45260239e-01 3.24891321e-02
-2.04787299e-01 1.09539175e+00 -1.13113701e+00 -7.64033318e-01
7.13389292e-02 -1.40215144e-01 -4.33081955e-01 2.92686343e-01
-6.31568909e-01 -1.79652989e-01 -9.20368373e-01 3.64420980e-01
3.59111011e-01 -4.79484111e-01 2.25131452e-01 -3.06443155e-01
3.83484125e-01 8.87963176e-01 -9.47364986e-01 1.90519011e+00
-7.59934545e-01 4.09918755e-01 -1.52922884e-01 -1.30750585e+00
7.85974264e-01 2.97898680e-01 3.20509881e-01 -7.93007076e-01
-5.00092059e-02 2.99542576e-01 -4.73461717e-01 -5.05306363e-01
5.75776398e-01 -1.90943599e-01 -3.73228252e-01 7.11887300e-01
-2.10322589e-01 -7.92632177e-02 2.42264941e-01 2.85949528e-01
1.10987580e+00 -1.43980697e-01 4.86742169e-01 -1.73053354e-01
9.25400019e-01 -5.62603138e-02 3.88386339e-01 9.46962893e-01
1.73176900e-01 2.26968050e-01 2.58886576e-01 -1.42861187e-01
-5.07486403e-01 -8.83676648e-01 1.72428731e-02 1.19499898e+00
6.24917388e-01 -5.98795116e-01 -4.53547567e-01 -6.44802690e-01
2.50665486e-01 8.55347037e-01 -2.94565648e-01 -3.51404041e-01
-6.72780454e-01 -3.81399304e-01 6.45085871e-01 5.78286231e-01
6.47051513e-01 -9.55400109e-01 -5.69981635e-01 4.43644851e-01
-5.89628637e-01 -1.29068065e+00 -7.18342841e-01 -1.79318622e-01
-8.93049657e-01 -1.02572596e+00 -5.11550188e-01 -6.58056676e-01
4.21716273e-01 6.22146189e-01 1.43095827e+00 2.87037343e-01
-1.16781481e-01 2.95914680e-01 -4.80169773e-01 -1.96133375e-01
-3.45423907e-01 1.68137193e-01 -4.36850518e-01 -2.64209695e-03
9.29069698e-01 -6.30374610e-01 -7.43341744e-01 4.95852768e-01
-1.23840249e+00 1.11858286e-01 4.26804751e-01 6.54881954e-01
2.38204464e-01 -1.73400298e-01 5.19568384e-01 -9.67965186e-01
8.42005968e-01 -6.56049728e-01 -2.12973669e-01 7.82900095e-01
-5.98167896e-01 4.07554805e-01 6.47725105e-01 -3.36352646e-01
-1.04855227e+00 -6.09057009e-01 -2.38143310e-01 -6.10698611e-02
-2.19955280e-01 5.40531576e-01 -2.53418654e-01 2.14336067e-01
3.22904110e-01 3.86035949e-01 -4.08632189e-01 -6.02729738e-01
4.53499079e-01 9.76437569e-01 5.24896264e-01 -1.01695156e+00
6.11089885e-01 4.54228073e-01 -1.44616291e-01 -3.09619904e-01
-1.06660604e+00 -1.13781178e+00 -6.26343563e-02 2.57040709e-01
6.25370860e-01 -7.77636647e-01 -8.11836183e-01 3.39870900e-01
-1.20823503e+00 1.67145386e-01 -3.49835679e-02 2.18172505e-01
-2.81908989e-01 7.54750669e-01 -4.96764839e-01 -3.42288882e-01
-6.70668542e-01 -6.60676003e-01 1.33039689e+00 3.04807365e-01
-5.51900089e-01 -1.12880540e+00 -2.46145397e-01 6.32191718e-01
4.68146622e-01 -3.51672769e-01 1.34680736e+00 -1.14240599e+00
-4.05614376e-01 -3.20410460e-01 -4.97267514e-01 -1.02778740e-01
3.07137519e-01 -4.84273851e-01 -6.49758458e-01 -1.00635387e-01
-4.53937687e-02 -3.02656680e-01 1.04261470e+00 -2.90042669e-01
1.36183870e+00 -5.67327797e-01 -6.52578592e-01 1.45500466e-01
1.14377713e+00 6.27708510e-02 5.31951845e-01 3.03934246e-01
4.01469618e-01 7.01571882e-01 5.91456771e-01 3.47474754e-01
5.83353519e-01 8.80827487e-01 2.08709449e-01 2.18286932e-01
7.92143047e-02 -5.70106566e-01 -2.26963442e-02 4.42513019e-01
5.27571142e-01 -3.31986964e-01 -6.83934689e-01 4.60130841e-01
-2.19699502e+00 -1.09126878e+00 1.75280005e-01 2.21315694e+00
1.08577371e+00 1.10286176e-01 -3.33456397e-01 1.38339236e-01
6.89818323e-01 4.36911136e-01 -5.61561525e-01 -2.18291841e-02
1.57321215e-01 3.27760339e-01 1.49759725e-01 4.00308192e-01
-7.72887349e-01 8.67180586e-01 5.23759937e+00 1.24220204e+00
-9.12194967e-01 -8.64580448e-04 2.93385774e-01 2.27924660e-01
-8.56840551e-01 2.36080214e-01 -9.11593020e-01 5.18502176e-01
4.76913631e-01 -5.55333614e-01 1.68640912e-01 6.06488466e-01
-4.05705087e-02 -3.54465246e-02 -1.32549286e+00 9.41570401e-01
1.43036798e-01 -1.38922715e+00 4.98404235e-01 -5.98941624e-01
3.77100050e-01 -3.57665986e-01 -4.06048983e-01 4.94238496e-01
7.35402554e-02 -7.12815344e-01 6.31695807e-01 2.77095616e-01
2.96686113e-01 -1.59380019e-01 6.01469278e-01 4.79287505e-01
-1.37812662e+00 8.75770953e-03 -3.03531438e-01 -1.34014003e-02
1.60690080e-02 6.04442656e-01 -3.54654580e-01 9.93500113e-01
4.44774985e-01 7.49108493e-01 -4.46830750e-01 9.38358903e-01
-3.48562002e-01 4.49862003e-01 -1.74572214e-01 -4.50879157e-01
2.77764559e-01 -1.35560244e-01 3.76951456e-01 1.23091805e+00
2.40193874e-01 -1.17928244e-03 1.62320256e-01 7.62192190e-01
-4.03616667e-01 4.04802501e-01 -1.28306776e-01 -7.54629746e-02
7.73207128e-01 9.07825768e-01 -5.85884273e-01 -7.63159454e-01
-8.17602694e-01 1.00856352e+00 2.00977162e-01 3.34919006e-01
-6.35148883e-01 -5.64581335e-01 7.90248930e-01 1.77965596e-01
1.37020618e-01 1.88304082e-01 -9.54195783e-02 -1.40788448e+00
3.97810489e-01 -8.21288407e-01 8.32687974e-01 -7.52221704e-01
-1.52171838e+00 2.45295972e-01 2.84807652e-01 -1.10656393e+00
-3.08313787e-01 -4.12060767e-01 -8.26465666e-01 8.76960695e-01
-1.61612725e+00 -9.51059818e-01 -4.92779791e-01 4.89827871e-01
6.79936945e-01 1.73482180e-01 8.06346416e-01 4.31690753e-01
-3.33591849e-01 6.57677770e-01 -1.31511077e-01 1.66338444e-01
7.37429738e-01 -8.10714364e-01 6.39547586e-01 6.29042327e-01
3.46092671e-01 8.85459006e-01 7.97395587e-01 -4.88709718e-01
-1.21444917e+00 -1.05323195e+00 1.56236064e+00 -1.11513853e-01
8.57535243e-01 -2.34381109e-01 -1.19450033e+00 4.56591934e-01
9.55171287e-02 -3.31558764e-01 6.54679596e-01 9.52168033e-02
-8.35505426e-01 -2.20730931e-01 -9.22241211e-01 1.00864363e+00
1.11648524e+00 -8.91683280e-01 -1.11762249e+00 4.13662314e-01
1.00107110e+00 -2.80340761e-01 -5.67423105e-01 4.38229829e-01
4.52460408e-01 -7.50749350e-01 1.10584223e+00 -9.83924448e-01
3.85932952e-01 -1.73198611e-01 -2.64531225e-01 -7.34691441e-01
1.09888971e-01 -5.75231671e-01 -2.33262092e-01 1.17612875e+00
2.32006550e-01 -8.33879590e-01 7.28627861e-01 9.61549640e-01
1.24040134e-01 -6.71731293e-01 -8.08598816e-01 -6.50893569e-01
-1.07841872e-01 -3.69478494e-01 9.65007305e-01 8.97406459e-01
4.22809601e-01 4.36285853e-01 1.88843548e-01 -1.63600385e-01
4.20415848e-01 7.25608110e-01 4.40921247e-01 -1.36436796e+00
-3.10544401e-01 -7.73938894e-01 -3.53204697e-01 -1.68873775e+00
5.21811128e-01 -1.11052251e+00 -2.70853758e-01 -1.81349623e+00
4.33384627e-01 -4.77883458e-01 -2.42676273e-01 5.83399475e-01
-7.15767622e-01 -2.57245094e-01 -5.41415662e-02 3.09609503e-01
-6.27131402e-01 5.54087222e-01 8.69477987e-01 -4.96191770e-01
2.27465555e-01 2.30025932e-01 -7.40253091e-01 6.20020330e-01
7.16154218e-01 -6.17237866e-01 -5.34493506e-01 -4.55222517e-01
3.64065766e-01 2.50346243e-01 4.93589908e-01 -5.64912915e-01
6.44397259e-01 -4.25729394e-01 -3.75340611e-01 -3.50625068e-01
1.64609477e-01 -9.31718051e-01 -1.09731756e-01 3.71597290e-01
-7.98383594e-01 1.19600939e-02 1.10063881e-01 6.30586982e-01
-6.28331304e-01 -7.22961903e-01 2.82792956e-01 -2.21464872e-01
-6.90267920e-01 2.78549075e-01 -1.13100953e-01 5.90680957e-01
6.28463268e-01 9.33455862e-03 -4.69983548e-01 -4.05305773e-01
-4.85495701e-02 4.28536624e-01 3.92427027e-01 7.40738153e-01
4.82150257e-01 -1.15054047e+00 -5.96878231e-01 -1.39412403e-01
4.68089759e-01 -1.86609268e-01 1.75424784e-01 4.04537410e-01
-7.28407800e-02 7.15885699e-01 3.05463195e-01 -1.78097516e-01
-1.21437001e+00 5.33032835e-01 1.72370866e-01 -4.30723548e-01
-5.14671326e-01 4.34681773e-01 2.07172811e-01 -3.37561488e-01
4.27033335e-01 -3.52591932e-01 -1.56998783e-01 -1.70461077e-03
6.81587100e-01 2.01669648e-01 9.95152518e-02 -2.48991176e-01
-4.62475359e-01 5.81125796e-01 -6.91979453e-02 5.87230586e-02
6.32182539e-01 -8.40418860e-02 -3.99170488e-01 2.00910851e-01
1.25747633e+00 -1.25445545e-01 -3.84017736e-01 -8.81423712e-01
4.71202880e-01 -4.95745540e-01 -1.39997467e-01 -4.41351295e-01
-6.43488169e-01 9.69705403e-01 -6.46625459e-02 1.45603701e-01
1.13183165e+00 2.24297494e-01 1.21647251e+00 7.79103577e-01
3.46728623e-01 -8.59566808e-01 5.79158776e-02 2.36406878e-01
4.94540721e-01 -1.08391500e+00 -1.32752940e-01 -5.66951036e-01
-2.25205660e-01 9.82543766e-01 4.98080313e-01 2.98486084e-01
4.07863438e-01 -8.04627612e-02 -1.68102086e-01 -2.22860396e-01
-8.14182878e-01 -2.08547741e-01 5.54369032e-01 -3.88199687e-02
5.08374214e-01 -1.19171634e-01 -6.46190166e-01 6.43361509e-01
1.99703261e-01 -8.20858777e-02 -1.03902459e-01 9.82049108e-01
-6.20867014e-01 -1.29494095e+00 -2.77335234e-02 5.46968043e-01
-5.16424417e-01 -4.43301409e-01 -2.89983422e-01 2.93989718e-01
-2.87665814e-01 9.82177079e-01 1.69017643e-01 -8.10427964e-02
3.77390951e-01 2.92557269e-01 2.54340410e-01 -7.02499330e-01
-8.49109828e-01 -5.81806839e-01 1.95953816e-01 -5.53515375e-01
-5.21686375e-01 -4.64520514e-01 -1.40685701e+00 -2.82082558e-01
-1.73735574e-01 4.95561302e-01 3.73471260e-01 1.30198419e+00
4.33865249e-01 1.99054271e-01 3.42026442e-01 5.48686050e-02
-7.22301841e-01 -7.70735383e-01 -6.85043558e-02 8.99029493e-01
-1.00118764e-01 -4.51415807e-01 -2.93626696e-01 -2.30731443e-02] | [10.963156700134277, 8.14614200592041] |
8b68dcc5-6975-4ede-a85d-bd7ddbbcb637 | drt-a-lightweight-single-image-deraining | 2204.11385 | null | https://arxiv.org/abs/2204.11385v1 | https://arxiv.org/pdf/2204.11385v1.pdf | DRT: A Lightweight Single Image Deraining Recursive Transformer | Over parameterization is a common technique in deep learning to help models learn and generalize sufficiently to the given task; nonetheless, this often leads to enormous network structures and consumes considerable computing resources during training. Recent powerful transformer-based deep learning models on vision tasks usually have heavy parameters and bear training difficulty. However, many dense-prediction low-level computer vision tasks, such as rain streak removing, often need to be executed on devices with limited computing power and memory in practice. Hence, we introduce a recursive local window-based self-attention structure with residual connections and propose deraining a recursive transformer (DRT), which enjoys the superiority of the transformer but requires a small amount of computing resources. In particular, through recursive architecture, our proposed model uses only 1.3% of the number of parameters of the current best performing model in deraining while exceeding the state-of-the-art methods on the Rain100L benchmark by at least 0.33 dB. Ablation studies also investigate the impact of recursions on derain outcomes. Moreover, since the model contains no deliberate design for deraining, it can also be applied to other image restoration tasks. Our experiment shows that it can achieve competitive results on desnowing. The source code and pretrained model can be found at https://github.com/YC-Liang/DRT. | ['Yang Liu', 'Saeed Anwar', 'Yuanchu Liang'] | 2022-04-25 | null | null | null | null | ['single-image-deraining'] | ['computer-vision'] | [ 1.42139435e-01 -2.03727365e-01 6.71385899e-02 -3.21204782e-01
-5.69996834e-01 5.30210920e-02 3.36836159e-01 -4.49794918e-01
-4.94769573e-01 4.59750146e-01 -1.30434290e-01 -6.44462645e-01
1.13856927e-01 -6.24997437e-01 -8.58602345e-01 -1.10236156e+00
3.24783713e-01 -1.56580769e-02 3.34086150e-01 -1.57844394e-01
-7.97565579e-02 4.90419567e-01 -1.47031248e+00 -2.10818518e-02
1.25764787e+00 1.08542538e+00 5.92412829e-01 4.17806476e-01
1.49615288e-01 8.98023665e-01 -5.97856820e-01 -3.27823371e-01
2.56833762e-01 -2.80112535e-01 -3.75041425e-01 -1.44497484e-01
9.04860258e-01 -6.93252444e-01 -6.81203067e-01 8.96232426e-01
9.00516212e-01 1.63077384e-01 2.94619679e-01 -8.16460252e-01
-8.20582807e-01 3.90747488e-01 -5.84833443e-01 5.97893298e-01
-3.50990146e-01 2.85659999e-01 8.76713455e-01 -1.10603201e+00
-3.57142836e-02 9.65249538e-01 7.97176957e-01 5.73849797e-01
-1.09170043e+00 -9.49365556e-01 2.86056072e-01 5.55804968e-01
-1.18541455e+00 -5.57030678e-01 5.97143710e-01 -4.64162380e-02
1.02939653e+00 1.65850461e-01 6.45843565e-01 1.05291307e+00
-2.12610811e-02 7.93434560e-01 1.14295352e+00 -1.05709121e-01
1.07592084e-01 -1.96515415e-02 1.72116950e-01 6.13789260e-01
2.55025387e-01 -1.63407847e-02 -3.78928632e-01 3.34468246e-01
7.16967046e-01 1.58527493e-01 -6.07888162e-01 1.15218230e-01
-7.07498491e-01 8.00289452e-01 9.52418268e-01 1.69845283e-01
-2.56762594e-01 3.18632156e-01 2.17684343e-01 3.25235575e-01
7.08727300e-01 2.55018413e-01 -3.85829896e-01 1.17994763e-01
-1.01717341e+00 7.63586760e-02 3.24282199e-01 6.64617717e-01
6.84841275e-01 5.01286447e-01 -2.39868745e-01 1.01204646e+00
-3.20403813e-03 5.99618495e-01 3.75812441e-01 -6.95088267e-01
4.54375386e-01 3.10735673e-01 -1.61589712e-01 -7.36137927e-01
-3.96812946e-01 -9.71671164e-01 -1.37686789e+00 2.63995916e-01
2.46840745e-01 -6.98349550e-02 -1.21319032e+00 1.60840321e+00
1.51199967e-01 5.77327013e-01 -6.14999197e-02 1.11195421e+00
1.02934730e+00 8.00398886e-01 -1.03354491e-01 -2.14057639e-01
1.25495470e+00 -1.13866317e+00 -4.73882645e-01 -6.67598724e-01
3.60481769e-01 -6.76480174e-01 1.47115898e+00 3.83466750e-01
-9.72750962e-01 -6.14818335e-01 -1.08223438e+00 -4.06837851e-01
1.89084724e-01 3.41271341e-01 8.20959032e-01 5.14584720e-01
-1.05139625e+00 7.00256884e-01 -9.04250801e-01 -2.24632114e-01
8.95233393e-01 1.90770164e-01 6.51813298e-02 -3.70213896e-01
-9.66827512e-01 8.75788629e-01 -6.39299536e-03 5.81247091e-01
-1.08858418e+00 -9.93717194e-01 -5.52175343e-01 2.53388047e-01
3.56315076e-01 -7.64249086e-01 1.17874718e+00 -7.18516290e-01
-1.58393657e+00 5.71263194e-01 -2.43419528e-01 -6.55576468e-01
4.24329400e-01 -6.68937206e-01 -5.30072413e-02 6.66180700e-02
-3.21282923e-01 4.18170065e-01 1.28408420e+00 -9.02620792e-01
-3.63571167e-01 -2.53715694e-01 3.03843558e-01 2.25377500e-01
-5.54369986e-01 -2.76067168e-01 -5.78296363e-01 -7.79718697e-01
1.36530921e-01 -9.19101298e-01 -2.97205776e-01 1.72355741e-01
-2.81475335e-01 -2.75322199e-01 7.68796086e-01 -6.92106009e-01
1.11784136e+00 -2.23154807e+00 -4.89167124e-03 -2.75644720e-01
2.96868443e-01 6.26621306e-01 -1.59164891e-01 -6.62990436e-02
-6.71881512e-02 -3.37689416e-03 -2.98189372e-01 -4.12984461e-01
-2.37020567e-01 2.42194280e-01 -5.07059574e-01 5.38323998e-01
2.31705257e-03 7.97977269e-01 -4.26225603e-01 -9.95848998e-02
2.34972030e-01 7.06220865e-01 -6.34968936e-01 3.68563205e-01
-8.69086459e-02 2.81772524e-01 -3.14869821e-01 4.91663992e-01
8.12071741e-01 -4.36130464e-01 -2.17095345e-01 -4.64244545e-01
3.39474156e-02 4.44071501e-01 -7.20921457e-01 1.51052845e+00
-8.99832904e-01 8.59688818e-01 -6.50078580e-02 -1.16210508e+00
7.46316731e-01 1.32711247e-01 -9.10290107e-02 -8.69140208e-01
1.53455645e-01 1.39510199e-01 7.26903207e-04 -4.59093988e-01
1.82274863e-01 -1.09484591e-01 4.87187415e-01 1.99880257e-01
-1.49872169e-01 2.92751901e-02 -3.29870284e-02 2.18355671e-01
1.04037106e+00 -1.36735097e-01 -1.51066408e-01 -1.50618687e-01
2.65597552e-01 -4.66308475e-01 7.56844461e-01 7.04106271e-01
5.97933754e-02 5.52757144e-01 2.48024210e-01 -4.16266561e-01
-1.01593077e+00 -8.80054414e-01 -2.77945936e-01 1.07526922e+00
8.23617354e-02 -1.76973552e-01 -6.22235000e-01 -3.16699892e-01
-2.74525791e-01 7.20136166e-01 -4.92548466e-01 -2.35339612e-01
-5.43161809e-01 -1.16328692e+00 4.96766448e-01 5.18778384e-01
9.46684361e-01 -9.47351158e-01 -7.55639374e-01 -4.57442217e-02
-1.20618157e-01 -1.20057964e+00 -5.98123729e-01 2.16689035e-01
-1.09959769e+00 -8.99140477e-01 -8.17110121e-01 -7.75142014e-01
5.70314646e-01 6.52011991e-01 1.13987279e+00 4.28754240e-01
-3.79743993e-01 -1.30359247e-01 -3.41483593e-01 -2.86951810e-01
1.90190405e-01 2.07483977e-01 -1.15520187e-01 -1.14010394e-01
9.26641077e-02 -9.48857546e-01 -9.19780970e-01 2.25909650e-01
-8.24117482e-01 2.39536569e-01 8.33950877e-01 1.06248772e+00
4.20369536e-01 8.23096335e-02 5.07256091e-01 -8.98848116e-01
3.62194806e-01 -1.56087548e-01 -7.27831006e-01 5.98028451e-02
-6.94658160e-01 1.16081156e-01 7.61938155e-01 -5.58542252e-01
-1.18264961e+00 -2.66814798e-01 -2.00426295e-01 -7.18031347e-01
2.31050819e-01 4.38307285e-01 -2.22147927e-01 -1.51692733e-01
6.59162164e-01 3.67155671e-01 -2.09585860e-01 -7.15678751e-01
2.33813211e-01 3.94004792e-01 4.57549185e-01 -2.55667269e-01
1.08933043e+00 3.35589439e-01 -1.72713205e-01 -8.97119582e-01
-1.30012584e+00 -1.12210535e-01 -1.30213156e-01 3.08336485e-02
4.78615165e-01 -1.29293227e+00 -8.35302949e-01 8.48056793e-01
-8.87142956e-01 -7.30103612e-01 -6.27157092e-02 4.04395372e-01
-1.23424970e-01 2.72582769e-01 -7.73849607e-01 -6.00503743e-01
-8.13418567e-01 -1.08792305e+00 6.70303524e-01 3.76905560e-01
4.35262382e-01 -6.52418673e-01 -3.88840109e-01 6.06562793e-01
8.44802797e-01 -1.71931818e-01 8.77514184e-01 -1.20906439e-02
-8.47496331e-01 1.66651845e-01 -5.08921683e-01 6.31692648e-01
-1.01356566e-01 -3.11469376e-01 -1.25617063e+00 -6.50180161e-01
1.82740286e-01 -5.09798765e-01 1.42999446e+00 5.21269441e-01
1.57375193e+00 -3.30589265e-01 3.72589752e-02 1.17870855e+00
1.42103124e+00 -9.56886485e-02 8.76891971e-01 3.25579345e-01
8.93165290e-01 2.58996665e-01 3.33098263e-01 3.93355399e-01
2.44940430e-01 5.85063457e-01 6.64392710e-01 -3.48535180e-01
-3.31995010e-01 -7.01851323e-02 3.44071686e-01 8.23517144e-01
-3.26909870e-01 -2.24902496e-01 -6.94174170e-01 5.19662797e-01
-1.63748860e+00 -9.77236927e-01 2.36406699e-02 2.15376806e+00
9.43047881e-01 2.14937434e-01 -3.29337329e-01 1.72397047e-01
3.69183511e-01 4.46541965e-01 -9.49019253e-01 -1.65867656e-01
-9.23356861e-02 5.15601397e-01 6.77698731e-01 4.27795380e-01
-9.46494043e-01 1.13096404e+00 5.25465441e+00 1.08444333e+00
-1.23404193e+00 2.30694458e-01 7.28976429e-01 -3.88619453e-01
-1.51965514e-01 -1.01668857e-01 -8.70516777e-01 3.78121376e-01
8.34858179e-01 8.61281622e-03 6.39256835e-01 7.67226815e-01
3.83745909e-01 -1.21420376e-01 -8.61535609e-01 1.14506721e+00
-8.28804728e-03 -1.25479531e+00 2.06123535e-02 -1.34477109e-01
4.75287914e-01 3.41802031e-01 2.76160508e-01 4.67721909e-01
1.85509622e-02 -1.22675061e+00 4.07211393e-01 2.18570963e-01
8.68039012e-01 -6.36574924e-01 7.46365249e-01 2.82875031e-01
-8.78953159e-01 -1.05124258e-01 -8.00870180e-01 -1.09450139e-01
-1.25684246e-01 1.05035186e+00 -5.23148715e-01 3.33117634e-01
1.13728845e+00 8.74181211e-01 -6.25891030e-01 1.13001883e+00
-5.71251869e-01 1.08184958e+00 -4.24199700e-01 1.71856597e-01
1.00119106e-01 -1.23387285e-01 3.76408964e-01 9.26931977e-01
5.07089019e-01 1.56136736e-01 -6.80925921e-02 5.18237352e-01
-3.90922129e-01 -1.24901935e-01 -2.16938406e-01 3.79608363e-01
4.47435945e-01 1.12188900e+00 -2.59271055e-01 -2.38309905e-01
-4.28256571e-01 1.07494390e+00 4.31632161e-01 5.04764259e-01
-1.08651400e+00 -2.83820480e-01 8.94162536e-01 2.06536099e-01
5.72918117e-01 -1.22573882e-01 -2.47089103e-01 -1.27567542e+00
2.87996143e-01 -8.64679515e-01 4.39990312e-02 -9.16700721e-01
-1.19964945e+00 9.32821274e-01 -3.37085187e-01 -1.08211136e+00
2.60006875e-01 -4.69303638e-01 -7.91619718e-01 9.09406900e-01
-1.92585301e+00 -9.78462696e-01 -8.36492240e-01 7.51053154e-01
6.47933602e-01 -9.60125327e-02 5.53840578e-01 5.28071880e-01
-1.01442790e+00 8.42941940e-01 1.66401476e-01 -1.01902142e-01
7.57158399e-01 -1.02001798e+00 4.80829805e-01 1.09643924e+00
1.03060454e-01 4.76534396e-01 8.32562327e-01 -1.40677929e-01
-1.28479481e+00 -1.25853622e+00 6.21556520e-01 -4.60985936e-02
5.45023739e-01 -4.19412732e-01 -1.31547415e+00 5.74799478e-01
2.16148153e-01 4.02310044e-01 2.95976818e-01 2.91096032e-01
-5.99401712e-01 -6.21434927e-01 -7.48626411e-01 6.49595380e-01
1.29575622e+00 -4.53705937e-01 -2.69360363e-01 4.42589015e-01
7.72168338e-01 -5.42929173e-01 -5.42025268e-01 3.67960662e-01
3.25651556e-01 -1.20907891e+00 1.02244282e+00 -2.12467566e-01
4.79965657e-01 -2.70898193e-01 -1.26601234e-02 -1.31458509e+00
-4.07546967e-01 -5.48373342e-01 -1.89259544e-01 1.10314786e+00
3.41263622e-01 -7.30542600e-01 8.69654059e-01 3.81340921e-01
-2.97924131e-01 -1.06874919e+00 -9.57752943e-01 -7.33333111e-01
8.98783952e-02 -5.01115561e-01 2.59928793e-01 6.38730943e-01
-7.09890187e-01 6.26844347e-01 -6.93526447e-01 3.57503027e-01
7.48566329e-01 2.37497419e-01 6.65771604e-01 -1.13890839e+00
-3.81023347e-01 -1.67015105e-01 -1.06403269e-01 -1.45057797e+00
7.05607012e-02 -7.66585231e-01 4.69026528e-02 -1.53356624e+00
1.74222395e-01 -6.16896629e-01 -2.22788185e-01 8.04153800e-01
-3.49089026e-01 3.95276219e-01 3.77221376e-01 3.69286597e-01
-3.00671726e-01 1.06529653e+00 1.33875299e+00 -2.42595986e-01
-1.57559156e-01 2.45163232e-01 -8.14667523e-01 7.50399590e-01
1.06117547e+00 -5.20856202e-01 -6.19231164e-01 -9.18873131e-01
1.20857731e-01 -1.82069063e-01 5.64907908e-01 -1.08102190e+00
2.77441174e-01 1.34457663e-01 2.79901266e-01 -3.49802762e-01
5.28524518e-01 -6.22405648e-01 -7.75767639e-02 4.30160820e-01
-3.82388160e-02 -1.23599939e-01 3.51680100e-01 4.70213324e-01
-1.37340993e-01 -7.02952445e-02 1.19186163e+00 1.49994195e-01
-5.55115461e-01 6.57482386e-01 -2.21566975e-01 9.12000835e-02
5.71833909e-01 -1.16278917e-01 -5.58278978e-01 -3.34067583e-01
-5.15091956e-01 1.40046939e-01 2.61077762e-01 2.81293601e-01
7.25675404e-01 -9.55414832e-01 -8.17380488e-01 1.13546893e-01
-3.11423630e-01 2.95967430e-01 6.54637635e-01 1.00778890e+00
-4.29760963e-01 2.24972025e-01 -3.30129713e-02 -4.70622629e-01
-1.26963782e+00 3.61457318e-01 4.61496651e-01 -1.95991054e-01
-1.21584439e+00 1.06789279e+00 5.90599239e-01 3.16744819e-02
3.88047755e-01 -3.85807425e-01 -7.59139284e-02 -9.51704979e-02
6.40168548e-01 2.22064435e-01 2.45165199e-01 -4.86881249e-02
-2.24230334e-01 6.02829278e-01 -3.67138058e-01 4.76927608e-01
1.57934678e+00 3.35433409e-02 -1.27850873e-02 1.13078617e-01
9.93134737e-01 -2.71904647e-01 -1.67860544e+00 -5.64394593e-01
-4.18614894e-01 -5.52162230e-01 6.07621312e-01 -6.12201154e-01
-1.65616298e+00 1.15833044e+00 8.07076097e-01 -2.23250657e-01
1.62158728e+00 -1.59028828e-01 9.40637410e-01 6.82341576e-01
2.23075852e-01 -6.84231341e-01 1.13105021e-01 4.95973647e-01
1.06029379e+00 -1.31063616e+00 1.36356071e-01 -3.30974847e-01
-5.10589957e-01 7.37350404e-01 7.39693820e-01 -2.47182325e-01
6.82435632e-01 3.32834184e-01 7.40822107e-02 4.89866212e-02
-9.33234751e-01 -2.19497383e-01 4.45846543e-02 3.93225074e-01
3.19667459e-01 -2.59806991e-01 -6.07352294e-02 4.11536932e-01
-2.25193903e-01 -1.41596898e-01 5.35405755e-01 4.38572049e-01
-5.75766087e-01 -7.05401182e-01 -1.09377868e-01 5.88185191e-01
-5.43200612e-01 -6.56991303e-01 2.54021764e-01 5.62086344e-01
-7.98457041e-02 8.81093502e-01 -5.78512847e-02 -2.27490216e-01
3.71780753e-01 -3.95732611e-01 3.91042471e-01 -4.03747290e-01
-5.31701982e-01 -1.44072045e-02 -1.44580022e-01 -6.07620955e-01
-4.84489590e-01 -2.64144957e-01 -8.56224179e-01 -5.88577211e-01
-3.62179965e-01 -1.02067105e-01 3.00457597e-01 7.70588934e-01
3.36319476e-01 7.93401420e-01 5.83971858e-01 -9.48632896e-01
-5.24004519e-01 -1.23668730e+00 -4.42495883e-01 2.18677707e-02
5.21252990e-01 -6.15689933e-01 -5.80899775e-01 -1.81976020e-01] | [11.092427253723145, -2.2643020153045654] |
2d507d4e-b99d-4ef6-98cb-b370e4868e3b | basic-tasks-of-sentiment-analysis | 1710.06536 | null | http://arxiv.org/abs/1710.06536v1 | http://arxiv.org/pdf/1710.06536v1.pdf | Basic tasks of sentiment analysis | Subjectivity detection is the task of identifying objective and subjective
sentences. Objective sentences are those which do not exhibit any sentiment.
So, it is desired for a sentiment analysis engine to find and separate the
objective sentences for further analysis, e.g., polarity detection. In
subjective sentences, opinions can often be expressed on one or multiple
topics. Aspect extraction is a subtask of sentiment analysis that consists in
identifying opinion targets in opinionated text, i.e., in detecting the
specific aspects of a product or service the opinion holder is either praising
or complaining about. | ['Soujanya Poria', 'Iti Chaturvedi', 'Erik Cambria'] | 2017-10-18 | null | null | null | null | ['aspect-extraction'] | ['natural-language-processing'] | [ 2.64767528e-01 2.62106866e-01 -3.53838921e-01 -6.65881932e-01
-3.15320849e-01 -9.30299520e-01 3.92001033e-01 7.81865656e-01
3.78854247e-03 6.29111469e-01 5.30418456e-01 -3.65983099e-01
3.34200233e-01 -5.91553569e-01 3.45554315e-02 -7.98952162e-01
4.84491736e-01 8.58924389e-02 -9.67717618e-02 -4.56273317e-01
6.39392495e-01 1.79241523e-01 -1.19495344e+00 2.47576728e-01
3.73917669e-01 1.31078112e+00 -2.52543509e-01 5.88724792e-01
-3.47141504e-01 1.20142436e+00 -1.04960871e+00 -5.64569652e-01
-4.14485574e-01 -5.43523073e-01 -6.52472377e-01 7.99741149e-01
-1.25950621e-02 2.09874004e-01 7.17866659e-01 1.52137220e+00
1.89386874e-01 -3.42911720e-01 5.23829877e-01 -1.07337332e+00
-5.06366074e-01 4.16459173e-01 -6.14028215e-01 5.22011638e-01
6.27950430e-01 -3.29283565e-01 1.34060621e+00 -9.23722029e-01
4.68152672e-01 9.07329679e-01 -6.03023581e-02 2.13473976e-01
-9.06218171e-01 -2.30828121e-01 5.22419393e-01 -2.04899848e-01
-8.31510842e-01 -5.27412415e-01 1.10830581e+00 -6.69929504e-01
6.23267412e-01 4.86630261e-01 4.97947812e-01 8.44875157e-01
6.67999029e-01 9.24953878e-01 1.16487753e+00 -2.09124729e-01
7.25717664e-01 1.03719580e+00 5.89089036e-01 1.69691533e-01
3.97083730e-01 -6.91370964e-01 -6.05318129e-01 -3.68511409e-01
-2.50775397e-01 -1.04038596e-01 -2.92041928e-01 -3.64545994e-02
-7.06254959e-01 1.00565934e+00 -2.68138468e-01 3.20303231e-01
-9.50011671e-01 -7.51918674e-01 6.81744397e-01 6.05518937e-01
8.63605022e-01 6.09604537e-01 -9.59561288e-01 -2.80921996e-01
-6.89279318e-01 4.00566347e-02 1.09024096e+00 6.72228932e-01
7.54231811e-01 -7.23848119e-02 -9.21451394e-03 6.17174447e-01
3.13642740e-01 6.80253148e-01 5.22940755e-01 -3.54526997e-01
2.53166318e-01 1.30325711e+00 2.35102728e-01 -1.19956744e+00
-3.02323073e-01 -2.51854300e-01 -5.07465005e-01 1.03003770e-01
-3.80430311e-01 -5.75891435e-01 -9.37899470e-01 1.26387119e+00
2.98723549e-01 -1.05503392e+00 4.94000643e-01 9.51172173e-01
1.33000457e+00 7.49040246e-01 1.05448827e-01 -7.80359447e-01
1.87424099e+00 -4.58850771e-01 -9.77268457e-01 -6.53060913e-01
2.98556983e-01 -1.28799284e+00 7.53269553e-01 3.98285031e-01
-1.00824976e+00 -8.98825750e-02 -1.06743526e+00 2.58406550e-01
-4.91026849e-01 1.09390765e-01 5.35156548e-01 6.78346574e-01
-4.34705913e-01 -8.16722289e-02 -3.60201061e-01 -1.81693695e-02
1.92733645e-01 2.08607182e-01 -3.07774156e-01 5.13935566e-01
-1.15635276e+00 6.00956321e-01 9.87809151e-02 1.35278419e-01
-1.99201390e-01 -1.77622288e-01 -1.01970899e+00 8.55961367e-02
4.81724799e-01 -2.00780421e-01 1.49764776e+00 -1.88942671e+00
-1.15308917e+00 1.01550341e+00 -7.60689855e-01 6.86702058e-02
-2.61029303e-01 4.52917330e-02 -9.49938238e-01 1.43702745e-01
5.69772422e-01 -2.45571449e-01 1.06426692e+00 -9.78304803e-01
-9.63473916e-01 -7.13147521e-01 3.45071405e-01 3.35266560e-01
-5.46364129e-01 6.42067015e-01 -2.27237970e-01 -4.42827672e-01
3.48334491e-01 -7.62480497e-01 -2.65005492e-02 -7.23235011e-01
-7.54024386e-01 -4.23809439e-01 1.11329973e+00 -5.43402076e-01
1.09536517e+00 -2.15532017e+00 -1.01633184e-01 4.83168304e-01
4.32099223e-01 1.01213604e-01 3.02034050e-01 1.11773789e-01
-2.55871117e-01 2.94960022e-01 2.03167975e-01 1.76675320e-01
-7.12692663e-02 -2.44079396e-01 -5.34722865e-01 3.27453345e-01
5.52830875e-01 5.65396547e-01 -7.66627014e-01 -5.13216257e-01
-3.73683929e-01 -6.47872640e-03 -6.57105260e-03 2.34526470e-01
-2.28701621e-01 3.41148049e-01 -9.55348074e-01 9.70390916e-01
4.74447995e-01 -4.74803716e-01 2.17527732e-01 -1.75551161e-01
-3.45259249e-01 9.39489901e-01 -9.34611380e-01 5.71680307e-01
-4.52923506e-01 8.98405075e-01 3.19872290e-01 -8.84730220e-01
9.20596540e-01 6.51825190e-01 2.89593428e-01 -7.07452893e-01
4.50280964e-01 2.36387432e-01 -4.36094888e-02 -4.95404035e-01
9.23610032e-01 -4.69053656e-01 -4.58759427e-01 5.18516243e-01
-4.53379571e-01 -1.93019703e-01 7.00501621e-01 1.57356441e-01
8.58217239e-01 -7.88246095e-01 6.81019068e-01 -3.03598076e-01
8.59487653e-01 2.19204322e-01 6.24298096e-01 -1.06791012e-01
-1.15521543e-01 2.72580862e-01 1.03523088e+00 -1.99291065e-01
-5.33058167e-01 -6.12508714e-01 2.33551282e-02 9.88994777e-01
1.36727288e-01 -4.39594001e-01 -3.16750824e-01 -7.77180076e-01
-4.02362943e-01 4.77115899e-01 -5.96659660e-01 1.07405663e-01
-1.04280472e-01 -7.15796947e-01 -5.11131704e-01 2.78504550e-01
1.55566648e-01 -1.09944069e+00 -5.17426193e-01 1.38109714e-01
-2.11654440e-01 -9.74605083e-01 -4.58460599e-01 3.18152130e-01
-7.34614909e-01 -1.00903773e+00 -6.05169177e-01 -8.51725519e-01
1.05044723e+00 4.31384444e-01 1.35442269e+00 -1.46699429e-01
5.21680892e-01 5.87415509e-03 -6.55618250e-01 -9.26702917e-01
-3.30463678e-01 1.83735281e-01 -5.32447621e-02 3.04823041e-01
9.89728332e-01 -1.00622341e-01 -5.45088947e-01 1.61495417e-01
-8.22728813e-01 -4.68508244e-01 6.26618087e-01 2.09960610e-01
4.51234341e-01 5.53884387e-01 5.69525063e-01 -1.16646290e+00
1.29970086e+00 -5.16681552e-01 -3.56675446e-01 5.17900512e-02
-6.42851710e-01 -3.17468017e-01 7.37850606e-01 -2.83794165e-01
-1.02212584e+00 -3.35084498e-01 -1.91526070e-01 5.20736814e-01
-1.69127151e-01 1.06650031e+00 -3.93303186e-01 5.25870800e-01
4.04367924e-01 2.56287962e-01 -4.25909877e-01 4.92238551e-02
-3.69941592e-01 1.24097264e+00 -2.08561629e-01 1.04442611e-01
3.55789214e-01 5.43851733e-01 -4.04107779e-01 -1.00893843e+00
-1.50793839e+00 -9.09986377e-01 -1.57445669e-01 -3.53638351e-01
8.13998163e-01 -7.41304457e-01 -5.70351779e-01 2.14130044e-01
-1.10658789e+00 6.85150683e-01 -2.20491350e-01 3.82393867e-01
-6.76437020e-02 1.17531434e-01 -1.94688722e-01 -9.86721277e-01
-6.24051988e-01 -1.16106570e+00 9.15309846e-01 6.97595537e-01
-1.09292996e+00 -1.04668784e+00 9.14627612e-02 5.20919681e-01
1.98370576e-01 -5.99701367e-02 8.15345287e-01 -9.51840103e-01
2.87447095e-01 -8.17419469e-01 1.57594576e-01 6.74128056e-01
5.96642792e-01 2.09239691e-01 -7.15248168e-01 -8.46174955e-02
6.36283159e-01 -3.68160158e-01 3.93024087e-01 3.50763708e-01
3.84044021e-01 -5.14341235e-01 1.93569511e-02 -3.37953866e-01
9.85326588e-01 6.08453393e-01 4.57411081e-01 3.36898148e-01
2.72742391e-01 8.60849321e-01 8.36915314e-01 3.36062044e-01
2.01214686e-01 2.55915672e-01 1.31996557e-01 -2.93696225e-01
6.41539454e-01 2.06798196e-01 7.63182819e-01 8.94984424e-01
2.74436563e-01 -3.55976135e-01 -6.61148548e-01 7.41280794e-01
-1.57283914e+00 -8.64532590e-01 -3.89298737e-01 1.70253932e+00
8.40563774e-01 6.56180382e-01 -8.45439732e-02 2.94969469e-01
8.40615690e-01 5.00140846e-01 -6.47146881e-01 -8.92021120e-01
-1.70954794e-01 5.42856380e-02 7.68831074e-02 2.53336400e-01
-1.15635717e+00 5.98688781e-01 5.79402828e+00 3.36818218e-01
-1.40478778e+00 -1.61623195e-01 8.42359841e-01 8.94613862e-02
-5.53483248e-01 7.72604197e-02 -9.51810241e-01 4.06595469e-01
6.13904536e-01 -5.03334463e-01 -4.56053376e-01 1.14320850e+00
6.06790841e-01 -6.05067074e-01 -7.24584460e-01 4.76769686e-01
2.51302242e-01 -8.36984158e-01 -2.35314742e-01 -5.55072725e-02
9.47210193e-01 -3.62960786e-01 8.03124309e-02 3.98040526e-02
-2.91844923e-03 -5.51228464e-01 8.44356358e-01 1.51474133e-01
5.57231009e-02 -7.25879967e-01 1.30913341e+00 2.14014500e-01
-9.27898467e-01 2.54076153e-01 -2.11149395e-01 -2.92799741e-01
3.88435602e-01 1.28136837e+00 -4.46933061e-01 1.85014993e-01
6.94166005e-01 7.66682804e-01 -2.38237247e-01 4.96134549e-01
-8.07211936e-01 8.52235556e-01 1.44722044e-01 -5.37665486e-01
2.23846242e-01 -2.39849225e-01 6.64308012e-01 8.50316286e-01
-1.02956276e-02 4.39337045e-02 5.38659059e-02 4.53025222e-01
-1.21275768e-01 5.25798440e-01 -3.89790326e-01 -6.23606205e-01
-1.12900719e-01 1.66097975e+00 -1.06953204e+00 -4.67363983e-01
-5.50709248e-01 7.97598004e-01 -2.42128417e-01 3.55698377e-01
-2.48513371e-01 -4.82125789e-01 6.78095341e-01 -9.33193415e-02
5.34118056e-01 2.39876673e-01 -5.60009778e-01 -1.25549972e+00
5.87891281e-01 -1.02670896e+00 1.19931161e-01 -6.41841114e-01
-1.22067976e+00 8.72141659e-01 -6.93933129e-01 -1.19574499e+00
-2.49539137e-01 -5.92670023e-01 -1.15365171e+00 9.44498420e-01
-1.53910863e+00 -5.10624170e-01 -3.79811451e-02 2.50260979e-01
5.67279637e-01 -1.98035017e-01 5.81876040e-01 -1.43625304e-01
-5.57375908e-01 -6.61534294e-02 -3.17922205e-01 7.45654926e-02
4.16933656e-01 -1.19378901e+00 -1.20523185e-01 7.72841454e-01
-2.32803375e-01 7.75397956e-01 1.29778326e+00 -5.55768669e-01
-1.14477396e+00 -7.99726307e-01 1.62376082e+00 -2.57064968e-01
8.75294864e-01 -1.48559764e-01 -5.25791347e-01 2.82624751e-01
1.61824629e-01 -7.35464394e-01 1.17587411e+00 3.77595007e-01
-3.80234793e-02 -1.43641204e-01 -9.36685801e-01 6.27545893e-01
1.55129265e-02 -6.03923976e-01 -8.45267057e-01 3.00294131e-01
4.99688238e-01 3.64901265e-03 -5.35781205e-01 2.14429125e-01
4.12634999e-01 -9.28177178e-01 1.61746845e-01 -5.95215976e-01
7.97905922e-01 -3.53447914e-01 5.91887766e-03 -1.43612075e+00
-1.21754198e-03 -3.94869775e-01 -5.89490756e-02 1.44370401e+00
1.00114310e+00 -5.36444843e-01 9.96999979e-01 7.53046930e-01
1.67372838e-01 -4.53731775e-01 -4.62807536e-01 -2.78157413e-01
-3.88277858e-01 -2.81842440e-01 4.23121780e-01 8.22408497e-01
5.72936475e-01 1.30602038e+00 1.65576246e-02 1.76171079e-01
4.70607057e-02 6.81924284e-01 5.03418565e-01 -1.13456166e+00
1.69957027e-01 -4.17690575e-01 -3.20657909e-01 -9.91288781e-01
1.04736216e-01 -2.68646359e-01 1.09437875e-01 -1.67157555e+00
1.85681269e-01 3.44456166e-01 -1.19498521e-01 1.21976621e-01
-3.05480421e-01 8.35446566e-02 -3.42076302e-01 3.00376564e-01
-7.66845286e-01 3.71276557e-01 1.21724057e+00 -4.75502402e-01
-4.91818935e-01 6.92578316e-01 -1.75460219e+00 1.19002664e+00
1.14872706e+00 -4.29882020e-01 -4.60180372e-01 2.09208772e-01
1.23438883e+00 -7.35608265e-02 -3.66778553e-01 -2.26439118e-01
8.67047682e-02 -4.07350987e-01 5.15571348e-02 -9.08314884e-01
1.74258389e-02 -7.63578355e-01 -5.10505378e-01 2.86133587e-01
-1.68293223e-01 2.40418524e-01 -1.42076418e-01 1.58197358e-01
-8.34667265e-01 -4.58625197e-01 5.00222385e-01 -8.98033977e-02
-6.10888004e-01 3.04375924e-02 -1.10321617e+00 -1.26159890e-02
8.49565685e-01 -1.30562752e-01 -3.46997291e-01 -7.17733800e-01
-7.18708277e-01 2.54102826e-01 3.60023588e-01 3.85124236e-01
8.42608333e-01 -9.25506353e-01 -6.28496826e-01 -6.08402453e-02
5.67560673e-01 -2.00673088e-01 -1.44896388e-01 1.02891409e+00
-4.99579348e-02 3.06293011e-01 3.23396206e-01 -2.02450290e-01
-1.67742503e+00 2.98218429e-01 -2.45202631e-02 -3.21103662e-01
1.24453686e-01 6.96092665e-01 1.95106611e-01 -2.54874639e-02
-9.11441520e-02 -6.68273941e-02 -1.10132289e+00 7.38973260e-01
8.39129090e-01 3.07808649e-02 3.35911810e-01 -1.08983207e+00
-6.93851054e-01 2.91096330e-01 -5.53184330e-01 -1.02527894e-01
1.19576621e+00 -1.99532345e-01 -7.64091015e-01 7.31094241e-01
1.30964220e+00 4.98511553e-01 -4.50251430e-01 2.80483980e-02
1.42661914e-01 -8.32281858e-02 3.45140576e-01 -7.73963630e-01
-1.12909746e+00 5.28721750e-01 1.54553875e-01 1.11894941e+00
1.21236348e+00 4.12241399e-01 6.12083554e-01 2.71923691e-01
2.95989998e-02 -1.47357988e+00 -1.83579251e-02 7.59478033e-01
8.02332997e-01 -1.58941495e+00 1.39528751e-01 -3.69439960e-01
-1.06170177e+00 1.07102048e+00 3.28235328e-01 2.62409598e-01
9.59147871e-01 6.53873011e-02 7.02697098e-01 -1.09236479e+00
-6.70212030e-01 -2.62385607e-01 5.88338733e-01 5.76272570e-02
8.21170688e-01 2.20602080e-01 -7.98845768e-01 7.35693634e-01
-3.86044741e-01 -3.97358298e-01 6.65498495e-01 1.20294690e+00
-6.73932076e-01 -8.11730564e-01 -2.02735439e-01 6.10315800e-01
-1.15976763e+00 -1.47752255e-01 -1.19654989e+00 -8.15683380e-02
-1.33554876e-01 1.83876634e+00 -9.85078514e-02 -2.80332446e-01
6.20481968e-01 -2.13820592e-01 -4.27994609e-01 -9.35484767e-01
-7.36595273e-01 1.69662118e-01 4.47066724e-01 -2.35207185e-01
-9.46803808e-01 -6.68674052e-01 -1.11418414e+00 -1.40465543e-01
-4.86262321e-01 7.79890060e-01 9.08463717e-01 1.20320845e+00
5.90548180e-02 3.45337212e-01 1.11210227e+00 -6.19720556e-02
8.17006454e-04 -9.38516617e-01 -8.17867815e-01 4.03055668e-01
6.34835482e-01 -1.31637290e-01 -7.52395153e-01 1.35933280e-01] | [11.313948631286621, 6.728762149810791] |
560e8fc6-7605-4b9a-9395-4a62a82221b5 | proactively-reducing-the-hate-intensity-of | 2206.04007 | null | https://arxiv.org/abs/2206.04007v1 | https://arxiv.org/pdf/2206.04007v1.pdf | Proactively Reducing the Hate Intensity of Online Posts via Hate Speech Normalization | Curbing online hate speech has become the need of the hour; however, a blanket ban on such activities is infeasible for several geopolitical and cultural reasons. To reduce the severity of the problem, in this paper, we introduce a novel task, hate speech normalization, that aims to weaken the intensity of hatred exhibited by an online post. The intention of hate speech normalization is not to support hate but instead to provide the users with a stepping stone towards non-hate while giving online platforms more time to monitor any improvement in the user's behavior. To this end, we manually curated a parallel corpus - hate texts and their normalized counterparts (a normalized text is less hateful and more benign). We introduce NACL, a simple yet efficient hate speech normalization model that operates in three stages - first, it measures the hate intensity of the original sample; second, it identifies the hate span(s) within it; and finally, it reduces hate intensity by paraphrasing the hate spans. We perform extensive experiments to measure the efficacy of NACL via three-way evaluation (intrinsic, extrinsic, and human-study). We observe that NACL outperforms six baselines - NACL yields a score of 0.1365 RMSE for the intensity prediction, 0.622 F1-score in the span identification, and 82.27 BLEU and 80.05 perplexity for the normalized text generation. We further show the generalizability of NACL across other platforms (Reddit, Facebook, Gab). An interactive prototype of NACL was put together for the user study. Further, the tool is being deployed in a real-world setting at Wipro AI as a part of its mission to tackle harmful content on online platforms. | ['Tanmoy Chakraborty', 'Md Shad Akhtar', 'Mohammad Aflah Khan', 'Manjot Bedi', 'Sarah Masud'] | 2022-06-08 | null | null | null | null | ['hate-speech-normalization', 'hate-span-identification', 'hate-intensity-prediction'] | ['natural-language-processing', 'natural-language-processing', 'natural-language-processing'] | [ 6.97716996e-02 -1.75286923e-02 7.47996569e-02 1.32025003e-01
-5.23805320e-01 -8.06328118e-01 5.93984187e-01 1.67428270e-01
-1.90882310e-01 4.00255978e-01 5.91062307e-01 -1.12220071e-01
1.10260911e-01 -4.25865710e-01 -3.53712022e-01 -6.36488080e-01
2.02268630e-01 -2.45457292e-01 -5.82092926e-02 -4.62519228e-01
4.48841095e-01 3.46791446e-01 -1.27695572e+00 1.36875868e-01
9.10419703e-01 1.97114080e-01 -2.71770775e-01 7.87731767e-01
3.62558573e-01 1.10527647e+00 -1.09258497e+00 -7.51176834e-01
5.46207419e-03 -4.36426461e-01 -6.95301950e-01 -4.16181842e-03
6.86133623e-01 -5.67650080e-01 -3.99922699e-01 1.03066874e+00
6.31181359e-01 8.98653865e-02 4.76416111e-01 -1.42926717e+00
-9.15707111e-01 5.92401087e-01 -6.06174767e-01 2.83950686e-01
2.94362843e-01 3.97766441e-01 8.67248118e-01 -5.38543105e-01
5.42464554e-01 1.10524905e+00 7.01252639e-01 6.42715812e-01
-1.26883256e+00 -8.87553990e-01 -3.89843076e-01 -1.94284350e-01
-1.23289454e+00 -4.36020613e-01 8.04012775e-01 -8.37486386e-01
7.27478027e-01 4.25512850e-01 4.69465941e-01 1.49508107e+00
-3.19767483e-02 5.39383709e-01 1.07154119e+00 -3.22460622e-01
-3.73520590e-02 4.16598648e-01 3.09388906e-01 4.57436085e-01
2.10130334e-01 -3.77601653e-01 -6.23376369e-01 -2.56646752e-01
1.83756515e-01 -3.03227026e-02 -1.76368341e-01 3.36580664e-01
-6.84325814e-01 1.02776289e+00 1.58247665e-01 6.63738370e-01
-1.21798329e-01 -3.28504741e-02 5.57691157e-01 1.93153191e-02
7.25865722e-01 8.97732317e-01 1.28142059e-01 -6.43814981e-01
-1.02529168e+00 4.71215159e-01 7.74432600e-01 5.43954492e-01
4.37583029e-01 -1.24352694e-01 -4.48768258e-01 8.67583752e-01
-1.31942540e-01 6.20984077e-01 2.94237852e-01 -8.38394940e-01
3.74614358e-01 5.14171720e-01 1.05852611e-01 -1.31402647e+00
-3.26374561e-01 -3.40281487e-01 -6.51809335e-01 -9.12388563e-02
4.81143802e-01 -4.43772614e-01 -7.99418330e-01 1.79464233e+00
4.96838279e-02 -4.48756553e-02 -5.48999786e-01 7.18469381e-01
3.28817338e-01 7.19163954e-01 3.05521876e-01 -1.33035854e-01
1.28429580e+00 -9.51232970e-01 -1.19429016e+00 -2.01048985e-01
1.06287324e+00 -1.10265970e+00 1.38921988e+00 4.83510822e-01
-8.64205480e-01 -5.32389805e-02 -1.08652103e+00 -1.67626470e-01
-6.62167847e-01 -1.86694443e-01 1.92795098e-01 1.22134829e+00
-7.49501169e-01 5.74110448e-01 -2.80047089e-01 -5.83641827e-01
2.54157931e-01 -6.34195954e-02 -3.68027449e-01 1.50205404e-01
-1.31612897e+00 1.08525729e+00 -5.11185499e-03 -2.27375671e-01
-5.76831460e-01 -9.36143100e-01 -7.81668663e-01 1.11474931e-01
2.18925580e-01 4.88110855e-02 1.12422478e+00 -1.02615130e+00
-1.11577225e+00 9.02993560e-01 1.29409507e-01 -2.64162987e-01
3.04561973e-01 -5.78151882e-01 -4.76454765e-01 -1.60430163e-01
1.05629899e-01 2.95863539e-01 1.00786471e+00 -1.22038341e+00
-2.57498413e-01 -3.62818629e-01 2.78239995e-02 -4.16581854e-02
-1.04730511e+00 4.40386266e-01 4.50765565e-02 -8.26187789e-01
-8.34515631e-01 -1.12284636e+00 4.62687194e-01 -6.12156034e-01
-6.28114641e-01 5.14999852e-02 1.25135231e+00 -1.36525416e+00
2.21657038e+00 -2.44651842e+00 -1.06987707e-01 2.00421393e-01
3.88396472e-01 7.22824574e-01 1.94842853e-02 6.59863055e-01
-7.34836832e-02 6.68888211e-01 -2.43011057e-01 -3.91751796e-01
1.32275864e-01 -8.34159181e-02 -1.98147610e-01 6.66315675e-01
1.78485855e-01 4.58114117e-01 -8.62978578e-01 -2.04786435e-01
2.08387282e-02 5.90409935e-01 -6.48485065e-01 2.81260759e-01
2.80601531e-01 -3.07362173e-02 2.37408817e-01 4.54387248e-01
5.81922472e-01 2.98777640e-01 -9.62577388e-02 7.65608624e-02
-4.63419706e-01 1.94955215e-01 -6.31893218e-01 1.17305529e+00
-2.21395567e-01 1.00793076e+00 1.10880814e-01 -1.40916273e-01
7.48053551e-01 1.99663386e-01 2.69800484e-01 -6.36102557e-01
2.33332068e-01 -3.16175632e-02 6.69679269e-02 -7.08899677e-01
7.66696811e-01 -1.70958007e-03 -3.15868944e-01 4.50274229e-01
-1.52322557e-02 1.45632168e-02 2.86528140e-01 4.68978852e-01
1.31941414e+00 -1.61633924e-01 1.54596165e-01 -1.44579396e-01
3.33120733e-01 -2.62661487e-01 8.53175893e-02 4.98841345e-01
-6.99686229e-01 4.28546488e-01 9.69154835e-01 -1.28642634e-01
-1.48864663e+00 -8.08134258e-01 1.69346601e-01 1.45395660e+00
-3.41182560e-01 -7.26265192e-01 -1.43258762e+00 -7.50471294e-01
-5.55607714e-02 1.31797647e+00 -9.21022117e-01 -4.31140721e-01
-4.65165645e-01 -5.87381363e-01 1.03472686e+00 -5.53404093e-02
4.06741619e-01 -9.98882473e-01 -5.34611285e-01 -2.61403531e-01
-3.37949872e-01 -9.57991362e-01 -8.99815917e-01 -1.12505332e-01
9.08622220e-02 -7.43003190e-01 -5.43461859e-01 -3.62993717e-01
2.97681302e-01 4.82493728e-01 5.41657984e-01 3.17887098e-01
-1.75200343e-01 1.79536924e-01 -5.78802466e-01 -4.95898366e-01
-5.96872866e-01 4.07042861e-01 3.96411940e-02 1.61931515e-02
5.20689011e-01 -4.96586472e-01 -3.95730019e-01 1.70299634e-01
-1.20313323e+00 -1.51549429e-01 2.91462868e-01 5.57254851e-01
-4.21227872e-01 5.89953810e-02 3.23861033e-01 -9.66197371e-01
1.00752974e+00 -8.05273294e-01 2.37440169e-02 -1.25305176e-01
-4.74035442e-01 -4.33741659e-01 7.15315819e-01 -5.32728136e-01
-1.03828526e+00 -3.02173018e-01 -1.13796107e-01 -1.73183918e-01
-1.49658620e-01 1.81464925e-01 6.41821772e-02 1.82720423e-01
8.04169476e-01 -8.29779953e-02 1.06982179e-01 -4.54014659e-01
3.68488878e-01 1.04991925e+00 4.54211593e-01 -2.10680351e-01
1.25401080e+00 5.30457944e-02 -4.02206510e-01 -1.24451828e+00
-9.87848222e-01 -6.18149519e-01 -5.35794556e-01 -4.32009935e-01
1.02397645e+00 -5.84842026e-01 -6.82753265e-01 6.65986598e-01
-1.25773120e+00 -3.97308171e-01 3.05822790e-01 -1.17427714e-01
7.38741755e-02 4.67592478e-01 -6.02088332e-01 -1.16827130e+00
-5.58229029e-01 -7.18339562e-01 7.04540908e-01 3.17597836e-01
-6.10817373e-01 -8.55407059e-01 3.89281273e-01 6.70769930e-01
4.20203567e-01 5.58021605e-01 9.07962441e-01 -9.33062077e-01
2.86109120e-01 -2.87406445e-01 -2.99487442e-01 7.36764848e-01
1.27187699e-01 5.17552137e-01 -1.13763249e+00 -3.31428707e-01
-1.92201212e-01 -3.21080923e-01 1.74258918e-01 -2.25117341e-01
7.66452849e-01 -7.54186690e-01 1.37451679e-01 1.23648241e-01
1.22365069e+00 5.13208210e-02 9.26691353e-01 5.67968607e-01
7.71829665e-01 6.47824824e-01 3.48824114e-01 7.62237966e-01
5.15526533e-02 6.51601613e-01 2.95523226e-01 -1.20914698e-01
-1.62640959e-01 -6.32230818e-01 7.60965407e-01 9.26422894e-01
1.66396555e-02 -2.57486850e-01 -9.94466186e-01 5.61331153e-01
-1.56114054e+00 -1.33072817e+00 -4.44879770e-01 2.18985033e+00
6.92100465e-01 2.31170982e-01 6.88999057e-01 9.97834429e-02
8.28845859e-01 2.48519748e-01 -4.60420251e-02 -9.61679101e-01
2.62886304e-02 -2.93163564e-02 5.31946540e-01 4.61092919e-01
-1.07625532e+00 9.65357542e-01 5.72735023e+00 8.25724959e-01
-9.81141448e-01 3.29169154e-01 6.59532547e-01 -2.60799289e-01
5.39667048e-02 -2.64814526e-01 -4.89360392e-01 8.40403616e-01
1.20812297e+00 -2.20034495e-01 7.97537684e-01 7.16643870e-01
5.85963368e-01 -1.04785144e-01 -7.55629957e-01 8.25399995e-01
4.29355204e-01 -8.23505461e-01 -2.25063190e-01 2.54079729e-01
5.45234144e-01 -2.62309283e-01 2.96503693e-01 5.37443757e-01
-3.09674740e-02 -1.05839229e+00 8.13526452e-01 1.96298212e-01
5.58896482e-01 -1.00566888e+00 7.97925830e-01 3.76503378e-01
-5.38018227e-01 -1.69091105e-01 -1.92486227e-03 -3.38013351e-01
-9.58281457e-02 3.09253067e-01 -9.12050188e-01 -9.52107161e-02
6.28186405e-01 1.39088646e-01 -8.80994797e-01 6.94685817e-01
-3.18207145e-01 8.72626662e-01 8.31333250e-02 -1.08732373e-01
3.99973035e-01 -7.14799240e-02 6.18674159e-01 1.78097463e+00
1.17173314e-01 -1.21443786e-01 -3.24730992e-01 7.45381117e-01
-2.53414541e-01 1.99829623e-01 -7.79168606e-01 -4.97443378e-01
7.21466482e-01 1.56858265e+00 -2.31603667e-01 1.33315817e-01
-3.35191399e-01 1.15034330e+00 4.56169575e-01 9.86699760e-02
-1.32549155e+00 -7.99189210e-01 8.05848658e-01 2.98344016e-01
-1.90721691e-01 -1.63885117e-01 -3.70000899e-01 -7.49759734e-01
-2.63156742e-01 -1.25693274e+00 6.43224120e-02 -8.05902839e-01
-1.21342385e+00 1.35051355e-01 -1.92420021e-01 -6.22996569e-01
5.86035959e-02 -4.21618998e-01 -6.11155748e-01 8.12792242e-01
-9.51347113e-01 -1.20102060e+00 -3.82862926e-01 2.84519196e-01
4.30358857e-01 3.25628042e-01 6.59540176e-01 4.95954275e-01
-1.11488986e+00 8.08972776e-01 -6.61447272e-02 4.69357997e-01
1.13578606e+00 -1.19251561e+00 2.58274198e-01 1.07317030e+00
-4.17021573e-01 8.44137788e-01 1.04805326e+00 -7.85177648e-01
-1.03130352e+00 -9.64104950e-01 1.10270321e+00 -8.04950476e-01
1.32602870e+00 -6.97714508e-01 -9.40705597e-01 4.88147587e-01
6.67797208e-01 -7.35944510e-01 9.77149963e-01 1.16186760e-01
-6.41660988e-01 2.05950007e-01 -1.23726797e+00 8.42946351e-01
8.01867008e-01 -5.84897161e-01 -4.05788392e-01 3.47212225e-01
8.96227002e-01 -5.10424785e-02 -9.40529108e-01 -2.95493722e-01
6.27201736e-01 -1.10414755e+00 3.36968273e-01 -6.20806813e-01
8.37994576e-01 3.06244418e-02 1.29090557e-02 -1.31426823e+00
-7.14833319e-01 -1.19971025e+00 -6.72370717e-02 1.84153414e+00
3.82455975e-01 -1.93320200e-01 4.38415021e-01 8.35751474e-01
-1.00337625e-01 -4.56550330e-01 -5.01041174e-01 -5.66975236e-01
1.90512463e-01 -1.79022089e-01 2.25419804e-01 1.41321421e+00
2.77659088e-01 5.93044817e-01 -8.63831460e-01 9.32169631e-02
5.10482609e-01 -9.02465165e-01 1.04880333e+00 -8.86663258e-01
1.05222069e-01 -5.19675255e-01 -1.44186258e-01 -1.25919253e-01
5.92543706e-02 -5.33683538e-01 -1.89550057e-01 -9.77343321e-01
6.52914405e-01 2.07780287e-01 6.87055737e-02 7.36175060e-01
-1.62529007e-01 3.88831705e-01 6.81767941e-01 1.19173944e-01
-4.67365295e-01 1.25039369e-01 9.07838404e-01 -1.46064693e-02
-2.80606389e-01 -4.27806914e-01 -8.92184138e-01 6.16765916e-01
9.28371489e-01 -5.22596955e-01 -2.87978083e-01 -6.83733970e-02
2.84890383e-01 -5.65549433e-01 2.85173684e-01 -9.85924482e-01
-2.32461229e-01 -1.34875447e-01 -5.74117806e-03 -2.29834884e-01
1.02506913e-01 -5.62277675e-01 -1.61064461e-01 4.76559579e-01
-3.27172190e-01 1.70021176e-01 1.74238443e-01 4.83527817e-02
2.25397646e-01 -4.28728253e-01 1.00051022e+00 2.79069334e-01
-1.14672653e-01 -1.28559202e-01 -6.72875822e-01 2.15144783e-01
1.07643592e+00 -4.60736305e-02 -9.09015834e-01 -5.70759833e-01
-2.33960539e-01 -1.09575182e-01 6.36661947e-01 5.01455307e-01
1.67377219e-01 -1.13878703e+00 -6.30763471e-01 -4.32982631e-02
1.61201760e-01 -9.83939052e-01 2.47947589e-01 1.00180614e+00
-5.50692976e-01 2.72693038e-01 -1.79758981e-01 1.07180893e-01
-1.60610485e+00 6.80708826e-01 1.35116816e-01 -2.24919736e-01
-2.41865918e-01 4.22006398e-01 -3.85363325e-02 -1.13715068e-01
1.41103879e-01 3.27295870e-01 -1.15629382e-01 2.60581285e-01
8.30858111e-01 8.05861533e-01 2.06484292e-02 -9.54230964e-01
-2.13743523e-01 -1.01103775e-01 -2.05884337e-01 -2.80285385e-02
1.30058706e+00 -1.14863001e-01 -3.60103995e-01 3.30488831e-01
1.42160547e+00 5.40396988e-01 -8.90063047e-01 2.88020760e-01
1.00467250e-01 -6.39473200e-01 3.08502875e-02 -9.85141039e-01
-4.38740432e-01 5.67846775e-01 4.59464073e-01 7.67618656e-01
9.71558452e-01 -3.72989565e-01 1.03513539e+00 6.62028193e-02
-2.63017476e-01 -1.33867192e+00 2.95759737e-01 5.97892880e-01
8.51259470e-01 -8.40699732e-01 -6.35510981e-02 -2.54312247e-01
-7.91622102e-01 9.33276713e-01 6.84194624e-01 1.18967243e-01
1.70228750e-01 1.64526522e-01 -3.72013971e-02 -1.73606664e-01
-4.76397127e-01 -1.05295107e-02 1.51981801e-01 6.44872785e-01
7.80239761e-01 9.47899520e-02 -4.39037472e-01 6.15785778e-01
-4.43900406e-01 -4.60037887e-01 9.07834768e-01 7.50063598e-01
-5.57556093e-01 -6.13277256e-01 -6.30967498e-01 2.17669591e-01
-6.97314918e-01 -1.35655597e-01 -1.02906263e+00 1.00119328e+00
3.52600634e-01 1.10616851e+00 -3.13175112e-01 -8.43293309e-01
4.05758202e-01 2.34409422e-01 1.61223218e-01 -4.59627926e-01
-1.08344507e+00 4.35551889e-02 4.10397202e-01 -3.20974588e-01
-9.72811133e-03 -5.96539974e-01 -7.49912977e-01 -1.09777868e+00
-2.16923475e-01 2.82908771e-02 7.86215425e-01 7.61492074e-01
3.53757262e-01 3.09160382e-01 8.21961522e-01 -6.09421015e-01
-3.50053936e-01 -1.19465649e+00 -6.28983617e-01 1.07276964e+00
2.44010195e-01 -5.10272741e-01 -7.00658202e-01 2.24820860e-02] | [8.742560386657715, 10.552167892456055] |
80f81fb8-c837-410b-b5b5-e17dd4c5b6c8 | towards-hardware-aware-tractable-learning-of | null | null | http://papers.nips.cc/paper/9525-towards-hardware-aware-tractable-learning-of-probabilistic-models | http://papers.nips.cc/paper/9525-towards-hardware-aware-tractable-learning-of-probabilistic-models.pdf | Towards Hardware-Aware Tractable Learning of Probabilistic Models | Smart portable applications increasingly rely on edge computing due to privacy and latency concerns. But guaranteeing always-on functionality comes with two major challenges: heavily resource-constrained hardware; and dynamic application conditions. Probabilistic models present an ideal solution to these challenges: they are robust to missing data, allow for joint predictions and have small data needs. In addition, ongoing efforts in field of tractable learning have resulted in probabilistic models with strict inference efficiency guarantees. However, the current notions of tractability are often limited to model complexity, disregarding the hardware's specifications and constraints. We propose a novel resource-aware cost metric that takes into consideration the hardware's properties in determining whether the inference task can be efficiently deployed. We use this metric to evaluate the performance versus resource trade-off relevant to the application of interest, and we propose a strategy that selects the device-settings that can optimally meet users' requirements. We showcase our framework on a mobile activity recognition scenario, and on a variety of benchmark datasets representative of the field of tractable learning and of the applications of interest. | ['Guy Van Den Broeck', 'Wannes Meert', 'Marian Verhelst', 'Laura I. Galindez Olascoaga', 'Nimish Shah'] | 2019-12-01 | null | null | null | neurips-2019-12 | ['small-data'] | ['computer-vision'] | [ 3.50504547e-01 -1.92651153e-01 -7.53691792e-01 -3.85766923e-01
-9.66374159e-01 -5.42147756e-01 1.28918812e-01 -1.35231659e-01
-2.96162874e-01 7.52666771e-01 -3.71034332e-02 -6.44305170e-01
-3.93540919e-01 -5.27057707e-01 -8.53829682e-01 -5.99594176e-01
-2.89589614e-01 2.11983427e-01 6.29261509e-02 5.10613739e-01
1.49943665e-01 3.12975556e-01 -1.66375291e+00 5.44400848e-02
4.76142615e-01 1.54010332e+00 8.95342156e-02 7.81063974e-01
2.99517483e-01 6.84665442e-01 -2.85228193e-01 -2.90273041e-01
3.04701656e-01 1.08885109e-01 -3.92121702e-01 -4.69816625e-02
1.16868742e-01 -3.16770613e-01 -3.80645961e-01 9.72265303e-01
5.50633252e-01 -1.74682364e-01 3.50707859e-01 -1.76734114e+00
-2.30475053e-01 5.95480978e-01 -5.05694151e-01 4.21299547e-01
1.82506353e-01 -2.65392900e-01 1.04820585e+00 -4.61675048e-01
3.49457800e-01 4.82103914e-01 7.62677729e-01 4.95765895e-01
-1.16414165e+00 -5.96923411e-01 3.60237867e-01 3.44925195e-01
-1.62247217e+00 -9.11227345e-01 5.03125250e-01 -1.39795154e-01
1.00131226e+00 7.12783515e-01 1.59953594e-01 1.14496601e+00
2.05407709e-01 1.08057106e+00 8.45328152e-01 -3.19977760e-01
7.56074011e-01 3.48969281e-01 2.43505135e-01 4.03470427e-01
5.19900978e-01 -5.98129779e-02 -8.44350696e-01 -5.01985908e-01
2.29056552e-01 1.26073927e-01 -3.87057334e-01 -6.49823368e-01
-6.74746931e-01 2.86713362e-01 -2.75274187e-01 2.82293987e-02
-3.13120037e-01 5.47474444e-01 3.32212687e-01 8.88778120e-02
2.41523474e-01 -8.82789493e-02 -9.17956173e-01 -5.55376887e-01
-1.19107008e+00 -8.94704014e-02 9.89412367e-01 1.44772768e+00
4.35507745e-01 -2.32355535e-01 -2.38608852e-01 2.70796746e-01
2.97981679e-01 3.16642404e-01 2.56514788e-01 -7.34013379e-01
5.92013359e-01 1.76470652e-01 2.49205932e-01 -7.92906046e-01
-9.38481912e-02 -4.52459455e-01 -5.35447299e-01 -1.48367137e-01
3.47943544e-01 -2.94517845e-01 -4.59449351e-01 1.84092355e+00
2.02992514e-01 6.37324154e-01 -1.45669401e-01 6.46992743e-01
1.69948444e-01 3.28903794e-01 1.39864713e-01 -4.23619807e-01
1.46455610e+00 -5.40623248e-01 -6.41137242e-01 -3.09288055e-01
8.32185805e-01 -3.96216959e-01 1.07775319e+00 3.97149622e-01
-9.75350916e-01 2.29546316e-02 -1.27246428e+00 1.63633555e-01
-1.86166272e-01 3.59518677e-01 1.03925765e+00 1.45158923e+00
-9.57562923e-01 3.98877114e-01 -1.08774793e+00 -1.82337105e-01
5.24802327e-01 7.59180129e-01 -1.31082097e-02 -3.84613238e-02
-6.78993344e-01 4.61918324e-01 5.64312376e-02 -3.20334844e-02
-8.01627636e-01 -7.23614573e-01 -3.74615699e-01 4.26230669e-01
7.05323040e-01 -7.72291541e-01 1.25871074e+00 -7.68207848e-01
-1.44158828e+00 5.85107088e-01 -2.16640010e-01 -5.38030028e-01
7.51630247e-01 -2.62743264e-01 -5.82584679e-01 -3.76463622e-01
-1.79525092e-01 -4.25525680e-02 6.90757334e-01 -8.03641975e-01
-1.07487357e+00 -4.86964554e-01 1.82977498e-01 9.05440897e-02
-8.35557103e-01 -1.11432217e-01 -8.80199313e-01 -1.99135169e-01
-1.96944147e-01 -1.21038604e+00 -2.58820593e-01 1.11292712e-01
-4.48859662e-01 1.24941841e-01 9.62452412e-01 -2.49901548e-01
1.44106209e+00 -2.27370930e+00 -3.41926575e-01 2.52696842e-01
8.55473652e-02 -3.58929969e-02 3.54285479e-01 -5.28118424e-02
4.59603071e-01 1.43566102e-01 8.76539424e-02 -5.63482344e-01
1.42969280e-01 1.51747212e-01 -1.80331603e-01 5.37543774e-01
-3.15025568e-01 6.04410410e-01 -5.82730770e-01 -5.03683805e-01
3.70131284e-02 4.70331997e-01 -6.90878689e-01 1.08225584e-01
-2.32902840e-01 -9.66375172e-02 -7.63092041e-01 8.73043060e-01
5.96447706e-01 -4.58331108e-01 4.74467278e-01 -1.97319895e-01
2.46964276e-01 1.48941398e-01 -1.49015737e+00 1.48825586e+00
-7.99469411e-01 4.51225281e-01 2.85643429e-01 -8.03810656e-01
3.99151683e-01 3.21891516e-01 4.50751930e-01 -2.83748001e-01
3.60839635e-01 1.34608075e-01 -3.23535770e-01 -3.78191829e-01
3.84466797e-01 3.10465485e-01 -1.84365019e-01 5.42194724e-01
-3.00765485e-01 6.00106776e-01 -3.98930222e-01 -4.02699038e-02
1.41796291e+00 1.63643867e-01 4.29780364e-01 -4.75881845e-01
6.97645769e-02 -5.83447516e-01 7.68436372e-01 9.08807516e-01
-5.39502859e-01 2.81213254e-01 3.70147705e-01 -3.79572362e-01
-6.91492975e-01 -8.53468180e-01 -4.48454656e-02 1.18743980e+00
1.94473371e-01 -5.07213354e-01 -7.88685977e-01 -7.09115922e-01
1.94165334e-02 6.16268873e-01 -5.52494466e-01 -9.44820493e-02
-1.36415660e-02 -1.01520193e+00 4.50890511e-01 5.74741066e-01
1.82741091e-01 -3.00024658e-01 -1.10788786e+00 -7.02224150e-02
1.38235530e-02 -1.63563335e+00 -5.51569343e-01 4.61450458e-01
-8.51008952e-01 -7.99963892e-01 -2.57069290e-01 -3.52864265e-01
6.14279151e-01 3.26044053e-01 1.16020262e+00 -1.04080908e-01
-3.76809448e-01 6.39342248e-01 -5.14903590e-02 -3.33265305e-01
1.80092528e-01 3.28109324e-01 3.58305216e-01 6.44438639e-02
4.18278515e-01 -7.80890286e-01 -7.51282215e-01 3.37625265e-01
-6.02101505e-01 -2.00665742e-01 6.93749011e-01 5.65140426e-01
9.16774213e-01 3.90814722e-01 4.31801766e-01 -1.15875924e+00
2.98045665e-01 -7.70034671e-01 -7.03225434e-01 6.15321100e-01
-9.93019402e-01 2.54445255e-01 5.44406056e-01 -6.16090298e-01
-9.13353264e-01 5.88307321e-01 2.77319461e-01 -4.27275777e-01
8.86373892e-02 2.69877046e-01 -7.81626821e-01 -2.80216813e-01
5.65744698e-01 1.79568991e-01 -6.74562514e-01 -1.56268626e-01
1.51847258e-01 9.00758088e-01 4.99960452e-01 -9.40137327e-01
3.11177492e-01 5.36461294e-01 2.67850637e-01 -7.86843598e-01
-6.95839763e-01 -4.43940490e-01 -1.96859092e-01 1.74687002e-02
2.70254165e-01 -1.03130841e+00 -1.13791430e+00 -1.06244296e-01
-5.33659399e-01 -5.97614311e-02 -3.48979980e-02 3.82195055e-01
-7.75488436e-01 4.14478153e-01 -1.96156770e-01 -1.29510593e+00
-4.96007890e-01 -1.23898399e+00 1.10746372e+00 2.13543504e-01
-8.85536373e-02 -6.98870599e-01 -3.15351069e-01 2.21830398e-01
4.09410149e-01 1.06885538e-01 6.21476293e-01 -5.23905277e-01
-9.55257475e-01 -4.79337901e-01 -2.20155150e-01 -3.42285223e-02
-2.33109090e-02 -2.41766900e-01 -1.38116062e+00 -4.10595715e-01
6.64271135e-03 5.67489527e-02 3.77380967e-01 6.14440441e-01
1.82169545e+00 -4.45328176e-01 -6.93244934e-01 8.91909420e-01
1.60604000e+00 -7.92101473e-02 4.67241615e-01 1.03565864e-01
5.71835518e-01 4.32817526e-02 6.35459065e-01 8.54420066e-01
3.78840476e-01 8.19953501e-01 3.70976448e-01 3.38159502e-01
3.20800364e-01 -3.57354730e-01 3.09313804e-01 3.74436557e-01
7.16473535e-02 -4.45401013e-01 -6.32018268e-01 4.97197568e-01
-2.14683914e+00 -7.04377830e-01 1.23937674e-01 2.74770617e+00
6.04867458e-01 5.92461109e-01 2.42377743e-01 4.45703059e-01
3.39987308e-01 -1.46763250e-01 -7.80196607e-01 -2.59344250e-01
1.84047088e-01 -9.77600738e-03 1.24940550e+00 1.79832056e-01
-1.04619753e+00 4.68704432e-01 6.89738512e+00 1.08847046e+00
-9.61900473e-01 2.96841025e-01 9.90250766e-01 -4.09623265e-01
-2.85073757e-01 1.18893050e-01 -1.06064391e+00 7.25371540e-01
1.44782364e+00 -3.23093206e-01 4.72129554e-01 1.45624387e+00
2.23686639e-03 -9.47373286e-02 -1.52242529e+00 1.44081223e+00
-1.20388716e-01 -1.23856187e+00 -5.41636467e-01 2.74669111e-01
5.52575052e-01 7.47014359e-02 2.92783678e-01 1.69701636e-01
1.53964639e-01 -8.80636871e-01 7.37107158e-01 4.00846928e-01
9.30706501e-01 -8.05391848e-01 5.32521486e-01 5.17009497e-01
-1.19021368e+00 -2.50409722e-01 -1.90046966e-01 -1.26817450e-01
-8.95956066e-04 8.83759737e-01 -8.50870073e-01 1.51289776e-01
8.90201509e-01 1.44819379e-01 -2.54118145e-01 1.01483071e+00
1.70639545e-01 6.08093381e-01 -7.34247327e-01 -3.12059581e-01
-3.12531710e-01 3.40755105e-01 1.55148298e-01 1.23203385e+00
5.75871527e-01 2.17395164e-02 2.25594804e-01 4.58440065e-01
-1.89584553e-01 1.17160358e-01 -5.18787324e-01 -4.30219583e-02
9.46030796e-01 1.20190632e+00 -7.72889793e-01 -1.38014173e-02
-4.94179130e-01 9.65558529e-01 1.39418364e-01 7.26236254e-02
-9.67039227e-01 -5.74934483e-02 8.91261816e-01 2.06216395e-01
2.69194752e-01 -2.41072759e-01 -5.68522155e-01 -1.12514675e+00
2.84976214e-01 -6.23651981e-01 5.81149697e-01 -2.29980290e-01
-8.72065067e-01 3.93140376e-01 -1.04500964e-01 -1.09160948e+00
-4.42630090e-02 -7.11670756e-01 -3.65190923e-01 4.67728376e-01
-1.45922244e+00 -1.19389749e+00 -1.68289378e-01 5.41557968e-01
2.90148556e-01 8.56034160e-02 1.02865243e+00 7.68523455e-01
-7.07962155e-01 1.34464633e+00 2.05352277e-01 -5.39159656e-01
3.11887056e-01 -1.06151795e+00 2.77823597e-01 9.52183068e-01
2.99550414e-01 3.17809224e-01 9.12683606e-01 -2.99207836e-01
-2.05319643e+00 -8.97299588e-01 7.63173580e-01 -6.49812400e-01
5.05270302e-01 -6.37359381e-01 -3.62195969e-01 8.34263742e-01
-5.74651957e-01 4.65883762e-01 1.13362420e+00 5.71152806e-01
-3.25749338e-01 -2.57965803e-01 -1.28234756e+00 6.81576848e-01
1.23565781e+00 -7.13163793e-01 4.43243027e-01 3.64134789e-01
4.32853490e-01 -3.72072726e-01 -8.29560757e-01 3.44624162e-01
8.44859958e-01 -8.29674780e-01 7.66404748e-01 -5.90819716e-01
-3.61888677e-01 -2.40299478e-01 -6.31354928e-01 -4.70083088e-01
-2.34907381e-02 -1.05083740e+00 -8.17936361e-01 1.18673086e+00
6.32249832e-01 -1.68886676e-01 1.45298851e+00 1.22921848e+00
2.47674480e-01 -8.88152242e-01 -1.08884430e+00 -9.03035879e-01
-6.67295456e-01 -9.53523695e-01 6.52693033e-01 7.44015396e-01
1.56673342e-02 4.62587215e-02 -8.37888718e-01 5.06747782e-01
7.00235307e-01 3.27746943e-02 6.30313873e-01 -1.02355206e+00
-8.16685438e-01 -1.71276629e-01 -6.21651113e-01 -9.83565807e-01
-4.68300022e-02 -4.85922813e-01 7.65196458e-02 -7.82472670e-01
4.17681366e-01 -7.63608992e-01 -5.83588541e-01 3.57711196e-01
1.59354098e-02 1.33176446e-02 -1.08015567e-01 1.85807869e-01
-9.26550925e-01 6.95007816e-02 2.83279330e-01 1.67506918e-01
-1.86635330e-01 6.46424174e-01 -9.61653054e-01 6.53734863e-01
8.13024998e-01 -4.09871608e-01 -8.31253946e-01 -4.22709644e-01
5.68677187e-01 -2.71785874e-02 -1.00683659e-01 -1.10504842e+00
4.93507326e-01 -3.19604039e-01 1.04403816e-01 -1.57245070e-01
3.94197464e-01 -1.37521362e+00 4.47728276e-01 3.26110244e-01
-1.68780670e-01 -1.53800383e-01 -2.34238250e-04 9.89732802e-01
4.50688154e-01 -9.84408036e-02 4.91563290e-01 4.43939358e-01
-7.51075208e-01 6.90689564e-01 -5.82934693e-02 -1.43220738e-01
1.35730135e+00 -2.87411898e-01 2.57367473e-02 -4.52490240e-01
-5.50733328e-01 1.38681218e-01 4.85386997e-01 2.03135908e-01
1.34102479e-01 -1.15916872e+00 -1.60115555e-01 -2.53150985e-02
1.92466274e-01 -3.82820904e-01 2.16810897e-01 7.47504354e-01
-2.45752916e-01 3.72558385e-01 3.44726630e-02 -5.16473353e-01
-1.36153126e+00 8.17469537e-01 9.82674658e-02 -3.34565490e-01
-3.49206567e-01 7.57488251e-01 -8.83430988e-02 6.84816539e-02
6.82116389e-01 -4.58970875e-01 4.30161148e-01 -4.44027632e-01
6.17063582e-01 4.39929932e-01 4.51841772e-01 -2.55764037e-01
-6.18435800e-01 1.24840274e-01 3.95607464e-02 3.64388414e-02
9.76608694e-01 -3.85348529e-01 6.41517580e-01 2.67094165e-01
1.02963221e+00 1.36572361e-01 -1.40369284e+00 -3.69459301e-01
3.19810331e-01 -5.56882083e-01 4.71693099e-01 -7.62591422e-01
-1.01395833e+00 6.41919076e-01 1.04265451e+00 9.98298675e-02
1.33062661e+00 -2.00154319e-01 7.38984644e-01 4.00666416e-01
8.16231489e-01 -1.20498478e+00 -5.65850079e-01 -4.92619239e-02
-7.81133249e-02 -1.13263679e+00 2.18502045e-01 -6.99736357e-01
-3.17618221e-01 5.85878313e-01 3.19116175e-01 1.94864482e-01
1.14137757e+00 8.51676524e-01 -3.76079112e-01 1.66656524e-01
-1.01408708e+00 1.09137580e-01 1.03541024e-01 7.16431677e-01
3.10603470e-01 5.54371595e-01 -1.10750169e-01 1.20398724e+00
-3.70551869e-02 3.70709389e-01 1.92043409e-01 1.12754953e+00
-2.44127482e-01 -1.11785293e+00 -1.47402078e-01 7.16116011e-01
-7.72269905e-01 7.16128349e-02 5.97585738e-03 3.16363931e-01
5.59635751e-04 8.40712011e-01 -1.67673215e-01 -5.60103118e-01
6.25008419e-02 -8.92725512e-02 4.21014279e-01 -2.86243379e-01
-4.92423475e-02 -2.01656416e-01 2.57068843e-01 -7.07361937e-01
-3.46663296e-02 -8.01477492e-01 -9.67398882e-01 -4.30153251e-01
-4.66032296e-01 -1.32890546e-03 9.56584930e-01 1.09700334e+00
8.71510088e-01 4.07344788e-01 6.83269083e-01 -5.03437698e-01
-8.76390219e-01 -4.16630924e-01 -6.99409246e-01 -8.67247581e-02
1.16340585e-01 -6.15447998e-01 -2.51521736e-01 -4.63211574e-02] | [5.945688247680664, 6.050426483154297] |
6dae7119-8ef8-421f-ac6a-e2edc01274da | the-cmu-submission-for-the-shared-task-on | null | null | https://aclanthology.org/W14-3909 | https://aclanthology.org/W14-3909.pdf | The CMU Submission for the Shared Task on Language Identification in Code-Switched Data | null | ['Chu-Cheng Lin', 'Waleed Ammar', 'Chris Dyer', 'Lori Levin'] | 2014-10-01 | null | null | null | ws-2014-10 | ['learning-word-embeddings'] | ['methodology'] | [-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] | [-7.332990646362305, 3.7051520347595215] |
4b1a0294-99bf-4692-b3ab-dfcf07b03019 | fast-inference-and-transfer-of-compositional | 2205.12648 | null | https://arxiv.org/abs/2205.12648v1 | https://arxiv.org/pdf/2205.12648v1.pdf | Fast Inference and Transfer of Compositional Task Structures for Few-shot Task Generalization | We tackle real-world problems with complex structures beyond the pixel-based game or simulator. We formulate it as a few-shot reinforcement learning problem where a task is characterized by a subtask graph that defines a set of subtasks and their dependencies that are unknown to the agent. Different from the previous meta-rl methods trying to directly infer the unstructured task embedding, our multi-task subtask graph inferencer (MTSGI) first infers the common high-level task structure in terms of the subtask graph from the training tasks, and use it as a prior to improve the task inference in testing. Our experiment results on 2D grid-world and complex web navigation domains show that the proposed method can learn and leverage the common underlying structure of the tasks for faster adaptation to the unseen tasks than various existing algorithms such as meta reinforcement learning, hierarchical reinforcement learning, and other heuristic agents. | ['Honglak Lee', 'Aleksandra Faust', 'Izzeddin Gur', 'lyubing qiang', 'Jongwook Choi', 'Hyunjae Woo', 'Sungryull Sohn'] | 2022-05-25 | null | null | null | null | ['hierarchical-reinforcement-learning'] | ['methodology'] | [ 4.39542145e-01 5.97937047e-01 -7.73373339e-03 -7.35518038e-02
-6.16470993e-01 -6.15140021e-01 6.38812065e-01 -1.46248728e-01
-5.73483884e-01 1.01845384e+00 8.84479731e-02 -2.13217214e-01
-2.84075081e-01 -5.88078797e-01 -9.20635402e-01 -7.32566357e-01
-2.39036471e-01 9.01599765e-01 7.24601686e-01 -5.78943312e-01
2.70647109e-01 -2.87096739e-01 -1.59764743e+00 1.85246974e-01
9.63562489e-01 6.86156750e-01 8.64143133e-01 7.98662364e-01
-2.51701213e-02 1.45289278e+00 -5.84546685e-01 -1.77803665e-01
2.17614993e-01 -4.13777798e-01 -1.08404291e+00 1.19817704e-01
1.72232259e-02 -3.68491828e-01 -3.88285935e-01 1.00578296e+00
3.98666769e-01 6.99832618e-01 7.62978315e-01 -1.22600126e+00
-4.39283729e-01 4.09174800e-01 -4.43837196e-01 1.83057234e-01
3.81337464e-01 5.80081165e-01 1.14627731e+00 -3.76948923e-01
7.03037858e-01 1.26158237e+00 2.51627713e-01 7.31405377e-01
-1.18772399e+00 -4.92715150e-01 7.08392382e-01 6.00879371e-01
-7.74595976e-01 3.84016521e-02 8.18300366e-01 -4.66944903e-01
1.28109181e+00 -4.13718998e-01 8.57538819e-01 1.66924560e+00
3.28776866e-01 9.32880819e-01 1.24788439e+00 -2.41404369e-01
7.02044129e-01 -3.48006159e-01 2.52814919e-01 1.10994112e+00
4.29059640e-02 3.68475139e-01 -5.35685718e-01 -2.78020531e-01
8.48984778e-01 -3.20196927e-01 -1.96062461e-01 -8.96065414e-01
-1.12177157e+00 9.11887825e-01 1.96304232e-01 -1.10479400e-01
-5.09864092e-01 4.81114388e-01 4.09901559e-01 5.53437829e-01
2.75291324e-01 5.92471719e-01 -7.44874001e-01 -9.46842227e-03
-2.69053757e-01 4.22979712e-01 1.14866364e+00 1.05918288e+00
1.10541129e+00 2.39632010e-01 -2.76929617e-01 5.25915086e-01
9.82181504e-02 8.30358937e-02 5.68281651e-01 -1.09942186e+00
7.86694288e-01 3.16198766e-01 2.33639479e-01 -2.29862839e-01
-5.39841354e-01 -3.29055309e-01 -2.86563605e-01 5.49860060e-01
4.34048146e-01 -3.37328106e-01 -1.35252631e+00 1.86868095e+00
5.29650569e-01 7.31826127e-01 8.00840184e-02 9.06940281e-01
6.03865027e-01 3.11189711e-01 3.31931323e-01 1.66582093e-01
1.21658313e+00 -1.35834193e+00 -7.42709160e-01 -9.19363201e-01
7.28273988e-01 9.37964842e-02 1.19001174e+00 4.09951955e-01
-7.86896408e-01 -9.29746151e-01 -1.21526647e+00 -4.77390550e-02
-4.99990225e-01 -5.02691627e-01 6.21020854e-01 4.47326154e-01
-1.13631868e+00 3.99978638e-01 -6.80515289e-01 -4.61461879e-02
4.29112643e-01 3.93657804e-01 -4.87178378e-02 -3.39886695e-01
-1.39592409e+00 8.26439798e-01 8.21252644e-01 -1.13775328e-01
-1.97262621e+00 -6.20848536e-01 -1.18911183e+00 6.58775419e-02
1.28421497e+00 -1.05317354e+00 1.49795246e+00 -7.82865942e-01
-1.74647391e+00 7.43438125e-01 2.24263594e-01 -3.67701083e-01
1.61270231e-01 -1.81064099e-01 1.85177594e-01 -1.71235338e-01
5.46575636e-02 6.08903050e-01 1.43941164e+00 -1.33834887e+00
-9.15059626e-01 -5.35004795e-01 6.95715487e-01 7.53808439e-01
2.76853472e-01 -7.82857656e-01 -4.22688365e-01 -3.05001646e-01
-2.88739502e-01 -1.05117178e+00 -5.35813749e-01 -5.30351102e-01
-3.07463765e-01 -3.97623986e-01 5.86550057e-01 -4.77706879e-01
6.30820096e-01 -1.88476634e+00 8.35542262e-01 -1.71299547e-01
5.99121153e-01 -1.01878934e-01 -5.56749642e-01 3.26343060e-01
1.99705929e-01 -2.87900895e-01 -1.65765375e-01 -1.17203325e-01
1.69603631e-01 5.89186430e-01 -4.36627716e-02 8.44502300e-02
-1.24555528e-01 1.22868526e+00 -1.43687344e+00 -4.31146771e-01
1.60631061e-01 -1.03015751e-01 -4.07653898e-01 6.74274445e-01
-8.94684017e-01 6.22260451e-01 -1.09862041e+00 3.33137840e-01
1.83938235e-01 -4.58097607e-01 4.75548297e-01 1.01165690e-01
5.13345301e-01 2.18387082e-01 -1.09675777e+00 2.16345620e+00
-5.68148434e-01 1.68140516e-01 1.14927702e-01 -1.34263873e+00
5.80105960e-01 1.34125277e-01 2.78676987e-01 -9.72506285e-01
-1.70324802e-01 -3.04363161e-01 1.26698852e-01 -7.00575113e-01
2.24585533e-01 1.22048549e-01 -2.48573154e-01 5.26103556e-01
5.92426360e-01 -4.37407017e-01 4.40747999e-02 1.80502206e-01
1.52183342e+00 7.76171327e-01 4.80975956e-01 -1.65657908e-01
3.14271927e-01 3.25961530e-01 3.97537649e-01 1.43008697e+00
-2.86487281e-01 8.85813460e-02 4.41097111e-01 -6.17851615e-01
-8.72925758e-01 -1.14199960e+00 6.44085348e-01 1.72660124e+00
1.81500688e-01 -3.70672107e-01 -6.82706237e-01 -1.15805793e+00
-1.29604219e-02 7.22043037e-01 -1.06013048e+00 -3.05747151e-01
-4.32680279e-01 -4.88127977e-01 1.98483486e-02 3.44874382e-01
5.48484981e-01 -1.47074056e+00 -7.82736063e-01 4.69841897e-01
-6.93740621e-02 -1.25919080e+00 -3.57610494e-01 6.40428662e-01
-7.97704041e-01 -1.49736500e+00 -4.23835963e-01 -1.04447114e+00
5.05280972e-01 1.43933550e-01 1.50353062e+00 2.47821920e-02
-2.82000482e-01 7.11440027e-01 -3.56885701e-01 -4.78419781e-01
-2.48547062e-01 1.73482880e-01 -4.08573210e-01 -3.08253139e-01
8.36208165e-02 -5.82678974e-01 -4.40675348e-01 7.78885707e-02
-7.41593063e-01 2.80842394e-01 6.04072034e-01 1.20290005e+00
7.35880494e-01 2.35832766e-01 5.62500894e-01 -1.54771852e+00
8.92586827e-01 -6.40480399e-01 -8.67017269e-01 3.15884620e-01
-5.48046410e-01 5.32398224e-01 6.41606271e-01 -5.85728288e-01
-1.27740765e+00 3.09373051e-01 3.12713146e-01 -2.96134025e-01
-3.23967814e-01 5.87647736e-01 -2.25283317e-02 5.19085042e-02
8.05544674e-01 4.06092346e-01 -1.30257398e-01 -2.68778712e-01
4.32719111e-01 -6.22158796e-02 2.12241918e-01 -9.37087238e-01
7.75397480e-01 3.49749953e-01 8.36473256e-02 -5.35789430e-01
-1.22044241e+00 -3.20713669e-01 -6.10625148e-01 -1.59093887e-01
1.23697460e+00 -9.99163270e-01 -8.42167556e-01 4.76240933e-01
-9.07656193e-01 -1.48112679e+00 -4.85932082e-01 2.05042213e-01
-1.26298881e+00 1.40083462e-01 -4.63001400e-01 -6.96219981e-01
2.15417594e-01 -1.32342482e+00 1.24353814e+00 1.80000603e-01
2.02287659e-01 -1.32551467e+00 4.92901891e-01 4.71610785e-01
1.06527358e-01 9.38976333e-02 1.34930837e+00 -6.01276875e-01
-7.86446571e-01 5.26969016e-01 7.22408071e-02 -2.24946335e-01
6.86514527e-02 -9.45867300e-01 -8.47388566e-01 -4.59359944e-01
2.57486165e-01 -1.06253266e+00 7.96931803e-01 4.74818945e-01
1.24920392e+00 -1.51184471e-02 -1.84783876e-01 5.63452899e-01
1.42012012e+00 3.01812112e-01 4.68882889e-01 5.71484983e-01
7.36827791e-01 6.01606727e-01 8.46195459e-01 3.78630459e-01
5.84212482e-01 4.64428723e-01 9.90956783e-01 2.02797785e-01
-1.00470506e-01 -4.58878070e-01 3.78964752e-01 3.29325140e-01
-2.18209699e-02 -1.59022361e-01 -8.03990841e-01 4.46753055e-01
-2.35878181e+00 -7.39063561e-01 4.28059191e-01 1.88228607e+00
6.89624488e-01 4.26114321e-01 1.47467464e-01 -4.66939926e-01
4.87017810e-01 6.79453671e-01 -1.19568002e+00 -1.65449306e-01
3.59602779e-01 2.85463482e-01 2.89987952e-01 5.83281696e-01
-9.20856774e-01 1.35804009e+00 6.39070272e+00 7.00431168e-01
-3.19495231e-01 2.74085671e-01 3.59750152e-01 1.81154922e-01
-7.71050230e-02 -4.89767678e-02 -8.07965159e-01 -1.62192285e-02
8.37227523e-01 -6.80152699e-02 1.06766355e+00 8.69395316e-01
-2.14829341e-01 -1.56038404e-01 -1.37188101e+00 5.79181671e-01
5.19494340e-02 -1.07264781e+00 -1.82572789e-02 6.31932691e-02
8.97015333e-01 2.55353063e-01 1.05295151e-01 1.20990956e+00
1.35506392e+00 -1.00087762e+00 4.78660047e-01 3.76269192e-01
4.76875514e-01 -4.37648892e-01 3.99040222e-01 8.31961334e-01
-1.20592558e+00 -3.94894421e-01 -4.94493932e-01 -1.27406687e-01
-1.34098649e-01 -1.41601041e-01 -8.66044044e-01 4.75679755e-01
5.39802432e-01 6.65244162e-01 -6.28672481e-01 7.49704957e-01
-6.94849968e-01 4.44913715e-01 2.16794237e-01 -2.03842461e-01
5.40602684e-01 -2.54081696e-01 4.29829091e-01 6.06430411e-01
-1.98927268e-01 1.90713689e-01 7.52703428e-01 7.59437621e-01
7.25093633e-02 -2.66282231e-01 -9.25239205e-01 -4.76933680e-02
7.11102113e-02 1.13674116e+00 -4.73341644e-01 -3.41388881e-01
-6.29236102e-01 1.21833086e+00 1.05976641e+00 9.03544068e-01
-8.81195247e-01 -8.71749520e-02 6.12886548e-01 -2.16573894e-01
5.87226212e-01 -4.38574702e-01 3.12761068e-02 -1.11395347e+00
-3.22377503e-01 -1.06215596e+00 7.61933446e-01 -9.79925752e-01
-1.12656629e+00 5.37300646e-01 4.67462167e-02 -8.89762521e-01
-7.35819876e-01 -7.93786228e-01 -5.74353993e-01 9.16339934e-01
-1.93372464e+00 -1.10710776e+00 -4.22986358e-01 1.16588366e+00
1.06080985e+00 -4.75301713e-01 8.32646012e-01 -7.43421972e-01
-3.48756671e-01 3.54449041e-02 -1.45650372e-01 -3.34481597e-02
2.09302485e-01 -1.58781505e+00 5.94452143e-01 3.64896953e-01
3.32412094e-01 -1.05419517e-01 5.39997995e-01 -7.85766125e-01
-1.62616134e+00 -9.75780070e-01 -3.42416279e-02 -6.89249575e-01
8.42578351e-01 -7.19627857e-01 -7.73084939e-01 1.04655159e+00
3.31118017e-01 -1.42993382e-03 2.16775700e-01 3.86366665e-01
-3.28027129e-01 2.13786721e-01 -7.26044178e-01 7.32546926e-01
1.51186717e+00 -6.51476681e-01 -9.74627018e-01 5.58925748e-01
9.98823106e-01 -7.09877491e-01 -4.05181557e-01 7.03784972e-02
1.23095751e-01 -6.51750267e-01 1.12110567e+00 -1.37109101e+00
2.82343835e-01 1.03228405e-01 1.50657684e-01 -1.87895024e+00
-7.23059714e-01 -6.35680199e-01 -5.18241644e-01 3.21340770e-01
4.63396013e-01 -4.54983443e-01 1.19288504e+00 2.42899939e-01
-2.14974523e-01 -5.32229543e-01 -9.36979592e-01 -7.53336191e-01
-1.40502810e-01 -2.53864408e-01 3.82997096e-01 7.60647953e-01
-2.48201400e-01 8.32302034e-01 -3.74946624e-01 2.96597421e-01
1.13289833e+00 2.08292648e-01 8.14233959e-01 -1.35757363e+00
-9.13685739e-01 4.00566421e-02 3.33658457e-02 -1.03512335e+00
7.76935399e-01 -8.85591626e-01 3.38527441e-01 -1.62299359e+00
1.83605894e-01 -1.41355246e-01 -6.22743666e-01 4.61096317e-01
-4.31518137e-01 -4.56979245e-01 -3.11619826e-02 -1.63858443e-01
-1.18201602e+00 6.45348847e-01 1.84376717e+00 -4.12693918e-01
-3.02394718e-01 5.93459327e-03 -5.95829189e-01 5.79232752e-01
6.21654212e-01 -6.09396100e-01 -1.13949776e+00 -5.60811937e-01
4.38208103e-01 4.82393175e-01 2.35389948e-01 -9.35599625e-01
4.20531392e-01 -5.20724356e-01 2.24489272e-01 -1.65109441e-01
4.78364438e-01 -8.28115821e-01 -2.28007048e-01 6.03853524e-01
-5.22636533e-01 -4.18139882e-02 1.01081870e-01 1.17663157e+00
8.41552839e-02 -3.90489876e-01 3.94584775e-01 -5.73501766e-01
-1.34312308e+00 5.66401243e-01 -4.24455673e-01 7.01586962e-01
1.13801908e+00 -2.51616031e-01 -4.81480181e-01 -5.27432919e-01
-1.28498423e+00 6.88073277e-01 6.00092718e-03 3.81012291e-01
6.93959296e-01 -8.99889767e-01 -5.98835111e-01 9.45176929e-02
2.04412073e-01 1.98316991e-01 4.11628932e-01 3.11924219e-01
1.33940712e-01 2.34917819e-01 -5.88964581e-01 -4.72127080e-01
-7.68963277e-01 8.71158838e-01 3.63155007e-01 -1.04697287e+00
-5.63447356e-01 6.92965925e-01 8.46866250e-01 -5.43696940e-01
3.06917399e-01 -1.45674095e-01 -4.54987079e-01 -1.58845559e-01
1.48472995e-01 7.24173933e-02 -1.48058072e-01 1.52549848e-01
2.21023649e-01 1.74027041e-01 -4.42031950e-01 -2.89150983e-01
1.35785460e+00 -5.68692312e-02 5.38871467e-01 5.21139085e-01
6.89356863e-01 -7.92323411e-01 -1.96026683e+00 -4.28219408e-01
1.66451663e-01 -1.40057325e-01 7.59377852e-02 -9.24669325e-01
-7.38406718e-01 7.64353037e-01 3.64016205e-01 2.10058048e-01
9.13667619e-01 1.55291215e-01 3.89318347e-01 9.32851195e-01
8.86423886e-01 -1.46370196e+00 8.02269697e-01 9.07268047e-01
8.42144728e-01 -1.27595961e+00 -2.86309391e-01 -2.94184476e-01
-8.99791420e-01 8.77746999e-01 1.18104911e+00 -3.47368270e-01
6.03572071e-01 1.54503852e-01 -3.16650659e-01 -5.52808285e-01
-1.21046269e+00 -8.26092482e-01 1.14949957e-01 1.31763864e+00
-2.00364798e-01 -2.88131833e-01 2.25373164e-01 5.41181803e-01
1.64246872e-01 5.25192358e-02 3.73762101e-01 1.09798098e+00
-5.49225450e-01 -1.01333213e+00 1.23327516e-01 5.54295897e-01
-7.71551346e-03 2.34849798e-03 -2.21803546e-01 9.92745399e-01
-1.38496846e-01 6.13828361e-01 -1.59999073e-01 -4.16250736e-01
4.56640184e-01 3.55510086e-01 8.18889499e-01 -1.23657656e+00
-5.59257388e-01 -2.52326369e-01 1.78052783e-01 -9.69495237e-01
-1.05375715e-01 -4.31760669e-01 -1.23003066e+00 4.87375498e-01
1.37898654e-01 1.11760080e-01 3.84482354e-01 1.19589186e+00
1.73397571e-01 9.56287622e-01 4.17365491e-01 -8.60877216e-01
-8.15215409e-01 -7.54208982e-01 -7.09402204e-01 6.07374549e-01
3.94537061e-01 -1.04631793e+00 -2.13736013e-01 4.62289294e-03] | [4.163809299468994, 1.3641831874847412] |
fce9866d-e914-4753-b1e0-7073860d5be0 | sparql-query-generation-for-complex-question | null | null | http://www.dialog-21.ru/media/5088/evseevdaplusarkhipov-myu-048.pdf | http://www.dialog-21.ru/media/5088/evseevdaplusarkhipov-myu-048.pdf | SPARQL query generation for complex question answering with BERT and BiLSTM-based model | In this paper we describe question answering system for answering of complex questions over Wikidata knowledge base. Unlike simple questions, which require extraction of single fact from the knowledge base, complex questions are based on more than one triplet and need logical or comparative reasoning. The proposed question answering system translates a natural language question into a query in SPARQL language, execution of which gives an answer. The system includes the models which define the SPARQL query template corresponding to the question and then fill the slots in the template with entities, relations and numerical values. For entity detection we use BERT-based sequence labelling model. Ranking of candidate relations is performed in two steps with BiLSTM and BERT-based models. The proposed models are the first solution for LC-QUAD2.0 dataset. The system is capable of answering complex questions which involve comparative or boolean reasoning. | ['Arkhipov M. Yu.', 'Evseev D. A.'] | 2020-06-17 | null | null | null | proceedings-of-the-international-conference | ['knowledge-base-question-answering'] | ['natural-language-processing'] | [-3.19253534e-01 7.35213995e-01 1.24584779e-01 -4.52683926e-01
-6.47597849e-01 -9.98248518e-01 4.89384025e-01 7.70994604e-01
-5.49084842e-01 1.18500376e+00 1.43702462e-01 -6.80078506e-01
-7.05311716e-01 -1.51437747e+00 -4.01868314e-01 3.27997178e-01
2.77571678e-01 1.21968091e+00 1.10390437e+00 -7.03631282e-01
1.34389639e-01 2.49903589e-01 -1.69305611e+00 7.57340372e-01
8.49911153e-01 1.06551266e+00 8.49581137e-02 9.08102453e-01
-1.02092004e+00 1.46489346e+00 -4.26061988e-01 -4.89658028e-01
-5.35620609e-04 -3.08064461e-01 -1.72369719e+00 -6.10590637e-01
3.47064704e-01 -2.61251051e-02 2.04476207e-01 8.54780257e-01
3.28988642e-01 1.35512382e-01 5.54282308e-01 -1.36018264e+00
-2.53243983e-01 2.81960189e-01 3.97508234e-01 2.78375238e-01
1.24112821e+00 -4.71999049e-01 1.29788995e+00 -6.19879663e-01
8.02720606e-01 1.39049149e+00 4.48649883e-01 2.03569382e-01
-4.99916255e-01 -8.21807161e-02 -3.70329112e-01 8.56019318e-01
-1.15463841e+00 -8.70787725e-02 3.08188219e-02 -2.57309705e-01
1.55330563e+00 7.70253658e-01 3.42318445e-01 -5.01872540e-01
-4.47154343e-02 1.86029375e-01 1.02578890e+00 -7.10654557e-01
2.40814358e-01 5.49268961e-01 8.08856249e-01 9.30229664e-01
2.95645386e-01 -5.71994364e-01 -2.33152732e-01 -3.36841285e-01
2.55783379e-01 -5.47331691e-01 2.30542317e-01 6.41317517e-02
-7.00910866e-01 7.01621711e-01 9.21959206e-02 3.45312119e-01
-5.29959738e-01 -1.70791730e-01 4.26580459e-01 4.85513657e-01
-3.91087800e-01 5.36015451e-01 -8.27353239e-01 1.71847984e-01
-4.79618311e-01 6.63526833e-01 1.70284450e+00 1.13781190e+00
1.02361131e+00 -8.37638974e-01 -3.02570492e-01 4.97186303e-01
4.90883172e-01 6.35665432e-02 3.51255506e-01 -8.21330845e-01
4.07569438e-01 1.39891183e+00 4.32819575e-01 -7.83266544e-01
-5.99036634e-01 -4.89577875e-02 1.00349225e-01 6.23657694e-03
5.59527338e-01 -2.56115615e-01 -6.73787177e-01 1.21293271e+00
1.03497565e+00 -3.25440556e-01 7.33250797e-01 5.35615742e-01
1.64205647e+00 6.34213984e-01 3.01311165e-01 -1.42730221e-01
2.36337090e+00 -6.02964103e-01 -9.39769924e-01 2.28183642e-01
7.72675991e-01 -9.55516279e-01 6.54240668e-01 6.48814440e-02
-9.64823663e-01 -3.16505730e-01 -7.29572952e-01 -7.51286924e-01
-1.03411782e+00 7.42908791e-02 4.79621977e-01 4.48388487e-01
-9.25338984e-01 -1.06716000e-01 -2.12988421e-01 -9.65724826e-01
-1.97850794e-01 4.55494761e-01 -3.96609306e-01 -6.84825480e-02
-1.60954416e+00 1.29208446e+00 1.02868283e+00 -3.53234820e-02
-2.52928048e-01 -4.60736811e-01 -6.89714313e-01 1.82515442e-01
7.60855675e-01 -7.17795193e-01 1.14242125e+00 -1.90253720e-01
-9.21052277e-01 1.00280225e+00 -2.92138249e-01 -6.25504613e-01
1.12587295e-01 1.17267475e-01 -6.28867924e-01 3.76594812e-01
4.14291859e-01 3.26540917e-01 -1.21370085e-01 -9.18187916e-01
-1.09247255e+00 -4.18755859e-01 8.24364543e-01 4.05973732e-01
5.07136941e-01 4.23498690e-01 -5.58500171e-01 3.22534174e-01
1.98170841e-01 -3.27426225e-01 1.43097103e-01 -2.62983233e-01
-2.42486894e-01 -8.56824815e-01 6.64573967e-01 -1.08489871e+00
1.50687575e+00 -1.23722363e+00 -5.67061245e-01 3.68771136e-01
5.37750386e-02 3.20375234e-01 3.40261757e-01 8.94420624e-01
9.88298804e-02 4.90056872e-02 1.67535603e-01 9.46937978e-01
3.44209045e-01 5.39157867e-01 -1.30257055e-01 -4.46825564e-01
4.89207581e-02 9.02555406e-01 -5.29181480e-01 -1.17100191e+00
-7.87830278e-02 -1.97872758e-01 -5.62193334e-01 2.56781518e-01
-1.06917048e+00 -3.42829794e-01 -7.46808887e-01 6.42949760e-01
6.08851075e-01 9.70055759e-02 3.69799614e-01 -5.46908915e-01
-6.81617334e-02 6.08587027e-01 -1.58350015e+00 1.20472300e+00
-3.60716671e-01 7.19924830e-03 -7.42501393e-02 -9.86679673e-01
1.02542722e+00 6.79350495e-01 4.32370044e-02 -7.86965251e-01
-6.72394484e-02 4.27744061e-01 -1.41504616e-01 -1.41160536e+00
4.04650331e-01 -3.43106151e-01 -5.58507927e-02 2.61075795e-01
3.53401929e-01 -3.72958720e-01 8.56847227e-01 3.74818265e-01
1.17319167e+00 3.62329334e-01 6.58911049e-01 -3.85641098e-01
1.25632501e+00 6.76159084e-01 3.12722772e-01 5.04781485e-01
2.59179860e-01 -3.30138445e-01 9.92289066e-01 -4.27098423e-01
-8.11387599e-01 -8.56840968e-01 -4.21294942e-02 8.67066741e-01
7.08821416e-02 -3.54407638e-01 -5.60426831e-01 -7.91463315e-01
-2.65651047e-02 8.30006361e-01 -3.26795667e-01 5.02576351e-01
-4.14180458e-01 -2.61638343e-01 6.72078788e-01 -1.62542954e-01
5.65477669e-01 -1.23259604e+00 -8.19069862e-01 3.30499560e-01
-3.79396051e-01 -1.14500260e+00 3.63329411e-01 1.29128769e-01
-6.04084373e-01 -1.69456244e+00 1.74876213e-01 -1.09976876e+00
4.18751776e-01 -5.23851633e-01 1.27880788e+00 5.56642532e-01
-2.43905857e-01 3.84117216e-01 -5.31917691e-01 -5.86542785e-01
-1.66549221e-01 1.63610607e-01 -6.47537589e-01 -3.91291469e-01
8.60583186e-01 -2.65832394e-01 -3.20418775e-01 1.38080657e-01
-1.18879926e+00 -2.64909655e-01 1.69149309e-01 3.20582986e-01
5.14299095e-01 1.84708416e-01 1.01183248e+00 -1.13812661e+00
9.16472316e-01 -7.47153401e-01 -7.91300833e-01 9.86125588e-01
-4.35855806e-01 5.24579704e-01 4.64675248e-01 3.89172822e-01
-1.26929331e+00 1.41004054e-02 -5.15747488e-01 7.71053195e-01
-2.61927247e-01 1.07299471e+00 -4.91555274e-01 1.20369516e-01
7.51734674e-01 -1.71299800e-01 -2.70457000e-01 -4.80589926e-01
4.16737258e-01 4.56146330e-01 5.11868954e-01 -9.42083240e-01
6.38412893e-01 4.22376730e-02 1.54634371e-01 -5.24491966e-01
-6.92405164e-01 -6.80937290e-01 -6.22582495e-01 -1.91918761e-01
1.10869813e+00 -6.07653320e-01 -1.10293531e+00 -2.74714798e-01
-1.25492179e+00 2.21808136e-01 -4.80479628e-01 2.73847491e-01
-4.51054096e-01 1.92188591e-01 -2.51505375e-01 -9.55017328e-01
-4.56871003e-01 -5.39460480e-01 5.24057388e-01 4.20611143e-01
-1.95583805e-01 -8.99648190e-01 3.14187706e-01 6.07057869e-01
1.60142422e-01 1.74892947e-01 1.46506178e+00 -1.18421376e+00
-6.67829037e-01 -4.10846919e-01 -3.71255606e-01 5.68769267e-03
7.96126015e-03 -1.79007038e-01 -4.06841874e-01 6.30022705e-01
-3.13524753e-01 -3.52527261e-01 1.51472315e-01 -3.92154723e-01
2.48473093e-01 -4.76915359e-01 -1.05482444e-01 -1.15832560e-01
1.92281353e+00 4.43838775e-01 1.00717330e+00 4.61643875e-01
5.01550809e-02 9.84702826e-01 8.33159804e-01 -1.03666019e-02
1.00937116e+00 3.49033773e-01 1.82352796e-01 3.94202262e-01
6.35285955e-03 -2.79078245e-01 -4.16860044e-01 4.16723996e-01
1.48584023e-01 -2.10628472e-02 -1.14504206e+00 7.08350897e-01
-1.83861196e+00 -1.07102633e+00 -7.90957272e-01 1.86355376e+00
1.18950450e+00 -1.47584170e-01 3.89501229e-02 2.72423536e-01
3.67475480e-01 -5.52200854e-01 -1.72551915e-01 -5.93573034e-01
-1.31204054e-01 7.43870616e-01 1.99351043e-01 1.17427766e+00
-3.80543590e-01 1.11280239e+00 5.37482405e+00 5.98548472e-01
-3.23931217e-01 1.43655673e-01 -1.16654851e-01 4.07398909e-01
-5.64561903e-01 6.04242682e-01 -1.01550627e+00 -7.12853372e-02
7.92701423e-01 -4.61502433e-01 3.20754617e-01 2.97844887e-01
-6.37638271e-02 -7.69327819e-01 -8.03968668e-01 3.74748588e-01
-2.00555861e-01 -1.47215164e+00 3.16221535e-01 -3.77787888e-01
3.10779542e-01 -4.07670408e-01 -8.54348540e-01 4.54409599e-01
3.72032374e-01 -9.28682089e-01 5.92484176e-01 1.11333644e+00
3.71758610e-01 -5.77405870e-01 9.82728302e-01 3.96332949e-01
-1.30214202e+00 2.55219247e-02 -4.57481951e-01 -5.71714491e-02
7.28413463e-02 2.82269180e-01 -1.01584053e+00 1.02912951e+00
6.16579294e-01 -3.28499138e-01 -7.13760734e-01 1.11164534e+00
-2.12183103e-01 3.65555704e-01 -6.83862805e-01 -3.99860591e-01
3.57777462e-03 -2.45876774e-01 -1.42319093e-03 1.10459769e+00
3.11175764e-01 6.67176187e-01 -1.37210026e-01 5.30670166e-01
8.04969072e-02 8.06580186e-01 -2.29435146e-01 1.86257496e-01
5.45930624e-01 1.15700722e+00 -6.04733288e-01 -7.65829146e-01
-4.18720812e-01 4.03538346e-01 2.45402440e-01 3.29346359e-01
-4.34953481e-01 -8.91877115e-01 -1.03481926e-01 2.67431825e-01
1.76553681e-01 1.81614935e-01 -1.13119101e-02 -8.00593257e-01
3.51407230e-01 -1.06560910e+00 9.54940200e-01 -1.04454410e+00
-6.39644444e-01 5.76847076e-01 1.10743046e-01 -4.74643916e-01
-3.71598661e-01 -4.98973519e-01 -4.45694089e-01 1.22219372e+00
-1.59972751e+00 -1.17042530e+00 -4.60854769e-01 8.23815942e-01
8.77469219e-03 1.55158684e-01 1.11954880e+00 6.24360383e-01
-8.59746411e-02 -1.29343405e-01 -6.06278062e-01 -5.30955940e-03
4.47724640e-01 -1.41134501e+00 -1.70924634e-01 4.74337935e-01
-9.65958089e-02 6.97347105e-01 8.05036426e-01 -8.20664167e-01
-1.24969196e+00 -6.08759999e-01 1.68466187e+00 -5.28876245e-01
5.64894199e-01 3.97987366e-02 -1.03865004e+00 8.41135204e-01
4.16698396e-01 -2.08390534e-01 6.90597773e-01 -1.58502325e-01
-3.34687978e-01 -2.61942178e-01 -1.59885442e+00 1.01046183e-03
4.43904638e-01 -6.78296804e-01 -1.20798194e+00 4.96961892e-01
6.05738044e-01 -2.65573591e-01 -1.10034609e+00 2.80330181e-01
5.50693095e-01 -8.12987864e-01 7.60529816e-01 -1.04594898e+00
-5.91428652e-02 -1.03662729e+00 -3.61801326e-01 -4.55149502e-01
2.38046482e-01 -2.58070350e-01 2.53320877e-02 1.26614487e+00
1.06786776e+00 -4.51494753e-01 7.22307682e-01 8.20569515e-01
3.66970986e-01 -5.08507729e-01 -9.52865660e-01 -1.84674472e-01
-2.42998183e-01 -3.64945292e-01 8.15063655e-01 8.23778450e-01
3.43785524e-01 8.40729117e-01 3.36975276e-01 4.84067053e-01
1.97689593e-01 -6.00527599e-03 5.84447145e-01 -1.49339461e+00
-1.02608070e-01 1.70205057e-01 -4.07500327e-01 -5.67728698e-01
-2.44346976e-01 -8.95583570e-01 -1.98017851e-01 -2.55797815e+00
-3.65841299e-01 -6.08054459e-01 3.01392257e-01 4.78460491e-01
-1.38114318e-01 -1.55392781e-01 -8.29567015e-02 -2.77817249e-01
-6.69609964e-01 -1.56779706e-01 7.97386765e-01 2.09226862e-01
1.00698538e-01 -1.02645881e-01 -5.29458702e-01 4.20619130e-01
8.25882018e-01 -3.82027209e-01 -6.20485425e-01 -1.05034389e-01
1.31471908e+00 5.14335692e-01 2.73790061e-01 -8.89448583e-01
7.22631752e-01 -2.91779131e-01 -1.36604577e-01 -8.17372799e-01
-4.97633242e-04 -1.07246995e+00 2.88580298e-01 3.76378387e-01
-2.99297810e-01 2.86644250e-01 2.12876797e-01 -2.99228565e-03
-4.78524894e-01 -1.03288722e+00 3.75375658e-01 -4.27908778e-01
-6.91295922e-01 -1.07616514e-01 -2.47417510e-01 4.24566805e-01
9.92914557e-01 3.04226410e-02 -4.22483385e-01 -2.34948412e-01
-1.06105936e+00 8.13423216e-01 -1.54714108e-01 4.66776453e-03
2.67492324e-01 -1.09849215e+00 -5.66264272e-01 -4.57342595e-01
1.19058155e-01 8.71564522e-02 9.62892398e-02 4.83035445e-01
-1.33864748e+00 8.62125397e-01 -3.35219294e-01 2.67933786e-01
-1.24170148e+00 6.13660693e-01 7.45283902e-01 -6.12141550e-01
1.96396112e-01 5.72083473e-01 -4.64659572e-01 -1.07737958e+00
-7.20493961e-03 -5.30036449e-01 -1.06974697e+00 2.93596745e-01
4.25005734e-01 3.82441282e-01 2.93792933e-01 -5.39674759e-01
-5.49169064e-01 5.92629671e-01 4.08950627e-01 -4.31824118e-01
1.08145940e+00 -7.11128786e-02 -7.80466437e-01 1.04822412e-01
7.29880989e-01 2.65073061e-01 2.62547936e-03 -3.14908564e-01
6.87817812e-01 1.08716473e-01 -4.80125070e-01 -1.28669429e+00
-2.21685529e-01 3.94481629e-01 2.36463666e-01 6.60390019e-01
9.20300066e-01 2.44978249e-01 5.81717849e-01 8.72608900e-01
2.39806458e-01 -1.12616396e+00 -5.43790340e-01 7.94377387e-01
1.01215529e+00 -8.78390491e-01 6.91742674e-02 -6.31520629e-01
-2.67659932e-01 1.21538937e+00 7.12299645e-01 1.85745999e-01
6.48548305e-01 1.75957605e-01 1.44428179e-01 -8.35952997e-01
-1.10675693e+00 -8.00688982e-01 1.72448859e-01 5.56857109e-01
3.44585031e-01 -2.22994044e-01 -1.20583677e+00 5.88669419e-01
-6.13963366e-01 5.07169962e-01 3.99749130e-01 1.01321125e+00
-9.63464916e-01 -1.19918418e+00 -5.94818711e-01 6.00220859e-01
-6.44473851e-01 -2.11161748e-01 -4.26736206e-01 7.78617978e-01
6.16067290e-01 1.26371205e+00 -2.66631618e-02 -2.05391515e-02
7.90140629e-01 6.09923542e-01 5.01670957e-01 -7.87093222e-01
-9.13719118e-01 -6.75954938e-01 1.01438129e+00 -1.05850466e-01
-3.32477540e-01 -5.00767469e-01 -1.92031169e+00 8.25718865e-02
-3.60905558e-01 9.91747022e-01 6.75786793e-01 1.21802366e+00
1.55589700e-01 1.79835305e-01 -5.54478392e-02 7.70680189e-01
-1.76281393e-01 -9.13610458e-01 -2.48926982e-01 1.62428156e-01
3.40168923e-02 -2.68063068e-01 2.27825269e-01 1.28186956e-01] | [10.390316009521484, 8.035091400146484] |
928f8b53-e7de-45e8-a60f-6d340a05c759 | darts-deceiving-autonomous-cars-with-toxic | 1802.06430 | null | http://arxiv.org/abs/1802.06430v3 | http://arxiv.org/pdf/1802.06430v3.pdf | DARTS: Deceiving Autonomous Cars with Toxic Signs | Sign recognition is an integral part of autonomous cars. Any
misclassification of traffic signs can potentially lead to a multitude of
disastrous consequences, ranging from a life-threatening accident to even a
large-scale interruption of transportation services relying on autonomous cars.
In this paper, we propose and examine security attacks against sign recognition
systems for Deceiving Autonomous caRs with Toxic Signs (we call the proposed
attacks DARTS). In particular, we introduce two novel methods to create these
toxic signs. First, we propose Out-of-Distribution attacks, which expand the
scope of adversarial examples by enabling the adversary to generate these
starting from an arbitrary point in the image space compared to prior attacks
which are restricted to existing training/test data (In-Distribution). Second,
we present the Lenticular Printing attack, which relies on an optical
phenomenon to deceive the traffic sign recognition system. We extensively
evaluate the effectiveness of the proposed attacks in both virtual and
real-world settings and consider both white-box and black-box threat models.
Our results demonstrate that the proposed attacks are successful under both
settings and threat models. We further show that Out-of-Distribution attacks
can outperform In-Distribution attacks on classifiers defended using the
adversarial training defense, exposing a new attack vector for these defenses. | ['Chawin Sitawarin', 'Mung Chiang', 'Arsalan Mosenia', 'Arjun Nitin Bhagoji', 'Prateek Mittal'] | 2018-02-18 | null | null | null | null | ['traffic-sign-recognition'] | ['computer-vision'] | [ 4.96480644e-01 -1.58001959e-01 2.57479310e-01 -1.64824069e-01
-4.85999346e-01 -1.31252289e+00 8.90592098e-01 -7.23114073e-01
-3.35730672e-01 5.59361279e-01 -6.77991390e-01 -9.03154433e-01
4.05249037e-02 -9.10299897e-01 -9.80931520e-01 -8.64246726e-01
5.77288307e-02 -2.40217745e-02 6.65953159e-01 -4.49168473e-01
3.74708682e-01 1.22894168e+00 -1.58553505e+00 8.06425065e-02
5.73179305e-01 1.07035315e+00 -7.56903708e-01 1.03863442e+00
1.03472836e-01 5.07426500e-01 -1.07692957e+00 -8.83303106e-01
8.19051802e-01 -1.24676950e-01 -6.56847730e-02 -3.06015432e-01
8.05323005e-01 -5.94690800e-01 -4.48315799e-01 1.34716022e+00
2.45783687e-01 -1.84316695e-01 6.31038189e-01 -2.29000068e+00
-4.87197906e-01 -5.42538166e-02 -3.21273267e-01 -7.28829727e-02
8.29536319e-02 6.65446222e-01 2.62892693e-01 -3.89675319e-01
3.37728709e-01 1.24969530e+00 3.96036893e-01 1.10431731e+00
-6.81793571e-01 -1.36652303e+00 -1.00782260e-01 3.06712717e-01
-1.10895514e+00 -4.03958917e-01 8.24464619e-01 -2.21842349e-01
2.51989216e-01 7.54301429e-01 1.64756507e-01 1.36017084e+00
3.72093767e-01 5.19500077e-01 1.36699224e+00 -1.24737822e-01
2.58447886e-01 2.38028109e-01 2.17968374e-01 5.59967041e-01
8.78018677e-01 9.33606505e-01 1.00662932e-01 -7.31276214e-01
2.44736522e-01 -3.33417840e-02 -5.00956289e-02 -3.47426504e-01
-6.29922390e-01 7.41584539e-01 1.37183845e-01 -2.63439894e-01
-1.83009908e-01 4.44283098e-01 1.95825383e-01 3.72684598e-01
-2.83673555e-01 2.27279082e-01 -1.61858514e-01 1.64682239e-01
-3.19686383e-01 2.34053925e-01 7.06492364e-01 8.81603956e-01
3.48496079e-01 5.51949084e-01 -4.51182686e-02 -2.88408361e-02
3.15669388e-01 1.31609285e+00 9.94157642e-02 -7.45913208e-01
3.21598053e-01 1.10982902e-01 3.69336635e-01 -9.24761653e-01
2.67470963e-02 4.21827510e-02 -3.55173171e-01 9.84502792e-01
5.70722580e-01 -4.72632587e-01 -1.17126501e+00 1.64120579e+00
4.61231798e-01 5.19636512e-01 2.17914149e-01 6.27811432e-01
4.15537447e-01 4.13539737e-01 2.00369328e-01 1.71446800e-01
1.23948741e+00 -4.90487278e-01 -5.29980898e-01 1.56329215e-01
3.70980144e-01 -4.78494495e-01 7.55336761e-01 3.99855852e-01
-5.19778490e-01 -1.88521087e-01 -1.36523831e+00 5.81489742e-01
-9.78501260e-01 -2.67617345e-01 2.98878551e-01 1.82461643e+00
-6.97030723e-01 4.45768498e-02 -5.23374736e-01 -2.48370260e-01
6.00188792e-01 4.51607436e-01 -3.22798014e-01 -3.96373495e-02
-1.31453204e+00 9.65190113e-01 -1.54942021e-01 2.32777238e-01
-1.25362372e+00 -5.52747607e-01 -5.61273515e-01 -3.09934139e-01
4.37707633e-01 -3.02346647e-01 9.40074384e-01 -5.33235967e-01
-1.38592482e+00 6.40622854e-01 -3.85669582e-02 -6.40483081e-01
9.02928650e-01 -8.45970064e-02 -9.37966526e-01 1.60466373e-01
-3.37532133e-01 4.38707352e-01 1.21962452e+00 -1.61375439e+00
-4.92016852e-01 -1.54625908e-01 1.85478061e-01 -4.50384408e-01
-2.53643692e-01 3.81110162e-01 3.18565518e-01 -4.60187644e-01
-5.19264877e-01 -1.27356744e+00 -4.50788252e-02 2.77719796e-01
-5.30359149e-01 2.24972308e-01 1.84580016e+00 -1.60217017e-01
7.25118339e-01 -2.14225793e+00 -8.54801595e-01 5.95913589e-01
6.43265471e-02 1.00412250e+00 -1.88093647e-01 2.56743878e-01
3.65774594e-02 4.97107267e-01 -1.69091091e-01 1.55378997e-01
2.10864693e-01 4.11891311e-01 -1.01097083e+00 6.30029440e-01
3.11288595e-01 8.74504983e-01 -5.67990124e-01 -1.14224413e-02
1.97214514e-01 3.89569849e-01 -2.44170815e-01 -7.36439899e-02
3.83494571e-02 2.75133193e-01 -7.42005706e-01 7.88146496e-01
1.21154630e+00 6.71313226e-01 -3.44198614e-01 1.62057728e-01
4.02248502e-02 -2.65850127e-01 -8.30783546e-01 1.15109481e-01
-2.04906724e-02 6.77024662e-01 4.04566638e-02 -5.94570279e-01
7.76186764e-01 1.40666857e-01 -8.25421140e-02 -3.66670460e-01
4.54601586e-01 2.10865140e-01 3.04343045e-01 -3.55353296e-01
2.78406501e-01 -2.46576041e-01 -3.35357964e-01 5.86957455e-01
-5.91003180e-01 -1.12056032e-01 -1.27484307e-01 1.31704643e-01
1.28408122e+00 -2.16709763e-01 -1.85109541e-01 1.81510568e-01
8.18426490e-01 -2.31796443e-01 3.88082922e-01 1.00773251e+00
-8.26965451e-01 1.04096197e-01 6.94261551e-01 -1.59561932e-01
-7.87829816e-01 -1.24140322e+00 1.55623937e-02 6.00550354e-01
3.65801126e-01 2.18879238e-01 -6.87627554e-01 -1.36653674e+00
5.03119230e-01 1.05716515e+00 -6.30070627e-01 -7.12287068e-01
-6.22464776e-01 -4.03446138e-01 1.69883573e+00 4.55178767e-01
8.07260334e-01 -7.68001497e-01 -6.66708767e-01 -5.25847435e-01
3.50511760e-01 -1.35637891e+00 -2.32966632e-01 -3.33327442e-01
-1.90522239e-01 -1.36351168e+00 -3.16912562e-01 -3.24769855e-01
9.02931333e-01 5.40759325e-01 2.45947242e-01 4.27682638e-01
-3.09291840e-01 6.32970929e-01 -2.31079504e-01 -9.90115345e-01
-1.07460678e+00 -6.94703162e-01 2.73535252e-01 6.25046253e-01
5.92436075e-01 -8.27804059e-02 -2.97229052e-01 8.68715048e-01
-1.15561223e+00 -7.73471415e-01 4.84041154e-01 4.96899188e-01
-5.13487980e-02 1.42727062e-01 5.76196015e-01 -6.66486859e-01
6.35716915e-01 -3.42355996e-01 -1.02126038e+00 3.18425149e-01
-1.88785091e-01 -5.26024252e-02 7.87033021e-01 -8.26063156e-01
-8.73644352e-01 2.70216744e-02 -8.10371265e-02 -6.50580168e-01
-5.99166870e-01 -3.29670548e-01 -5.68853021e-01 -9.35087264e-01
4.33019251e-01 3.01550031e-01 1.10393263e-01 3.86729389e-02
4.88120466e-01 6.01969182e-01 3.94029111e-01 -4.85575765e-01
1.76560354e+00 8.35612118e-01 6.23335540e-01 -1.03213906e+00
-2.35949382e-01 4.67456765e-02 -3.02098364e-01 -5.26982725e-01
6.74425721e-01 -1.76188484e-01 -1.10938811e+00 1.00328636e+00
-1.12203813e+00 -7.22027794e-02 9.74187106e-02 1.27258599e-01
-3.29698771e-01 6.04079008e-01 -1.53242633e-01 -1.18215394e+00
1.60901859e-01 -1.14461267e+00 8.56764495e-01 9.76849571e-02
2.19327748e-01 -5.91934741e-01 -2.42757797e-01 3.43368560e-01
6.79233968e-01 6.58462226e-01 6.73130810e-01 -1.10475230e+00
-8.14133108e-01 -9.60471332e-01 -1.07535012e-01 6.43312991e-01
2.57321503e-02 2.54128456e-01 -1.04589713e+00 -1.89162299e-01
1.63430080e-01 -3.23948592e-01 6.24949038e-01 -1.92633435e-01
8.62825930e-01 -5.87067246e-01 -4.32257980e-01 4.96057898e-01
9.50849175e-01 7.31138825e-01 9.47785974e-01 1.99040949e-01
4.88913000e-01 5.98148406e-01 6.48439348e-01 7.16128200e-02
-2.26099998e-01 6.43849194e-01 7.07149684e-01 7.82541707e-02
1.17691532e-01 -2.20740661e-01 6.01121247e-01 -3.48789066e-01
1.82633057e-01 -4.29637909e-01 -6.25053585e-01 1.36593003e-02
-1.40383625e+00 -1.29974174e+00 -3.36112946e-01 2.39201260e+00
2.60410994e-01 3.29196155e-01 2.48706535e-01 2.05274448e-01
9.20863450e-01 -5.11940271e-02 -6.18098795e-01 -6.55311227e-01
-1.01440929e-01 1.47677034e-01 1.04531085e+00 5.67344666e-01
-1.14854348e+00 9.46908534e-01 6.29598856e+00 6.66519582e-01
-1.30388987e+00 1.69459462e-01 3.56411308e-01 3.57630961e-02
-1.37787804e-01 1.93696264e-02 -9.77611244e-01 6.63377643e-01
1.06444895e+00 -1.24739185e-01 8.35533440e-02 7.49774456e-01
1.10654525e-01 3.08080047e-01 -8.75228822e-01 4.12786156e-01
3.83650631e-01 -8.45787048e-01 3.36538255e-01 4.21650976e-01
2.66384184e-01 -3.58447939e-01 5.35792649e-01 3.36654902e-01
6.49026275e-01 -9.17093694e-01 7.11223364e-01 2.07603842e-01
7.77634144e-01 -8.78807425e-01 5.86138606e-01 3.12967479e-01
-8.83142591e-01 -2.25646034e-01 -5.59210256e-02 2.57744044e-01
3.34580004e-01 -1.77056134e-01 -6.04749620e-01 3.49912196e-01
1.38675317e-01 -2.38833889e-01 -5.30608773e-01 9.54226971e-01
-5.32385826e-01 8.77192914e-01 -6.37244225e-01 -2.93618441e-01
2.67530501e-01 1.51217341e-01 9.19129848e-01 9.92461324e-01
3.12975138e-01 1.03754513e-01 -1.35362253e-01 9.38569665e-01
9.18928236e-02 -6.82991683e-01 -1.22392023e+00 6.67393431e-02
4.38790917e-01 9.41890001e-01 -4.90006953e-01 -2.05826029e-01
-1.15616560e-01 8.15259457e-01 -4.80081439e-01 5.32275677e-01
-1.43308485e+00 -7.46357083e-01 1.05391884e+00 -3.42485234e-02
5.22101879e-01 -2.64633238e-01 -5.26569672e-02 -8.10992956e-01
1.26501068e-01 -8.47971737e-01 1.49213865e-01 -5.55722177e-01
-1.23477674e+00 3.69202077e-01 1.98023528e-01 -1.51705921e+00
-9.03010741e-02 -1.02263618e+00 -9.96705949e-01 5.88679969e-01
-1.67303741e+00 -1.30785501e+00 -9.53017622e-02 7.90467560e-01
-6.86802343e-02 -3.66237134e-01 4.07963008e-01 1.99921742e-01
-4.13757354e-01 1.07313418e+00 -4.77063702e-03 3.79035175e-01
5.19849956e-01 -5.45921445e-01 6.53926194e-01 1.37542844e+00
-1.97967649e-01 5.56994200e-01 6.01373792e-01 -8.07355046e-01
-1.48143435e+00 -1.21843958e+00 5.42087793e-01 -7.81697035e-01
8.40412378e-01 -6.17120266e-01 -5.08034110e-01 6.33023024e-01
-1.13416500e-01 2.61927187e-01 4.55609024e-01 -8.21988881e-01
-8.01044643e-01 -1.53312281e-01 -1.81676018e+00 8.38623583e-01
7.94919193e-01 -4.25203234e-01 -3.59834462e-01 9.87228528e-02
5.18415928e-01 4.56960388e-02 5.48091426e-04 4.00775492e-01
8.33055794e-01 -7.72280395e-01 1.03259695e+00 -1.07937980e+00
-1.17537998e-01 -6.41319811e-01 -2.18610406e-01 -8.52522433e-01
2.57635295e-01 -1.05364573e+00 -6.86839549e-03 1.17133760e+00
2.84940511e-01 -1.45048165e+00 7.03726351e-01 7.56839812e-01
3.08011007e-02 -2.68925071e-01 -1.15385675e+00 -1.42557573e+00
3.54962796e-01 -7.20574856e-01 7.92633057e-01 6.54709697e-01
-4.04750526e-01 -4.56156641e-01 -4.50393766e-01 8.30821753e-01
9.90646780e-01 -4.71981764e-01 1.20618916e+00 -8.72157872e-01
-2.14373227e-02 -5.33552706e-01 -9.02866662e-01 -7.61076152e-01
2.29721442e-01 -5.03235698e-01 8.77993628e-02 -4.76585865e-01
-2.12604254e-01 -3.45159829e-01 -2.82431126e-01 4.63111311e-01
6.76702857e-02 5.01029670e-01 5.19948900e-01 -4.48514037e-02
-1.32639587e-01 3.01671356e-01 8.44133377e-01 -3.48194152e-01
3.40011060e-01 4.71443623e-01 -5.25745213e-01 7.32538760e-01
8.10035706e-01 -6.87453151e-01 -3.24208170e-01 2.42439911e-01
-2.15269461e-01 -3.91157061e-01 9.09751892e-01 -1.03992963e+00
8.32117945e-02 -4.55787629e-01 -8.48603621e-02 -4.46040303e-01
3.78030121e-01 -1.22561777e+00 -1.09188244e-01 8.06096017e-01
-1.64992258e-01 -1.47212058e-01 4.63151187e-01 7.75529921e-01
1.34701684e-01 -1.78136528e-01 1.02486420e+00 3.33541065e-01
-4.36176807e-01 9.77661014e-02 -6.66131079e-01 -3.16522300e-01
1.64275301e+00 -6.76605344e-01 -1.04835975e+00 -5.12011886e-01
-2.87313551e-01 7.24953860e-02 4.53256458e-01 5.13393223e-01
7.82558858e-01 -1.30335546e+00 -5.43075442e-01 5.81794262e-01
9.85676423e-02 -8.57847989e-01 1.72512159e-02 5.33215880e-01
-5.31547427e-01 4.21352595e-01 -3.05334806e-01 -1.73972756e-01
-1.71756160e+00 8.60523880e-01 3.65289360e-01 8.51115957e-02
-1.69547915e-01 5.36280751e-01 4.54974025e-01 -2.76753217e-01
1.69998929e-01 -9.95431095e-02 7.69330785e-02 -4.94939446e-01
5.39840281e-01 5.04580379e-01 -5.89133278e-02 -9.11312759e-01
-5.29768527e-01 5.25572002e-01 -8.60477686e-02 -2.25606799e-01
5.10431826e-01 2.24339768e-01 2.08024830e-01 -3.15638155e-01
9.81707454e-01 3.22004884e-01 -1.13743079e+00 3.66600811e-01
-1.88558683e-01 -7.47300863e-01 -4.24569696e-01 -9.46647882e-01
-9.82353568e-01 8.20978403e-01 7.07499266e-01 4.08118755e-01
7.90043175e-01 -3.75341177e-01 1.04162073e+00 7.00881898e-01
5.21570802e-01 -5.54786682e-01 -9.40556526e-02 3.90235424e-01
7.37220943e-01 -1.01734591e+00 -3.68292511e-01 -5.86017787e-01
-6.99054003e-01 8.52134645e-01 5.82804024e-01 -1.97357565e-01
8.29072714e-01 4.11326885e-01 3.66489977e-01 1.21321958e-02
-5.25271952e-01 -3.56643870e-02 1.47291481e-01 1.07475138e+00
-5.48620105e-01 2.43435777e-03 -1.52642846e-01 3.82149071e-01
-9.87989828e-02 -1.47738785e-01 9.42165911e-01 1.39184964e+00
-2.79174656e-01 -1.16047454e+00 -9.02436972e-01 1.09016992e-01
-3.56627643e-01 2.36098453e-01 -1.04325175e+00 1.00672913e+00
1.11959383e-01 1.17834580e+00 -2.12158725e-01 -6.35794461e-01
4.69737560e-01 2.26348326e-01 1.56247079e-01 2.06816383e-02
-3.86767924e-01 -5.94774246e-01 2.32946381e-01 -5.48824072e-01
-9.28486660e-02 -7.17412233e-01 -1.02388036e+00 -5.21453738e-01
-3.80796760e-01 -2.11089209e-01 7.09507465e-01 8.60499799e-01
4.29275334e-01 8.22375864e-02 9.18807626e-01 -6.44402742e-01
-1.12619340e+00 -4.28177923e-01 -4.46293145e-01 5.06520152e-01
6.07283771e-01 -9.32269454e-01 -9.15792763e-01 -1.48446038e-01] | [5.4310302734375, 7.7906880378723145] |
74ae6e50-bcb6-4028-9d5c-c0a13ec56e2b | interface-adjustable-angular-margin-inter | 2210.02018 | null | https://arxiv.org/abs/2210.02018v2 | https://arxiv.org/pdf/2210.02018v2.pdf | InterFace:Adjustable Angular Margin Inter-class Loss for Deep Face Recognition | In the field of face recognition, it is always a hot research topic to improve the loss solution to make the face features extracted by the network have greater discriminative power. Research works in recent years has improved the discriminative power of the face model by normalizing softmax to the cosine space step by step and then adding a fixed penalty margin to reduce the intra-class distance to increase the inter-class distance. Although a great deal of previous work has been done to optimize the boundary penalty to improve the discriminative power of the model, adding a fixed margin penalty to the depth feature and the corresponding weight is not consistent with the pattern of data in the real scenario. To address this issue, in this paper, we propose a novel loss function, InterFace, releasing the constraint of adding a margin penalty only between the depth feature and the corresponding weight to push the separability of classes by adding corresponding margin penalties between the depth features and all weights. To illustrate the advantages of InterFace over a fixed penalty margin, we explained geometrically and comparisons on a set of mainstream benchmarks. From a wider perspective, our InterFace has advanced the state-of-the-art face recognition performance on five out of thirteen mainstream benchmarks. All training codes, pre-trained models, and training logs, are publicly released \footnote{$https://github.com/iamsangmeng/InterFace$}. | ['Shan Zhao', 'Yang Yang', 'Anning Pan', 'Pan Tan', 'Mengzhen Li', 'Jiaxuan Chen', 'Meng Sang'] | 2022-10-05 | null | null | null | null | ['face-model'] | ['computer-vision'] | [ 7.56476298e-02 -8.90114605e-02 -3.46699446e-01 -9.71861064e-01
-1.70139119e-01 -4.97220196e-02 4.22779024e-01 -1.55622870e-01
-4.95476216e-01 4.52892810e-01 4.53163907e-02 -1.22673690e-01
-1.12816691e-01 -7.21446574e-01 -5.39328873e-01 -7.17484593e-01
-2.72768624e-02 2.71183671e-03 6.42375350e-02 2.44748630e-02
1.76209375e-01 5.78803897e-01 -1.44411445e+00 1.29956976e-01
7.28291333e-01 1.30822718e+00 4.05890197e-02 -5.58139645e-02
-7.15824366e-02 2.21599057e-01 -4.66615945e-01 -6.56550348e-01
6.12926185e-01 -2.77211338e-01 -4.90966469e-01 -5.95131181e-02
6.03132606e-01 -3.57903183e-01 -5.05967259e-01 1.20534170e+00
7.13547170e-01 9.21175033e-02 4.72382784e-01 -1.53366029e+00
-6.35614932e-01 3.76153141e-01 -7.73471534e-01 1.50936753e-01
9.75681171e-02 -8.10095817e-02 8.73629034e-01 -9.65722442e-01
2.49715582e-01 1.15901756e+00 5.41579783e-01 5.94834566e-01
-1.04052925e+00 -1.13461089e+00 1.97144896e-01 3.48809153e-01
-1.55979598e+00 -5.82075059e-01 9.04747844e-01 -2.00612545e-01
8.11245918e-01 2.56045491e-01 2.65235662e-01 9.58989620e-01
-4.00966592e-02 5.44420540e-01 7.43858635e-01 -4.26954389e-01
4.87632304e-02 2.49009430e-01 1.75787732e-01 7.03388095e-01
8.76380801e-02 6.57647103e-02 -3.83633107e-01 -9.32836160e-02
6.97747827e-01 2.72244543e-01 -3.91826689e-01 -3.09735090e-01
-6.05601490e-01 7.94442594e-01 5.64432323e-01 3.94171089e-01
1.25509826e-02 -2.53796969e-02 2.60271639e-01 1.80617765e-01
4.17725593e-01 1.68511048e-02 -4.30402607e-01 -3.04766628e-03
-9.81602013e-01 -1.21168517e-01 6.28572106e-01 7.20863402e-01
8.24121177e-01 -8.02957043e-02 -3.23216021e-01 1.12911856e+00
4.18572605e-01 2.60377288e-01 4.10095930e-01 -7.21458733e-01
5.98183155e-01 7.42419958e-01 -3.66059273e-01 -1.10724449e+00
-1.56622246e-01 -5.92170417e-01 -8.70475471e-01 1.16559915e-01
4.61294115e-01 -1.45103186e-01 -8.88201118e-01 1.89313269e+00
2.73782879e-01 3.56466293e-01 -4.76218313e-01 9.53286767e-01
5.76842308e-01 5.58829486e-01 -1.05556034e-01 -1.32418901e-01
1.28537464e+00 -7.46327400e-01 -5.34478128e-01 -2.97786653e-01
4.45730329e-01 -8.33524644e-01 1.06652343e+00 9.13019180e-02
-6.98237717e-01 -6.35825813e-01 -1.09272921e+00 1.61622003e-01
-2.28463098e-01 3.40146005e-01 5.26136756e-01 7.59525001e-01
-7.45015681e-01 6.29329681e-01 -7.95246363e-01 -2.17805624e-01
5.76314688e-01 5.74100375e-01 -5.60666740e-01 -2.44202852e-01
-1.15767622e+00 6.50554180e-01 1.33874208e-01 1.80921599e-01
-4.22391623e-01 -6.10553861e-01 -7.54452646e-01 2.82258660e-01
4.36572045e-01 -1.42400667e-01 7.91986048e-01 -1.07207704e+00
-1.45880437e+00 6.78251624e-01 -2.04364419e-01 -8.03701356e-02
4.12937969e-01 -2.10093990e-01 -5.11731088e-01 -1.67268261e-01
-2.26714000e-01 6.87607169e-01 6.82864606e-01 -9.26493824e-01
-3.99644881e-01 -5.85149646e-01 -2.01741289e-02 -8.96904245e-02
-9.82218146e-01 9.99119654e-02 -7.35025644e-01 -7.41952896e-01
-2.90644243e-02 -8.70190740e-01 6.72013834e-02 1.75178170e-01
-2.70663768e-01 -4.41662461e-01 9.07987237e-01 -5.20104527e-01
1.42371368e+00 -2.52794766e+00 -1.61358461e-01 3.34483892e-01
-1.09769642e-01 4.98959839e-01 -2.96079367e-01 2.27409720e-01
-2.17593551e-01 3.37190628e-02 -3.98882717e-01 -4.32332873e-01
1.29554588e-02 9.24966857e-02 -1.70769989e-01 5.41505754e-01
3.48603070e-01 3.63449037e-01 -5.01395881e-01 -1.92976162e-01
-1.10561252e-01 8.01220179e-01 -8.37785661e-01 1.35103151e-01
1.90049633e-01 6.22724108e-02 -3.29051167e-01 5.17312706e-01
1.06484807e+00 -9.42986086e-02 7.89697543e-02 -3.55683476e-01
1.40953287e-01 1.87661171e-01 -1.49377751e+00 1.36501336e+00
-2.18780160e-01 4.62048233e-01 2.86770761e-02 -1.14409328e+00
1.05860579e+00 1.63830072e-01 5.03875196e-01 -6.62678957e-01
1.58458307e-01 1.47411093e-01 2.63462931e-01 -2.85059094e-01
1.04027607e-01 -2.83159073e-02 3.52895349e-01 2.05130577e-01
5.65630011e-02 4.41162020e-01 -1.56153506e-02 -1.58580780e-01
8.12549889e-01 -4.94183134e-03 3.61545593e-04 -2.20064238e-01
6.51670396e-01 -7.05920756e-01 9.16130483e-01 2.20375136e-01
-3.34719181e-01 6.20895147e-01 4.57408130e-01 -3.22668016e-01
-6.40576005e-01 -8.24129581e-01 -4.69243407e-01 1.02183712e+00
-6.08569290e-03 -5.52350521e-01 -7.89496601e-01 -9.49322045e-01
2.75190920e-01 3.27747613e-01 -6.89477861e-01 -2.76941657e-01
-6.01127625e-01 -8.44435275e-01 5.66077411e-01 5.60144603e-01
8.77477169e-01 -8.12225640e-01 -1.42032996e-01 -1.40439495e-01
1.72001988e-01 -9.55951571e-01 -8.41284990e-01 1.58402815e-01
-5.16284227e-01 -8.76466393e-01 -6.06294811e-01 -9.21327412e-01
9.05905664e-01 1.67700872e-01 6.86094344e-01 3.27988386e-01
-4.07881975e-01 -1.81816339e-01 -2.51393229e-01 -2.53815949e-01
2.43574739e-01 1.27384797e-01 3.85637805e-02 2.07868174e-01
4.31068480e-01 -4.48122561e-01 -7.25632370e-01 5.89615941e-01
-8.98086607e-01 -2.74701267e-01 4.03775901e-01 8.95407677e-01
2.74427086e-01 2.07623038e-02 6.05251729e-01 -5.54507136e-01
4.39925492e-01 -4.32617992e-01 -4.76752192e-01 2.10612059e-01
-6.94685638e-01 4.04313430e-02 5.68421066e-01 -5.21318734e-01
-7.95526028e-01 -1.84019152e-02 -3.89053822e-01 -4.48755473e-01
9.70187262e-02 3.75358909e-01 -5.90058267e-01 -1.31208286e-01
2.53963798e-01 1.27379552e-01 3.99829336e-02 -5.09808362e-01
5.38542569e-02 8.07909727e-01 1.81847692e-01 -5.21557629e-01
6.38585865e-01 3.47737014e-01 -1.24925850e-02 -6.11431003e-01
-5.57197392e-01 -3.11579585e-01 -4.38822925e-01 1.08126163e-01
4.11736310e-01 -6.67602479e-01 -6.96100295e-01 4.38532591e-01
-8.46722126e-01 -1.94967270e-01 1.07000386e-02 6.55812442e-01
4.22565043e-02 4.90596443e-01 -4.62401152e-01 -5.21979749e-01
-2.96276182e-01 -1.27122712e+00 7.10402071e-01 4.30259436e-01
1.79986143e-03 -7.16793060e-01 -1.57593548e-01 1.47942483e-01
5.74756265e-01 -6.13067746e-02 7.05525756e-01 -6.43531561e-01
-1.75527021e-01 -2.30296597e-01 -4.27868694e-01 8.11947763e-01
4.29611623e-01 1.10746138e-01 -9.07870471e-01 -6.38581812e-01
9.14765447e-02 -6.59168139e-02 1.10472548e+00 1.05489798e-01
1.42855036e+00 -3.12016129e-01 -3.26809615e-01 8.26870799e-01
1.30590463e+00 2.76396394e-01 7.52916873e-01 2.07050040e-01
5.22814095e-01 4.60747212e-01 2.51887172e-01 3.78668427e-01
2.82053441e-01 9.73143458e-01 4.24327314e-01 1.86232198e-02
-3.38178836e-02 -9.72742289e-02 4.44515824e-01 3.69707853e-01
2.15050839e-02 -1.68890342e-01 -6.72075093e-01 3.23593408e-01
-1.48699725e+00 -9.37800050e-01 4.01047736e-01 2.42593217e+00
9.10078049e-01 1.85870633e-01 4.74510677e-02 1.84982732e-01
8.24066520e-01 1.65641949e-01 -4.03985500e-01 -2.05978423e-01
8.51968974e-02 2.30547503e-01 4.05216575e-01 5.12024820e-01
-1.06057262e+00 7.33392119e-01 5.08247757e+00 1.15009856e+00
-1.55410182e+00 -6.95483759e-02 7.59477675e-01 -2.62097210e-01
8.34418368e-03 8.85835364e-02 -1.24842680e+00 7.56438196e-01
5.98173022e-01 7.56418630e-02 3.66569668e-01 9.10398304e-01
-1.55718881e-03 7.90311024e-02 -1.26041675e+00 1.00691974e+00
2.58108944e-01 -9.01305914e-01 -5.98199992e-03 2.15609059e-01
4.07749623e-01 -5.23245893e-02 1.75767466e-01 3.69512230e-01
-2.55080462e-01 -1.14714384e+00 5.23063123e-01 2.84685254e-01
8.07338834e-01 -7.22574055e-01 8.17408979e-01 1.76745251e-01
-1.15748990e+00 -1.87121317e-01 -4.07776356e-01 -8.86022002e-02
-1.83346018e-01 7.37334549e-01 -5.94845176e-01 3.98212582e-01
7.06040621e-01 5.57690084e-01 -6.23448730e-01 1.11828399e+00
-9.76471826e-02 6.11678958e-01 -4.70636576e-01 2.15782225e-01
1.95884220e-02 -1.77385822e-01 3.12082559e-01 1.23662281e+00
3.42039168e-01 -4.85592075e-02 1.21381320e-01 6.93904817e-01
-3.37857962e-01 1.60834000e-01 -2.68292993e-01 1.80729628e-01
5.26415229e-01 1.25619853e+00 -5.23097336e-01 -7.60602653e-02
-5.73037267e-01 8.40871692e-01 3.18417042e-01 2.11490348e-01
-9.46820736e-01 -6.78120136e-01 8.18120599e-01 3.43439907e-01
3.50208759e-01 -2.32677594e-01 -3.25460285e-01 -9.36629713e-01
4.11072552e-01 -7.33612895e-01 4.50676382e-01 -1.36631146e-01
-1.32782197e+00 6.66743934e-01 -1.05618797e-01 -1.07462704e+00
1.08926527e-01 -5.67748785e-01 -7.24690139e-01 1.03773236e+00
-1.46274889e+00 -8.84524524e-01 -3.78221005e-01 5.63556790e-01
3.10227692e-01 -2.11845800e-01 5.88472247e-01 7.65405297e-01
-1.04676235e+00 1.18730664e+00 -5.14248796e-02 4.62315261e-01
1.03532994e+00 -5.68848431e-01 1.16396919e-01 7.02289701e-01
-8.34588632e-02 9.71116841e-01 4.86400396e-01 -3.84865075e-01
-1.05124044e+00 -9.31177437e-01 8.21249604e-01 5.55115826e-02
2.68722296e-01 -4.80475605e-01 -1.22581565e+00 5.13237119e-01
2.29605511e-02 4.59295422e-01 7.32293904e-01 1.36300297e-02
-6.07941866e-01 -6.02550328e-01 -1.26890302e+00 4.69035774e-01
1.06655645e+00 -4.27245826e-01 -2.26126149e-01 2.91647017e-02
2.77823150e-01 -3.00903190e-02 -7.32393503e-01 7.76697338e-01
8.39376271e-01 -8.48466992e-01 8.55824292e-01 -4.20964539e-01
2.52923578e-01 -4.30491388e-01 -2.88465291e-01 -1.12920415e+00
-3.25592905e-01 -1.90417647e-01 4.70174365e-02 1.51526833e+00
3.09212327e-01 -8.19031060e-01 9.07185078e-01 6.63771689e-01
-1.81753524e-02 -1.27620530e+00 -1.12508154e+00 -8.44646156e-01
2.88261753e-02 -2.30962947e-01 6.77924633e-01 9.32734907e-01
-2.84851119e-02 1.98620558e-02 -3.70151728e-01 1.11589000e-01
3.47896308e-01 -1.45666942e-01 5.18152237e-01 -1.09530663e+00
-2.80902833e-01 -6.82412982e-01 -4.30423051e-01 -1.10639763e+00
1.88705221e-01 -9.66312289e-01 -2.52227485e-01 -1.21890581e+00
3.40484798e-01 -5.31497359e-01 -5.18045723e-01 8.52021873e-01
-2.02696010e-01 4.17165995e-01 3.87457669e-01 1.18240029e-01
-1.83676437e-01 7.39636004e-01 1.06311512e+00 -1.40428320e-01
-1.46304846e-01 -3.06554325e-02 -7.82830536e-01 5.23330271e-01
8.32512438e-01 -4.93210286e-01 -3.55814427e-01 -6.25459969e-01
-1.16411753e-01 -3.91811430e-01 2.58614033e-01 -9.64928567e-01
3.71724904e-01 -1.71812043e-01 5.75343251e-01 -1.38338745e-01
3.58689576e-01 -8.84167135e-01 1.60199869e-02 4.28063542e-01
-2.13530242e-01 -1.66661322e-01 2.82203615e-01 2.50562876e-01
-2.59555548e-01 -1.47177100e-01 1.02403522e+00 2.97727287e-01
-4.81283575e-01 5.11878550e-01 2.77166039e-01 -1.62699729e-01
1.01934445e+00 -3.15517634e-01 -2.89461613e-01 -1.38458773e-01
-4.54375118e-01 2.36447036e-01 3.78865331e-01 6.94081664e-01
5.31655788e-01 -1.39197171e+00 -7.39546537e-01 6.88336015e-01
-8.18673298e-02 -3.60530674e-01 1.66618913e-01 7.58602023e-01
-1.10332966e-01 3.42591763e-01 -3.26826632e-01 -3.60179245e-01
-1.40856135e+00 2.97673702e-01 5.35153866e-01 -1.16354823e-01
-3.91550928e-01 1.00546992e+00 2.08804175e-01 -2.85779238e-01
5.48165619e-01 -1.45649120e-01 -1.27057478e-01 -2.05332208e-02
5.76142490e-01 2.60282427e-01 2.14575931e-01 -6.14135861e-01
-6.61241233e-01 6.11636460e-01 -1.59832865e-01 2.28471592e-01
1.34603429e+00 8.17683265e-02 -6.29925355e-02 -8.52474049e-02
1.48938942e+00 1.43839568e-01 -1.41925728e+00 -2.79654205e-01
-1.17594309e-01 -7.20319450e-01 -1.24544026e-02 -7.05441058e-01
-1.56690717e+00 8.17300797e-01 8.61868739e-01 -2.08020676e-02
1.22036564e+00 -1.39351726e-01 6.54003561e-01 3.68898585e-02
1.29837468e-01 -9.88919854e-01 1.61259919e-01 6.15291059e-01
8.93046200e-01 -1.31576335e+00 -6.15007710e-03 -5.19717872e-01
-3.35378349e-01 1.05765450e+00 8.23119998e-01 -7.52867460e-02
1.07565069e+00 3.52442265e-01 -1.74151495e-01 1.00819826e-01
-4.08443958e-01 1.07874073e-01 3.79847050e-01 1.79941565e-01
6.96010590e-01 -3.81625593e-02 -5.30849040e-01 8.14453423e-01
-1.21288598e-01 8.97595137e-02 -1.07393079e-01 8.72540295e-01
-2.96154290e-01 -1.38705707e+00 -1.98546141e-01 4.25583333e-01
-6.62346423e-01 -7.30879158e-02 -4.90723252e-02 4.94371623e-01
4.80740696e-01 9.04548228e-01 1.74964383e-01 -4.57769245e-01
4.35659796e-01 1.34120286e-01 3.82981777e-01 -5.54475546e-01
-3.42120260e-01 -1.04872905e-01 -1.76095277e-01 -3.57706845e-01
-7.33831972e-02 -3.81204933e-01 -1.21126449e+00 -3.50735128e-01
-3.67156118e-01 1.49102628e-01 6.42210841e-01 8.41195762e-01
4.87843931e-01 2.47296304e-01 8.33039224e-01 -6.29022419e-01
-7.85272598e-01 -1.16316330e+00 -5.03787756e-01 3.60377550e-01
7.25352913e-02 -8.06319237e-01 -4.97310847e-01 -1.64442092e-01] | [13.243943214416504, 0.7299585938453674] |
6846fec6-846e-45cc-9654-631e5600c231 | deep-learning-for-real-time-crime-forecasting-1 | 1707.03340 | null | http://arxiv.org/abs/1707.03340v1 | http://arxiv.org/pdf/1707.03340v1.pdf | Deep Learning for Real Time Crime Forecasting | Accurate real time crime prediction is a fundamental issue for public safety,
but remains a challenging problem for the scientific community. Crime
occurrences depend on many complex factors. Compared to many predictable
events, crime is sparse. At different spatio-temporal scales, crime
distributions display dramatically different patterns. These distributions are
of very low regularity in both space and time. In this work, we adapt the
state-of-the-art deep learning spatio-temporal predictor, ST-ResNet [Zhang et
al, AAAI, 2017], to collectively predict crime distribution over the Los
Angeles area. Our models are two staged. First, we preprocess the raw crime
data. This includes regularization in both space and time to enhance
predictable signals. Second, we adapt hierarchical structures of residual
convolutional units to train multi-factor crime prediction models. Experiments
over a half year period in Los Angeles reveal highly accurate predictive power
of our models. | ['Andrea L. Bertozzi', 'Duo Zhang', 'P. Jeffery Brantingham', 'Duanhao Zhang', 'Bao Wang'] | 2017-07-09 | null | null | null | null | ['crime-prediction'] | ['miscellaneous'] | [ 4.76967469e-02 -6.85606122e-01 -1.49592841e-02 -5.83852470e-01
-9.61416721e-01 -2.94664055e-01 5.48611879e-01 4.67025250e-01
-5.91297030e-01 8.30401778e-01 8.38812113e-01 -2.67128319e-01
-3.73164028e-01 -1.02745557e+00 -6.93501294e-01 -2.56424308e-01
-3.06469411e-01 9.46014822e-02 -4.09968905e-02 -2.56454706e-01
2.65526384e-01 4.30803031e-01 -9.35096204e-01 5.71863592e-01
8.08846474e-01 4.37809616e-01 -2.42295831e-01 5.20397604e-01
2.22118109e-01 1.02971530e+00 -2.78572232e-01 -3.28758240e-01
9.09293070e-02 -8.53655189e-02 -7.42647171e-01 -3.31338555e-01
2.13446513e-01 -3.30562919e-01 -8.81850779e-01 7.73593128e-01
3.31003904e-01 4.38102156e-01 7.46281326e-01 -7.62141287e-01
-7.67301261e-01 6.27953649e-01 -7.41881669e-01 8.15896690e-01
3.65665078e-01 1.83507532e-01 8.02321315e-01 -6.82467997e-01
4.32960302e-01 1.12633288e+00 1.12005210e+00 1.81974128e-01
-1.31849909e+00 -8.13321173e-01 8.91496539e-02 4.54508364e-01
-1.36790144e+00 -3.55810970e-01 8.23728800e-01 -8.85566175e-01
1.41528010e+00 -1.29969865e-01 6.07101262e-01 1.50781429e+00
5.15813708e-01 7.50949860e-01 8.86550546e-01 1.62219614e-01
1.82898849e-01 -6.24415100e-01 4.01263684e-01 2.47630790e-01
8.00204277e-03 1.24520317e-01 -5.04898787e-01 -2.44644165e-01
6.43665969e-01 7.16878712e-01 2.16743186e-01 5.40215015e-01
-6.06381416e-01 9.32372868e-01 4.64067668e-01 5.26361227e-01
-4.61931139e-01 2.60512382e-01 5.08605480e-01 -1.23414025e-01
1.07278061e+00 1.71837777e-01 -3.03662658e-01 -6.09013498e-01
-1.11561084e+00 5.52470267e-01 3.78007591e-02 1.03150621e-01
5.47465622e-01 8.21033046e-02 -2.27844924e-01 9.87815917e-01
-1.85150474e-01 5.57943165e-01 1.32750943e-01 -4.51136857e-01
1.00606859e+00 5.05938470e-01 -1.79045603e-01 -1.52430713e+00
-8.58429790e-01 -1.79325417e-01 -1.36628103e+00 -2.04276055e-01
4.98618454e-01 -3.88091743e-01 -7.70159602e-01 1.61069572e+00
-4.74199876e-02 8.16298783e-01 -4.30844784e-01 6.98923826e-01
3.76699388e-01 7.65800178e-01 5.41652560e-01 8.44369605e-02
7.84189761e-01 -3.09644341e-01 -5.06577134e-01 -3.58377904e-01
6.68648779e-01 -2.28597552e-01 7.77103245e-01 1.38569385e-01
-9.48724210e-01 -4.32799965e-01 -5.42422414e-01 -2.51875460e-01
-4.31660414e-01 2.65599731e-02 7.42425561e-01 3.15930665e-01
-7.60455668e-01 8.92614484e-01 -1.24778473e+00 -2.66146868e-01
9.80898499e-01 1.92228165e-02 -3.72756541e-01 -7.20676631e-02
-1.33600426e+00 6.36940360e-01 -9.99609753e-02 5.11569530e-02
-5.44694781e-01 -9.42371130e-01 -1.13531327e+00 2.41884083e-01
-2.80224353e-01 -9.01337937e-02 6.26015544e-01 -4.16889638e-01
-6.86661899e-01 6.28440082e-01 -6.21031225e-01 -4.56698239e-01
2.85135895e-01 -3.13596129e-01 -6.39211535e-01 -2.72949517e-01
7.75258005e-01 -1.36386663e-01 5.31519353e-01 -5.66641808e-01
-6.92022860e-01 -7.19689310e-01 -3.02660823e-01 -3.07831198e-01
-4.28042442e-01 3.07890564e-01 -1.96334422e-01 -8.57291043e-01
-1.00443073e-01 -6.05747879e-01 -7.18756855e-01 -6.98527098e-01
-2.47389257e-01 -3.81025195e-01 2.99963117e-01 -9.18041825e-01
1.80919600e+00 -2.36471534e+00 -2.59740561e-01 1.79636762e-01
1.60981372e-01 5.42308018e-02 -2.18196273e-01 4.29121017e-01
-3.49333882e-01 1.89631134e-01 -4.22925681e-01 -6.98402762e-01
-2.07378939e-01 -1.11225583e-01 -7.31931806e-01 8.44448924e-01
5.12888074e-01 9.00689781e-01 -9.42921877e-01 -9.07172039e-02
2.65007585e-01 6.50461912e-01 -7.99909592e-01 -1.90349862e-01
4.67802644e-01 4.97874409e-01 -5.12835622e-01 4.01404232e-01
6.22180343e-01 -2.01542974e-01 -2.48781160e-01 5.01071513e-01
-2.99108207e-01 5.74829578e-01 -8.04108262e-01 1.48528230e+00
-3.66854310e-01 9.35920238e-01 -3.76537263e-01 -1.23464668e+00
7.61831403e-01 2.07324877e-01 9.42087591e-01 -1.13015485e+00
-4.22209837e-02 -1.00586288e-01 -3.28039825e-01 -5.99401116e-01
6.15530431e-01 8.95130485e-02 -4.56285357e-01 6.65982366e-01
-3.05002570e-01 1.81077138e-01 1.91480353e-01 -1.48466835e-02
1.66702020e+00 -2.93689370e-01 1.61164161e-02 -7.05680773e-02
-1.51633918e-02 1.44208316e-03 9.31730628e-01 7.81367779e-01
-2.21443594e-01 7.72818267e-01 6.24857306e-01 -1.33891714e+00
-1.11334777e+00 -9.81606066e-01 -4.36695606e-01 9.50547159e-01
-3.68132293e-01 -4.18063611e-01 -5.66881180e-01 -2.87740171e-01
3.19593027e-03 5.38331151e-01 -1.04142225e+00 -7.51078352e-02
-1.01203203e+00 -1.35725784e+00 6.71353161e-01 7.29797184e-01
4.85433877e-01 -1.10294640e+00 -5.31638801e-01 5.43158054e-01
-2.35135138e-01 -1.04205334e+00 -2.71354645e-01 -8.23327899e-02
-4.92112994e-01 -1.06658041e+00 -2.81986326e-01 -2.54032850e-01
2.95228571e-01 2.48431310e-01 1.12886953e+00 -1.16434330e-02
-5.41195571e-01 -2.62035038e-02 -2.57657558e-01 -4.21207070e-01
3.47359896e-01 1.89147398e-01 4.73171584e-02 2.16650322e-01
9.83752847e-01 -1.11230886e+00 -5.04645586e-01 -3.22087467e-01
-6.35032356e-01 -1.83503062e-01 -8.75835493e-02 6.77148581e-01
4.68460977e-01 3.47909629e-01 4.98070210e-01 -8.78844321e-01
9.33513105e-01 -9.98207033e-01 -4.83776122e-01 -1.00414321e-01
-9.41526741e-02 -3.31486046e-01 7.71298945e-01 -1.68109149e-01
-9.52378750e-01 1.67858183e-01 -2.00342700e-01 -1.36520296e-01
-6.10105813e-01 6.71414495e-01 5.16533792e-01 6.72774196e-01
7.68880486e-01 1.44401431e-01 -1.03141022e+00 -4.43821579e-01
-2.63085783e-01 3.17943484e-01 4.05003101e-01 -5.44964075e-01
5.72949111e-01 6.48185790e-01 1.61452405e-02 -9.29007471e-01
-9.73444581e-01 -5.16872287e-01 -8.75084519e-01 -7.06939399e-02
1.29313922e+00 -9.04171765e-01 -6.90277278e-01 7.07032561e-01
-1.52838683e+00 -6.50464773e-01 6.71700835e-02 4.75944757e-01
-9.53850523e-02 -9.28262547e-02 -8.94591272e-01 -9.46755171e-01
-1.96391970e-01 -6.77200258e-01 1.22812510e+00 2.70863362e-02
-3.54321867e-01 -1.07119584e+00 6.49797440e-01 3.89905125e-01
2.82634288e-01 5.09123862e-01 7.35071599e-01 -2.47282252e-01
-1.52203813e-01 -2.01693565e-01 -3.67743611e-01 -3.23431820e-01
1.59262955e-01 1.22068087e-02 -9.91204739e-01 5.17712682e-02
-3.53195727e-01 -1.12141028e-01 1.37921131e+00 8.54447961e-01
1.70413327e+00 -2.80917674e-01 -2.31081158e-01 9.27510679e-01
1.10822308e+00 1.22767329e-01 6.67756081e-01 3.80575180e-01
7.57948518e-01 6.05923772e-01 1.61854506e-01 8.18944633e-01
6.89551532e-01 4.38581735e-01 -4.15774658e-02 -2.83193588e-01
4.91332293e-01 -4.21152592e-01 1.03884816e-01 2.28580907e-01
-4.19408888e-01 6.36788234e-02 -1.58402789e+00 1.06860876e+00
-2.15495038e+00 -1.78269839e+00 -3.84653866e-01 1.74715495e+00
4.71050888e-01 6.44684434e-02 1.37880474e-01 1.23769499e-01
5.08485317e-01 4.32853907e-01 -3.80190164e-01 -4.96115357e-01
-2.97588050e-01 4.73145604e-01 5.55157006e-01 4.97102946e-01
-1.51298273e+00 1.04701209e+00 6.47357321e+00 8.46370935e-01
-1.12098229e+00 1.73598796e-01 1.17933321e+00 -5.20093918e-01
-8.38169605e-02 -2.72845358e-01 -5.75577259e-01 7.10112512e-01
1.01670337e+00 -1.22478478e-01 5.62786639e-01 6.70529485e-01
7.65293241e-01 -6.96452558e-02 -6.47924602e-01 1.13064718e+00
-1.34470999e-01 -1.69169080e+00 -5.59341908e-01 5.01487255e-02
8.98737967e-01 4.22573268e-01 2.08602697e-01 4.03915644e-01
8.01810682e-01 -1.53268552e+00 4.38519895e-01 7.21559405e-01
6.14094317e-01 -1.08755624e+00 4.81712788e-01 4.46664751e-01
-1.30148411e+00 -3.26790422e-01 -3.69320005e-01 -6.37415230e-01
5.58678389e-01 8.15112412e-01 -2.56866485e-01 -5.28524518e-02
1.23652029e+00 1.51102757e+00 -5.55054843e-01 6.27767026e-01
-2.38927528e-01 1.00338924e+00 -1.63550735e-01 2.41454497e-01
3.93134862e-01 -2.15291783e-01 7.69533515e-02 1.49517751e+00
3.47698599e-01 7.68658459e-01 7.79314265e-02 8.75958920e-01
1.25336245e-01 -2.31436864e-01 -1.05139673e+00 2.53236532e-01
2.84416914e-01 8.39582443e-01 -2.58455008e-01 3.85373719e-02
-4.51672941e-01 7.14890957e-01 8.05623829e-01 4.33144480e-01
-7.80323744e-01 -1.97902128e-01 9.37397301e-01 2.42466569e-01
-1.08356752e-01 -4.60155636e-01 -7.78263330e-01 -1.44897580e+00
-1.23247202e-03 -2.87067890e-01 4.58288193e-01 -4.54234034e-01
-1.62559056e+00 6.25161469e-01 -1.04046896e-01 -1.06835473e+00
-2.66575128e-01 -2.24739119e-01 -1.15706646e+00 9.12331104e-01
-1.29451835e+00 -1.16854370e+00 1.61805794e-01 9.97430682e-01
5.89245498e-01 -2.07820460e-01 6.42919004e-01 5.98051965e-01
-9.95819092e-01 3.53603870e-01 1.11621112e-01 6.98597193e-01
2.76225865e-01 -7.92349339e-01 9.15291607e-01 1.12922740e+00
4.18654643e-02 5.80548108e-01 2.96561450e-01 -8.03953111e-01
-8.06457698e-01 -1.52764666e+00 1.54982233e+00 -6.19271219e-01
9.91326272e-01 -5.43117940e-01 -1.05651271e+00 7.67605245e-01
-2.53440499e-01 1.27522320e-01 7.70098686e-01 6.88925087e-01
-2.98680186e-01 5.09677231e-02 -9.96857166e-01 6.22424006e-01
9.59447324e-01 -9.53720450e-01 -2.35735029e-01 3.29743475e-01
3.40971470e-01 -1.05327904e-01 -7.19576359e-01 1.75409436e-01
3.80440921e-01 -1.02825677e+00 9.69362497e-01 -1.00270700e+00
1.09033370e+00 4.16113645e-01 -9.20473561e-02 -1.32358503e+00
-8.15210044e-01 -3.12702537e-01 3.04346204e-01 1.03649569e+00
4.06644493e-01 -4.54874367e-01 8.76279652e-01 1.11604297e+00
-4.00898121e-02 -7.09288001e-01 -1.40606856e+00 -4.60648656e-01
5.27613282e-01 -1.18483615e+00 8.62799227e-01 1.28180373e+00
4.91819173e-01 3.31792116e-01 -7.15479732e-01 2.35458121e-01
5.73255241e-01 -3.76840644e-02 5.69890082e-01 -1.13862467e+00
-3.75143811e-02 -5.41196525e-01 -4.46859986e-01 -5.63314259e-01
5.42898417e-01 -6.89701974e-01 -4.03655022e-01 -1.28579736e+00
3.70386124e-01 -5.73691964e-01 -2.83169717e-01 7.22423136e-01
-3.87067646e-01 2.29270354e-01 -1.50282219e-01 -5.58402687e-02
-4.49335545e-01 5.76136470e-01 6.43740952e-01 -2.81375885e-01
-4.20116931e-01 -1.93589758e-02 -2.97206134e-01 8.31008196e-01
1.11173570e+00 -6.31870151e-01 -1.06482551e-01 -1.07821655e+00
5.46130955e-01 2.57234961e-01 4.44734097e-01 -1.00859082e+00
2.65395284e-01 -7.15393305e-01 6.55046046e-01 -6.01767778e-01
3.09163481e-01 -5.85584998e-01 -1.17852539e-01 1.58128336e-01
-3.71565163e-01 2.90781766e-01 3.75183880e-01 4.12103176e-01
-2.73854077e-01 4.51099753e-01 5.68272769e-01 9.58727449e-02
-3.50132883e-01 8.54864001e-01 -6.66040063e-01 4.47347313e-02
7.60609090e-01 -5.53486170e-03 -4.25042629e-01 -3.46477985e-01
-5.86897433e-01 2.51870900e-01 -6.23992868e-02 4.36872452e-01
7.29828894e-01 -1.34132349e+00 -1.08585382e+00 2.97796994e-01
-3.05715770e-01 -6.85601234e-02 8.03254187e-01 6.96451545e-01
-1.83225930e-01 5.19778669e-01 -2.13323355e-01 -2.97767997e-01
-8.80699337e-01 3.77566844e-01 4.26226780e-02 -6.23553813e-01
-6.92878008e-01 6.80341184e-01 3.74683887e-02 -4.47203100e-01
-3.52667570e-01 -1.54989764e-01 -5.35145879e-01 -6.76188245e-02
8.76709163e-01 5.93048453e-01 -9.90049616e-02 -8.55772913e-01
-3.14400434e-01 5.41576385e-01 2.05474809e-01 -3.80052440e-02
2.45423198e+00 1.24961980e-01 -9.95144397e-02 3.17201287e-01
1.19488776e+00 -1.36946067e-01 -1.15401924e+00 -1.24963753e-01
4.63341288e-02 -6.70961022e-01 1.05224818e-01 -3.07381958e-01
-1.17138553e+00 1.21133137e+00 2.38555118e-01 2.21438557e-01
1.13479710e+00 -1.29413724e-01 1.00666714e+00 2.01916873e-01
3.19047123e-01 -1.32835829e+00 -2.22996578e-01 1.00687242e+00
8.17053735e-01 -1.44736052e+00 -2.73244083e-02 3.10883410e-02
-5.07766962e-01 8.77373457e-01 5.03858268e-01 -6.91886723e-01
1.12115777e+00 3.24510932e-01 -4.28980231e-01 -3.62819761e-01
-7.94936478e-01 -1.66968610e-02 1.74210116e-01 4.42842275e-01
5.63009262e-01 2.81744450e-01 -5.98714761e-02 1.07910419e+00
-2.63092369e-01 -8.09318870e-02 4.09379303e-01 5.11974871e-01
-1.21494450e-01 -6.84519410e-01 -3.72425854e-01 7.96671748e-01
-8.88351023e-01 -2.99771875e-01 -1.09212101e-01 3.41604888e-01
2.06688181e-01 1.23145199e+00 4.17684674e-01 -5.47351062e-01
3.35828036e-01 -3.65647137e-01 -1.40898854e-01 -5.61504960e-01
-7.61353552e-01 -1.54497907e-01 -1.39669240e-01 -8.74774754e-01
-4.84700948e-02 -1.19447780e+00 -1.05525935e+00 -7.53793180e-01
4.87116456e-01 -1.88753903e-01 3.41224641e-01 1.11143947e+00
3.63539577e-01 4.57303345e-01 7.65802503e-01 -9.12465453e-01
2.18225718e-01 -9.94118690e-01 -8.05828869e-01 7.32329249e-01
6.45663977e-01 -5.20848930e-01 -1.51253000e-01 2.36030575e-02] | [6.741401672363281, 1.9824378490447998] |
a8c47f6f-4a55-4a11-9466-83c164b35e28 | a-pre-training-oracle-for-predicting | 2106.03233 | null | https://arxiv.org/abs/2106.03233v1 | https://arxiv.org/pdf/2106.03233v1.pdf | A Pre-training Oracle for Predicting Distances in Social Networks | In this paper, we propose a novel method to make distance predictions in real-world social networks. As predicting missing distances is a difficult problem, we take a two-stage approach. Structural parameters for families of synthetic networks are first estimated from a small set of measurements of a real-world network and these synthetic networks are then used to pre-train the predictive neural networks. Since our model first searches for the most suitable synthetic graph parameters which can be used as an "oracle" to create arbitrarily large training data sets, we call our approach "Oracle Search Pre-training" (OSP). For example, many real-world networks exhibit a Power law structure in their node degree distribution, so a Power law model can provide a foundation for the desired oracle to generate synthetic pre-training networks, if the appropriate Power law graph parameters can be estimated. Accordingly, we conduct experiments on real-world Facebook, Email, and Train Bombing networks and show that OSP outperforms models without pre-training, models pre-trained with inaccurate parameters, and other distance prediction schemes such as Low-rank Matrix Completion. In particular, we achieve a prediction error of less than one hop with only 1% of sampled distances from the social network. OSP can be easily extended to other domains such as random networks by choosing an appropriate model to generate synthetic training data, and therefore promises to impact many different network learning problems. | ['Rasika Karkare', 'Anura Jayasumana', 'Randy Paffenroth', 'Gunjan Mahindre'] | 2021-06-06 | null | null | null | null | ['low-rank-matrix-completion'] | ['methodology'] | [ 3.08344334e-01 6.89221501e-01 -3.68643492e-01 -4.27409947e-01
-4.92417961e-01 -5.61075151e-01 3.90879571e-01 1.23575449e-01
-2.83107877e-01 8.61384571e-01 -2.11049885e-01 -6.08937025e-01
-5.58018565e-01 -1.21769440e+00 -1.03823626e+00 -5.42231739e-01
-5.15135109e-01 1.07015073e+00 3.18114579e-01 -4.10200238e-01
9.69245061e-02 6.65077507e-01 -8.48047316e-01 -1.87801763e-01
1.15529478e+00 5.31790257e-01 -6.91578165e-03 7.04412520e-01
4.21356171e-01 5.38350999e-01 -6.08723283e-01 -6.47993982e-01
4.50655252e-01 -2.35491112e-01 -8.33732605e-01 -1.17363110e-01
3.76104377e-02 -3.87886286e-01 -8.90137017e-01 1.12115538e+00
4.05729532e-01 9.03356001e-02 6.40726328e-01 -1.39854693e+00
-4.40138251e-01 1.24535263e+00 -2.85787791e-01 -1.03875443e-01
4.10169393e-01 -6.21303543e-02 1.03122139e+00 -5.79306483e-01
5.50918639e-01 1.14564705e+00 1.02705550e+00 4.47772056e-01
-1.44018233e+00 -7.65440822e-01 -2.55784601e-01 1.83661193e-01
-1.60783958e+00 -2.53332824e-01 9.57977057e-01 -1.26305535e-01
1.90665543e-01 2.35237360e-01 5.55386722e-01 1.45008385e+00
-2.73421794e-01 4.10080075e-01 5.90420425e-01 -1.07466578e-01
2.54875630e-01 1.99868068e-01 -1.93461955e-01 7.95237720e-01
4.50539768e-01 1.24878295e-01 6.84990138e-02 -5.05193651e-01
8.01592529e-01 9.00344774e-02 -2.25985661e-01 -4.60954666e-01
-1.34709513e+00 9.73655105e-01 8.68659616e-01 -1.04526160e-02
-2.34373704e-01 1.52819321e-01 1.84397083e-02 6.21314645e-01
3.58950615e-01 5.94133377e-01 -2.40946725e-01 1.35638207e-01
-7.20881522e-01 -5.86528853e-02 1.21188498e+00 7.41809189e-01
9.53515291e-01 -1.30725935e-01 2.98602611e-01 9.03601408e-01
1.58291698e-01 4.37777787e-01 1.16357498e-01 -1.07570183e+00
6.97688878e-01 5.74831605e-01 5.40723577e-02 -1.44207156e+00
-5.11151135e-01 -6.18853450e-01 -1.14919472e+00 -4.11736041e-01
7.14905679e-01 -5.64544499e-01 -4.35832560e-01 1.67233276e+00
4.58436847e-01 5.39882123e-01 -9.08296730e-04 5.69124043e-01
3.37481886e-01 6.72114670e-01 -6.53182149e-01 5.81543259e-02
4.69306111e-01 -8.68779600e-01 1.08367853e-01 -2.59107500e-01
1.02541184e+00 -7.88960084e-02 9.08621371e-01 3.07617515e-01
-7.57170439e-01 -2.85415113e-01 -1.14488387e+00 5.62807977e-01
-2.45123729e-01 -1.16127744e-01 6.90638602e-01 7.14141726e-01
-1.12524962e+00 1.22269893e+00 -6.54587686e-01 -3.43821883e-01
3.07677358e-01 5.02861619e-01 -3.29786628e-01 -3.73867005e-01
-1.36963391e+00 3.52016091e-01 5.13963342e-01 1.45497218e-01
-1.07157314e+00 -5.69129944e-01 -4.84949082e-01 -1.39169708e-01
5.86043954e-01 -4.95452225e-01 6.90395832e-01 -6.73707545e-01
-1.25828600e+00 3.72357786e-01 3.60060453e-01 -5.76533973e-01
5.30064762e-01 3.73395264e-01 -5.24742007e-01 2.56176531e-01
-4.30485681e-02 4.03012812e-01 8.46310019e-01 -1.15329278e+00
-7.61754215e-02 -1.60542369e-01 4.36733782e-01 -2.06772014e-01
-5.66709578e-01 -3.27299386e-01 -2.40981519e-01 -3.90536606e-01
2.41200790e-01 -1.40226078e+00 -7.07830846e-01 7.94155076e-02
-8.26812983e-01 4.21721078e-02 6.02357566e-01 -3.44434857e-01
1.04738808e+00 -1.83426547e+00 1.61008939e-01 1.01142490e+00
5.91465831e-01 1.20555945e-01 -4.78538662e-01 5.54081261e-01
-1.03109397e-01 3.56548429e-01 -7.70298541e-02 -6.95052072e-02
-2.16177896e-01 3.60899687e-01 -1.42629957e-02 4.39407974e-01
-1.41225129e-01 5.36235392e-01 -1.18999004e+00 -2.92619228e-01
-2.24764735e-01 3.28436315e-01 -6.40881836e-01 3.02886087e-02
-1.61583751e-01 4.94603395e-01 -5.03895581e-01 2.52536803e-01
6.20313048e-01 -7.72721231e-01 4.52558100e-01 1.64339673e-02
7.18773901e-01 1.34340137e-01 -1.16144967e+00 1.10207701e+00
-2.50761092e-01 4.89590347e-01 1.38024297e-02 -1.28551161e+00
1.21742249e+00 -1.17192436e-02 5.95575511e-01 -1.77725986e-01
1.28320798e-01 1.98503643e-01 2.56685048e-01 -2.01254711e-01
1.04936861e-01 3.67149115e-01 4.80657220e-02 5.49420357e-01
-2.27069363e-01 1.57701895e-02 3.53231728e-01 5.95213354e-01
1.52697277e+00 -6.80436373e-01 -2.82880843e-01 1.04708523e-01
4.99659419e-01 -2.12909728e-01 3.47227573e-01 8.83194327e-01
1.28535241e-01 5.11105061e-01 9.78531063e-01 -3.71662557e-01
-1.19331408e+00 -1.07905889e+00 5.92980087e-02 8.26023817e-01
2.13716514e-02 -4.06839132e-01 -8.61734331e-01 -1.10704255e+00
2.50762925e-02 5.05202174e-01 -6.34911478e-01 -3.55417877e-01
-3.99442345e-01 -7.71510959e-01 8.63966405e-01 1.05566584e-01
3.79834473e-01 -6.51298702e-01 6.89180613e-01 1.43162817e-01
-1.20875783e-01 -1.32977271e+00 -4.57018912e-01 -1.75453424e-01
-6.90980434e-01 -1.32253444e+00 -3.71154636e-01 -6.20101571e-01
1.16935110e+00 4.57682431e-01 8.01502466e-01 4.47109193e-01
2.55671620e-01 9.09306016e-03 -2.86283016e-01 1.11723848e-01
-7.68725634e-01 5.85756838e-01 3.33265394e-01 2.50106364e-01
-2.49688178e-02 -1.18442512e+00 -5.75298488e-01 7.06808746e-01
-7.48896897e-01 -1.69711083e-01 7.59685218e-01 7.92967975e-01
3.42701584e-01 2.75827914e-01 7.37654209e-01 -1.28557992e+00
9.22172546e-01 -9.19766545e-01 -6.87659621e-01 3.11475009e-01
-7.19247937e-01 1.00098789e-01 1.12548721e+00 -9.49441373e-01
-3.04531246e-01 -2.83786599e-02 -5.45135513e-02 -4.43406880e-01
2.30689868e-01 7.11595356e-01 -5.76891378e-02 -5.15603721e-01
1.20354223e+00 1.13810990e-02 2.10204482e-01 -2.76700795e-01
3.27746570e-01 5.52629411e-01 2.27566808e-01 -5.95026016e-01
1.48739970e+00 2.05885395e-01 4.00379986e-01 -8.01236928e-01
-9.21091735e-01 -6.75743297e-02 -7.52903640e-01 -3.92136835e-02
7.00996742e-02 -6.34024858e-01 -7.43916154e-01 2.72832334e-01
-8.93117547e-01 -4.74193633e-01 -1.38447106e-01 5.18586099e-01
-2.29797304e-01 4.97922957e-01 -6.60025835e-01 -3.94128710e-01
-2.06806526e-01 -7.99892366e-01 4.99955535e-01 3.71324420e-02
-2.42763814e-02 -1.48833156e+00 -3.94988153e-03 3.96580338e-01
4.08424735e-01 3.99267942e-01 8.91343057e-01 -1.05718327e+00
-5.92260838e-01 -5.89019537e-01 -3.49980265e-01 3.79086316e-01
-5.34937792e-02 1.39632091e-01 -3.98715466e-01 -5.09176672e-01
-4.84203994e-01 -3.38016033e-01 3.32058549e-01 -3.63760442e-02
1.53229582e+00 -6.98962331e-01 -6.27350867e-01 8.17047238e-01
1.05696297e+00 -3.08713228e-01 5.04062235e-01 -1.55670434e-01
8.44577968e-01 3.89667183e-01 4.25012171e-01 4.13707197e-01
5.63175201e-01 3.62193823e-01 4.73454714e-01 3.26056778e-02
3.17167401e-01 -9.03171659e-01 1.41937330e-01 9.87360060e-01
9.18746926e-03 -2.86927670e-01 -1.08263171e+00 2.62506843e-01
-1.75693500e+00 -9.07652617e-01 -4.25516963e-02 2.34822440e+00
7.99774826e-01 4.93273258e-01 3.27535510e-01 1.83379740e-01
8.43227804e-01 1.08049899e-01 -8.42919767e-01 1.13007151e-01
3.46828625e-02 -4.12410386e-02 8.89817715e-01 5.83644986e-01
-7.51216888e-01 7.03164101e-01 6.27453136e+00 8.61437917e-01
-1.04648423e+00 -2.16086656e-01 7.35312283e-01 2.09917709e-01
-5.53108454e-01 3.74420673e-01 -6.48079336e-01 5.44920564e-01
1.33145392e+00 -3.38871211e-01 9.09868896e-01 9.19432938e-01
1.43442050e-01 4.44973767e-01 -1.15171409e+00 8.36739361e-01
-2.11605787e-01 -1.40874159e+00 -3.74468900e-02 3.92919958e-01
8.94027650e-01 3.57277811e-01 -1.58329949e-01 3.69548589e-01
7.28055239e-01 -1.11504281e+00 -1.09798402e-01 5.20518482e-01
7.61829317e-01 -8.14682305e-01 5.40404975e-01 9.49797273e-01
-9.00585115e-01 -1.73439354e-01 -8.08441520e-01 2.09409252e-01
-1.40475720e-01 9.42194521e-01 -1.48954713e+00 4.09601450e-01
2.19412103e-01 7.55797207e-01 -8.11086774e-01 1.16624606e+00
-9.62269455e-02 8.62455308e-01 -8.02392781e-01 -3.60918522e-01
1.17605381e-01 -4.28184867e-01 4.19080675e-01 4.90085870e-01
4.41949785e-01 -2.85367817e-01 2.42583081e-01 6.22127116e-01
-5.57088554e-01 7.03898966e-02 -7.92907059e-01 -2.51571596e-01
9.59768832e-01 1.22618496e+00 -5.24106324e-01 -7.54593983e-02
-1.36081055e-01 7.11247146e-01 7.14039564e-01 4.91532087e-01
-1.01954162e+00 -7.89817512e-01 2.77635813e-01 5.63352227e-01
3.98316607e-02 -3.58840585e-01 2.08211854e-01 -1.08209300e+00
-1.26738638e-01 -9.14736688e-01 1.35476351e-01 -5.29976666e-01
-1.39687061e+00 6.43545032e-01 -1.47672310e-01 -1.23464394e+00
-4.76828694e-01 -3.78655434e-01 -8.20789397e-01 4.19470787e-01
-1.08302450e+00 -7.48969674e-01 -3.88897717e-01 6.66475892e-01
-2.68532544e-01 -1.75695583e-01 6.61519647e-01 4.02048230e-01
-7.24027812e-01 8.83504450e-01 4.76325542e-01 4.22720164e-01
4.79630113e-01 -1.00341630e+00 5.29266655e-01 5.78588068e-01
3.22980404e-01 3.53897780e-01 5.64363420e-01 -6.76356435e-01
-1.42610323e+00 -1.25812697e+00 4.22209412e-01 -3.59709829e-01
1.11832392e+00 -4.54960316e-01 -8.47533107e-01 7.66093850e-01
-6.07994735e-01 4.31643039e-01 5.01260459e-01 2.72378772e-01
-3.28191191e-01 -4.52383757e-01 -1.29593098e+00 7.46227086e-01
1.45689750e+00 -4.59194183e-01 2.34904006e-01 9.95687902e-01
9.65862870e-01 -1.87778428e-01 -1.21085286e+00 3.56906772e-01
2.46462569e-01 -7.11679399e-01 1.14512980e+00 -7.36708224e-01
1.10739149e-01 1.11046657e-01 1.36687592e-01 -1.67034602e+00
-1.80811927e-01 -1.03574574e+00 -3.39831412e-01 9.33101892e-01
1.00437927e+00 -9.75614607e-01 1.39920890e+00 5.26016533e-01
5.42613804e-01 -9.13636804e-01 -6.03454232e-01 -7.31406927e-01
-8.90029818e-02 -3.06534797e-01 7.88790822e-01 1.15253544e+00
-1.24421470e-01 5.13389289e-01 -6.31302118e-01 3.36097330e-01
9.41494465e-01 -3.82984102e-01 1.17355740e+00 -1.70177162e+00
-5.01431465e-01 -2.64336448e-02 -3.55387568e-01 -1.46446693e+00
3.72031003e-01 -1.09399366e+00 -4.83665228e-01 -1.20140302e+00
-2.07412928e-01 -1.19620109e+00 1.76888570e-01 2.10455090e-01
1.79443747e-01 1.43881980e-02 -3.06294739e-01 1.23307958e-01
-6.17821574e-01 4.70078766e-01 1.42366970e+00 -1.19347624e-01
-1.69529170e-01 7.51034975e-01 -7.47962534e-01 5.87676466e-01
9.50390577e-01 -7.08217621e-01 -8.77286971e-01 -9.89952609e-02
6.35413527e-01 5.12763739e-01 2.13369891e-01 -1.11463320e+00
4.91601586e-01 -2.61925519e-01 2.34871268e-01 -3.47668678e-01
4.10964489e-01 -8.88450146e-01 2.31470719e-01 3.54007483e-01
-3.21550161e-01 -7.69583657e-02 -5.74565828e-01 8.99976432e-01
2.16564730e-01 -3.51804346e-01 6.08154893e-01 1.42781779e-01
2.26557557e-03 7.91878700e-01 1.04638331e-01 3.37572783e-01
9.73296881e-01 -2.20403016e-01 -5.12058794e-01 -9.78624463e-01
-7.97982991e-01 4.34612274e-01 4.58736658e-01 2.01678410e-01
7.65460551e-01 -1.29202104e+00 -6.97070718e-01 2.33232930e-01
-1.17939256e-01 5.49891181e-02 -7.30070099e-02 8.06976020e-01
-6.16795361e-01 1.45435622e-02 2.03752369e-01 -4.98737544e-01
-7.69660354e-01 6.36887729e-01 3.98427457e-01 -2.97167420e-01
-2.87731498e-01 5.86659729e-01 -6.42238021e-01 -9.56099868e-01
2.96416700e-01 6.90366626e-02 1.21431969e-01 -3.90470028e-01
2.93854952e-01 3.26352090e-01 -2.62410492e-01 -4.94850129e-01
6.76321834e-02 2.20842183e-01 -5.58852740e-02 3.19755733e-01
1.44383156e+00 -4.89087496e-03 -3.54083255e-02 1.29139811e-01
1.58789039e+00 1.51870981e-01 -1.04614604e+00 -6.19089603e-01
6.91906214e-02 -4.88291472e-01 -3.54642600e-01 -2.00989425e-01
-1.27049148e+00 4.89441693e-01 -2.91569112e-03 7.51074195e-01
6.84923291e-01 1.43828392e-01 7.17889786e-01 1.10229552e+00
6.40575886e-01 -9.06308770e-01 3.28078449e-01 4.96701449e-01
5.93907237e-01 -1.19029117e+00 -1.24093838e-01 -7.09134221e-01
-3.06720465e-01 1.02775705e+00 6.62518799e-01 -3.59679550e-01
1.16531384e+00 3.57969810e-04 -5.13982713e-01 -4.61772010e-02
-7.09490657e-01 2.45884150e-01 1.48170903e-01 7.28700876e-01
-1.23517506e-01 4.27700616e-02 2.09178582e-01 3.58137310e-01
-6.52145684e-01 -3.63673031e-01 6.99687958e-01 1.72092050e-01
-5.40792167e-01 -1.33407187e+00 1.09937135e-02 8.91093135e-01
-7.22490475e-02 1.43640533e-01 -5.40984154e-01 5.78021228e-01
-4.53037858e-01 8.67769122e-01 -3.22973609e-01 -9.99597967e-01
2.09309265e-01 -4.24674809e-01 2.05834404e-01 -4.75381017e-01
-1.04362011e-01 -8.11881006e-01 4.93907005e-01 -4.60090488e-01
2.01548059e-02 -1.80870593e-01 -9.55617428e-01 -9.95728433e-01
-3.87383997e-01 5.19942939e-01 5.74606776e-01 7.31039107e-01
3.90961647e-01 2.25033406e-02 1.36333835e+00 -7.54299521e-01
-9.18172598e-01 -9.82880414e-01 -4.84941036e-01 3.65524739e-01
7.92798921e-02 -4.26663548e-01 -8.42550933e-01 -7.41452157e-01] | [6.8495259284973145, 5.725025177001953] |
ecf3c231-f561-42cd-9a08-8acdd9c1a598 | ai-enhanced-3d-rf-representation-using-low | null | null | https://doi.org/10.1145/3274783.3275210 | https://www.semanticscholar.org/paper/AI-Enhanced-3D-RF-Representation-Using-Low-Cost-Fang-Nirjon/36075af2129b590dfa128f5de5a159395952549e | AI-Enhanced 3D RF Representation Using Low-Cost mmWave Radar | This paper introduces a system that takes radio frequency (RF) signals from an off-the-shelf, low-cost, 77 GHz mm Wave radar and produces an enhanced 3D RF representation of a scene. Such a system can be used in scenarios where camera and other types of sensors do not work, or their performance is impacted due to bad lighting conditions and occlusions, or an alternate RF sensing system like synthetic aperture radar (SAR) is too large, inconvenient, and costly. The enhanced RF representation can be used in numerous applications such as robot navigation, human-computer interaction, and patient monitoring. We use off-the-shelf parts to capture RF signals and collect our own data set for training and testing of the approach. The novelty of the system lies in its use of AI to generate a fine-grained 3D representation of an RF scene from its sparse RF representation which a mm Wave radar of the same class cannot achieve. | ['Shiwei Fang', 'Shahriar Nirjon'] | 2018-11-04 | null | null | null | sensys-18-proceedings-of-the-16th-acm | ['rf-based-pose-estimation'] | ['computer-vision'] | [ 6.42049134e-01 2.46747985e-01 3.98355544e-01 -4.60563451e-01
-4.73294914e-01 -5.79914331e-01 3.28920543e-01 -4.68743831e-01
-3.06933701e-01 8.90865803e-01 5.17934598e-02 -4.11259741e-01
-3.85236025e-01 -9.88405824e-01 -9.99276713e-02 -7.69675910e-01
3.69595550e-02 6.59071267e-01 7.88418110e-03 -1.29369333e-01
-1.52899437e-02 9.24317658e-01 -1.43962550e+00 9.51794311e-02
4.19924408e-01 1.01221919e+00 6.57444477e-01 6.71466589e-01
6.82359338e-02 5.68072677e-01 -8.62029552e-01 5.83634853e-01
6.30193889e-01 -2.42912129e-01 -1.00190379e-01 -3.69692683e-01
2.08333820e-01 -1.52794227e-01 -2.46602505e-01 7.34285414e-01
6.21998310e-01 -1.76308036e-01 7.25040197e-01 -6.20384872e-01
9.91830379e-02 1.31536812e-01 -5.72271407e-01 -2.53331009e-02
6.43601775e-01 -1.16893306e-01 -8.51612464e-02 -4.07958597e-01
6.46725833e-01 1.00512362e+00 5.95478594e-01 3.27312499e-01
-9.96161282e-01 -8.88133764e-01 -3.83269131e-01 -3.16428393e-01
-1.31829464e+00 -4.57008153e-01 8.21677804e-01 -1.51527092e-01
6.22038782e-01 3.78830820e-01 5.84174514e-01 1.16194272e+00
8.40043008e-01 -2.21491575e-01 1.31286848e+00 -4.69882786e-01
4.48151618e-01 2.42058352e-01 -2.27797359e-01 6.48334146e-01
5.26168048e-01 5.41846812e-01 -1.61152750e-01 -3.84785026e-01
8.43116879e-01 2.98957288e-01 -5.54991782e-01 -4.55126762e-01
-1.16845381e+00 8.49114954e-01 4.47380066e-01 9.58839238e-01
-7.28028595e-01 1.22237824e-01 -6.77763939e-01 4.89074379e-01
1.22255318e-01 1.04115188e+00 -3.46127301e-01 -2.13674493e-02
-9.82963741e-01 -2.34484002e-02 8.30263138e-01 7.07831025e-01
7.05987871e-01 2.90175378e-01 2.19656125e-01 5.26361406e-01
4.49919850e-01 1.20228970e+00 4.59938318e-01 -4.50009942e-01
-6.43443540e-02 2.34018788e-01 4.49142456e-02 -7.27110922e-01
-9.90079284e-01 -7.41317809e-01 -8.16865206e-01 4.04718429e-01
-1.35525718e-01 -6.18061543e-01 -1.43138301e+00 1.26740992e+00
5.67486975e-03 2.08112285e-01 3.53859782e-01 9.73648548e-01
1.03016889e+00 4.09519702e-01 -3.61520171e-01 -1.75705314e-01
1.39375174e+00 -1.56940892e-01 -4.89567965e-01 -8.10930610e-01
3.43389034e-01 -9.38269436e-01 2.70595521e-01 3.73534769e-01
-6.19016349e-01 -6.07253671e-01 -1.52081311e+00 7.71011114e-01
-1.08485542e-01 -1.03811547e-01 8.31426144e-01 1.05281925e+00
-6.86365008e-01 8.11121836e-02 -5.28900445e-01 -3.89410079e-01
1.60104454e-01 4.87805367e-01 -2.41723374e-01 -4.86957282e-01
-9.58010972e-01 1.18349099e+00 -2.73158669e-01 1.16779752e-01
-4.65627313e-01 -8.00278008e-01 -7.85207510e-01 -1.51077569e-01
-1.01332895e-01 -7.46984303e-01 8.66547287e-01 -5.99196672e-01
-1.47770858e+00 3.75138581e-01 2.93079644e-01 -2.05025703e-01
7.28342449e-03 -2.70549208e-02 -9.82401967e-01 4.10428852e-01
2.08858978e-02 1.33408487e-01 1.03896856e+00 -9.29470718e-01
-2.05879748e-01 -6.27755582e-01 -8.11861828e-02 -9.63659361e-02
5.70078552e-01 -3.67493391e-01 4.81382370e-01 -3.94075304e-01
6.28740311e-01 -1.04080582e+00 -6.34871960e-01 -4.68252987e-01
-2.91540306e-02 8.77116561e-01 1.16103780e+00 6.60117529e-03
4.83818501e-01 -2.09388614e+00 -3.79706591e-01 6.11974418e-01
-1.96426108e-01 7.71093890e-02 -1.25113606e-01 4.39531863e-01
-1.29697904e-01 -3.20590824e-01 -2.98208505e-01 5.29223680e-01
-4.86269265e-01 -2.15564221e-01 -1.59482792e-01 6.53076410e-01
-8.96575004e-02 4.20385450e-01 -5.69227695e-01 -2.67901272e-03
3.52392793e-01 6.36373937e-01 -1.03430979e-01 5.24071418e-02
1.39369428e-01 7.50497401e-01 -9.19300079e-01 6.45472646e-01
9.41312015e-01 2.37763122e-01 1.77136153e-01 -4.77519542e-01
-5.98854609e-02 -3.85162756e-02 -1.37399840e+00 1.70499563e+00
-8.31446111e-01 5.60009122e-01 1.52709216e-01 -9.06408906e-01
1.45092392e+00 3.98237467e-01 5.08283496e-01 -1.05516088e+00
2.32171759e-01 2.70861804e-01 -3.03452499e-02 -6.07699156e-01
2.78032809e-01 -7.16178000e-01 -3.08396995e-01 6.81847155e-01
-1.98555455e-01 -7.18766272e-01 -4.32505250e-01 -2.48160213e-01
1.76873553e+00 1.32074699e-01 4.00582343e-01 -3.82498085e-01
3.74573946e-01 2.41430923e-01 9.19214785e-02 8.74319434e-01
3.96656722e-01 6.46600842e-01 -5.20337939e-01 -5.10040343e-01
-4.54161406e-01 -1.09770703e+00 -4.22471583e-01 1.96668908e-01
2.35486671e-01 3.30363840e-01 -5.68148941e-02 -3.60788614e-01
-7.31662661e-02 6.35138810e-01 -2.03223079e-01 -2.42907465e-01
-6.73493922e-01 -7.87900448e-01 3.05944115e-01 1.57703981e-01
5.25005341e-01 -8.73373032e-01 -1.40969551e+00 3.65487248e-01
2.78724283e-02 -1.06740499e+00 3.40175211e-01 3.87569845e-01
-9.12282705e-01 -9.71305013e-01 -5.51127315e-01 -6.57425463e-01
7.66576767e-01 7.05807388e-01 1.04713523e+00 -2.33767629e-01
-8.40654850e-01 4.98020858e-01 -5.02957523e-01 -6.86980426e-01
-1.10469781e-01 -3.22714150e-01 8.08510855e-02 -2.35856295e-01
4.18378770e-01 -7.98188210e-01 -6.62831247e-01 3.56134832e-01
-7.04759657e-01 -4.09699291e-01 1.18026650e+00 4.59229767e-01
2.86730766e-01 1.83529019e-01 9.00741637e-01 -1.06237555e+00
3.32906216e-01 -2.62767911e-01 -5.54442406e-01 -2.72669286e-01
-2.15130880e-01 2.34545007e-01 4.80786562e-01 -3.49582464e-01
-9.81045842e-01 4.53462660e-01 -5.45685329e-02 -1.29806213e-02
-3.04487616e-01 4.29456770e-01 -1.33728474e-01 -4.11190063e-01
1.04538798e+00 -1.42514542e-01 -5.00100628e-02 -4.03245687e-01
1.00548975e-01 1.00175190e+00 4.84403610e-01 1.48899317e-01
1.18412912e+00 7.05246925e-01 3.74231696e-01 -1.17486978e+00
-6.42344296e-01 -5.22317350e-01 -3.47034246e-01 -2.90739369e-02
6.47370160e-01 -9.51568782e-01 -2.02688694e-01 -1.88845426e-01
-9.16948438e-01 -9.95438695e-02 -1.91690400e-01 1.17333663e+00
-5.71733296e-01 -8.75201672e-02 1.85037181e-01 -8.52408111e-01
-3.45559537e-01 -7.21196771e-01 9.55952048e-01 4.18049932e-01
-3.38587284e-01 -3.93917829e-01 3.46966237e-01 2.64604747e-01
8.24741960e-01 3.90829384e-01 7.11332560e-01 -1.32462054e-01
-5.82577705e-01 -4.29815650e-01 5.19443862e-02 -2.74365157e-01
2.50479639e-01 -8.21338236e-01 -1.13651872e+00 -4.03345495e-01
3.49033862e-01 -7.63429701e-02 6.68365300e-01 6.52755141e-01
5.74065745e-01 1.99457556e-01 -8.13294351e-01 7.90726900e-01
1.69534731e+00 7.25823939e-01 8.96561563e-01 2.34491453e-01
1.26206493e-02 5.42098582e-01 1.05351019e+00 1.96456790e-01
-3.73576492e-01 6.85887456e-01 3.33540052e-01 -2.49337226e-01
-2.67165489e-02 3.08577299e-01 1.42437220e-01 -9.99480113e-02
5.73339574e-02 -8.83625001e-02 -6.23359799e-01 1.53937921e-01
-1.24824512e+00 -1.13429391e+00 1.12039298e-01 2.14211178e+00
1.05897998e-02 -1.22046404e-01 -5.22376239e-01 1.21781573e-01
3.33626598e-01 -4.91937809e-02 -2.78138429e-01 -3.65969658e-01
1.73561141e-01 9.16145742e-01 7.24705040e-01 3.72446805e-01
-8.16957295e-01 3.09440702e-01 5.93698597e+00 1.34246334e-01
-1.40524745e+00 -1.57613412e-01 1.59862135e-02 5.74346399e-03
-3.04784268e-01 -1.49251223e-01 -3.87307942e-01 1.59388483e-02
9.73585188e-01 2.75354654e-01 1.09158948e-01 6.58045113e-01
3.78914014e-03 -5.76389730e-01 -9.36526120e-01 1.27117789e+00
2.74167866e-01 -1.20106375e+00 -5.89640498e-01 3.50791097e-01
7.25840926e-02 6.25631064e-02 -2.11954743e-01 1.91209644e-01
-9.91590880e-03 -1.22887802e+00 1.95765942e-01 5.99321306e-01
8.87650669e-01 -8.45937610e-01 1.08232820e+00 4.33348298e-01
-9.75467086e-01 -7.53194168e-02 -4.73710924e-01 -2.93495823e-02
2.90201247e-01 9.42521811e-01 -1.18222511e+00 5.89892149e-01
5.36253035e-01 -3.25901568e-01 -1.72959939e-01 9.12972748e-01
7.52014220e-02 1.30796030e-01 -5.65000117e-01 -1.75176352e-01
-4.34195448e-04 -2.23687753e-01 6.60699189e-01 9.64016140e-01
9.85517144e-01 4.15608943e-01 1.01680480e-01 4.48028326e-01
4.34853047e-01 -4.19639796e-01 -1.20859313e+00 1.56337008e-01
4.25263464e-01 1.58909523e+00 -7.40224421e-01 1.54808298e-01
-2.61113226e-01 5.61761439e-01 -5.67316949e-01 2.38982171e-01
-6.13365650e-01 -8.29896152e-01 4.88120049e-01 4.96206045e-01
5.51711977e-01 -3.59024614e-01 -5.37822507e-02 -7.18709707e-01
-2.67203599e-01 -5.78993976e-01 5.32972021e-03 -1.17341137e+00
-8.36695790e-01 1.21868050e+00 9.49083921e-03 -1.58441079e+00
-8.31715584e-01 -4.38985676e-01 -3.10082734e-01 8.89015317e-01
-1.48599064e+00 -7.49693692e-01 -8.36115718e-01 4.94647324e-01
2.99819056e-02 -4.71141845e-01 1.28496242e+00 2.02860698e-01
4.69761342e-01 -6.66947365e-02 -3.02577913e-01 -2.52887547e-01
6.72488511e-01 -6.07113481e-01 -1.60751939e-01 4.49192971e-01
2.18464509e-01 2.64204443e-01 9.32482719e-01 -6.14319146e-01
-1.84622741e+00 -9.83721018e-01 5.02947986e-01 -1.33069828e-01
6.61839917e-02 -5.52230060e-01 -8.39125067e-02 2.59012401e-01
-1.27270967e-01 -1.27842464e-02 9.76533234e-01 8.97662640e-02
5.85825481e-02 -3.56962085e-01 -1.76661086e+00 2.51018733e-01
8.49238634e-01 -7.92850107e-02 -7.30622113e-01 3.27304929e-01
2.73587286e-01 -1.27078727e-01 -5.82444370e-01 7.21866846e-01
9.20605838e-01 -9.25056756e-01 1.04228568e+00 1.10664152e-01
-2.04684973e-01 -4.68513012e-01 -5.39180994e-01 -1.52186477e+00
-4.76854831e-01 -3.76378894e-01 3.78428340e-01 4.20226961e-01
5.07224977e-01 -9.32494819e-01 9.96574342e-01 -2.61842430e-01
-4.52607404e-03 -3.33599746e-01 -8.38584006e-01 -7.17679262e-01
-7.30512261e-01 -4.18812603e-01 7.22607315e-01 5.86706460e-01
-2.02200741e-01 6.77858174e-01 -4.40168679e-02 5.18179119e-01
6.12697721e-01 6.98757887e-01 8.77037466e-01 -1.74019086e+00
-5.70195436e-01 4.56996650e-01 -8.08252037e-01 -7.84892023e-01
-2.94061691e-01 -5.34056842e-01 2.57239223e-01 -1.79673779e+00
-3.85152280e-01 -9.47574854e-01 7.56478682e-02 1.05848916e-01
8.27351451e-01 6.98056221e-01 1.69820622e-01 4.28085588e-02
1.91510886e-01 1.77985653e-01 9.77747619e-01 -1.93364754e-01
-3.38019192e-01 3.34763527e-01 -8.13959539e-01 7.89632082e-01
6.48811758e-01 -5.86114705e-01 -5.14381766e-01 -9.04984325e-02
1.48866221e-01 5.19904792e-01 1.23251833e-01 -1.61015058e+00
8.07575211e-02 6.18104935e-02 9.09583509e-01 -5.74593067e-01
7.68096626e-01 -1.44077373e+00 7.05670595e-01 7.42935896e-01
6.16749346e-01 -1.02050364e-01 5.15332483e-02 3.68847281e-01
-7.68845947e-03 -4.28420514e-01 8.96344781e-01 -4.12078410e-01
-3.36785704e-01 -2.48899255e-02 -7.69431233e-01 -4.34661001e-01
1.05962312e+00 -6.80264175e-01 -4.78956938e-01 -5.45923114e-01
-4.12898749e-01 -4.13450062e-01 3.48089278e-01 1.42992556e-01
7.53888071e-01 -1.11382198e+00 -6.62538350e-01 7.98643827e-01
3.81608084e-02 -2.15101093e-01 2.49859422e-01 4.95029420e-01
-5.95958650e-01 6.69210970e-01 -4.75276798e-01 -6.09130323e-01
-1.07910931e+00 2.79702663e-01 2.29391500e-01 -2.42561385e-01
-1.01272202e+00 3.67357254e-01 2.04712316e-01 -3.77963275e-01
-3.90257597e-01 -2.14344382e-01 -3.11905593e-01 -1.40954778e-01
6.88833773e-01 -1.39102057e-01 3.82745385e-01 -3.45620394e-01
-7.85484910e-01 1.02063584e+00 4.63746786e-01 -4.84440893e-01
1.62233472e+00 1.66353300e-01 2.65115738e-01 5.84846688e-03
7.12484360e-01 5.18257558e-01 -5.46683013e-01 9.09763575e-02
-4.50476050e-01 -4.23844814e-01 3.91771197e-01 -9.35611129e-01
-9.77834582e-01 3.98789644e-01 1.00227523e+00 1.68844610e-01
1.43462813e+00 1.17983229e-01 3.78701270e-01 7.65397847e-01
1.14906383e+00 -6.78789496e-01 -2.68447787e-01 2.15516761e-01
7.83423007e-01 -5.19447267e-01 5.51403701e-01 -4.25650120e-01
-3.48686576e-01 1.14905608e+00 1.16611989e-02 -3.09820145e-01
9.73206103e-01 9.92190659e-01 4.42105949e-01 -4.26887780e-01
-4.95934367e-01 -2.47060001e-01 2.52211536e-03 1.19514227e+00
3.51193100e-01 2.59972811e-02 -1.12817250e-01 3.50035995e-01
-6.59647405e-01 -9.26044304e-04 9.50633645e-01 1.25570989e+00
-8.53064358e-01 -9.85475838e-01 -5.96177995e-01 7.42281497e-01
-3.54376137e-01 1.42990157e-01 2.68413164e-02 7.23353088e-01
1.25502795e-01 1.14465225e+00 1.27640933e-01 -2.94403583e-01
4.82411295e-01 -1.38620555e-01 9.01990652e-01 -6.59688711e-01
-3.44329208e-01 8.23743641e-02 3.43051016e-01 -5.69201708e-01
-6.37511432e-01 -3.69852185e-01 -1.16207743e+00 3.11694115e-01
-2.34525293e-01 2.38233820e-01 1.27575254e+00 7.39927530e-01
4.67480123e-01 6.52909815e-01 8.12276602e-01 -8.24368656e-01
-8.66160914e-02 -9.64460909e-01 -9.00433004e-01 -3.80042821e-01
4.03252244e-01 -9.15196896e-01 -4.31579977e-01 -4.77780342e-01] | [6.725404739379883, 0.7338322997093201] |
60b71a66-f195-49b8-88b0-2b6c77714fe2 | adversarial-domain-adaptation-for-duplicate | 1809.02255 | null | http://arxiv.org/abs/1809.02255v1 | http://arxiv.org/pdf/1809.02255v1.pdf | Adversarial Domain Adaptation for Duplicate Question Detection | We address the problem of detecting duplicate questions in forums, which is
an important step towards automating the process of answering new questions. As
finding and annotating such potential duplicates manually is very tedious and
costly, automatic methods based on machine learning are a viable alternative.
However, many forums do not have annotated data, i.e., questions labeled by
experts as duplicates, and thus a promising solution is to use domain
adaptation from another forum that has such annotations. Here we focus on
adversarial domain adaptation, deriving important findings about when it
performs well and what properties of the domains are important in this regard.
Our experiments with StackExchange data show an average improvement of 5.6%
over the best baseline across multiple pairs of domains. | ['Tao Lei', 'Darsh J Shah', 'Preslav Nakov', 'Salvatore Romeo', 'Alessandro Moschitti'] | 2018-09-07 | adversarial-domain-adaptation-for-duplicate-1 | https://aclanthology.org/D18-1131 | https://aclanthology.org/D18-1131.pdf | emnlp-2018-10 | ['question-similarity'] | ['natural-language-processing'] | [-8.90399888e-02 2.88970739e-01 1.50765553e-01 -3.95967394e-01
-1.30192411e+00 -1.24180508e+00 5.91960371e-01 3.37236375e-01
-4.88761604e-01 1.08528626e+00 3.52481455e-01 -4.36297119e-01
2.94531047e-01 -6.06781244e-01 -8.47159505e-01 -6.56611696e-02
5.16365528e-01 8.78332973e-01 6.10416889e-01 -5.23291647e-01
2.24184245e-01 5.34829730e-03 -1.00628865e+00 5.02332747e-01
1.35375357e+00 4.28339362e-01 -5.55277169e-02 5.59990108e-01
-7.42409229e-01 8.61136138e-01 -1.23308992e+00 -1.15450144e+00
1.53236747e-01 -5.35417140e-01 -1.52874005e+00 -1.82790413e-01
6.59493327e-01 -3.78121972e-01 -4.73409928e-02 1.29831231e+00
4.40490425e-01 8.05484280e-02 3.96735489e-01 -9.85310435e-01
-7.19401360e-01 5.30600905e-01 -1.80943340e-01 3.76927108e-01
6.18960798e-01 -7.80829564e-02 1.13749421e+00 -6.20253921e-01
9.48517442e-01 1.01896679e+00 6.00771248e-01 8.13360155e-01
-1.32032502e+00 -5.92962682e-01 3.66120189e-02 -8.38500797e-04
-9.53902721e-01 -4.84900773e-01 4.96741444e-01 -5.32734871e-01
6.77926958e-01 1.98854536e-01 -2.82481313e-01 1.25779688e+00
-1.46911949e-01 6.08803988e-01 1.03878975e+00 -6.14745975e-01
2.28167504e-01 6.69450343e-01 1.61989614e-01 1.61001071e-01
2.72626400e-01 -4.79661345e-01 -3.29298258e-01 -6.78309023e-01
2.29299203e-01 -4.77937669e-01 -2.91854411e-01 -1.07543945e-01
-7.68009603e-01 9.45373356e-01 1.01832755e-01 4.04865742e-01
-1.82040915e-01 -3.47969264e-01 5.00279009e-01 7.86233306e-01
6.57937050e-01 1.17704403e+00 -6.66147232e-01 -1.55117556e-01
-8.31988394e-01 6.80653572e-01 1.08312190e+00 9.85789537e-01
6.47783160e-01 -6.48097456e-01 -1.58013299e-01 9.38888848e-01
-6.77556992e-02 9.21565369e-02 4.93902087e-01 -1.28929293e+00
8.71909857e-01 7.83280313e-01 7.64231324e-01 -5.95085979e-01
2.32798923e-02 -7.38255680e-02 -7.86174759e-02 -2.72685681e-02
9.44968939e-01 -5.32164872e-01 -3.96864027e-01 1.74289286e+00
4.06693250e-01 -1.90100208e-01 5.18953279e-02 6.63015604e-01
7.22258031e-01 4.26551074e-01 2.31917903e-01 1.50311276e-01
1.54264653e+00 -8.46344173e-01 -6.80904984e-01 -4.70006466e-01
7.04766929e-01 -1.13756394e+00 1.23942792e+00 6.46784306e-02
-1.07527673e+00 -3.33365947e-01 -6.36459231e-01 -1.39732569e-01
-3.00089389e-01 -1.90859020e-01 2.82106578e-01 6.97674692e-01
-8.68075430e-01 4.41302478e-01 -4.49302435e-01 -5.78298509e-01
-5.73326275e-02 2.48026967e-01 -4.44637865e-01 -1.79530621e-01
-1.63875008e+00 9.15952563e-01 1.96196139e-01 -5.73577642e-01
-2.66094118e-01 -6.65826976e-01 -4.90317851e-01 8.04306716e-02
3.72584105e-01 -5.51100373e-01 1.88718700e+00 -1.13361228e+00
-1.13969243e+00 1.10135293e+00 -4.69916552e-01 -4.88069028e-01
6.68078184e-01 -3.13282251e-01 -5.16086400e-01 1.81258500e-01
5.52144408e-01 4.78229851e-01 5.45882702e-01 -8.98244500e-01
-7.51878023e-01 -1.75917417e-01 5.18135846e-01 2.43233759e-02
-5.59712887e-01 5.18421412e-01 -1.63158983e-01 -5.21610081e-01
-1.76793307e-01 -9.01170731e-01 -1.65306747e-01 -1.50116935e-01
-1.36563212e-01 -5.87901354e-01 5.97561061e-01 -1.14587808e+00
1.16261065e+00 -1.91178942e+00 -3.63941759e-01 -1.44821644e-01
2.70533979e-01 5.59999526e-01 -1.46096945e-01 5.57637930e-01
4.20240313e-03 4.71491843e-01 -7.56464228e-02 -1.46996200e-01
2.55522504e-02 -4.58419174e-02 -4.68303502e-01 6.04750775e-02
5.30276179e-01 5.93412161e-01 -1.11068141e+00 -4.62737173e-01
-4.82367724e-01 7.35419709e-03 -6.52625680e-01 5.38246930e-01
-6.68244243e-01 6.19449854e-01 -5.02243936e-01 2.86390901e-01
7.24433362e-01 -3.78823042e-01 2.48700887e-01 3.88763756e-01
1.12163067e-01 1.04009020e+00 -1.15895987e+00 1.39671385e+00
-4.75524485e-01 7.10645437e-01 2.48080418e-01 -7.06641078e-01
9.52774763e-01 5.70686460e-01 7.18829334e-02 -5.83587170e-01
-2.50865191e-01 4.89399433e-01 -1.34914115e-01 -7.12815225e-01
6.61686957e-01 -2.55621046e-01 -1.73154458e-01 6.00320518e-01
8.36001933e-02 4.60789353e-02 3.47880989e-01 4.79753971e-01
1.66559386e+00 -1.40558809e-01 7.53056630e-02 -3.79490070e-02
5.79050839e-01 4.31711763e-01 7.71299124e-01 6.84767485e-01
-4.77123737e-01 6.25055969e-01 7.64035523e-01 -9.01326686e-02
-1.01807165e+00 -8.75290215e-01 4.23989333e-02 1.10625172e+00
-3.45810503e-02 -1.88634530e-01 -8.55074823e-01 -1.16257477e+00
-4.07907106e-02 8.28987718e-01 -3.67683053e-01 -6.25669435e-02
-6.55385494e-01 -1.83791324e-01 6.66510761e-01 5.16681969e-01
4.81253177e-01 -7.43497252e-01 2.93144770e-02 4.71935034e-01
-9.16923046e-01 -1.33675718e+00 -5.00847697e-01 -1.25299022e-02
-7.38756716e-01 -1.10853803e+00 -8.76530707e-01 -8.26031208e-01
5.15994966e-01 3.40078354e-01 1.59882605e+00 1.15779646e-01
2.53635496e-01 4.06213760e-01 -3.95022213e-01 -3.32362622e-01
-9.21759427e-01 4.12276447e-01 -2.91631073e-01 -3.96746397e-01
8.82124543e-01 -3.67545307e-01 -4.58498776e-01 4.68249649e-01
-7.82753766e-01 -7.75182128e-01 3.31192434e-01 9.03880954e-01
8.02966133e-02 -1.11571707e-01 9.10411477e-01 -1.35027385e+00
1.02396286e+00 -6.91930056e-01 -6.59975111e-01 5.58316588e-01
-2.08317205e-01 1.36194065e-01 7.94636786e-01 -2.65838832e-01
-1.54679060e+00 -1.50218550e-02 -1.47782326e-01 -2.50039715e-02
-4.04882163e-01 2.46626243e-01 -2.16207519e-01 -6.34147786e-03
1.17207897e+00 -4.37387794e-01 -1.67499408e-01 -6.05241001e-01
2.20773131e-01 9.73233044e-01 3.19099516e-01 -6.82899833e-01
8.81342530e-01 2.49033287e-01 -7.40364850e-01 -6.86959445e-01
-8.71315062e-01 -9.21935618e-01 -4.94937927e-01 1.31554917e-01
7.16914177e-01 -8.99108887e-01 -2.09510386e-01 -6.38060346e-02
-1.69860876e+00 4.01759706e-03 -2.47082077e-02 1.98967591e-01
-8.61935094e-02 5.29507518e-01 -5.82546830e-01 -5.19518495e-01
-2.04861090e-01 -8.63157153e-01 7.72305846e-01 2.56658375e-01
-9.57845509e-01 -1.11659479e+00 4.47124094e-01 9.23505962e-01
4.09753829e-01 -5.34746647e-02 9.02793050e-01 -1.34735942e+00
-4.18661118e-01 -4.45600688e-01 8.27752228e-04 4.89984512e-01
-3.24322213e-03 -7.75745958e-02 -9.87277389e-01 -6.98316172e-02
1.51468739e-01 -4.90515143e-01 4.56476837e-01 -2.34629959e-01
7.07102716e-01 -2.96450317e-01 -3.31771553e-01 -1.43347487e-01
1.10988390e+00 -4.42506000e-02 4.84517783e-01 5.55412650e-01
9.68543962e-02 1.09207106e+00 6.77229404e-01 1.14994302e-01
3.04125875e-01 5.94456077e-01 1.43438205e-01 2.58658618e-01
-3.77992839e-02 -3.55936080e-01 4.27959710e-01 4.82725441e-01
5.23500979e-01 -4.28448200e-01 -9.18459535e-01 8.52942467e-01
-1.54735351e+00 -1.02614057e+00 -4.05424237e-01 2.23206043e+00
1.06913662e+00 1.94166526e-01 2.12868273e-01 -2.43729442e-01
1.23703957e+00 -1.57637239e-01 -4.36113358e-01 -4.66536552e-01
-7.11707026e-02 4.81160760e-01 2.62357295e-01 3.36454898e-01
-1.07076073e+00 7.20169187e-01 6.45775318e+00 3.02036434e-01
-8.04156959e-01 3.09101433e-01 3.55980992e-01 1.16775386e-01
-5.44443369e-01 1.90612346e-01 -8.58767271e-01 7.07908690e-01
1.14850497e+00 -3.08762670e-01 1.64583191e-01 8.95532668e-01
-1.12808617e-02 7.69383982e-02 -9.82983947e-01 2.27497280e-01
-1.30358517e-01 -1.08306599e+00 -3.05600703e-01 -1.05068348e-01
9.14443493e-01 2.22308598e-02 -2.36158162e-01 4.99801934e-01
8.22511733e-01 -6.17257476e-01 4.29856062e-01 -7.75394663e-02
3.18306208e-01 -5.10983825e-01 9.25176561e-01 5.51695943e-01
-5.73943019e-01 5.07645905e-02 -5.62580764e-01 8.89409240e-03
3.34468670e-02 6.22587979e-01 -1.10781634e+00 2.31103182e-01
8.76753926e-01 -2.83494908e-02 -6.90372705e-01 1.17800820e+00
-5.90392888e-01 8.27596188e-01 -1.98183537e-01 -7.88321942e-02
1.33627325e-01 1.57816574e-01 4.71890837e-01 9.83556569e-01
2.68141091e-01 -1.24059208e-01 -8.11401531e-02 1.03063357e+00
-6.04443491e-01 7.15348870e-02 -6.00067377e-01 -1.46278173e-01
8.32202852e-01 1.11304188e+00 -3.39096367e-01 -2.44706154e-01
-8.42982709e-01 1.10246253e+00 5.85819066e-01 1.90673321e-01
-7.99278736e-01 -7.50597298e-01 7.39463031e-01 2.72854775e-01
2.23347068e-01 6.95503503e-02 -1.40465587e-01 -1.22970533e+00
5.52708805e-01 -1.22548568e+00 6.66169643e-01 -4.03833538e-01
-1.63355231e+00 4.05703902e-01 -3.82949322e-01 -1.25331414e+00
-4.03418124e-01 -4.12198454e-01 -9.00029659e-01 1.08690035e+00
-1.48241627e+00 -7.18948066e-01 -2.74407923e-01 4.01867926e-01
4.26141918e-01 1.46691635e-01 7.36963212e-01 6.03572547e-01
-3.72996777e-01 9.30958450e-01 4.07596201e-01 3.77385616e-01
1.48394012e+00 -1.46888769e+00 8.40657532e-01 8.19407582e-01
1.42292067e-01 9.20771241e-01 9.62947011e-01 -6.78653359e-01
-7.11594045e-01 -1.02009594e+00 1.52185583e+00 -7.95226097e-01
8.53530467e-01 -3.41131955e-01 -1.48923576e+00 6.25116467e-01
4.84858125e-01 -2.50312626e-01 8.31268787e-01 2.52429098e-01
-4.59342867e-01 2.60215193e-01 -1.30776739e+00 4.66430157e-01
7.72203028e-01 -7.26885974e-01 -1.07025671e+00 6.91371083e-01
9.50627029e-01 -4.57894325e-01 -7.84697771e-01 -1.37104660e-01
-5.31596951e-02 -9.44273472e-01 7.15157032e-01 -8.58278215e-01
3.51737618e-01 -1.47659197e-01 1.95945054e-01 -1.19333613e+00
-9.08320770e-02 -7.66615510e-01 2.62699604e-01 1.89195180e+00
6.43669248e-01 -7.35543013e-01 1.02216554e+00 1.24815130e+00
1.09173588e-01 -4.01241146e-02 -9.22499120e-01 -1.10084605e+00
5.21317303e-01 5.29530384e-02 6.72622085e-01 1.26919723e+00
6.33515120e-02 6.47059143e-01 4.90766615e-02 1.81478605e-01
2.04426274e-01 -2.14790806e-01 9.22563970e-01 -1.42079425e+00
-2.51021922e-01 -1.38893247e-01 -3.31286490e-02 -1.07040405e+00
6.96221530e-01 -6.72986865e-01 -2.81925835e-02 -1.28077912e+00
-1.04370534e-01 -3.68967086e-01 1.22764714e-01 5.48118874e-02
-6.50829732e-01 -2.21593484e-01 -9.09912884e-02 2.41239935e-01
-7.04399228e-01 1.44455075e-01 8.31800997e-01 -1.61247075e-01
-1.27619743e-01 3.74118268e-01 -1.07269800e+00 7.07844138e-01
9.66553628e-01 -7.08491027e-01 -1.57056198e-01 -7.00210750e-01
1.39973745e-01 -7.60602653e-02 1.14149690e-01 -8.26536834e-01
3.60839754e-01 -3.87999117e-02 -7.63561530e-03 -1.33599058e-01
2.63867285e-02 -6.76179588e-01 -3.68432462e-01 2.06822887e-01
-5.66084623e-01 1.49450973e-01 1.51444063e-01 6.21082544e-01
-5.97864509e-01 -9.77416813e-01 5.89570165e-01 -5.45918226e-01
-7.01705098e-01 -3.07480752e-01 -3.89226645e-01 8.72425437e-01
8.45030248e-01 1.77318249e-02 -5.72757542e-01 -6.95325553e-01
-4.19015884e-01 3.32479507e-01 6.57944739e-01 3.45216513e-01
-2.86644939e-02 -1.00510919e+00 -7.50039339e-01 -3.22552115e-01
3.09725940e-01 -9.27740112e-02 -1.00802695e-02 2.30259523e-01
-3.82911593e-01 4.59391356e-01 1.09986164e-01 -1.51711881e-01
-1.33883369e+00 3.52179021e-01 2.74969846e-01 -6.17978334e-01
-9.10766125e-02 1.07735443e+00 -2.78082993e-02 -7.76627719e-01
1.21395707e-01 2.03797504e-01 -1.63677260e-01 7.30231479e-02
5.51557779e-01 5.32369018e-01 4.74509090e-01 -2.99044162e-01
-4.55749214e-01 3.28619480e-02 -3.74065310e-01 -2.41908327e-01
9.60452020e-01 -2.22038224e-01 -2.56115705e-01 -7.41125196e-02
1.19579828e+00 2.87368685e-01 -8.24315250e-01 -4.41582888e-01
6.39295638e-01 -4.99299765e-01 -3.98606926e-01 -8.38330448e-01
-4.96692896e-01 1.01069272e+00 1.57861724e-01 4.52283353e-01
7.67936945e-01 9.97766629e-02 1.05484152e+00 6.00635946e-01
2.71683753e-01 -1.21668959e+00 2.17893720e-01 8.09116781e-01
7.47037113e-01 -1.53230798e+00 -2.74259031e-01 -5.01346290e-01
-6.47056639e-01 1.00595725e+00 1.01376832e+00 7.47472048e-02
9.93222669e-02 -2.17634112e-01 3.05592149e-01 -1.27457529e-02
-7.58752286e-01 -1.18171461e-01 2.74697747e-02 6.22934520e-01
8.55568111e-01 -3.00437987e-01 -4.94716823e-01 7.56778419e-01
-1.10921919e-01 -1.71153575e-01 6.53755009e-01 1.03866112e+00
-4.19990212e-01 -1.60096109e+00 -6.93338096e-01 3.81105721e-01
-1.02214277e+00 6.07064106e-02 -9.71253157e-01 6.12666726e-01
-1.77341968e-01 1.33798337e+00 -3.42754275e-02 -9.00757685e-02
6.05058908e-01 5.08310556e-01 1.13033526e-01 -9.80541527e-01
-1.08346117e+00 -4.65045840e-01 4.52053249e-01 -9.30916071e-02
-2.14624465e-01 -6.57727182e-01 -9.49758172e-01 -3.69508445e-01
-6.58564687e-01 7.20314860e-01 3.23636502e-01 8.97908390e-01
5.35795271e-01 1.15273252e-01 5.46441615e-01 -6.04962781e-02
-1.02808428e+00 -7.38384366e-01 -8.59834552e-02 7.91170061e-01
1.11650281e-01 -3.73703301e-01 -5.13326168e-01 1.73836946e-02] | [11.360297203063965, 8.158501625061035] |
ebf426fa-67fc-4704-8c56-60d6211687c0 | piqa-reasoning-about-physical-commonsense-in | 1911.11641 | null | https://arxiv.org/abs/1911.11641v1 | https://arxiv.org/pdf/1911.11641v1.pdf | PIQA: Reasoning about Physical Commonsense in Natural Language | To apply eyeshadow without a brush, should I use a cotton swab or a toothpick? Questions requiring this kind of physical commonsense pose a challenge to today's natural language understanding systems. While recent pretrained models (such as BERT) have made progress on question answering over more abstract domains - such as news articles and encyclopedia entries, where text is plentiful - in more physical domains, text is inherently limited due to reporting bias. Can AI systems learn to reliably answer physical common-sense questions without experiencing the physical world? In this paper, we introduce the task of physical commonsense reasoning and a corresponding benchmark dataset Physical Interaction: Question Answering or PIQA. Though humans find the dataset easy (95% accuracy), large pretrained models struggle (77%). We provide analysis about the dimensions of knowledge that existing models lack, which offers significant opportunities for future research. | ['Yejin Choi', 'Rowan Zellers', 'Ronan Le Bras', 'Jianfeng Gao', 'Yonatan Bisk'] | 2019-11-26 | null | null | null | null | ['physical-commonsense-reasoning'] | ['reasoning'] | [ 2.38308072e-01 4.09445047e-01 -4.76056308e-01 -1.14617705e-01
-5.14583945e-01 -8.10313344e-01 8.06738079e-01 2.37664834e-01
-3.69560242e-01 1.12959075e+00 3.46223146e-01 -8.15358818e-01
-5.07240772e-01 -1.06981432e+00 -7.51295865e-01 -2.65420407e-01
3.74422640e-01 4.78790700e-01 3.65453988e-01 -6.87295020e-01
5.25347412e-01 3.62651646e-01 -1.15772665e+00 1.06924087e-01
8.25598001e-01 9.73255098e-01 2.47086510e-02 7.22418606e-01
-2.89529383e-01 1.70664930e+00 -6.32872045e-01 -3.42387617e-01
-5.57231493e-02 -4.31918561e-01 -1.65755105e+00 -5.60903430e-01
7.94906199e-01 -5.80287874e-01 -7.29677200e-01 7.67630637e-01
9.40138772e-02 3.87162328e-01 8.17529023e-01 -1.41436422e+00
-1.51073170e+00 2.87773609e-01 -1.54374868e-01 5.47940195e-01
9.93559480e-01 5.57529867e-01 1.32752597e+00 -3.01340669e-01
7.14519024e-01 1.30316198e+00 6.35233819e-01 7.65813351e-01
-1.03894925e+00 -3.70093465e-01 -5.41951954e-02 7.59700418e-01
-8.61991286e-01 -1.35102794e-01 8.82361472e-01 -3.02891999e-01
1.23462903e+00 5.23768365e-01 4.94708925e-01 1.45272100e+00
4.49644029e-01 6.78358436e-01 1.36925149e+00 -2.90081829e-01
3.26643795e-01 -1.03712358e-01 3.66734087e-01 5.62825680e-01
4.74476159e-01 1.55067280e-01 -9.55513418e-01 -3.60753745e-01
7.90876150e-01 -2.00232476e-01 -1.78310767e-01 -2.35773638e-01
-1.34664690e+00 5.59351265e-01 5.12744963e-01 5.11699855e-01
-3.13638419e-01 6.46624863e-01 2.30884835e-01 5.85738719e-01
6.22591600e-02 1.11675334e+00 -6.55676782e-01 -3.71671528e-01
-4.73982334e-01 8.89457941e-01 1.18985319e+00 6.85895026e-01
4.14281279e-01 -3.30514669e-01 2.54974037e-01 5.90961516e-01
1.42104015e-01 6.26055956e-01 3.46186966e-01 -1.16902936e+00
5.18529773e-01 4.96268809e-01 5.44475257e-01 -1.17071807e+00
-4.78812307e-01 2.22947702e-01 -2.04862848e-01 6.23663403e-02
8.52850199e-01 -1.53819919e-01 -7.10410476e-01 1.87452137e+00
3.92761469e-01 -9.21393782e-02 1.70494244e-01 9.79493678e-01
1.06695306e+00 5.11014938e-01 4.30552244e-01 2.77187586e-01
1.59570754e+00 -4.13383245e-01 -7.57680357e-01 -6.91937864e-01
5.73606074e-01 -4.53017652e-01 1.61690557e+00 3.55374277e-01
-1.05848038e+00 -2.56359786e-01 -1.08271182e+00 -9.63251233e-01
-7.61849821e-01 -4.46232140e-01 9.61760461e-01 3.90108019e-01
-6.47003353e-01 7.35808849e-01 -4.87481773e-01 -7.92465270e-01
3.50013107e-01 -9.32928268e-03 -3.46677929e-01 -1.92245975e-01
-1.58043003e+00 1.83579624e+00 2.22421110e-01 -2.95695752e-01
-6.91281617e-01 -8.46369565e-01 -9.77910697e-01 -1.30814806e-01
7.54677713e-01 -8.10878396e-01 1.41107535e+00 -3.91342580e-01
-1.49444592e+00 9.55527365e-01 1.79445207e-01 -4.20942485e-01
2.99090177e-01 -6.00460589e-01 -5.08222997e-01 3.53079081e-01
1.01574689e-01 5.45947611e-01 4.65438306e-01 -1.07689703e+00
-9.45496336e-02 -2.69357949e-01 8.79894316e-01 -2.51910668e-02
-1.05707191e-01 -1.03862599e-01 5.06999850e-01 -4.48574394e-01
4.03495617e-02 -8.49770665e-01 2.34560236e-01 4.44663823e-01
-3.75793219e-01 -6.40042484e-01 8.87508571e-01 -7.47836113e-01
9.10990238e-01 -1.72479153e+00 1.54543191e-01 -3.22268873e-01
3.10882062e-01 -1.25360310e-01 -2.84766015e-02 5.42126238e-01
2.70075239e-02 2.36043826e-01 -1.58092499e-01 5.18395185e-01
5.33602774e-01 4.54331845e-01 -6.88411117e-01 1.82852626e-01
3.03163946e-01 1.16758645e+00 -1.24854743e+00 -6.51434243e-01
2.81005949e-01 3.74353565e-02 -4.22358841e-01 -2.31590327e-02
-8.57903123e-01 -3.33768800e-02 -6.14302099e-01 6.54237688e-01
2.41673663e-01 -5.96301198e-01 -1.16117232e-01 -3.45406570e-02
3.38141888e-01 1.01537466e+00 -7.72688985e-01 1.75996697e+00
-5.06764591e-01 7.30859637e-01 -2.48308897e-01 -6.45398915e-01
4.54596400e-01 2.18261436e-01 1.07008144e-01 -7.93607950e-01
-9.48837250e-02 -5.96430749e-02 1.64126739e-01 -1.03808033e+00
5.35357475e-01 -7.66948581e-01 -3.08076233e-01 5.90306699e-01
-7.46336728e-02 -1.08059549e+00 -1.71054155e-02 4.76226807e-01
1.40842032e+00 4.67376947e-01 4.04690564e-01 -4.76966977e-01
1.75064102e-01 5.09477735e-01 7.49736428e-02 7.83574939e-01
-4.26871270e-01 1.85906425e-01 3.87425423e-01 -7.41550744e-01
-9.50493276e-01 -1.39394021e+00 -2.40792811e-01 1.22248876e+00
4.62250620e-01 -2.60373652e-01 -5.78089654e-01 -5.34634829e-01
1.05849683e-01 1.31270897e+00 -7.05748677e-01 -3.55656624e-01
-3.87099892e-01 -3.66293550e-01 6.32417798e-01 6.92820251e-01
6.13717914e-01 -1.51787972e+00 -9.77438509e-01 7.81871974e-02
-4.77508187e-01 -1.18174314e+00 9.56496969e-02 5.60407415e-02
-6.47414327e-01 -1.15855670e+00 -2.46963277e-01 -2.50413686e-01
-1.27740065e-02 6.61663786e-02 1.45463312e+00 2.82216102e-01
-4.35228825e-01 6.60595953e-01 -1.26227036e-01 -6.88661993e-01
-2.54621089e-01 -9.37674716e-02 2.26282198e-02 -8.76155019e-01
8.22810531e-01 -6.81775153e-01 -5.98964930e-01 9.16571692e-02
-9.60831583e-01 -1.97685227e-01 9.58667472e-02 5.86296558e-01
1.05881179e-02 -3.94392043e-01 7.19057500e-01 -5.60435534e-01
7.58584142e-01 -8.75046432e-01 -1.21279009e-01 2.27171496e-01
-3.05238944e-02 1.77741468e-01 5.72253585e-01 -6.01339340e-01
-1.08687472e+00 -6.56138539e-01 2.47458462e-02 2.13115245e-01
-3.12921196e-01 4.78787810e-01 3.88112217e-02 1.11974645e-02
1.38462293e+00 -6.01239055e-02 -2.88061380e-01 -3.39135617e-01
5.58447063e-01 2.34017611e-01 6.12516940e-01 -1.16542542e+00
8.79750431e-01 5.20200610e-01 1.40403852e-01 -9.25849259e-01
-1.48526859e+00 -2.27346376e-01 -3.45358968e-01 1.99526966e-01
1.20646238e+00 -3.58422488e-01 -9.88862872e-01 9.92134884e-02
-1.15862525e+00 -5.86820126e-01 -4.86725271e-01 3.61085474e-01
-8.59413505e-01 4.93228376e-01 -8.12421501e-01 -7.47319937e-01
5.35330102e-02 -4.04465556e-01 8.67478490e-01 2.37680167e-01
-9.89910483e-01 -1.14990723e+00 1.40335515e-01 6.64694667e-01
5.53399980e-01 4.56949919e-01 1.34801376e+00 -5.25911689e-01
-6.00042105e-01 1.55726820e-01 -2.76526004e-01 7.30234906e-02
2.28019997e-01 -4.44297880e-01 -9.46609974e-01 3.74889255e-01
1.47816211e-01 -1.16142654e+00 5.90239823e-01 -1.88419119e-01
1.37675357e+00 -3.16658556e-01 -2.23037302e-01 -1.11950599e-01
1.22043240e+00 -1.82974823e-02 8.70557129e-01 3.55910063e-01
4.41259086e-01 4.86125618e-01 4.12120134e-01 1.28350079e-01
7.73081720e-01 3.92051071e-01 3.16793621e-01 3.66531193e-01
-1.32871315e-01 -4.98116165e-01 2.32721433e-01 2.15764403e-01
-2.47913539e-01 -3.02345306e-01 -1.07693970e+00 7.47597575e-01
-1.51246238e+00 -1.32947135e+00 -5.99488541e-02 1.58657217e+00
1.20465302e+00 1.47709206e-01 -1.93944991e-01 2.26593718e-01
-5.49356900e-02 3.60593975e-01 -9.42366838e-01 -4.61043090e-01
5.86528778e-02 4.80484992e-01 1.51264081e-02 5.19121468e-01
-8.34738195e-01 8.47797871e-01 7.14728785e+00 6.73370540e-01
-7.12715209e-01 3.04836556e-02 2.44854882e-01 2.29494907e-02
-4.83911425e-01 1.27774820e-01 -1.64074063e-01 2.89694190e-01
9.67988193e-01 1.73385665e-01 5.38180053e-01 8.34018171e-01
-1.76390171e-01 -5.83325446e-01 -1.59887838e+00 9.32714343e-01
2.42503453e-02 -1.30942833e+00 5.17798439e-02 -1.02613315e-01
4.21741694e-01 -7.65990168e-02 -1.41958088e-01 3.94636214e-01
8.13518345e-01 -1.39378953e+00 6.09609544e-01 7.09014237e-01
5.27994275e-01 -4.38813455e-02 2.58447230e-01 5.55448890e-01
-5.33992708e-01 -5.65837361e-02 -5.09422660e-01 -6.93576336e-01
1.61896393e-01 3.20650011e-01 -4.92219776e-01 2.46980920e-01
6.71618342e-01 1.84993371e-01 -3.97188693e-01 5.15898287e-01
-4.58739907e-01 6.51079595e-01 -5.87007165e-01 -4.64691818e-01
2.09538862e-01 1.12874940e-01 3.49546015e-01 8.63042831e-01
-3.76183511e-04 7.71021008e-01 -1.87347293e-01 1.22247195e+00
-1.13248765e-01 -3.14952284e-01 -6.90964401e-01 -3.49955887e-01
3.41570646e-01 7.89611340e-01 -2.73892283e-01 -4.36712980e-01
-4.03511912e-01 8.28107178e-01 4.38738257e-01 3.38102847e-01
-8.23176205e-01 -3.33166867e-01 7.62772262e-01 2.53051788e-01
-4.05222654e-01 -3.91176075e-01 -5.02416193e-01 -1.22334623e+00
1.96741700e-01 -8.04667890e-01 3.70580614e-01 -1.53730309e+00
-1.70104098e+00 -2.65579104e-01 3.40910137e-01 -3.95157695e-01
-2.30570063e-01 -1.03657138e+00 -5.44267476e-01 5.07715285e-01
-1.40552795e+00 -1.14001131e+00 -4.10237342e-01 5.17742336e-01
1.84037775e-01 4.12600279e-01 8.44083726e-01 -1.37294963e-01
2.27574259e-01 6.08754009e-02 -3.24015617e-01 -2.68938728e-02
6.51746631e-01 -1.43026936e+00 4.49096441e-01 2.81466600e-02
1.98307201e-01 8.02388251e-01 1.04902494e+00 -5.98848641e-01
-1.70989215e+00 -2.91549385e-01 9.93263185e-01 -1.37988615e+00
1.14433670e+00 -1.52677327e-01 -1.05641925e+00 8.48405719e-01
1.86410889e-01 1.23087158e-02 6.93458319e-01 4.25164491e-01
-8.75114024e-01 3.81807238e-01 -1.44511974e+00 7.82010972e-01
1.48265386e+00 -1.11288381e+00 -1.56867731e+00 8.13073456e-01
8.09891403e-01 -4.33154523e-01 -9.36222553e-01 8.86458084e-02
5.54269433e-01 -5.87085366e-01 1.11069155e+00 -1.33492887e+00
9.96156156e-01 -1.82145894e-01 -1.94356740e-01 -1.12255251e+00
-1.53135210e-01 -5.43154955e-01 -3.12820047e-01 6.01588666e-01
2.95747072e-01 -5.56182861e-01 5.90654016e-01 1.20406306e+00
2.04309180e-01 -6.42804980e-01 -8.76267493e-01 -7.47654498e-01
6.34322822e-01 -5.10707974e-01 5.29156744e-01 1.18646169e+00
7.24650502e-01 6.46699369e-01 -1.29433423e-01 -1.22909032e-01
3.56510848e-01 2.74554025e-02 5.61299384e-01 -1.24896836e+00
-5.09865224e-01 -1.02104709e-01 -2.73265213e-01 -1.11511052e+00
2.21223056e-01 -6.55615747e-01 -2.10042462e-01 -1.76095939e+00
2.86237329e-01 -1.36080697e-01 -1.56740695e-01 5.87741733e-01
-1.50840119e-01 5.37141077e-02 3.03352535e-01 -2.33932763e-01
-6.11017525e-01 2.70558327e-01 1.48398709e+00 -1.26540288e-01
3.38594258e-01 -6.87119126e-01 -9.96066511e-01 1.01521838e+00
8.60326469e-01 -1.97695062e-01 -6.04548872e-01 -4.26545203e-01
7.62451351e-01 2.71498978e-01 1.13660288e+00 -8.41782629e-01
3.28269266e-02 -8.55998397e-01 4.28421676e-01 -2.26511970e-01
6.80735171e-01 -8.17021847e-01 -2.62128472e-01 3.34159881e-01
-5.49704492e-01 -6.54763803e-02 3.52686226e-01 5.29123902e-01
1.89717472e-01 -3.04579943e-01 6.74411833e-01 -3.02521855e-01
-6.71265244e-01 -6.90761507e-02 -2.19823197e-01 6.88558340e-01
7.51552880e-01 -2.94655412e-02 -8.25213134e-01 -6.81360364e-01
-6.62268817e-01 9.23293978e-02 4.41327125e-01 4.69055831e-01
3.53365034e-01 -1.15122807e+00 -2.20093057e-01 -7.75758803e-01
2.22955659e-01 -1.56857952e-01 2.57937968e-01 3.02838326e-01
-5.38273454e-01 5.53436220e-01 -1.63641676e-01 -3.81869450e-02
-5.85230350e-01 6.99856818e-01 3.53988022e-01 -1.11297287e-01
-5.67182720e-01 8.68155420e-01 2.08922803e-01 -5.79038620e-01
-2.76801884e-01 -8.86195838e-01 1.65245593e-01 -4.05207306e-01
4.58679318e-01 3.44699025e-01 -3.80347043e-01 -9.60856006e-02
-5.08395731e-01 4.33859348e-01 1.46180898e-01 -8.15941989e-02
1.25685596e+00 9.81605798e-02 -1.15598194e-01 6.39599323e-01
8.71172249e-01 -5.85002638e-02 -9.96167898e-01 -2.21849769e-01
5.91917261e-02 -3.21888179e-01 -3.48103732e-01 -1.33031058e+00
-7.35941008e-02 9.50088143e-01 -6.31691376e-03 4.35424775e-01
4.12826717e-01 5.71814418e-01 1.04012632e+00 1.17638803e+00
5.62762558e-01 -1.32068634e+00 5.26603758e-01 5.86762607e-01
1.25713193e+00 -1.31323850e+00 2.68128186e-01 -2.27682456e-01
-4.91499335e-01 8.21482241e-01 8.80186200e-01 -2.38991857e-01
5.55350423e-01 1.11226544e-01 -1.99375317e-01 -6.91047490e-01
-7.94499457e-01 2.64504701e-02 1.56850472e-01 7.10950732e-01
5.11630058e-01 2.49881204e-02 -2.37552553e-01 5.02868474e-01
-8.40321064e-01 2.10437819e-01 5.61583996e-01 9.89583194e-01
-6.45240843e-01 -5.91829002e-01 -3.90198529e-01 4.95053202e-01
-3.42380375e-01 -7.11161420e-02 -8.47033083e-01 1.02508795e+00
6.14797547e-02 1.10497582e+00 -1.95588797e-01 8.12613443e-02
2.63643891e-01 4.38898921e-01 9.79643404e-01 -4.70933169e-01
-3.00174475e-01 -1.16394782e+00 4.16793495e-01 -5.83631396e-01
-3.82256985e-01 -5.24084330e-01 -1.57902670e+00 -6.29760981e-01
-1.86630368e-01 -5.26774302e-02 3.43941569e-01 1.39971054e+00
1.25321135e-01 3.05750400e-01 -4.37734753e-01 -4.46265876e-01
-8.75867605e-01 -9.97749269e-01 -2.54018813e-01 7.46740282e-01
4.66633499e-01 -9.81757879e-01 -3.94058585e-01 4.47961763e-02] | [9.766639709472656, 7.540696144104004] |
f9f35c2a-6378-4d5b-b4f0-aaa95ca53e97 | efficient-cross-lingual-transfer-for-chinese | 2305.11540 | null | https://arxiv.org/abs/2305.11540v1 | https://arxiv.org/pdf/2305.11540v1.pdf | Efficient Cross-Lingual Transfer for Chinese Stable Diffusion with Images as Pivots | Diffusion models have made impressive progress in text-to-image synthesis. However, training such large-scale models (e.g. Stable Diffusion), from scratch requires high computational costs and massive high-quality text-image pairs, which becomes unaffordable in other languages. To handle this challenge, we propose IAP, a simple but effective method to transfer English Stable Diffusion into Chinese. IAP optimizes only a separate Chinese text encoder with all other parameters fixed to align Chinese semantics space to the English one in CLIP. To achieve this, we innovatively treat images as pivots and minimize the distance of attentive features produced from cross-attention between images and each language respectively. In this way, IAP establishes connections of Chinese, English and visual semantics in CLIP's embedding space efficiently, advancing the quality of the generated image with direct Chinese prompts. Experimental results show that our method outperforms several strong Chinese diffusion models with only 5%~10% training data. | ['Maosong Sun', 'Zhiyuan Liu', 'Wenhao Li', 'Yutong Chen', 'Xiaoyuan Yi', 'Xu Han', 'Jinyi Hu'] | 2023-05-19 | null | null | null | null | ['cross-lingual-transfer'] | ['natural-language-processing'] | [ 1.05366446e-01 -9.46475118e-02 -7.21585900e-02 -1.89040512e-01
-8.31277072e-01 -5.28714716e-01 7.35710621e-01 -5.33289909e-01
-5.32385349e-01 6.26082301e-01 5.99904478e-01 -1.62401125e-01
4.11555529e-01 -7.15566337e-01 -8.10536146e-01 -6.59412563e-01
5.49803793e-01 1.29774123e-01 3.10027182e-01 -1.86398268e-01
2.07719356e-01 -5.33381896e-03 -8.86445522e-01 4.82707292e-01
1.06331241e+00 6.75530493e-01 1.05480385e+00 7.02368975e-01
-3.83064151e-01 1.05323768e+00 -6.37993753e-01 -6.82679772e-01
2.07356289e-01 -8.38379264e-01 -5.55515826e-01 1.76925585e-01
3.35220188e-01 -5.84116161e-01 -6.65107906e-01 1.21559322e+00
6.11227274e-01 -1.82918504e-01 5.80446959e-01 -1.04246461e+00
-1.62785482e+00 8.30152512e-01 -6.60206854e-01 1.01573750e-01
1.16071336e-01 3.42482835e-01 7.71456301e-01 -9.95021522e-01
1.01714313e+00 1.20895827e+00 3.16333920e-01 8.74323487e-01
-9.58490908e-01 -6.67310238e-01 2.22467005e-01 3.89346555e-02
-1.21481073e+00 -4.17274356e-01 6.36688948e-01 -2.12812722e-01
8.45714808e-01 1.30482584e-01 7.00912237e-01 1.30341804e+00
1.52922064e-01 1.19322765e+00 1.00183094e+00 -4.04997408e-01
-1.41476065e-01 6.18566453e-01 -6.04895830e-01 6.56765103e-01
-3.16807836e-01 -1.07738696e-01 -5.00574172e-01 5.47063529e-01
9.36635911e-01 -1.16309777e-01 -3.72060895e-01 1.38767259e-02
-1.44469571e+00 7.51825869e-01 5.30503154e-01 3.36453468e-01
-2.88872272e-01 2.40979806e-01 1.28954083e-01 5.48713386e-01
5.21434784e-01 3.32468122e-01 -3.67156208e-01 -2.26769701e-01
-7.96468318e-01 -1.51599636e-02 3.35751414e-01 1.43423104e+00
6.44583702e-01 2.69121647e-01 -2.91345119e-01 8.93787265e-01
7.41689876e-02 9.59285557e-01 7.32535362e-01 -7.72364914e-01
7.25276768e-01 3.41091365e-01 2.23484952e-02 -9.93467331e-01
2.50316918e-01 -2.73410290e-01 -1.00655508e+00 -2.21860185e-01
-2.44762190e-02 -5.20422935e-01 -7.98269689e-01 1.64434648e+00
1.03337869e-01 -1.23063766e-03 1.29226014e-01 9.47324514e-01
4.87370163e-01 1.20685089e+00 -2.99985036e-02 -9.96319205e-02
9.46210623e-01 -1.68032146e+00 -8.98944974e-01 -3.38064611e-01
7.02857316e-01 -1.11234808e+00 1.61441648e+00 2.25311685e-02
-1.39087820e+00 -7.09334016e-01 -9.39415812e-01 -4.12366271e-01
-1.90443426e-01 4.96388733e-01 4.07945633e-01 3.91117692e-01
-1.35293257e+00 4.29634333e-01 -4.67355788e-01 -2.96010911e-01
3.41931611e-01 1.63703367e-01 -3.08885098e-01 -3.59954983e-01
-1.26526654e+00 8.10036838e-01 1.37236953e-01 7.33629242e-02
-8.72420728e-01 -6.73294604e-01 -6.72945440e-01 7.95848817e-02
2.31654838e-01 -6.88880324e-01 9.83554006e-01 -1.57853949e+00
-1.79670155e+00 7.18860567e-01 -1.16062261e-01 -2.66079485e-01
7.30878592e-01 -4.49540406e-01 -4.52791005e-01 1.68028310e-01
2.81326979e-01 1.25126266e+00 1.18859816e+00 -1.05915403e+00
-7.57239401e-01 1.12220302e-01 -8.53914544e-02 4.41285133e-01
-7.85766125e-01 2.85090238e-01 -1.10003972e+00 -8.99024904e-01
-4.32628930e-01 -8.44233155e-01 -3.16414803e-01 3.29536706e-01
-2.77871668e-01 4.38130535e-02 8.69682729e-01 -6.85749471e-01
1.21797574e+00 -2.45455360e+00 4.48993355e-01 -2.32066050e-01
1.65414423e-01 2.00098544e-01 -6.23836696e-01 3.77164155e-01
3.31309110e-01 1.75906524e-01 -1.39758483e-01 -4.53281432e-01
-6.95386976e-02 1.08803190e-01 -4.31455225e-01 9.88915116e-02
3.74992996e-01 1.27153432e+00 -8.93156588e-01 -6.20465219e-01
1.54124498e-01 5.06489456e-01 -7.51891613e-01 4.62728947e-01
-2.34073162e-01 3.16863209e-01 -5.29887617e-01 1.61408320e-01
7.97582090e-01 -4.98832703e-01 1.81619637e-02 2.21765158e-03
-1.84121162e-01 -2.80704983e-02 -7.05846727e-01 1.93704796e+00
-7.08595216e-01 9.12959397e-01 -2.43571009e-02 -6.28314853e-01
8.52235854e-01 1.18631519e-01 1.11183040e-01 -1.09772897e+00
-4.36117733e-03 2.16797620e-01 -2.30488300e-01 -6.35614991e-01
6.04395866e-01 -9.79638323e-02 5.36050089e-02 4.36440706e-01
7.14420825e-02 -4.82751995e-01 9.85654294e-02 5.73966920e-01
7.53417373e-01 1.48011535e-01 -3.85889918e-01 -2.87234426e-01
4.46806282e-01 -5.39193377e-02 3.47139299e-01 5.12731135e-01
-7.66635984e-02 9.51569319e-01 4.77358252e-01 -1.59620903e-02
-1.40540588e+00 -9.30850744e-01 4.89594966e-01 8.99832964e-01
3.04692119e-01 -5.01956820e-01 -1.27226281e+00 -7.27946401e-01
-4.04566765e-01 5.82217753e-01 -6.32790446e-01 -1.50072843e-01
-6.33545816e-01 -5.42293847e-01 3.80763441e-01 4.93005753e-01
8.37629259e-01 -1.24047077e+00 -5.13056852e-02 1.32949889e-01
-3.01909417e-01 -1.31120861e+00 -1.29265213e+00 -2.74497032e-01
-6.19919717e-01 -4.16540027e-01 -1.34928262e+00 -1.18307149e+00
8.86354506e-01 6.30617380e-01 9.23534095e-01 7.93571118e-03
-1.09155901e-01 1.19467884e-01 -4.42696810e-01 -1.24779642e-01
-4.05834794e-01 2.13630095e-01 -3.71455699e-01 3.08103058e-02
1.80836111e-01 -2.50684142e-01 -8.53156149e-01 2.93883920e-01
-1.08459759e+00 4.60415900e-01 7.65897810e-01 8.98811221e-01
3.69675636e-01 -3.45650204e-02 3.75290364e-01 -8.43351305e-01
9.15389717e-01 -3.56241047e-01 -5.28647900e-01 4.14042473e-01
-5.86737692e-01 7.03416392e-02 9.67386544e-01 -6.98463321e-01
-1.28301346e+00 -1.30864292e-01 -1.96784750e-01 -5.32991409e-01
3.49316239e-01 4.82746452e-01 -4.47008349e-02 2.12043449e-01
4.47937518e-01 7.14410365e-01 -2.25336626e-01 -2.71056831e-01
7.16027558e-01 7.79668629e-01 3.25719863e-01 -4.76763994e-01
7.60588527e-01 4.36861813e-01 -6.57956123e-01 -8.48985374e-01
-6.45789742e-01 -3.03797405e-02 -6.30424023e-01 -1.35862574e-01
1.16111350e+00 -1.15636611e+00 -1.63350761e-01 7.76624382e-01
-1.24301028e+00 -6.46071851e-01 -3.00337046e-01 4.46442366e-01
-3.04122776e-01 3.36118102e-01 -9.13986266e-01 -3.49037290e-01
-2.42125586e-01 -1.25867200e+00 1.07785559e+00 1.46269321e-01
8.42192098e-02 -1.01855481e+00 -1.75608024e-01 1.83420494e-01
7.23370969e-01 -4.29944128e-01 7.73783982e-01 1.78097591e-01
-9.85787570e-01 3.58373709e-02 -6.73820078e-01 6.52658045e-01
2.83822287e-02 1.76470309e-01 -6.27219081e-01 -2.03130901e-01
9.17109177e-02 -5.49908102e-01 9.02605355e-01 3.32889259e-01
9.31842148e-01 -2.45017558e-01 -1.20915957e-01 7.12158203e-01
1.47922647e+00 2.06314564e-01 8.03017378e-01 1.45999655e-01
7.46836841e-01 4.78218883e-01 4.54668969e-01 1.13283955e-01
4.54357654e-01 3.53875935e-01 -9.69832316e-02 -4.76171553e-01
-6.87021613e-01 -6.99330986e-01 6.87965572e-01 1.52834296e+00
2.34351769e-01 -5.06159365e-01 -4.60647583e-01 6.43059254e-01
-1.66773391e+00 -8.57798576e-01 1.16371992e-03 1.76667905e+00
1.06266177e+00 1.61624849e-01 -3.25030237e-01 -5.29770494e-01
6.52210474e-01 2.87632495e-01 -4.44649696e-01 -1.32787868e-01
-6.65553868e-01 -1.29244506e-01 3.86187136e-01 6.23179018e-01
-6.52832031e-01 1.55608308e+00 6.28151989e+00 1.17122519e+00
-1.45651603e+00 4.71270472e-01 9.16955233e-01 -1.55181870e-01
-6.72793746e-01 -1.44122764e-02 -6.94827437e-01 5.98201871e-01
7.09248066e-01 -1.55236721e-01 6.38037503e-01 7.20946193e-01
3.39085639e-01 4.63443920e-02 -6.67191863e-01 9.39190269e-01
2.13474959e-01 -1.52813542e+00 3.36639553e-01 -1.34563565e-01
1.30503559e+00 1.49729073e-01 4.68125910e-01 2.17793703e-01
4.67230648e-01 -8.34440708e-01 1.04707503e+00 2.89509505e-01
1.00143576e+00 -4.31406796e-01 2.35545710e-01 3.47525984e-01
-1.00390255e+00 -3.52695547e-02 -7.23666847e-01 2.12134197e-01
3.35497230e-01 3.61950487e-01 -1.21033974e-01 1.16135649e-01
5.05006135e-01 9.02587891e-01 -4.63837922e-01 4.78203535e-01
-4.27658170e-01 4.47785765e-01 -4.81380001e-02 -1.66542441e-01
7.10491240e-01 -3.49468917e-01 7.60889873e-02 1.25144553e+00
6.54941142e-01 9.87073928e-02 4.97801043e-02 9.81563330e-01
-3.70815694e-01 3.57553899e-01 -7.09619641e-01 -2.18766659e-01
2.29641825e-01 1.09402835e+00 -5.86283028e-01 -6.13373399e-01
-7.02432156e-01 1.74308157e+00 3.66860151e-01 5.99231899e-01
-1.09059334e+00 -4.12353903e-01 3.60178083e-01 -4.18370077e-03
3.88785422e-01 -2.42687359e-01 -1.12943174e-02 -1.48274612e+00
1.08481325e-01 -8.58253539e-01 -2.09828407e-01 -1.38446379e+00
-1.15221524e+00 9.64888513e-01 -4.90569532e-01 -1.12273669e+00
1.39717594e-01 -4.09577489e-01 -4.59746778e-01 1.09796202e+00
-1.64548385e+00 -1.36629856e+00 -7.93319941e-02 8.35612833e-01
1.11018753e+00 -6.96450248e-02 3.29245567e-01 4.91219282e-01
-5.74119806e-01 7.12835133e-01 3.74758571e-01 1.08045012e-01
1.00286412e+00 -1.02275717e+00 6.53222620e-01 8.85151923e-01
5.53189628e-02 5.33340931e-01 3.43124479e-01 -7.30674148e-01
-1.37784660e+00 -1.19316626e+00 1.01150501e+00 -3.14969808e-01
8.18988264e-01 -3.83454591e-01 -7.40046799e-01 5.38144052e-01
9.55195129e-01 -3.00265968e-01 2.62041509e-01 -4.57572103e-01
-1.79703832e-01 -1.21189497e-01 -5.61884880e-01 1.09239137e+00
1.14322901e+00 -7.47138560e-01 -6.52169809e-02 4.43497151e-01
1.07970726e+00 -2.20844656e-01 -6.12363517e-01 -1.64314017e-01
3.25846136e-01 -8.33761752e-01 7.41317391e-01 -4.29483831e-01
1.05303717e+00 -7.92775825e-02 -6.82613254e-02 -1.43256617e+00
-4.40297544e-01 -9.38867569e-01 3.92723441e-01 1.22862554e+00
7.88353860e-01 -2.71735370e-01 6.74570680e-01 3.90036017e-01
-1.57886922e-01 -4.92221087e-01 -4.44847137e-01 -5.70166707e-01
3.03429991e-01 -8.16733465e-02 3.61223370e-01 1.08976758e+00
-1.77675057e-02 7.01199114e-01 -8.06485116e-01 -2.03737080e-01
2.19753414e-01 -3.04462202e-02 7.89757073e-01 -2.69626528e-01
-2.28632331e-01 -5.28820038e-01 2.30231434e-01 -1.73884928e+00
1.21309020e-01 -7.52640128e-01 8.20420682e-02 -1.37776303e+00
3.25285435e-01 -5.07384002e-01 -9.80613530e-02 4.13654074e-02
-4.38034356e-01 2.50099570e-01 4.03094649e-01 3.39730382e-01
-5.62596679e-01 8.59245121e-01 1.99039483e+00 -3.91505510e-01
-1.89055630e-03 -5.06550789e-01 -7.31436789e-01 3.37150306e-01
6.63384378e-01 -4.75616097e-01 -8.66869748e-01 -1.30810916e+00
2.87346154e-01 1.48621723e-01 -1.52116567e-01 -7.33917773e-01
4.30335969e-01 -2.59107113e-01 3.59052122e-01 -2.08243802e-01
2.23834500e-01 -6.40988588e-01 -7.57176653e-02 2.84188211e-01
-6.98055863e-01 2.98237741e-01 -4.66044471e-02 6.01981521e-01
-5.07121742e-01 -2.26605684e-01 7.64280677e-01 -3.48086178e-01
-6.71299219e-01 3.92847806e-01 -3.11624199e-01 5.04679903e-02
9.92325366e-01 -5.49614429e-02 -2.93624729e-01 -6.37704909e-01
-3.91371906e-01 2.46903852e-01 5.69455802e-01 6.98163748e-01
7.87427485e-01 -1.37921321e+00 -8.57656896e-01 3.49486768e-01
-1.26782328e-01 -2.08867192e-01 5.58190823e-01 7.11317062e-01
-6.91170335e-01 5.03689945e-01 -2.17791528e-01 -2.83326328e-01
-1.01774812e+00 9.13847327e-01 1.55438278e-02 -1.97183862e-01
-6.29193664e-01 1.08178961e+00 7.11435318e-01 -2.11348593e-01
1.00259848e-01 -2.86154270e-01 1.58494696e-01 -1.80319384e-01
6.73255205e-01 -1.27669901e-01 -4.82839614e-01 -4.89459008e-01
2.57178754e-01 7.40494013e-01 -3.99401456e-01 -4.79280293e-01
1.06662428e+00 -6.67582750e-01 -1.40964597e-01 1.24770269e-01
1.41834104e+00 2.33011946e-01 -1.66403413e+00 -1.90931767e-01
-4.19719934e-01 -6.50415778e-01 2.00743116e-02 -6.60956085e-01
-1.52250278e+00 1.34447634e+00 4.24840868e-01 -1.00409061e-01
1.33763397e+00 -1.21038482e-01 1.19687736e+00 2.35640094e-01
8.57093744e-03 -1.31597304e+00 4.98191804e-01 3.02957118e-01
9.94753838e-01 -1.18540120e+00 -3.64440918e-01 -3.57197523e-01
-1.14567173e+00 8.27587783e-01 8.76357317e-01 -1.84566826e-01
5.64791262e-01 2.50316679e-01 3.41782510e-01 2.12952152e-01
-1.04555845e+00 -5.51514551e-02 -7.19817355e-02 6.03707075e-01
4.42530364e-01 -1.72061995e-01 -2.91623175e-01 2.80236602e-01
8.18832219e-02 9.47713926e-02 5.35715044e-01 6.25144660e-01
-3.73200625e-01 -1.14272392e+00 8.85815620e-02 8.77546892e-02
-4.26779240e-01 -6.34079993e-01 -3.47709477e-01 5.93891621e-01
-8.03549737e-02 6.15588188e-01 7.90426135e-02 -3.57478112e-01
1.11080244e-01 -3.83631945e-01 2.95427024e-01 -2.97231197e-01
-5.42492270e-01 5.15337527e-01 -1.66386008e-01 -3.11818719e-01
-3.09794217e-01 -3.18814695e-01 -8.50610852e-01 -4.20639962e-01
-3.77109021e-01 9.46666226e-02 6.29884183e-01 5.93750596e-01
5.60229838e-01 4.82443869e-01 7.78851330e-01 -3.94918889e-01
-4.57524449e-01 -9.40489173e-01 -5.15548110e-01 4.86516356e-01
5.48440926e-02 5.15711792e-02 7.99332559e-03 4.89224911e-01] | [11.469698905944824, -0.2290440946817398] |
79c73c7a-9db1-42bf-8b6d-3293d7b17d7a | an-understanding-oriented-robust-machine | 2207.00187 | null | https://arxiv.org/abs/2207.00187v1 | https://arxiv.org/pdf/2207.00187v1.pdf | An Understanding-Oriented Robust Machine Reading Comprehension Model | Although existing machine reading comprehension models are making rapid progress on many datasets, they are far from robust. In this paper, we propose an understanding-oriented machine reading comprehension model to address three kinds of robustness issues, which are over sensitivity, over stability and generalization. Specifically, we first use a natural language inference module to help the model understand the accurate semantic meanings of input questions so as to address the issues of over sensitivity and over stability. Then in the machine reading comprehension module, we propose a memory-guided multi-head attention method that can further well understand the semantic meanings of input questions and passages. Third, we propose a multilanguage learning mechanism to address the issue of generalization. Finally, these modules are integrated with a multi-task learning based method. We evaluate our model on three benchmark datasets that are designed to measure models robustness, including DuReader (robust) and two SQuAD-related datasets. Extensive experiments show that our model can well address the mentioned three kinds of robustness issues. And it achieves much better results than the compared state-of-the-art models on all these datasets under different evaluation metrics, even under some extreme and unfair evaluations. The source code of our work is available at: https://github.com/neukg/RobustMRC. | ['Qi Ma', 'Chunchao Liu', 'Jiaqi Wang', 'Bingchao Wang', 'Shilei Liu', 'Bochao Li', 'Yongkang Liu', 'Feiliang Ren'] | 2022-07-01 | null | null | null | null | ['machine-reading-comprehension'] | ['natural-language-processing'] | [ 1.69583634e-01 3.73441502e-02 1.27279554e-02 -4.58927691e-01
-1.10199273e+00 -4.37037438e-01 4.77526754e-01 3.72045189e-01
-3.66169721e-01 3.77768040e-01 5.01360774e-01 -4.40130562e-01
-1.56008378e-01 -7.24976003e-01 -9.10100639e-01 -2.34562516e-01
4.84124839e-01 2.83413351e-01 5.33467352e-01 -5.52112281e-01
7.35278785e-01 -3.07091296e-01 -1.45252156e+00 4.02135670e-01
1.52460325e+00 9.05402958e-01 7.43518248e-02 6.49134517e-01
-3.93649749e-02 9.27611351e-01 -4.71624672e-01 -5.63827455e-01
-3.26535076e-01 -4.38261956e-01 -1.33409846e+00 -3.05719584e-01
4.69636291e-01 -3.23095471e-01 -1.13321587e-01 1.18243313e+00
5.96044481e-01 3.31210285e-01 5.32608688e-01 -1.02478051e+00
-1.31659448e+00 7.66749978e-01 -5.65337539e-01 3.42643768e-01
8.35372984e-01 2.75692999e-01 1.17831266e+00 -7.99504042e-01
1.83456158e-03 1.62603104e+00 3.97610575e-01 5.56174874e-01
-7.38583207e-01 -5.36021888e-01 5.25936842e-01 7.85884857e-01
-1.08567894e+00 -4.04932976e-01 6.79280937e-01 -8.37043598e-02
8.06238770e-01 4.25346553e-01 -9.82704610e-02 1.26943696e+00
3.00218642e-01 1.10677207e+00 1.21521759e+00 -3.58000278e-01
3.41647640e-02 -1.08758301e-01 8.42261016e-01 6.45505846e-01
-5.15463352e-02 -9.60474387e-02 -3.35555911e-01 1.67783037e-01
1.06144689e-01 -2.07116559e-01 -6.31730914e-01 1.53852314e-01
-1.20877111e+00 9.05502856e-01 4.77769554e-01 2.54315108e-01
1.33910647e-03 -2.29947597e-01 3.32923293e-01 4.74475652e-01
3.55263948e-01 5.14935374e-01 -6.84078813e-01 -4.18175980e-02
-3.83528709e-01 3.42666924e-01 7.44312346e-01 1.04049325e+00
3.45028251e-01 -3.74995470e-01 -5.29288292e-01 9.82620776e-01
5.27271807e-01 3.71737987e-01 7.49563456e-01 -5.90026319e-01
8.44019532e-01 7.33939469e-01 -2.26992935e-01 -1.14353216e+00
-6.68670177e-01 -4.51378316e-01 -9.15722668e-01 -4.14199799e-01
3.24518681e-01 1.14809223e-01 -6.65480554e-01 2.02478957e+00
8.49285796e-02 1.03407681e-01 2.57417440e-01 7.43252158e-01
1.27991450e+00 7.55200624e-01 9.58487913e-02 -8.04892741e-03
1.64089143e+00 -1.45747912e+00 -7.81588495e-01 -3.44441026e-01
4.95394677e-01 -6.42111599e-01 1.91861463e+00 3.88189673e-01
-1.15832472e+00 -6.73631966e-01 -1.01991463e+00 -4.42844778e-01
-5.78122079e-01 -1.38748407e-01 -7.66150057e-02 4.36504632e-01
-7.30260670e-01 2.63320744e-01 -3.44242096e-01 -3.93579572e-01
2.17042848e-01 -2.19795838e-01 -4.54413481e-02 -2.59093612e-01
-1.60244250e+00 1.14717805e+00 4.98526096e-01 8.08206294e-03
-8.04557681e-01 -5.23089826e-01 -9.48310316e-01 1.85337231e-01
6.38282895e-01 -7.95176685e-01 1.42912078e+00 -6.00274742e-01
-1.43652785e+00 6.89525127e-01 -1.59126803e-01 -2.41275541e-02
3.53081018e-01 -7.39202619e-01 -4.88825858e-01 -4.37366106e-02
5.20393513e-02 3.97208333e-01 6.75904751e-01 -1.12748468e+00
-2.91818976e-01 -4.72550064e-01 3.33796710e-01 2.53084064e-01
-2.54376024e-01 1.77697092e-02 -4.53798234e-01 -8.14620316e-01
1.04310207e-01 -6.12269998e-01 1.33014202e-01 -3.96715969e-01
-6.44301713e-01 -7.31001258e-01 5.02607942e-01 -9.89438355e-01
1.56283677e+00 -1.86449671e+00 3.72822046e-01 -3.01922113e-01
9.64210331e-02 3.17395627e-01 -5.53199768e-01 2.89199114e-01
3.60763841e-03 3.81001949e-01 -2.86738813e-01 -2.05100894e-01
4.40575182e-02 -8.58789831e-02 -4.44615066e-01 7.12691322e-02
2.61055946e-01 1.16142571e+00 -8.91559660e-01 -2.48723477e-01
-1.31331399e-01 7.49180466e-02 -4.84218866e-01 5.07295012e-01
-5.64466834e-01 3.05035561e-01 -5.61100602e-01 5.56986153e-01
7.96775281e-01 -4.48470354e-01 -4.32725579e-01 -5.80240674e-02
3.01569074e-01 5.84201694e-01 -9.58416998e-01 1.89780915e+00
-3.90755296e-01 1.24716088e-01 -4.42556709e-01 -8.23781312e-01
8.09545934e-01 1.13598667e-02 -5.51577866e-01 -1.01937938e+00
2.90058821e-01 4.12881142e-04 -1.96098164e-02 -9.07143235e-01
5.41684985e-01 6.82036877e-02 -2.07032755e-01 6.20908916e-01
-1.09120458e-01 -5.02371900e-02 1.13634299e-02 4.52834040e-01
8.86479318e-01 7.02038454e-03 1.88077867e-01 -4.20729339e-01
9.59520638e-01 -4.11718547e-01 4.81588036e-01 7.71580577e-01
-2.55994320e-01 7.00474620e-01 6.27452314e-01 -8.71926770e-02
-4.93311554e-01 -1.04441500e+00 7.69431517e-02 1.57786620e+00
4.75017548e-01 -4.60772395e-01 -9.66525495e-01 -7.03825653e-01
-3.38259608e-01 1.17614472e+00 -6.76748157e-01 -5.54981112e-01
-4.44896400e-01 -8.23020756e-01 6.59939826e-01 5.89307666e-01
9.26028609e-01 -9.99971688e-01 -4.01975095e-01 -6.39683381e-02
-7.16314375e-01 -9.37101662e-01 -5.28962553e-01 -3.49331409e-01
-5.81094444e-01 -1.17948353e+00 -3.84246945e-01 -8.25846136e-01
3.94073367e-01 3.70864719e-01 1.27026582e+00 5.48524737e-01
2.57546097e-01 2.15313032e-01 -8.39320898e-01 -6.42672598e-01
-4.09412444e-01 2.88481146e-01 -1.50812343e-01 -2.14239955e-01
3.95410866e-01 -2.85548389e-01 -3.31594080e-01 5.21241844e-01
-1.09805405e+00 2.77989537e-01 5.20165026e-01 7.88811743e-01
3.50987881e-01 -1.02833726e-01 1.07110918e+00 -6.95462525e-01
1.18998218e+00 -7.37718821e-01 -2.44456947e-01 8.36490691e-01
-5.79351604e-01 2.20247105e-01 7.48642802e-01 -3.76659602e-01
-1.05145192e+00 -7.48541534e-01 -3.70955348e-01 6.96401447e-02
-2.18151718e-01 6.09195352e-01 -6.67302251e-01 4.17670071e-01
4.98445272e-01 3.43273461e-01 -1.24994606e-01 -7.23022103e-01
5.03647506e-01 6.74402952e-01 5.67802489e-01 -8.12053740e-01
7.91337311e-01 -2.02959985e-01 -7.11732090e-01 -5.43026507e-01
-1.45762527e+00 -1.25982508e-01 -2.73769885e-01 6.10041730e-02
7.82649279e-01 -7.13538408e-01 -6.52706623e-01 8.13737273e-01
-1.31264365e+00 -1.73506334e-01 2.60101885e-01 7.98638836e-02
-5.69098055e-01 5.27837455e-01 -6.60321295e-01 -4.49292541e-01
-5.70368946e-01 -1.28819144e+00 8.20556164e-01 5.46547472e-01
-2.35826209e-01 -9.84697104e-01 -1.14517428e-01 9.18276072e-01
6.56875789e-01 -3.37155089e-02 1.48506498e+00 -9.76678312e-01
-4.69552040e-01 6.96851462e-02 -3.59295696e-01 3.63493860e-01
-7.16394931e-02 -1.74412593e-01 -9.67537761e-01 -2.74822563e-01
2.59210974e-01 -7.34711528e-01 1.00325549e+00 -8.41678865e-03
1.69016552e+00 -4.20188457e-01 1.18628796e-02 4.72064435e-01
1.04551554e+00 -1.90435350e-01 8.60318184e-01 5.17811954e-01
5.47902703e-01 6.65399492e-01 6.93583369e-01 -1.70275401e-02
1.05318630e+00 4.48620021e-01 4.44599152e-01 2.59300143e-01
8.99380725e-03 -6.81810200e-01 4.12720293e-01 1.20444548e+00
2.64532298e-01 -5.30526757e-01 -9.95439768e-01 4.26532388e-01
-1.95916116e+00 -6.98009133e-01 -2.16636911e-01 1.91281438e+00
9.14463282e-01 2.28578806e-01 -1.24993317e-01 1.45247594e-01
5.96587658e-01 2.82595724e-01 -7.33155191e-01 -5.34878552e-01
-2.43185386e-01 -1.13917626e-01 -1.61151186e-01 6.34179533e-01
-1.02384198e+00 1.03568542e+00 5.70053673e+00 9.19760287e-01
-5.39666593e-01 1.33728668e-01 7.29934394e-01 2.09939748e-01
-6.12285733e-01 -8.32069144e-02 -6.58508658e-01 4.26688135e-01
8.15720677e-01 -2.68836856e-01 2.99141169e-01 3.40942264e-01
7.59309381e-02 -1.83926478e-01 -1.03182459e+00 6.37995958e-01
5.04985750e-01 -9.25235689e-01 4.24832046e-01 -6.11684978e-01
3.61412585e-01 -1.76419660e-01 1.86335802e-01 5.86213589e-01
7.52100199e-02 -1.22792828e+00 6.26525104e-01 7.63354719e-01
2.99589336e-01 -7.79581845e-01 8.13461125e-01 8.43939543e-01
-7.85200179e-01 -2.01078236e-01 -4.49812144e-01 -2.22176984e-01
8.78367200e-02 2.93581873e-01 5.37625402e-02 7.05932856e-01
8.93711090e-01 5.80907524e-01 -1.14877391e+00 7.89592922e-01
-8.04736912e-01 6.93036139e-01 1.08977869e-01 -3.29758257e-01
1.06523842e-01 1.78288922e-01 3.77219617e-01 8.98967326e-01
7.45018348e-02 4.02150303e-01 -7.25988671e-02 1.03884518e+00
-2.11900115e-01 2.90563673e-01 -2.21979275e-01 3.19703937e-01
5.15241325e-01 9.62788165e-01 -2.05611810e-01 -2.78864115e-01
-4.90453690e-01 9.06662405e-01 7.69833863e-01 2.83096910e-01
-9.49043870e-01 -6.38396204e-01 3.26597333e-01 -1.08273745e-01
-1.13941379e-01 4.42707650e-02 -5.01814246e-01 -1.55603063e+00
2.95325696e-01 -1.54700470e+00 6.76769435e-01 -9.86173570e-01
-1.39692760e+00 4.55135107e-01 -5.93747608e-02 -8.39115500e-01
1.62476018e-01 -4.74240094e-01 -9.77898061e-01 6.79649413e-01
-1.75052786e+00 -1.08589542e+00 -3.66359919e-01 5.36278486e-01
9.32703495e-01 -6.45211488e-02 7.43633270e-01 4.22190689e-02
-9.47422147e-01 9.73237336e-01 -1.18768834e-01 1.21580221e-01
7.55925834e-01 -1.13342357e+00 6.83843970e-01 1.05055642e+00
-1.30361378e-01 8.64327192e-01 5.26371181e-01 -4.01195735e-01
-1.29093850e+00 -9.23226118e-01 1.06891251e+00 -7.85868943e-01
6.32156193e-01 -1.52456805e-01 -1.63233042e+00 6.77062452e-01
6.08176172e-01 -7.01348007e-01 6.71547890e-01 2.14107320e-01
-7.42497325e-01 2.06921443e-01 -8.99155259e-01 5.23338079e-01
1.00896788e+00 -3.76928180e-01 -1.32342947e+00 3.66469324e-01
1.29521465e+00 -4.67599809e-01 -6.91510201e-01 5.59170544e-01
1.43258691e-01 -9.61214900e-01 7.73896456e-01 -8.72196734e-01
7.65418231e-01 -1.91786453e-01 -1.99137956e-01 -1.48232043e+00
-3.66035312e-01 -3.85714561e-01 -3.93279083e-02 1.39742386e+00
6.77756786e-01 -6.93264425e-01 3.70722152e-02 6.17699623e-01
-1.18199654e-01 -8.79527450e-01 -6.92865551e-01 -6.20928049e-01
6.36892915e-01 -3.67419869e-01 7.69973576e-01 8.82129490e-01
1.32096067e-01 9.21427369e-01 -2.90839940e-01 2.77702123e-01
3.83860826e-01 1.97886705e-01 6.63413227e-01 -9.49200511e-01
-2.39721924e-01 -6.44526064e-01 1.24223776e-01 -1.44697690e+00
3.03482801e-01 -8.46849442e-01 2.59053148e-02 -1.53988004e+00
4.32348102e-01 -1.19069917e-02 -3.93003643e-01 4.87838119e-01
-9.76905525e-01 -4.15863693e-01 2.15286776e-01 -6.57968074e-02
-9.93005812e-01 8.30952942e-01 1.35075521e+00 -1.51435614e-01
1.03861734e-01 -4.34773006e-02 -1.27977955e+00 7.00970888e-01
1.13031507e+00 -1.73096552e-01 -5.46235383e-01 -1.02500355e+00
2.81556785e-01 -1.54898837e-01 4.34848636e-01 -7.86433816e-01
3.25394988e-01 -1.56427130e-01 7.02861696e-02 -6.20571256e-01
-1.07642435e-01 -2.90854454e-01 -6.21833324e-01 4.28376287e-01
-8.68281841e-01 4.73304451e-01 2.69316524e-01 5.01389444e-01
-1.44986346e-01 -2.27870673e-01 8.82821262e-01 4.28364947e-02
-7.08008289e-01 1.04744405e-01 -4.84442823e-02 6.67164803e-01
7.85769761e-01 3.23023379e-01 -9.19731438e-01 -6.50304019e-01
-3.57176632e-01 8.80187035e-01 2.46002048e-01 1.13792408e+00
8.74366820e-01 -1.12176454e+00 -1.12268090e+00 2.01141521e-01
5.55884302e-01 6.45115376e-02 4.38916802e-01 6.95301116e-01
-1.36545181e-01 3.92716229e-01 5.69208199e-03 -3.23455155e-01
-1.02358484e+00 7.72880554e-01 3.88183177e-01 -2.55306244e-01
-3.66123557e-01 9.78436470e-01 1.56979054e-01 -8.41439009e-01
3.89764696e-01 -4.76964891e-01 -5.00524879e-01 -1.28036320e-01
1.03049576e+00 4.39864784e-01 -6.95933774e-02 -5.14810503e-01
-2.69232094e-01 8.53225887e-01 -3.45669776e-01 3.87971014e-01
9.75307703e-01 -4.72688675e-01 -2.09196046e-01 3.74521881e-01
1.09189057e+00 -2.21691027e-01 -8.30820501e-01 -4.81030434e-01
2.05722645e-01 -2.18067542e-01 -2.50799954e-01 -1.20635533e+00
-7.11941600e-01 1.23759365e+00 2.66101897e-01 1.18148081e-01
1.35981846e+00 1.56988353e-01 9.98536527e-01 4.35931176e-01
-2.12391522e-02 -1.04969132e+00 2.93136597e-01 9.25934315e-01
1.23038101e+00 -1.52882421e+00 -2.28168502e-01 -3.14952493e-01
-7.10756063e-01 8.35208714e-01 1.13838458e+00 2.32950777e-01
3.06896746e-01 -2.65374839e-01 1.71946153e-01 -6.76183254e-02
-1.03303647e+00 -7.00773969e-02 5.01418173e-01 3.87268752e-01
4.86116081e-01 3.05191558e-02 -4.55053985e-01 1.07116222e+00
-3.73755872e-01 -2.86314160e-01 5.49828470e-01 6.00880384e-01
-7.87935913e-01 -8.21277201e-01 -3.32094789e-01 2.35022545e-01
-3.50031406e-01 -2.77309090e-01 -4.57822025e-01 5.29885411e-01
-3.55232447e-01 1.45131481e+00 -3.92601967e-01 -6.03572786e-01
7.28744924e-01 8.65847692e-02 1.63934454e-01 -4.97498959e-01
-5.26155114e-01 -4.73777860e-01 -4.60812338e-02 -6.18244886e-01
-1.24907501e-01 -3.14814925e-01 -1.24952400e+00 -4.31433976e-01
-3.86409700e-01 7.92541951e-02 1.95577800e-01 1.15167749e+00
3.00961971e-01 5.62069476e-01 5.15589416e-01 -9.23834294e-02
-1.03674734e+00 -1.14151132e+00 -4.99340408e-02 5.57181001e-01
3.21840614e-01 -3.97562236e-01 -4.42248106e-01 -3.45448494e-01] | [11.266475677490234, 8.16019058227539] |
a479b21f-b69a-406a-90ff-c4ead739be29 | all-points-matter-entropy-regularized | 2305.15832 | null | https://arxiv.org/abs/2305.15832v1 | https://arxiv.org/pdf/2305.15832v1.pdf | All Points Matter: Entropy-Regularized Distribution Alignment for Weakly-supervised 3D Segmentation | Pseudo-labels are widely employed in weakly supervised 3D segmentation tasks where only sparse ground-truth labels are available for learning. Existing methods often rely on empirical label selection strategies, such as confidence thresholding, to generate beneficial pseudo-labels for model training. This approach may, however, hinder the comprehensive exploitation of unlabeled data points. We hypothesize that this selective usage arises from the noise in pseudo-labels generated on unlabeled data. The noise in pseudo-labels may result in significant discrepancies between pseudo-labels and model predictions, thus confusing and affecting the model training greatly. To address this issue, we propose a novel learning strategy to regularize the generated pseudo-labels and effectively narrow the gaps between pseudo-labels and model predictions. More specifically, our method introduces an Entropy Regularization loss and a Distribution Alignment loss for weakly supervised learning in 3D segmentation tasks, resulting in an ERDA learning strategy. Interestingly, by using KL distance to formulate the distribution alignment loss, it reduces to a deceptively simple cross-entropy-based loss which optimizes both the pseudo-label generation network and the 3D segmentation network simultaneously. Despite the simplicity, our method promisingly improves the performance. We validate the effectiveness through extensive experiments on various baselines and large-scale datasets. Results show that ERDA effectively enables the effective usage of all unlabeled data points for learning and achieves state-of-the-art performance under different settings. Remarkably, our method can outperform fully-supervised baselines using only 1% of true annotations. Code and model will be made publicly available. | ['DaCheng Tao', 'Chaoyue Wang', 'Shanshan Zhao', 'Zhe Chen', 'Liyao Tang'] | 2023-05-25 | null | null | null | null | ['pseudo-label'] | ['miscellaneous'] | [ 3.03898931e-01 4.45312023e-01 -5.46106994e-01 -5.88713288e-01
-1.24770129e+00 -8.17863166e-01 4.16656286e-01 -2.62361411e-02
-3.18260372e-01 6.79582715e-01 -2.01371461e-01 -2.04010651e-01
4.25250083e-01 -3.34729940e-01 -7.49423802e-01 -1.00189972e+00
3.69738489e-01 5.34721851e-01 4.08035517e-02 4.42991316e-01
1.35178089e-01 2.99727917e-01 -1.32057035e+00 -1.49832547e-01
1.08448434e+00 1.09841847e+00 8.33424404e-02 8.52485597e-02
-1.12062789e-01 6.55347645e-01 -5.61259985e-01 -5.23698449e-01
5.92579305e-01 -3.53718519e-01 -8.95897627e-01 4.64426070e-01
5.46692848e-01 -3.60359460e-01 2.08932068e-03 1.19289839e+00
4.49449450e-01 3.19606960e-02 9.68081415e-01 -1.06970918e+00
-3.99061531e-01 4.92761821e-01 -8.71246159e-01 -3.53306621e-01
2.44668107e-02 2.72980303e-01 1.10137153e+00 -9.52868760e-01
5.65456212e-01 1.13535762e+00 7.44547367e-01 7.45581746e-01
-1.52446759e+00 -7.23402560e-01 1.64484128e-01 -3.15595806e-01
-1.45470452e+00 -2.29811311e-01 8.87384355e-01 -4.84660387e-01
3.48461211e-01 3.19160782e-02 3.28581125e-01 1.19048309e+00
-3.76030803e-01 1.13200986e+00 1.30615139e+00 -4.05318886e-01
2.55083650e-01 2.96405286e-01 1.80719018e-01 8.39989901e-01
1.41753182e-01 2.35856369e-01 -4.27253604e-01 -9.04244035e-02
6.92518234e-01 -3.89712244e-01 -1.00104965e-01 -6.64992273e-01
-9.56440508e-01 7.39023030e-01 3.54339331e-01 -1.71199679e-01
-8.28976631e-02 -1.91809237e-01 2.96307951e-01 1.48901844e-03
9.05940533e-01 3.79825741e-01 -6.39398336e-01 6.84760809e-02
-9.83012438e-01 9.99957323e-02 5.89627326e-01 1.00899231e+00
9.91248190e-01 -2.66360790e-01 -3.48797113e-01 1.04084539e+00
4.35482621e-01 5.10779917e-01 1.88386172e-01 -1.12572336e+00
5.15066922e-01 6.59148991e-01 1.03518844e-01 -6.97415769e-01
-2.43899241e-01 -6.16010606e-01 -6.97499752e-01 9.73434895e-02
7.17149854e-01 -2.12194636e-01 -1.28181922e+00 1.89920115e+00
7.02907085e-01 4.29131061e-01 -1.61330745e-01 1.03271067e+00
5.34712553e-01 2.32681468e-01 1.46290600e-01 -2.21467376e-01
9.03519392e-01 -1.10888982e+00 -4.75415885e-01 -2.70558000e-01
9.25827563e-01 -6.46814048e-01 1.39791143e+00 1.43906087e-01
-8.90094221e-01 -4.47140187e-01 -9.45297480e-01 -5.73103316e-02
3.06857470e-02 2.87778407e-01 4.00954068e-01 5.84989488e-01
-6.55904531e-01 5.85634351e-01 -9.55275476e-01 3.91406678e-02
8.95290554e-01 2.37273142e-01 -1.99811995e-01 -6.00765757e-02
-8.72028172e-01 6.08937979e-01 4.71544355e-01 9.80703719e-03
-1.04933929e+00 -8.08107913e-01 -7.69148231e-01 -3.15099269e-01
6.77326977e-01 -3.31789523e-01 1.28211212e+00 -8.09698582e-01
-1.48372209e+00 1.48224902e+00 -6.92864368e-03 -3.76226574e-01
8.38825643e-01 -1.50267422e-01 1.63623527e-01 5.21603562e-02
3.70787591e-01 1.05297244e+00 9.53716993e-01 -1.73167157e+00
-5.95972717e-01 -3.57985079e-01 -9.18288752e-02 3.24604303e-01
-1.74997389e-01 -6.23330235e-01 -6.33472800e-01 -7.32007504e-01
4.44137037e-01 -1.37970376e+00 -3.10235351e-01 5.37472703e-02
-8.47908020e-01 -3.05600107e-01 6.43507302e-01 -1.75076589e-01
9.21671510e-01 -2.11743069e+00 -6.68455139e-02 1.86625123e-01
1.85472414e-01 3.35457474e-01 -1.51316762e-01 -3.15983087e-01
1.40856788e-01 3.46544653e-01 -6.44641399e-01 -6.58542991e-01
1.98222813e-03 3.37395519e-01 -3.11200321e-01 5.57656169e-01
2.83396453e-01 9.81523514e-01 -1.01371884e+00 -7.58527398e-01
2.14208096e-01 2.93262482e-01 -3.52864027e-01 3.18231851e-01
-4.25197870e-01 8.51972461e-01 -5.97999632e-01 8.07246864e-01
8.51826727e-01 -5.29661417e-01 -8.90095718e-03 -1.03253387e-01
3.95806551e-01 2.93185383e-01 -9.10451412e-01 1.86541200e+00
-3.80470484e-01 1.33319154e-01 2.66504418e-02 -1.05359900e+00
9.76692736e-01 8.28182548e-02 3.91046435e-01 -3.30809534e-01
1.06212735e-01 4.71885353e-01 -4.07270819e-01 -2.75354564e-01
8.60936269e-02 -3.23930651e-01 -6.37518093e-02 5.94243824e-01
4.70746942e-02 -4.53056842e-01 -1.99694410e-02 4.60941494e-02
7.34223664e-01 6.29007876e-01 -1.08863525e-01 -1.37865767e-01
3.71407300e-01 7.21977055e-02 7.76470304e-01 7.71014512e-01
-4.78379309e-01 7.08791256e-01 5.46567559e-01 -1.45293415e-01
-8.60587418e-01 -1.01870680e+00 -3.24360311e-01 8.71975124e-01
4.34935987e-01 -1.05712622e-01 -8.70858550e-01 -1.54907537e+00
4.48741876e-02 8.72589409e-01 -6.30518794e-01 -2.08480731e-01
-3.14258367e-01 -7.72153437e-01 6.23219967e-01 2.88250893e-01
4.75750238e-01 -8.68923366e-01 -7.49261752e-02 -9.10773501e-02
-2.55774438e-01 -1.15997076e+00 -5.89529872e-01 4.41403896e-01
-1.15273821e+00 -9.95046139e-01 -5.58109879e-01 -7.29990900e-01
1.06895173e+00 2.47467771e-01 1.13159430e+00 2.60041915e-02
7.52588958e-02 4.39114496e-02 -3.19410563e-01 -1.91185310e-01
-5.91849089e-01 4.40330178e-01 -5.65507784e-02 -1.26308620e-01
5.19413769e-01 -3.65355372e-01 -4.31735635e-01 6.24429286e-01
-7.75986850e-01 9.02615264e-02 4.67542738e-01 1.02181637e+00
1.04960740e+00 -1.70546815e-01 6.50374234e-01 -1.30745685e+00
1.02424428e-01 -3.79690945e-01 -7.16339052e-01 2.13635460e-01
-9.09761846e-01 1.73503146e-01 5.44685066e-01 -5.53519249e-01
-1.14478481e+00 3.78382295e-01 -6.42184839e-02 -7.42350221e-01
-3.74309301e-01 2.84870178e-01 -2.15432584e-01 -2.82119326e-02
6.21243596e-01 -1.00388508e-02 5.62838232e-03 -5.58438480e-01
3.74605745e-01 6.78508461e-01 3.08841556e-01 -7.40037799e-01
9.83725071e-01 5.22614419e-01 1.46378323e-01 -3.82477909e-01
-1.69751036e+00 -3.54837090e-01 -7.75607705e-01 -2.79137008e-02
6.49010122e-01 -1.03322673e+00 -2.60780454e-01 5.82363605e-01
-8.02110255e-01 -5.69803774e-01 -3.83701682e-01 3.83154780e-01
-5.22503138e-01 4.19258177e-01 -6.01060152e-01 -8.01817179e-01
-3.42804007e-02 -1.20718300e+00 1.41097462e+00 7.72031993e-02
-1.03871718e-01 -1.08712101e+00 -1.04417384e-01 7.64489114e-01
-1.66748151e-01 2.63819367e-01 8.11070025e-01 -8.07837963e-01
-5.54889739e-01 -2.35926867e-01 -3.22343349e-01 7.00930059e-01
2.60293305e-01 -2.46329531e-01 -1.25153875e+00 -2.16516539e-01
-7.25345910e-02 -9.86166477e-01 1.00284863e+00 3.48225445e-01
1.49042022e+00 -4.81162220e-02 -3.41835767e-01 6.02762938e-01
1.08092594e+00 -2.94469535e-01 2.57850826e-01 4.44695493e-03
1.04807484e+00 8.53382945e-01 1.12369549e+00 2.99741417e-01
3.47346693e-01 5.59902668e-01 4.57946867e-01 -1.40354320e-01
-1.62554905e-01 -6.13601744e-01 9.71349701e-02 6.31566882e-01
2.63295919e-01 -1.94992289e-01 -8.54018927e-01 3.90816838e-01
-1.78757930e+00 -4.96224701e-01 -3.51223126e-02 2.42703271e+00
1.32082069e+00 3.62389714e-01 -6.31475449e-02 -9.53838229e-02
8.33187401e-01 1.50564194e-01 -1.04220033e+00 1.42842233e-01
-1.34623125e-01 -9.07279700e-02 6.05437338e-01 5.47149062e-01
-1.46787930e+00 1.22407901e+00 5.85443115e+00 1.24020028e+00
-1.06163347e+00 1.49428040e-01 1.21925545e+00 -9.48458761e-02
-3.35390329e-01 -8.71497393e-02 -9.66412544e-01 5.45395315e-01
4.77274150e-01 3.15532237e-01 9.96559337e-02 9.60809588e-01
2.73949176e-01 -3.19930501e-02 -1.14047873e+00 1.01137745e+00
-1.39476672e-01 -9.41793740e-01 -4.18743081e-02 1.42093763e-01
1.09936035e+00 5.90423010e-02 1.58390597e-01 2.29202539e-01
3.98302794e-01 -8.61680865e-01 8.79120290e-01 3.92176285e-02
1.00449407e+00 -6.56694829e-01 6.87184393e-01 4.44966704e-01
-7.36741245e-01 1.56174883e-01 -2.91099757e-01 1.17119618e-01
1.98010325e-01 1.01555598e+00 -8.62627506e-01 3.36643249e-01
4.46247816e-01 8.38173091e-01 -4.30078954e-01 6.57300413e-01
-6.03702068e-01 9.15318012e-01 -4.37082320e-01 2.20485479e-01
3.88959885e-01 -2.97226399e-01 5.33233345e-01 9.19473350e-01
3.65919359e-02 -1.77230641e-01 4.61440563e-01 1.03293073e+00
-4.45661783e-01 2.09995229e-02 -4.12726402e-01 7.12216496e-02
6.48112893e-01 1.22029757e+00 -8.17595303e-01 -1.43416330e-01
-1.65556416e-01 8.43490481e-01 4.85247433e-01 4.14963633e-01
-8.61826420e-01 1.43112496e-01 4.34110075e-01 -1.64225344e-02
1.35689735e-01 8.31825100e-03 -7.06032574e-01 -1.25216568e+00
8.24087113e-02 -6.60823882e-01 2.37534136e-01 -3.74585718e-01
-1.63843167e+00 3.19388717e-01 -4.69702594e-02 -1.36631954e+00
-1.34210140e-01 -4.10342783e-01 -2.10328877e-01 6.91685081e-01
-1.57671487e+00 -1.25836217e+00 -9.29810032e-02 1.19617812e-01
4.74747390e-01 7.40787759e-02 5.62063098e-01 2.23884970e-01
-7.21140683e-01 8.03131759e-01 -3.38599794e-02 2.12555993e-02
9.45246577e-01 -1.52339792e+00 2.55487800e-01 6.45722985e-01
2.33214602e-01 2.03400701e-01 5.05238116e-01 -7.00096846e-01
-6.49930120e-01 -1.27816904e+00 6.61041915e-01 -5.25740385e-01
3.77129167e-01 -4.71232355e-01 -1.00296640e+00 5.27776480e-01
-4.35399801e-01 2.50714749e-01 8.16552401e-01 -2.14819983e-02
-4.56203371e-01 1.02283724e-01 -1.34454703e+00 5.79695821e-01
1.22563434e+00 -4.33624208e-01 -3.37097675e-01 4.43162531e-01
8.13372314e-01 -6.31086111e-01 -6.80456877e-01 6.82372093e-01
3.15351427e-01 -8.52930427e-01 8.49234283e-01 -3.13746005e-01
4.91378397e-01 -3.64267766e-01 1.50241470e-02 -1.19212270e+00
2.70341039e-02 -4.84825343e-01 -1.50430053e-01 1.46787405e+00
7.17460334e-01 -4.04068619e-01 1.23790812e+00 8.62450421e-01
-8.98674205e-02 -1.06181836e+00 -8.78599405e-01 -6.98461056e-01
2.42578432e-01 -4.50550705e-01 3.82711679e-01 1.11955094e+00
-4.22614783e-01 3.70922595e-01 -3.35136443e-01 3.34432945e-02
1.01223457e+00 1.45503938e-01 6.61415577e-01 -1.24355674e+00
-2.01095074e-01 -3.36571068e-01 1.27033433e-02 -1.52383435e+00
6.38652742e-01 -1.11051452e+00 3.66851658e-01 -1.04238033e+00
1.26216456e-01 -1.17848182e+00 -2.25510672e-01 8.36914659e-01
-4.31721061e-01 5.54488599e-01 -4.30137850e-02 4.67924476e-01
-6.87031984e-01 6.94423676e-01 1.62974918e+00 -1.37261003e-01
-3.83558631e-01 2.04030067e-01 -6.58244193e-01 8.53517950e-01
7.67138541e-01 -7.12523639e-01 -5.33891737e-01 -3.29848856e-01
-1.09672928e-02 -3.80267709e-01 3.40060174e-01 -4.93787289e-01
-2.23605856e-01 -9.36182067e-02 2.42487162e-01 -5.50647438e-01
1.32073149e-01 -7.25929022e-01 -3.39643240e-01 1.59436911e-01
-5.20912826e-01 -8.96509051e-01 -1.76195167e-02 6.97480381e-01
-1.37289986e-01 -4.87677932e-01 9.53009069e-01 -2.84332521e-02
-3.67725492e-01 4.98782039e-01 2.28791729e-01 4.34192777e-01
9.82037425e-01 -2.42439151e-01 -5.18782511e-02 -1.72153294e-01
-6.39448464e-01 4.82802212e-01 7.56696582e-01 1.96486011e-01
1.78560719e-01 -1.23808444e+00 -6.08295798e-01 1.22378133e-01
1.40069246e-01 6.75846457e-01 8.17316771e-02 6.85824394e-01
-2.61447936e-01 1.29814878e-01 3.23776990e-01 -9.34119165e-01
-9.72586274e-01 3.45280081e-01 3.31039369e-01 -4.28097963e-01
-3.77712607e-01 1.01152146e+00 3.94173920e-01 -8.17241728e-01
4.07072097e-01 -6.82465881e-02 1.38564423e-01 -9.59847867e-03
8.64896923e-02 2.95547038e-01 -1.38141036e-01 -6.35609388e-01
-2.67852306e-01 6.65292263e-01 -2.10873201e-01 7.27467313e-02
1.07090378e+00 -4.27184969e-01 1.42304704e-01 5.12919366e-01
1.31949246e+00 -1.31557077e-01 -1.94522202e+00 -3.15450102e-01
4.16624546e-02 -4.85246688e-01 6.13463297e-02 -9.16688740e-01
-1.12176442e+00 8.89634609e-01 4.10104334e-01 8.30375850e-02
1.02232194e+00 2.33454689e-01 8.75855148e-01 2.03960240e-01
3.88999283e-01 -1.16325331e+00 1.73344523e-01 3.62735480e-01
3.75852168e-01 -1.75994110e+00 -1.41187355e-01 -7.67003655e-01
-8.29263628e-01 6.64952457e-01 8.27293396e-01 6.15750700e-02
5.66745281e-01 4.15828283e-04 5.00200152e-01 6.07781932e-02
-3.62513930e-01 -1.98188186e-01 2.84807742e-01 5.05602479e-01
2.81885654e-01 1.12724425e-02 -1.74928471e-01 5.64397037e-01
2.84034926e-02 -1.41368032e-01 1.73217058e-01 6.34768128e-01
-1.75347239e-01 -1.32203686e+00 -1.87634364e-01 5.66223443e-01
-5.32199144e-01 8.92611369e-02 -4.04591203e-01 5.38001120e-01
2.39981949e-01 9.19746399e-01 -1.04088828e-01 -3.37706268e-01
4.00900990e-02 2.04901010e-01 3.92115176e-01 -7.70052314e-01
-2.43467197e-01 2.83923447e-01 -8.69180337e-02 -3.42269748e-01
-6.60067022e-01 -5.65850556e-01 -1.35527325e+00 1.34502416e-02
-7.20931292e-01 3.14795524e-02 4.12200928e-01 9.94536161e-01
4.04635429e-01 7.05397576e-02 8.39242339e-01 -8.19304824e-01
-8.92101943e-01 -9.03925180e-01 -7.42136002e-01 7.20495522e-01
2.56287813e-01 -8.93415451e-01 -6.90978944e-01 7.52337649e-02] | [9.443851470947266, 1.2663640975952148] |
c599fa29-6eb0-4310-aa01-05bca80c5759 | unsupervised-algorithm-for-disaggregating-low | 1908.10713 | null | https://arxiv.org/abs/1908.10713v1 | https://arxiv.org/pdf/1908.10713v1.pdf | Unsupervised algorithm for disaggregating low-sampling-rate electricity consumption of households | Non-intrusive load monitoring (NILM) has been extensively researched over the last decade. The objective of NILM is to identify the power consumption of individual appliances and to detect when particular devices are on or off from measuring the power consumption of an entire house. This information allows households to receive customized advice on how to better manage their electrical consumption. In this paper, we present an alternative NILM method that breaks down the aggregated power signal into categories of appliances. The ultimate goal is to use this approach for demand-side management to estimate potential flexibility within the electricity consumption of households. Our method is implemented as an algorithm combining NILM and load profile simulation. This algorithm, based on a Markov model, allocates an activity chain to each inhabitant of the household, deduces from the whole-house power measurement and statistical data the appliance usage, generate the power profile accordingly and finally returns the share of energy consumed by each appliance category over time. To analyze its performance, the algorithm was benchmarked against several state-of-the-art NILM algorithms and tested on three public datasets. The proposed algorithm is unsupervised; hence it does not require any labeled data, which are expensive to acquire. Although better performance is shown for the supervised algorithms, our proposed unsupervised algorithm achieves a similar range of uncertainty while saving on the cost of acquiring labeled data. Additionally, our method requires lower computational power compared to most of the tested NILM algorithms. It was designed for low-sampling-rate power measurement (every 15 min), which corresponds to the frequency range of most common smart meters. | ['Jordan Holweger', 'Marina Dorokhova', 'Lionel Bloch', 'Christophe Ballif', 'Nicolas Wyrsch'] | 2019-08-20 | null | null | null | null | ['non-intrusive-load-monitoring', 'non-intrusive-load-monitoring', 'non-intrusive-load-monitoring'] | ['knowledge-base', 'miscellaneous', 'time-series'] | [ 3.10167491e-01 1.96309522e-01 -1.74083650e-01 -2.80232042e-01
-4.29346353e-01 -5.93318939e-01 5.39396286e-01 3.47503424e-01
-1.00887872e-01 7.68535674e-01 -3.78009155e-02 -2.14863151e-01
-3.46917599e-01 -1.17676795e+00 -2.06906900e-01 -1.04850650e+00
3.84842642e-02 6.79337919e-01 -1.24328233e-01 4.07169044e-01
-1.82955563e-01 5.29028296e-01 -1.57594335e+00 -2.71049321e-01
8.32030118e-01 1.03065157e+00 2.41989702e-01 4.23208445e-01
3.73914063e-01 5.64806342e-01 -6.35437369e-01 1.60209566e-01
-9.42485705e-02 -5.04574716e-01 -5.50123990e-01 2.31886849e-01
-4.88328010e-01 -4.48669910e-01 3.28198671e-01 9.14812982e-01
4.83518809e-01 6.17627464e-02 8.73107433e-01 -1.48287249e+00
3.52188021e-01 9.94149566e-01 -3.80534947e-01 5.30887805e-02
5.17447889e-01 1.18537001e-01 7.04720199e-01 2.44361550e-01
-1.22145183e-01 7.03750193e-01 3.21589708e-01 -1.17610618e-02
-1.55662096e+00 -6.03971839e-01 -1.43869221e-01 1.96135506e-01
-1.46722794e+00 -1.38082087e-01 1.00127661e+00 -3.28948647e-01
1.10048330e+00 6.43301010e-01 8.79403472e-01 8.29129159e-01
3.70315015e-02 6.22437835e-01 1.36359131e+00 -5.55263340e-01
8.94311845e-01 3.11725080e-01 -4.29137088e-02 1.43211842e-01
5.52475870e-01 -1.41737759e-01 -7.82467518e-03 -3.87939751e-01
-1.17772490e-01 7.74663538e-02 -2.09543675e-01 -1.87382773e-01
-8.56544375e-01 5.14385402e-01 -5.55903465e-02 6.89910769e-01
-6.11525774e-01 -1.13828950e-01 3.13811481e-01 -2.58590877e-01
3.70464951e-01 4.50190008e-02 -4.38438535e-01 -2.68189937e-01
-1.34040380e+00 -1.24200910e-01 1.15186346e+00 6.33150458e-01
9.17171478e-01 -9.79022533e-02 2.52762735e-02 2.73400128e-01
5.78208089e-01 6.85621798e-01 5.17102838e-01 -7.16460228e-01
1.92023039e-01 8.85010839e-01 2.34781474e-01 -7.66115546e-01
-8.72236311e-01 -7.19371364e-02 -1.09734881e+00 -5.76892635e-03
2.08039612e-01 -4.55150574e-01 -2.18521908e-01 1.55241632e+00
3.64851326e-01 2.29339991e-02 -7.61355460e-02 3.10872644e-01
1.22326598e-01 8.09049606e-01 3.88735682e-01 -8.79166186e-01
1.60250235e+00 -8.77690166e-02 -8.49236310e-01 3.05919081e-01
5.17492175e-01 -3.62034768e-01 5.16467988e-01 3.90345395e-01
-7.95213580e-01 -2.27681115e-01 -1.07403207e+00 7.27156460e-01
-5.40087879e-01 1.90867811e-01 3.02762955e-01 9.19024408e-01
-5.65898955e-01 7.78111041e-01 -1.04944003e+00 -6.64565444e-01
2.35921502e-01 4.89695609e-01 5.97513914e-02 6.08868182e-01
-9.38960850e-01 8.48389447e-01 7.33681679e-01 -1.09200358e-01
-6.13248050e-01 -5.78528285e-01 -7.69056499e-01 3.11945379e-01
2.64403075e-01 -4.16707605e-01 1.29075110e+00 -5.65077186e-01
-1.69713485e+00 3.59565437e-01 8.07859749e-02 -7.34825075e-01
6.00264907e-01 7.39140734e-02 -6.33488715e-01 -5.71226068e-02
1.18776914e-02 3.71523425e-02 5.83143353e-01 -1.01694918e+00
-8.42666209e-01 -4.42103207e-01 -3.95807177e-01 -9.05967504e-02
-2.12259173e-01 -4.74315405e-01 8.94306600e-02 -2.19559908e-01
-1.78762734e-01 -9.13346052e-01 1.19671009e-01 -9.45614815e-01
-6.70407474e-01 -5.49768925e-01 9.74065185e-01 -5.55129886e-01
1.46514964e+00 -1.98380613e+00 -2.91907459e-01 5.64532340e-01
-4.04590845e-01 5.21544814e-02 7.35412419e-01 8.70648444e-01
-2.16187034e-02 -2.06073657e-01 -5.95542967e-01 -3.67720753e-01
5.01774728e-01 3.78105134e-01 -1.33521587e-01 7.90505886e-01
-3.84274632e-01 6.72121525e-01 -8.61830592e-01 -4.43393022e-01
1.08469129e+00 3.74973267e-01 2.38686744e-02 3.83578926e-01
-3.48073959e-01 3.56102198e-01 -3.47181529e-01 5.24549127e-01
4.86103952e-01 8.63609836e-03 6.67497218e-01 -5.03458500e-01
-2.57392943e-01 1.85801238e-01 -1.44806540e+00 1.29621327e+00
-7.20219135e-01 2.42061675e-01 -3.03552657e-01 -1.18838322e+00
7.61030912e-01 6.01568758e-01 9.76686120e-01 -4.83280659e-01
3.97403657e-01 1.54603243e-01 -3.32248777e-01 -4.02169973e-01
-4.03153710e-02 1.98745310e-01 -3.22812766e-01 7.23441839e-01
-1.09661445e-01 -9.35345218e-02 3.78427893e-01 -3.00706357e-01
1.15095174e+00 1.39889091e-01 9.87159252e-01 -5.84098458e-01
7.19388306e-01 -3.63038331e-01 4.58095074e-01 2.38491639e-01
6.78602085e-02 -4.35287058e-01 1.76338241e-01 -1.52984127e-01
-6.73342288e-01 -9.41440761e-01 -1.32406667e-01 6.79254830e-01
-6.22853190e-02 -2.28134587e-01 -1.04755712e+00 -5.42160690e-01
6.06693476e-02 1.33619964e+00 -3.44046593e-01 9.99118686e-02
-3.38706046e-01 -1.25880110e+00 1.60257846e-01 1.41589895e-01
7.99668372e-01 -9.67253983e-01 -1.49816835e+00 4.82728273e-01
-1.22590326e-01 -8.17456186e-01 -1.42908664e-02 5.41391850e-01
-6.22329831e-01 -1.24458814e+00 -4.81321327e-02 -2.94703722e-01
6.33570492e-01 -3.64785910e-01 1.24341238e+00 -4.54486310e-01
-1.88064963e-01 5.06403923e-01 -2.16471002e-01 -5.98276019e-01
-4.77720737e-01 3.52680802e-01 1.55278504e-01 3.18495601e-01
7.40772009e-01 -1.15689576e+00 -7.04176486e-01 4.10537302e-01
-7.12549627e-01 -2.25683734e-01 4.99371022e-01 9.37942639e-02
5.41340649e-01 9.76713657e-01 5.92459142e-01 -9.33745682e-01
3.49196106e-01 -7.57206976e-01 -1.16095412e+00 9.66152623e-02
-1.08449423e+00 6.93931729e-02 8.03572774e-01 -2.67314106e-01
-9.79171991e-01 4.79253650e-01 1.06426820e-01 3.84899378e-01
-5.42051613e-01 -1.63342923e-01 -7.19156027e-01 5.96405089e-01
-8.22193101e-02 3.22694212e-01 -4.50741678e-01 -7.75146186e-01
2.02815354e-01 7.79834092e-01 6.32672548e-01 -3.79040241e-01
8.79026413e-01 4.93160963e-01 2.38828242e-01 -8.92182648e-01
-3.36050749e-01 -5.18478274e-01 -6.03436172e-01 -2.31910065e-01
8.44969928e-01 -6.92932785e-01 -1.51065814e+00 4.76797193e-01
-7.35949099e-01 -3.10294628e-01 -5.21631241e-01 3.67087841e-01
-6.14637971e-01 3.57358038e-01 -3.80652547e-02 -1.40884447e+00
-5.64379096e-01 -9.71524894e-01 9.65654671e-01 3.03251266e-01
-7.19776213e-01 -1.00862134e+00 2.28224322e-01 6.71724603e-02
1.84999287e-01 6.02481484e-01 9.22504008e-01 -6.39832139e-01
-3.32938641e-01 -3.21565211e-01 2.64158487e-01 2.58970141e-01
5.60468137e-01 -1.69251844e-01 -1.05271673e+00 -4.36925590e-01
2.24072993e-01 2.38095596e-01 1.31044045e-01 4.23639417e-01
1.19203222e+00 -6.53334975e-01 -5.36762297e-01 2.62744993e-01
1.82362223e+00 4.25567538e-01 4.10163373e-01 2.20515236e-01
2.23218352e-01 5.35595119e-01 3.73290032e-01 8.84331942e-01
5.23632407e-01 8.41774464e-01 6.60080850e-01 2.00890396e-02
4.62659687e-01 -2.25320071e-01 5.73396564e-01 6.51481092e-01
1.01715609e-01 -4.28977072e-01 -5.19541442e-01 4.82116520e-01
-1.86318314e+00 -1.08919179e+00 7.62829036e-02 2.25336480e+00
5.42414010e-01 8.04989710e-02 5.85600495e-01 9.46557820e-01
5.46307981e-01 8.75346959e-02 -6.06268346e-01 -3.24578702e-01
2.75642216e-01 1.75727844e-01 9.23193336e-01 3.08623344e-01
-9.83234704e-01 -9.98666584e-02 5.48352861e+00 6.02308631e-01
-6.14758432e-01 2.01228514e-01 4.94599283e-01 -3.63233797e-02
1.25549719e-01 -1.93151072e-01 -7.48392940e-01 1.07908726e+00
1.35763025e+00 -3.31474006e-01 5.53107977e-01 8.64047706e-01
7.82981455e-01 -8.10600758e-01 -1.38073313e+00 9.10189986e-01
-2.60684252e-01 -5.44098377e-01 -5.03617227e-01 3.87914449e-01
6.50254726e-01 -4.77491207e-02 -6.18231058e-01 -1.41433135e-01
4.40897107e-01 -6.32927179e-01 4.49982822e-01 6.23065174e-01
3.63777459e-01 -1.02425945e+00 9.93085563e-01 7.61184216e-01
-1.51796675e+00 -3.29679847e-01 1.96213007e-01 -1.46303535e-01
3.42617035e-01 1.11031783e+00 -7.76883245e-01 7.19289243e-01
7.60350525e-01 1.99101597e-01 -3.33846122e-01 5.30386925e-01
-2.92909890e-01 8.19112360e-01 -8.64888966e-01 -1.30025625e-01
-1.60280928e-01 -5.14367402e-01 1.43311992e-01 1.22645903e+00
4.25706714e-01 5.30326692e-03 2.04429716e-01 7.99597263e-01
1.55701369e-01 1.41172353e-02 -5.96697927e-01 3.17270905e-01
8.20166111e-01 1.49162853e+00 -9.57568228e-01 -4.26538974e-01
-3.02277476e-01 8.21584702e-01 -3.76710087e-01 4.30673100e-02
-7.15104938e-01 -8.08519870e-02 3.79970759e-01 1.96612075e-01
1.37700751e-01 2.05411211e-01 -3.73417325e-02 -6.59199297e-01
-9.03732106e-02 -4.34643716e-01 5.33597469e-01 -3.97487223e-01
-1.21199715e+00 4.92172465e-02 4.98242617e-01 -9.53079104e-01
-9.49317634e-01 -1.82838932e-01 -7.37729013e-01 6.50003850e-01
-1.23246646e+00 -8.32078099e-01 -4.03324157e-01 5.39580584e-01
4.84979689e-01 1.61087751e-01 1.03144765e+00 2.22227395e-01
-7.31458664e-01 9.88637805e-02 4.04266447e-01 -1.40247839e-02
-1.38292462e-01 -1.49008346e+00 -2.25631762e-02 8.19392025e-01
7.87003525e-03 -1.57049745e-01 8.17360759e-01 -5.26677787e-01
-1.28626454e+00 -1.16859674e+00 7.77414382e-01 -2.89496124e-01
4.81986254e-01 -3.87997389e-01 -4.30104285e-01 4.84512389e-01
3.64937425e-01 -4.04721081e-01 8.24209809e-01 -3.72259140e-01
3.97946805e-01 -5.09379447e-01 -1.58791482e+00 -5.06692985e-03
4.56840575e-01 -3.28775883e-01 -5.20397365e-01 1.83202386e-01
-9.03122649e-02 2.71508843e-01 -1.20760000e+00 3.66867453e-01
4.47999656e-01 -1.20865846e+00 4.47967350e-01 4.41226363e-01
-5.30690730e-01 -6.49725199e-01 -3.78517598e-01 -1.30630815e+00
-2.11777970e-01 -7.44361579e-01 -5.41988015e-01 1.89551711e+00
-4.55089919e-02 -9.22059178e-01 5.55163920e-01 5.56110203e-01
5.35424948e-01 -3.89509261e-01 -1.06774950e+00 -7.02906132e-01
-5.93790233e-01 -4.37432081e-01 1.17657506e+00 6.30608082e-01
1.02481976e-01 1.64493352e-01 1.06833261e-02 5.05146027e-01
9.62664366e-01 2.13059083e-01 5.20169675e-01 -1.27271628e+00
-2.99574345e-01 -3.48379254e-01 -4.11397576e-01 -2.34139413e-01
1.26511410e-01 -5.84222555e-01 -7.12103769e-02 -1.51413691e+00
1.14474967e-01 -1.12668477e-01 -3.19473982e-01 5.56290507e-01
4.78003561e-01 2.95736611e-01 4.49059978e-02 2.61477213e-02
-1.33753017e-01 2.21576825e-01 -5.54465875e-02 -2.95414627e-01
-3.29181373e-01 4.31953311e-01 -2.00173780e-01 1.02682471e+00
1.08794987e+00 -3.32436770e-01 -6.93120301e-01 3.46733630e-01
-8.81979428e-03 -2.74274349e-01 2.67488360e-01 -1.36071301e+00
-2.68599782e-02 -2.67993007e-02 4.27604944e-01 -1.08225691e+00
-1.13375396e-01 -1.57491601e+00 1.10332251e+00 8.83056283e-01
2.78736055e-01 3.24488133e-02 -1.08434290e-01 3.33421320e-01
2.78545529e-01 -2.03499332e-01 7.85254180e-01 -1.12070955e-01
-3.53914529e-01 -1.27884030e-01 -5.92047036e-01 -6.02181792e-01
1.34584439e+00 1.20199807e-02 6.12818524e-02 -1.26267850e-01
-4.86709177e-01 2.39170402e-01 5.49771309e-01 3.65777314e-02
-3.44509155e-01 -1.27245831e+00 -2.76884764e-01 5.07522821e-02
-1.15267552e-01 -4.31704707e-02 2.04630010e-02 6.66286647e-01
-1.53396115e-01 5.92784822e-01 2.91073561e-01 -7.09869564e-01
-1.10006249e+00 7.37406552e-01 3.45332503e-01 -5.76552868e-01
-5.17664373e-01 -3.29254329e-01 -2.73501933e-01 1.09207988e-01
1.37706831e-01 -5.54709852e-01 -2.88435042e-01 5.58500409e-01
3.80996168e-01 1.03126109e+00 2.78293133e-01 -5.65903306e-01
-6.04542196e-01 6.66474223e-01 6.67520523e-01 1.54181853e-01
1.30855155e+00 -4.90321934e-01 -1.99920818e-01 8.06643367e-01
1.15916467e+00 -1.03840359e-01 -1.02066350e+00 2.17019409e-01
3.03080916e-01 1.04282774e-01 6.01775795e-02 -8.98822367e-01
-8.86716068e-01 2.78496087e-01 9.14700985e-01 9.00281131e-01
1.74740040e+00 1.96206849e-02 5.38264453e-01 7.18212575e-02
7.18272924e-01 -1.30661297e+00 -7.76017785e-01 -5.15892684e-01
1.47475049e-01 -9.29210901e-01 3.61774653e-01 -1.08988181e-01
1.05880737e-01 7.96041071e-01 3.25789372e-03 2.42498949e-01
7.83165753e-01 4.19890970e-01 -3.82321328e-01 -5.01423553e-02
-3.73775005e-01 -2.68122613e-01 -1.60604686e-01 5.66115022e-01
-2.65575889e-02 6.58022404e-01 -3.54227215e-01 4.92096126e-01
-3.00573081e-01 1.88769326e-01 2.41993681e-01 7.47487545e-01
-2.65805304e-01 -1.01506233e+00 -4.96606469e-01 5.37329972e-01
-4.11943465e-01 5.23950338e-01 1.28228381e-01 6.92224264e-01
5.89132547e-01 1.35792029e+00 1.13278091e-01 -8.37354064e-02
5.26166677e-01 2.76331216e-01 2.45923430e-01 -2.54486501e-01
-3.31273973e-01 -5.02341092e-02 -1.27190752e-02 -5.31930029e-01
-8.66557062e-01 -9.49307740e-01 -1.07326400e+00 -4.07175601e-01
-2.35728428e-01 3.75830859e-01 7.86695123e-01 1.13377571e+00
6.03550002e-02 5.81261039e-01 1.25948334e+00 -1.03233683e+00
-4.31848556e-01 -1.12517095e+00 -1.01843190e+00 3.56755674e-01
3.98923680e-02 -5.65400720e-01 -6.08999789e-01 7.03842863e-02] | [5.988373279571533, 2.5779459476470947] |
94b48805-e7e3-44b5-bf03-a21b9d2eef41 | rumor-detection-on-twitter-with-tree | null | null | https://aclanthology.org/P18-1184 | https://aclanthology.org/P18-1184.pdf | Rumor Detection on Twitter with Tree-structured Recursive Neural Networks | Automatic rumor detection is technically very challenging. In this work, we try to learn discriminative features from tweets content by following their non-sequential propagation structure and generate more powerful representations for identifying different type of rumors. We propose two recursive neural models based on a bottom-up and a top-down tree-structured neural networks for rumor representation learning and classification, which naturally conform to the propagation layout of tweets. Results on two public Twitter datasets demonstrate that our recursive neural models 1) achieve much better performance than state-of-the-art approaches; 2) demonstrate superior capacity on detecting rumors at very early stage. | ['Kam-Fai Wong', 'Wei Gao', 'Jing Ma'] | 2018-07-01 | null | null | null | acl-2018-7 | ['rumour-detection'] | ['natural-language-processing'] | [-1.54781103e-01 -1.34273171e-01 -4.11226839e-01 -5.79424977e-01
-3.00891817e-01 -6.85887039e-02 7.96832561e-01 3.86783689e-01
-9.66247246e-02 4.73954558e-01 7.49580562e-01 -6.56006813e-01
1.44554138e-01 -9.75638747e-01 -2.43659750e-01 -2.41240129e-01
-4.03159410e-01 4.50378567e-01 3.36674690e-01 -7.05537081e-01
6.87605977e-01 4.83651727e-01 -1.19656730e+00 8.02782118e-01
3.22016865e-01 9.15524900e-01 -4.44924206e-01 5.49116373e-01
-3.38338047e-01 1.85046983e+00 -5.23923039e-01 -9.80104357e-02
-3.01565886e-01 -5.46381712e-01 -1.11630774e+00 9.56158340e-02
7.08033815e-02 -5.74617624e-01 -6.87700808e-01 8.59175205e-01
2.20285967e-01 9.56474915e-02 8.28666389e-01 -7.70842135e-01
-8.86425495e-01 1.13460946e+00 -7.00702608e-01 8.70860457e-01
4.61882204e-01 -4.15182889e-01 9.12529051e-01 -9.20974135e-01
3.20361286e-01 1.11836290e+00 1.06685317e+00 2.88193792e-01
-8.97912681e-01 -8.90939772e-01 2.12927997e-01 5.55679321e-01
-1.17591405e+00 -3.65438133e-01 1.10168183e+00 -5.62467873e-01
6.67969108e-01 2.39555568e-01 2.13046297e-01 1.35627234e+00
3.07650566e-01 1.04895258e+00 1.26548421e+00 5.28313592e-02
-1.12125471e-01 3.15815955e-01 8.05081367e-01 9.71359551e-01
-9.68515128e-02 2.31852625e-02 -7.21955836e-01 -6.57720566e-01
4.54949498e-01 3.30993831e-01 -1.89087719e-01 2.78648198e-01
-6.05554640e-01 1.53964758e+00 8.64244461e-01 5.96900403e-01
-7.41369188e-01 -3.30590665e-01 6.36447966e-01 5.65993428e-01
9.30832744e-01 4.55950260e-01 5.97506799e-02 2.94530094e-01
-1.48072112e+00 2.34724909e-01 9.95866299e-01 4.03779685e-01
6.33657515e-01 3.48826379e-01 -2.53736854e-01 9.55291510e-01
9.23386961e-02 -2.34869361e-01 9.23559725e-01 -7.57216811e-02
2.44969532e-01 5.29881597e-01 1.19927578e-01 -1.50197303e+00
-1.00909984e+00 -8.66232395e-01 -1.13638937e+00 -5.89836389e-02
-2.71640867e-02 -1.41394492e-02 -7.56956160e-01 1.03349376e+00
-4.45860997e-02 1.92587093e-01 -1.90539420e-01 7.76632369e-01
1.19161117e+00 8.83938789e-01 -3.23277861e-01 -3.02423805e-01
1.40072703e+00 -1.25700343e+00 -8.25905263e-01 -1.29855976e-01
3.95543694e-01 -5.07374227e-01 5.82123876e-01 4.47159708e-01
-8.97491574e-01 -2.80634880e-01 -1.08501518e+00 1.47564709e-01
-5.54972112e-01 -2.63754487e-01 7.39313960e-01 6.88103139e-01
-6.87480748e-01 6.63920343e-01 -4.31179345e-01 2.40989421e-02
6.03819013e-01 -9.62553397e-02 9.93864983e-02 1.01504497e-01
-1.58312070e+00 8.45221341e-01 3.86820912e-01 1.21385425e-01
-1.32856083e+00 -1.31430656e-01 -6.75311446e-01 1.18002251e-01
2.57797465e-02 -4.47752923e-01 1.39033139e+00 -9.08411920e-01
-1.60022593e+00 7.46166945e-01 -1.34676144e-01 -9.34263468e-01
4.30407912e-01 -2.56505728e-01 -5.36984205e-01 1.03911392e-01
-3.38467048e-03 -2.95147687e-01 1.34579265e+00 -1.28572547e+00
-4.63842660e-01 -3.13566685e-01 -2.58707672e-01 -8.60717520e-02
-4.22588706e-01 6.79184556e-01 3.10743451e-01 -5.29802024e-01
2.65537500e-01 -4.79952782e-01 -2.28002176e-01 -7.54698813e-01
-7.32951939e-01 -3.71640831e-01 9.55836475e-01 -7.84806073e-01
1.59102190e+00 -1.52006686e+00 -3.92454743e-01 1.04214422e-01
7.42335379e-01 3.71867329e-01 1.79393172e-01 6.63413644e-01
3.47716361e-02 1.69678386e-02 1.23613790e-01 -3.50337207e-01
-2.15680510e-01 -2.88446690e-03 -7.35145032e-01 7.07088292e-01
-1.05795987e-01 6.90408885e-01 -1.05677569e+00 -3.01809937e-01
-2.42088899e-01 1.74805358e-01 -3.14604759e-01 5.91155887e-01
7.66472444e-02 8.42646360e-02 -6.32237434e-01 6.01183414e-01
6.75823569e-01 -3.99119258e-01 3.55696417e-02 1.46155104e-01
-1.01883925e-01 9.80235755e-01 -5.41708291e-01 6.60920024e-01
-4.80991781e-01 7.25334823e-01 2.02446759e-01 -1.10768759e+00
1.23587668e+00 3.60678524e-01 3.69451642e-02 -4.67797220e-01
4.68401581e-01 -4.58199531e-02 -4.38435465e-01 -5.31892896e-01
9.13861990e-01 -5.20120740e-01 -9.40419659e-02 8.81012321e-01
-1.00440100e-01 2.85264790e-01 -2.51978993e-01 2.97963053e-01
9.24394488e-01 -7.05597520e-01 5.81050158e-01 -2.40945548e-01
5.55892348e-01 -2.28228226e-01 2.61309773e-01 1.30740094e+00
-2.65240908e-01 6.47865057e-01 7.53331006e-01 -8.59213948e-01
-5.90567410e-01 -7.15377271e-01 -6.58844691e-03 1.73549354e+00
-1.54653072e-01 -2.38506913e-01 -4.03637558e-01 -9.00908232e-01
-2.53196135e-02 9.82904792e-01 -1.15032625e+00 -3.55290957e-02
-6.15232527e-01 -1.03028452e+00 9.75843310e-01 5.08158386e-01
4.76342678e-01 -1.05003178e+00 -2.48706460e-01 4.06060874e-01
-2.39097297e-01 -8.21573436e-01 -2.38637432e-01 3.02097112e-01
-8.35958004e-01 -8.22589934e-01 -4.51952010e-01 -8.26805472e-01
3.20646763e-01 5.61013579e-01 1.15056491e+00 2.67887264e-01
3.38020235e-01 -3.64789814e-01 -5.01227558e-01 -2.86032140e-01
-6.20929241e-01 1.59659714e-01 5.75077981e-02 3.15458834e-01
1.68690026e-01 -7.05584705e-01 -3.04694057e-01 4.08712298e-01
-7.22267151e-01 -8.64700675e-02 3.72812092e-01 1.01119053e+00
-1.05273359e-01 2.90071387e-02 8.13202381e-01 -1.29887199e+00
1.47806597e+00 -1.28533196e+00 -1.00358261e-03 -3.62623245e-01
-7.73167551e-01 -7.03952536e-02 8.21771443e-01 -3.50844830e-01
-9.67625976e-01 -5.83792388e-01 -3.59309584e-01 -3.07613790e-01
-1.67723402e-01 8.60878825e-01 8.86681616e-01 1.02174804e-01
1.12375307e+00 6.98641300e-01 -9.63150859e-02 -7.75620699e-01
2.76899874e-01 1.29827809e+00 5.13654709e-01 -6.27401248e-02
5.47716200e-01 5.72723866e-01 -4.65054750e-01 -8.97342205e-01
-1.41827726e+00 -8.67821515e-01 -4.78422523e-01 -7.47077242e-02
3.93059373e-01 -5.53502202e-01 -4.73565996e-01 6.32384658e-01
-1.27757657e+00 7.30844513e-02 3.26789707e-01 1.40319616e-01
-2.19643056e-01 4.12816018e-01 -1.40597141e+00 -9.50419605e-01
-8.82482111e-01 -5.09528816e-01 3.06487948e-01 -5.92047255e-03
-1.67442679e-01 -1.04860950e+00 3.06196809e-01 3.79246384e-01
1.02751172e+00 2.19427887e-02 7.20248222e-01 -1.41909504e+00
9.61604267e-02 -4.46847796e-01 -6.20217979e-01 1.94015145e-01
-7.76055381e-02 -4.06973124e-01 -9.35703039e-01 -3.84653121e-01
3.50361198e-01 -5.54449677e-01 1.49581158e+00 -1.78659353e-02
1.16372776e+00 -9.07991767e-01 -1.91809356e-01 4.46049482e-01
9.05024707e-01 -5.17628491e-01 3.64835054e-01 5.03091037e-01
5.25531292e-01 5.71664631e-01 -6.91213384e-02 7.29265928e-01
5.92527211e-01 3.98544341e-01 7.97210693e-01 3.96733508e-02
2.09841952e-01 -5.00301898e-01 5.48561573e-01 1.10399044e+00
4.38187942e-02 -1.16238482e-01 -9.69759703e-01 7.11360693e-01
-1.89014590e+00 -1.24466181e+00 -5.07292747e-01 1.81232584e+00
7.66435146e-01 2.62306631e-01 5.03965914e-01 -2.09961627e-02
6.42363667e-01 7.93835700e-01 1.19794160e-01 -8.85760665e-01
5.41310050e-02 -1.73870206e-01 3.01803470e-01 5.12510657e-01
-1.26227093e+00 8.30165982e-01 7.45885420e+00 4.44236219e-01
-1.36336434e+00 3.81516516e-01 3.96278888e-01 -1.62317473e-02
-4.29027751e-02 -3.89213741e-01 -7.27924407e-01 3.81173342e-01
1.13127124e+00 -2.13743791e-01 2.89097935e-01 9.09452617e-01
1.92819104e-01 4.97809649e-01 -6.45980120e-01 6.28183901e-01
4.27208304e-01 -1.68162549e+00 3.48785669e-01 -4.07210886e-01
7.27564037e-01 7.16024458e-01 -1.11040935e-01 7.16220975e-01
5.21866679e-01 -1.29998708e+00 6.32905185e-01 3.01478416e-01
1.15081193e-02 -8.00210238e-01 1.18887663e+00 8.18677306e-01
-7.61412859e-01 -3.33502531e-01 -6.00366116e-01 -4.08882976e-01
2.32647240e-01 5.90687096e-01 -1.41516781e+00 5.41449249e-01
6.14472628e-01 8.23504746e-01 -5.58466733e-01 8.76047552e-01
-4.53035176e-01 1.29997385e+00 5.51207960e-02 -3.87050003e-01
8.18950772e-01 3.83176118e-01 6.51307464e-01 2.12092257e+00
-2.25041479e-01 -1.15066230e-01 3.50984395e-01 9.19679046e-01
-3.34768176e-01 2.52869427e-01 -5.57821810e-01 1.38357719e-02
2.44866580e-01 1.12127233e+00 -3.37721139e-01 -4.20635730e-01
-2.62602389e-01 8.52051258e-01 6.96202636e-01 1.68181598e-01
-6.63521588e-01 -2.14725673e-01 3.39506626e-01 4.35420424e-01
3.84464115e-01 -9.20238942e-02 -2.73399472e-01 -1.14492309e+00
-6.04583979e-01 -7.78973937e-01 6.96189582e-01 -3.45340848e-01
-1.69221461e+00 8.99783611e-01 -2.83041209e-01 -9.00931239e-01
-4.35947418e-01 -3.72566700e-01 -1.24163473e+00 6.29783213e-01
-1.79136837e+00 -1.12320006e+00 -2.94390857e-01 8.59061182e-01
6.21511638e-01 -1.85631543e-01 8.58578980e-01 6.34882301e-02
-5.37693799e-01 5.40238261e-01 1.99689165e-01 6.12031460e-01
2.39115924e-01 -9.82414961e-01 3.16023171e-01 5.00341356e-01
1.13107964e-01 8.53881478e-01 8.37078512e-01 -3.66796672e-01
-9.18902576e-01 -1.11861086e+00 1.09457695e+00 -2.41891235e-01
1.09918988e+00 -1.67804986e-01 -1.46337581e+00 8.08924973e-01
4.29876745e-01 -2.54746079e-01 6.81762099e-01 7.64715672e-01
-9.36082244e-01 3.42900127e-01 -1.08958292e+00 1.40421420e-01
4.17959332e-01 -6.66401327e-01 -1.13219476e+00 4.45112973e-01
2.40862072e-01 -5.75774871e-02 -3.75345290e-01 1.28585130e-01
4.41096157e-01 -1.31077683e+00 6.48594797e-01 -1.22577703e+00
7.69284248e-01 2.05396816e-01 1.36448085e-01 -1.28963733e+00
-8.73899221e-01 -7.70617485e-01 -4.70174909e-01 7.99964488e-01
3.84734631e-01 -7.41146863e-01 8.01321089e-01 -6.34846747e-01
1.45484775e-01 -1.01720166e+00 -7.91956007e-01 -5.83995700e-01
2.91749567e-01 -3.23925704e-01 2.59568661e-01 1.39749634e+00
4.73981470e-01 8.59301567e-01 -9.48356092e-01 1.62651032e-01
6.00339293e-01 4.91361678e-01 5.56468785e-01 -1.33408773e+00
-1.31213650e-01 -8.94577026e-01 -3.33158910e-01 -1.59961522e+00
2.31455326e-01 -8.53276789e-01 3.26913185e-02 -1.40022898e+00
5.94007671e-01 -1.17923029e-01 -8.03241372e-01 3.87497127e-01
-7.52975419e-02 2.89466023e-01 -2.09166959e-01 5.87838829e-01
-6.81147158e-01 5.85393012e-01 7.27939844e-01 -3.14432919e-01
-1.33620158e-01 4.83562499e-01 -7.78576851e-01 9.80342567e-01
1.03520763e+00 -8.20846200e-01 -4.71294075e-02 -1.73931807e-01
2.93666452e-01 2.45059222e-01 3.15060884e-01 -4.73481715e-01
2.96200961e-01 -5.65009341e-02 -3.20977792e-02 -8.64329636e-01
1.43282965e-01 5.44508919e-02 -7.33179152e-01 3.81803006e-01
-5.56044340e-01 -6.12351745e-02 -6.09241240e-02 6.89231694e-01
-2.97850192e-01 -5.22573709e-01 9.27283823e-01 -2.86090612e-01
-3.48528892e-01 8.57384726e-02 -1.04513776e+00 -7.59647414e-02
3.76753747e-01 1.41792491e-01 -5.03778517e-01 -1.02670431e+00
-5.77998757e-01 3.30416448e-02 2.00776830e-02 6.02478266e-01
9.12597597e-01 -9.64840889e-01 -1.17241335e+00 1.49856210e-01
-1.98625885e-02 -7.40801156e-01 -1.37614965e-01 1.09899032e+00
-2.75653571e-01 4.36316073e-01 -2.21723810e-01 -2.13029593e-01
-1.17983210e+00 6.99544489e-01 4.36940879e-01 -7.74673820e-01
-7.72621930e-01 1.07856035e+00 2.16783769e-02 -6.16404235e-01
1.25791142e-02 2.03685954e-01 -7.70538926e-01 2.96433955e-01
1.00812805e+00 5.85115194e-01 -5.59010264e-03 -7.86821365e-01
-1.99523702e-01 -3.79735321e-01 -7.93220758e-01 3.50001782e-01
1.44385076e+00 -1.06967531e-01 -3.51835102e-01 4.77895290e-01
1.22857583e+00 9.28165093e-02 -5.66256583e-01 -7.24138379e-01
2.58017424e-02 -3.19687784e-01 5.48910737e-01 -4.71677840e-01
-8.43175650e-01 8.57972682e-01 -9.10415575e-02 9.63038862e-01
6.58824384e-01 4.92347516e-02 9.55385387e-01 4.44801509e-01
1.95562929e-01 -6.99131191e-01 2.44928226e-01 1.04394007e+00
1.03390265e+00 -1.07210636e+00 2.69014239e-01 -1.92049637e-01
-5.80691457e-01 1.41598153e+00 2.58515235e-02 -3.92637521e-01
7.72283614e-01 -4.53855880e-02 1.13889575e-01 -6.54974818e-01
-8.30352187e-01 4.65741986e-03 1.07314870e-01 2.75603771e-01
5.56742609e-01 1.13586068e-01 -3.07799220e-01 9.22985971e-01
-2.83561796e-01 -2.09332928e-01 9.28424478e-01 8.33613813e-01
-1.04952002e+00 -3.18946898e-01 -4.94802207e-01 6.52340412e-01
-8.05849791e-01 -2.38986492e-01 -6.71923518e-01 4.79640871e-01
-4.85732079e-01 1.13611150e+00 -2.57936299e-01 -7.48539925e-01
-6.80471435e-02 1.28944904e-01 -4.69479039e-02 -6.59077466e-01
-9.58387971e-01 -4.66489404e-01 4.99112844e-01 -4.37915355e-01
-5.58874011e-02 -4.29152071e-01 -8.31027865e-01 -7.40042508e-01
-4.76691574e-01 3.57090980e-01 3.51288140e-01 1.01734459e+00
1.86503157e-01 4.86947060e-01 1.17276990e+00 -1.05241203e+00
-1.21197629e+00 -1.53381634e+00 -7.39148855e-01 5.03623486e-01
7.58079171e-01 -5.90867102e-01 -5.39599597e-01 -3.55025440e-01] | [8.2064790725708, 10.164151191711426] |
8420cc08-fcf9-48be-8c5f-f6fe9388b4f4 | learning-view-priors-for-single-view-3d | 1811.10719 | null | http://arxiv.org/abs/1811.10719v2 | http://arxiv.org/pdf/1811.10719v2.pdf | Learning View Priors for Single-view 3D Reconstruction | There is some ambiguity in the 3D shape of an object when the number of
observed views is small. Because of this ambiguity, although a 3D object
reconstructor can be trained using a single view or a few views per object,
reconstructed shapes only fit the observed views and appear incorrect from the
unobserved viewpoints. To reconstruct shapes that look reasonable from any
viewpoint, we propose to train a discriminator that learns prior knowledge
regarding possible views. The discriminator is trained to distinguish the
reconstructed views of the observed viewpoints from those of the unobserved
viewpoints. The reconstructor is trained to correct unobserved views by fooling
the discriminator. Our method outperforms current state-of-the-art methods on
both synthetic and natural image datasets; this validates the effectiveness of
our method. | ['Tatsuya Harada', 'Hiroharu Kato'] | 2018-11-26 | learning-view-priors-for-single-view-3d-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Kato_Learning_View_Priors_for_Single-View_3D_Reconstruction_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Kato_Learning_View_Priors_for_Single-View_3D_Reconstruction_CVPR_2019_paper.pdf | cvpr-2019-6 | ['single-view-3d-reconstruction'] | ['computer-vision'] | [ 8.43233839e-02 5.15908659e-01 -6.32020310e-02 -5.11321247e-01
-5.39458752e-01 -7.83940017e-01 5.58271527e-01 -6.38340473e-01
4.28832024e-01 2.75761604e-01 8.83201733e-02 1.55218840e-01
1.13610454e-01 -8.05653036e-01 -1.08089030e+00 -4.37117249e-01
3.04132670e-01 1.04974103e+00 4.00876790e-01 3.91692162e-01
5.38269430e-02 7.91120291e-01 -1.43278670e+00 6.78213477e-01
4.08816457e-01 8.87758315e-01 3.86813223e-01 5.90718865e-01
5.46805598e-02 7.01753497e-01 -3.26021075e-01 -4.71477568e-01
6.81863904e-01 -4.89256650e-01 -5.09989321e-01 9.46339130e-01
8.20708573e-01 -8.96359384e-01 -3.75274062e-01 1.12515402e+00
-9.17901993e-02 -2.56357193e-01 7.42909551e-01 -1.17552090e+00
-9.02934611e-01 1.22181974e-01 -3.42318118e-01 -4.34846729e-01
5.98289788e-01 -1.75970923e-02 9.47337925e-01 -1.37608862e+00
9.66048419e-01 1.43960154e+00 3.29334319e-01 6.44273400e-01
-1.34821963e+00 -2.46968716e-01 4.48802650e-01 -1.32847726e-01
-1.23277617e+00 -7.13808835e-01 9.44873691e-01 -6.02142036e-01
6.80081069e-01 1.67791933e-01 6.79369688e-01 1.14554620e+00
3.39486331e-01 7.25614607e-01 9.95923936e-01 -1.34979084e-01
3.13815773e-01 3.23968798e-01 -4.25858974e-01 7.32375205e-01
4.91428614e-01 3.91325295e-01 -4.13874418e-01 -3.37758720e-01
1.19418824e+00 3.76814365e-01 -3.30094755e-01 -9.24803019e-01
-1.31012368e+00 5.43302536e-01 2.85455942e-01 -2.68526405e-01
-4.64147508e-01 -8.16030055e-02 -3.49078357e-01 3.01920176e-01
4.68733460e-01 4.04018760e-01 -4.70816553e-01 4.72881615e-01
-7.40351677e-01 1.32508233e-01 8.49027991e-01 1.23988426e+00
6.49788380e-01 2.72355527e-01 3.99323851e-01 4.40044612e-01
5.10170162e-01 9.18748796e-01 -2.88967669e-01 -1.50536156e+00
3.23838294e-01 6.85714304e-01 3.37966084e-01 -8.20178866e-01
3.13431472e-02 -2.22342715e-01 -6.05053604e-01 7.72806048e-01
2.93493330e-01 3.56898487e-01 -9.62871671e-01 1.39626622e+00
5.23233771e-01 -4.98938151e-02 1.61818385e-01 1.08788633e+00
7.13792145e-01 4.02374327e-01 -1.02021778e+00 -1.74447551e-01
9.31653321e-01 -8.61500680e-01 -5.12256503e-01 -3.92128319e-01
-2.22415775e-01 -1.01396680e+00 5.13647377e-01 6.69608593e-01
-1.53344059e+00 -6.93287313e-01 -9.64474678e-01 -6.56191036e-02
2.96449780e-01 2.93131441e-01 2.78222084e-01 2.01718912e-01
-9.50865030e-01 4.41199571e-01 -1.11962306e+00 -1.10243067e-01
2.69013733e-01 6.69055060e-02 -7.60691583e-01 -4.16318685e-01
-1.53103396e-01 1.03861570e+00 8.40985030e-02 6.80332631e-02
-1.68731451e+00 -4.13456589e-01 -7.15809286e-01 4.04203869e-02
5.72109818e-01 -9.17888939e-01 1.10391545e+00 -1.13671029e+00
-1.27810860e+00 9.89386916e-01 -1.95421100e-01 4.03017141e-02
6.63480639e-01 -1.05255380e-01 -1.73248410e-01 4.07378197e-01
6.37763739e-02 5.47809720e-01 9.97532308e-01 -2.10144901e+00
-3.17938983e-01 -6.08572006e-01 2.44526982e-01 2.24949613e-01
6.12296939e-01 -5.38011014e-01 -4.80936378e-01 -4.87199336e-01
1.10272789e+00 -1.01085758e+00 4.08551330e-03 5.66653192e-01
-5.18979907e-01 3.84171158e-01 9.35104370e-01 -4.85025555e-01
1.77882686e-01 -2.05688214e+00 3.30091149e-01 -1.67772427e-01
2.87770241e-01 -3.35744917e-01 -9.79375392e-02 5.76517761e-01
-1.53210545e-02 -2.47084349e-01 -2.23703086e-02 -6.62885249e-01
-2.07211420e-01 5.40446579e-01 -5.85362256e-01 7.28133678e-01
-7.18497932e-02 6.04281127e-01 -7.27891743e-01 9.47878808e-02
3.97225380e-01 3.78455907e-01 -6.31576657e-01 7.21720755e-01
-4.11343455e-01 8.09211612e-01 -4.22729433e-01 6.34120345e-01
1.27388155e+00 -5.50866425e-01 3.33900541e-01 -2.80439734e-01
2.73512881e-02 5.66509902e-01 -1.42992187e+00 1.62222862e+00
-1.77135468e-01 2.86069483e-01 3.61131877e-02 -7.52511859e-01
8.30795169e-01 5.22266269e-01 4.49574202e-01 -1.99692219e-01
-2.67792374e-01 2.46873394e-01 -1.61473691e-01 -5.51323175e-01
1.53292656e-01 -3.89686376e-01 3.35288316e-01 7.07982600e-01
1.48060441e-01 -4.56463695e-01 -5.23806632e-01 2.76042402e-01
7.27433205e-01 2.79802799e-01 4.39434648e-01 5.89026809e-02
1.34327441e-01 -2.73166478e-01 5.98071694e-01 6.43044591e-01
2.93317556e-01 1.24400616e+00 3.75293046e-01 -8.29719961e-01
-1.26295447e+00 -1.57842565e+00 4.99157608e-02 3.72769386e-01
5.14683008e-01 -7.36359134e-02 -2.58442551e-01 -1.05222452e+00
5.05855344e-02 8.30664217e-01 -5.62219203e-01 -2.89888680e-02
-3.84006977e-01 2.08664775e-01 -1.18886143e-01 5.58147252e-01
8.05809349e-02 -7.71915019e-01 -5.81144035e-01 -1.64071739e-01
-1.81144282e-01 -1.41706526e+00 -5.14932990e-01 -7.95556381e-02
-1.19365275e+00 -1.39705300e+00 -3.81581873e-01 -5.72554171e-01
1.37832773e+00 7.43085802e-01 1.52544808e+00 1.05575487e-01
3.18670541e-01 5.04581690e-01 -2.63826579e-01 -1.63742900e-01
-8.44367385e-01 -6.78654253e-01 1.07641101e-01 9.15924236e-02
-6.64337948e-02 -7.46663809e-01 -5.68450272e-01 6.12368464e-01
-7.72206962e-01 4.71244901e-01 3.18319947e-01 7.32228994e-01
7.61134803e-01 -2.54989088e-01 1.92579087e-02 -8.77850652e-01
-4.52257305e-01 -2.59819120e-01 -9.24655914e-01 3.39935511e-01
-3.29562128e-01 9.06615406e-02 7.21153975e-01 -5.00128388e-01
-1.13352919e+00 3.38570863e-01 1.28861889e-01 -8.84595811e-01
-2.34925196e-01 9.06944647e-03 -3.95477176e-01 1.68602869e-01
4.43376780e-01 2.25713432e-01 -7.96247125e-02 -7.73510575e-01
1.43708020e-01 1.66899860e-01 3.85570198e-01 -4.95877683e-01
7.20142245e-01 1.08820164e+00 -4.15282175e-02 -4.34480131e-01
-1.03575194e+00 -8.66777375e-02 -8.07032108e-01 -1.61661997e-01
6.79831624e-01 -1.04988825e+00 -5.88630617e-01 1.30560219e-01
-1.39160514e+00 1.44106939e-01 -6.87056303e-01 6.47722542e-01
-7.46280253e-01 3.47737163e-01 -1.74250111e-01 -8.29266071e-01
2.87295699e-01 -1.37500679e+00 1.48716617e+00 -1.15494363e-01
6.53291345e-02 -8.76365483e-01 -1.18995152e-01 3.00700426e-01
-1.74419090e-01 1.73506871e-01 9.19697583e-01 -4.43723738e-01
-1.35752237e+00 -4.63483259e-02 3.41382995e-02 4.22464073e-01
3.72145951e-01 6.17577285e-02 -1.12015975e+00 -5.13653100e-01
6.24474227e-01 -3.93264145e-01 6.61475718e-01 6.30148411e-01
8.10238302e-01 -5.12656689e-01 -3.44287127e-01 6.18921220e-01
1.43192875e+00 1.91635728e-01 4.44504231e-01 -3.16970050e-01
5.47474742e-01 5.22858143e-01 2.72506446e-01 3.83568197e-01
5.20452797e-01 5.86529195e-01 1.09701300e+00 2.26927679e-02
-3.14166576e-01 -6.37524664e-01 2.92057544e-01 7.89199412e-01
-1.14259645e-01 -5.28004587e-01 -5.37833512e-01 4.85629648e-01
-1.63426292e+00 -9.89381135e-01 5.45387752e-02 2.44404078e+00
1.54580429e-01 1.06257401e-01 -1.75177693e-01 -1.74780235e-01
3.68959904e-01 2.48373479e-01 -8.10809612e-01 -7.91148841e-02
-1.42172620e-01 -3.68931562e-01 1.07672542e-01 6.97084665e-01
-5.84640026e-01 5.08295715e-01 7.24645758e+00 9.44024399e-02
-8.84566069e-01 -7.58812428e-02 3.49635750e-01 3.05694621e-02
-9.40891743e-01 5.08677840e-01 -5.77125788e-01 8.21603909e-02
5.61439209e-02 3.00784141e-01 6.17066085e-01 1.00218678e+00
-1.03031822e-01 -1.22891732e-01 -1.74707675e+00 7.54838288e-01
4.65974361e-01 -1.47223389e+00 4.45442677e-01 1.52707413e-01
1.03547859e+00 -4.13039997e-02 6.23927116e-02 -1.65179759e-01
3.15526336e-01 -9.65998232e-01 1.15983284e+00 7.44029105e-01
6.25425279e-01 -2.64264047e-01 3.30243528e-01 1.03734064e+00
-9.50899780e-01 2.68396705e-01 -6.56092644e-01 -2.17129886e-02
2.58035570e-01 4.02044624e-01 -9.69603837e-01 6.20107889e-01
4.43003237e-01 7.92288899e-01 -2.46535107e-01 6.43322051e-01
-2.35331222e-01 1.95369899e-01 -2.09837615e-01 7.64747798e-01
-2.51457840e-01 -3.75285029e-01 8.78340721e-01 2.68917710e-01
6.36532247e-01 3.86046112e-01 5.09589553e-01 1.27505457e+00
-8.52903649e-02 -8.30939472e-01 -9.60635960e-01 2.85608917e-01
3.73669267e-01 8.03699672e-01 -6.23706102e-01 -3.75202566e-01
-6.88936591e-01 1.19301546e+00 2.73251861e-01 4.33808118e-01
-6.00965559e-01 6.99066460e-01 4.27201599e-01 4.59796131e-01
6.60887182e-01 -2.89682060e-01 -2.93937683e-01 -1.59499002e+00
3.44360232e-01 -9.14630592e-01 2.44018972e-01 -1.30780089e+00
-1.61248446e+00 7.88870335e-01 1.80153623e-01 -1.90855324e+00
-2.77703345e-01 -6.90006375e-01 -3.11266720e-01 6.17834449e-01
-1.17423022e+00 -1.34708560e+00 -3.03713024e-01 3.22046489e-01
9.25729990e-01 -1.64899409e-01 9.92965877e-01 -4.08415914e-01
4.08930480e-01 -2.44551618e-02 2.04217397e-02 -2.02652365e-01
4.75540012e-01 -1.11966157e+00 5.02905905e-01 7.16756225e-01
5.83184540e-01 4.19824511e-01 7.08911598e-01 -6.90653563e-01
-1.57249308e+00 -6.63545072e-01 6.96972370e-01 -7.55170107e-01
6.19983934e-02 -4.44417119e-01 -8.99471462e-01 1.24587953e+00
2.38241345e-01 6.41258657e-01 3.02939594e-01 -3.64106297e-01
-7.07903028e-01 8.31704736e-02 -1.22320473e+00 2.53839016e-01
1.07637835e+00 -5.96767902e-01 -7.05094576e-01 2.47182950e-01
5.51563442e-01 -8.62117290e-01 -5.48838913e-01 4.90241408e-01
8.49891245e-01 -1.55387855e+00 1.14769804e+00 -5.31777084e-01
4.67291862e-01 -3.84766310e-01 -5.64204276e-01 -1.28104532e+00
-1.72470242e-01 -1.47560731e-01 -4.68505263e-01 6.63001776e-01
2.35821307e-01 -5.47832549e-01 8.59437585e-01 5.50955951e-01
-1.50512382e-01 -5.32465816e-01 -9.35923040e-01 -8.75204623e-01
-1.77733108e-01 -1.14369854e-01 6.22127116e-01 8.31376314e-01
-6.10723019e-01 3.22366387e-01 -6.47987545e-01 6.25589788e-01
8.29146147e-01 9.14001048e-01 8.71195853e-01 -1.36534786e+00
-7.95918226e-01 2.60173380e-01 -1.67627141e-01 -1.58461249e+00
-2.50656307e-02 -7.38183141e-01 -2.54379641e-02 -1.47115362e+00
5.00847042e-01 -1.05291098e-01 3.11410517e-01 1.51214287e-01
1.28446132e-01 -6.02084324e-02 4.16073680e-01 4.34313267e-01
-4.68874812e-01 5.43252230e-01 1.84431124e+00 -1.48625463e-01
1.11626856e-01 4.16819513e-01 -4.66293246e-01 1.24678314e+00
1.02489218e-01 -8.48502338e-01 -6.07232392e-01 -1.09364307e+00
2.73618698e-01 7.25431383e-01 6.41473174e-01 -5.38421512e-01
5.29873883e-03 -3.91328514e-01 8.56376350e-01 -1.09147882e+00
8.53050232e-01 -1.36529207e+00 7.76636779e-01 2.56153017e-01
-3.10862958e-01 1.39991254e-01 -3.10105056e-01 7.86894500e-01
-6.81222305e-02 -2.52863079e-01 8.02481294e-01 -5.70015728e-01
2.17612423e-02 4.93710101e-01 -1.40933827e-01 -3.05194799e-02
8.29908907e-01 -5.33790588e-01 -2.09930882e-01 -5.90953708e-01
-8.45752060e-01 -2.37714916e-01 1.15393460e+00 2.84213275e-01
1.13412905e+00 -1.58966899e+00 -5.59017003e-01 6.78342462e-01
1.58895642e-01 4.51868981e-01 2.57158518e-01 5.05468488e-01
-2.83017188e-01 5.48037924e-02 -1.03499852e-01 -9.82419133e-01
-1.11972225e+00 7.22094417e-01 4.10870671e-01 4.76529673e-02
-9.17151034e-01 5.59464812e-01 8.00226092e-01 -6.85933948e-01
8.77150968e-02 -6.04703546e-01 2.15814173e-01 -5.60248733e-01
3.36761117e-01 2.40261525e-01 -1.03755631e-01 -7.78071523e-01
-3.52379382e-01 6.15121305e-01 1.18578225e-01 -1.21774040e-01
1.40061808e+00 -2.44211882e-01 2.84029115e-02 6.67113006e-01
1.07963109e+00 3.94951880e-01 -1.79659081e+00 -1.89536899e-01
-5.00727236e-01 -1.12472188e+00 -3.22931379e-01 -7.21041560e-01
-1.13285422e+00 9.93700385e-01 2.40129367e-01 2.89056152e-01
9.74689186e-01 4.11858916e-01 4.73063886e-01 3.85343522e-01
7.33920872e-01 -3.92224222e-01 4.27542359e-01 3.99981588e-01
1.33212829e+00 -1.41607666e+00 2.98873216e-01 -5.84817171e-01
-8.32868516e-01 1.18652380e+00 6.42562747e-01 -3.98310125e-01
6.81302369e-01 4.23077703e-01 -1.16838347e-02 -5.07057786e-01
-9.68889773e-01 2.88286000e-01 6.52888954e-01 6.08931303e-01
-2.37070955e-02 4.82762828e-02 5.79017639e-01 4.18927908e-01
1.08990081e-01 -2.85606444e-01 6.99883044e-01 6.94777668e-01
-3.10855806e-01 -8.37650418e-01 -5.01929641e-01 8.19463879e-02
-3.24068442e-02 2.76046753e-01 -6.36211812e-01 7.15423167e-01
1.56257555e-01 7.12537885e-01 8.66105705e-02 -1.79059908e-01
4.18197423e-01 -2.35006571e-01 8.89200330e-01 -1.01250541e+00
2.37576198e-02 4.60688412e-01 -1.01814441e-01 -7.23758459e-01
-6.20443523e-01 -5.89462578e-01 -1.02821314e+00 -9.65813994e-02
-5.27060509e-01 -1.90676734e-01 1.94232151e-01 8.78050208e-01
2.37111211e-01 3.83032672e-02 1.01976907e+00 -1.00852561e+00
-7.50430763e-01 -5.41712284e-01 -6.55281723e-01 5.02801478e-01
6.57883942e-01 -7.53112674e-01 -5.65187871e-01 4.71569270e-01] | [8.725756645202637, -2.9726388454437256] |
203f1be2-62a4-495e-b291-0f5ae08caad5 | relational-reasoning-via-set-transformers | 2209.09845 | null | https://arxiv.org/abs/2209.09845v3 | https://arxiv.org/pdf/2209.09845v3.pdf | Relational Reasoning via Set Transformers: Provable Efficiency and Applications to MARL | The cooperative Multi-A gent R einforcement Learning (MARL) with permutation invariant agents framework has achieved tremendous empirical successes in real-world applications. Unfortunately, the theoretical understanding of this MARL problem is lacking due to the curse of many agents and the limited exploration of the relational reasoning in existing works. In this paper, we verify that the transformer implements complex relational reasoning, and we propose and analyze model-free and model-based offline MARL algorithms with the transformer approximators. We prove that the suboptimality gaps of the model-free and model-based algorithms are independent of and logarithmic in the number of agents respectively, which mitigates the curse of many agents. These results are consequences of a novel generalization error bound of the transformer and a novel analysis of the Maximum Likelihood Estimate (MLE) of the system dynamics with the transformer. Our model-based algorithm is the first provably efficient MARL algorithm that explicitly exploits the permutation invariance of the agents. Our improved generalization bound may be of independent interest and is applicable to other regression problems related to the transformer beyond MARL. | ['Zhaoran Wang', 'Zhuoran Yang', 'Vincent Y. F. Tan', 'Kaixin Wang', 'Boyi Liu', 'Fengzhuo Zhang'] | 2022-09-20 | null | null | null | null | ['relational-reasoning'] | ['natural-language-processing'] | [-6.03361465e-02 1.92577407e-01 -1.18931299e-02 1.58681154e-01
-5.53925157e-01 -7.77799428e-01 7.47733057e-01 1.56331137e-01
-4.22798783e-01 8.49370599e-01 -3.96150947e-01 -3.65981162e-01
-1.25354600e+00 -6.41041636e-01 -9.69760656e-01 -1.10576057e+00
-6.21795356e-01 6.50491714e-01 2.76342541e-01 -6.22734129e-01
1.19829467e-02 3.87065440e-01 -1.23499525e+00 -4.69193012e-01
7.36975014e-01 7.43193269e-01 -6.35974333e-02 8.09109390e-01
6.04633272e-01 9.84275520e-01 -2.71538198e-01 -2.00296968e-01
8.19858372e-01 -5.45713365e-01 -4.19649005e-01 -1.25460252e-01
5.86058653e-04 -4.96547669e-02 -6.49434626e-01 1.23777115e+00
2.93939799e-01 3.92595708e-01 5.17275751e-01 -1.94617403e+00
-4.47673529e-01 9.51281667e-01 -7.63589144e-01 1.08999565e-01
-8.50076601e-02 -3.36568743e-01 1.10522199e+00 -3.36462080e-01
3.95425320e-01 1.43070555e+00 7.35089540e-01 2.75099486e-01
-1.13621867e+00 -7.01295316e-01 3.92706722e-01 2.33389422e-01
-1.37887931e+00 5.87662719e-02 5.73139191e-01 -4.72423643e-01
6.09852672e-01 1.02717660e-01 5.72902024e-01 7.65214503e-01
2.41984740e-01 5.88066161e-01 1.38022017e+00 -4.67965662e-01
3.79164785e-01 5.79389073e-02 3.32763195e-01 1.01306295e+00
7.28064597e-01 5.41998327e-01 -5.47381997e-01 -2.92306036e-01
9.49322462e-01 9.56833959e-02 -4.82385941e-02 -9.07842278e-01
-9.93871212e-01 9.97698247e-01 3.55552733e-01 4.32543177e-03
-2.94857472e-01 4.56845254e-01 1.20102726e-01 9.15517092e-01
1.88740224e-01 5.04676819e-01 -5.55973589e-01 -5.90092167e-02
-2.65891939e-01 4.25043434e-01 1.20349729e+00 9.34258223e-01
5.78475237e-01 3.63550000e-02 3.01126540e-01 4.28100079e-01
4.42575157e-01 6.38281345e-01 7.06894994e-02 -1.17422950e+00
5.08044124e-01 5.63238919e-01 3.66630971e-01 -9.44387138e-01
-6.68180883e-01 -7.28268623e-01 -7.83394635e-01 2.43068710e-01
6.50212348e-01 -3.54436755e-01 -1.64115280e-01 2.14057541e+00
3.77018631e-01 2.95800686e-01 2.29924664e-01 6.20087206e-01
-1.05309874e-01 4.08130914e-01 -3.71449500e-01 -4.88400221e-01
1.05163538e+00 -1.00968540e+00 -4.53026414e-01 -7.19015598e-02
7.30856657e-01 -1.08235300e-01 6.10477328e-01 3.42440844e-01
-8.54829133e-01 1.27136027e-02 -1.10361600e+00 4.09867436e-01
-9.93822739e-02 -2.11211607e-01 8.85339260e-01 7.37543702e-01
-1.05915332e+00 5.06924987e-01 -1.06166148e+00 -5.40657222e-01
2.91547686e-01 6.26551867e-01 -3.38370383e-01 2.26970702e-01
-9.30156291e-01 9.34509277e-01 1.01618044e-01 4.40766543e-01
-1.05886209e+00 -5.34709632e-01 -6.19706035e-01 -2.94039473e-02
1.04066205e+00 -4.82188225e-01 1.20591712e+00 -7.33456731e-01
-1.53222692e+00 1.75486371e-01 2.60559529e-01 -7.42834389e-01
9.16993082e-01 -7.12400228e-02 2.26710394e-01 1.00870989e-01
-2.11598612e-02 -1.93279147e-01 6.99963570e-01 -1.11139750e+00
-6.54465258e-01 -5.04728496e-01 5.97249687e-01 2.90636271e-01
-4.76551615e-02 -5.65213919e-01 3.17645222e-01 -3.90174389e-01
-1.22801356e-01 -1.14075053e+00 -5.19522786e-01 -9.51353908e-02
4.77612950e-02 -3.35631371e-01 4.78756458e-01 -1.55687243e-01
6.81322277e-01 -2.07667446e+00 4.81177658e-01 4.01636392e-01
2.84406960e-01 -1.27593175e-01 -4.82989341e-01 8.58694553e-01
1.47937506e-01 -1.91720817e-02 9.16867927e-02 -4.57471609e-02
5.11530519e-01 3.94220769e-01 -5.76725721e-01 9.33834016e-01
-2.33044535e-01 7.34053314e-01 -9.73231077e-01 -1.21646941e-01
2.45567970e-02 1.18449636e-01 -5.95993400e-01 1.36520535e-01
-7.58276135e-02 1.98406786e-01 -7.35553920e-01 9.87485573e-02
4.89888936e-01 -1.55815676e-01 4.59990859e-01 9.96051505e-02
-9.68302041e-02 -5.73725030e-02 -1.26009917e+00 1.06620514e+00
-1.90367162e-01 3.07579041e-01 3.30727518e-01 -1.44218969e+00
4.75833893e-01 1.35546505e-01 3.89478862e-01 -4.32765990e-01
1.57241642e-01 1.26489967e-01 2.27719858e-01 -3.12205672e-01
8.47406164e-02 -1.58494934e-01 -2.47645885e-01 6.55564427e-01
-1.01942249e-01 1.62412852e-01 5.26897550e-01 1.64623514e-01
1.15370142e+00 1.79216191e-01 4.86315817e-01 -4.64339495e-01
3.97859186e-01 -3.00921112e-01 5.38394809e-01 1.46324456e+00
-1.59974650e-01 -5.15226662e-01 8.74099731e-01 -9.23383757e-02
-9.38108027e-01 -1.08273602e+00 7.49889463e-02 1.09616876e+00
3.41826946e-01 -2.75310755e-01 -5.53278387e-01 -3.97470832e-01
1.75395101e-01 4.57656711e-01 -9.78918135e-01 -3.09462160e-01
-3.78296047e-01 -7.88688481e-01 5.00592113e-01 3.69420528e-01
4.47085172e-01 -4.27144855e-01 -6.78538084e-01 1.74528614e-01
2.41820365e-01 -1.02949274e+00 -1.04589239e-01 3.25690866e-01
-6.81244314e-01 -1.56976044e+00 -1.54708102e-01 -5.19949436e-01
6.25284016e-01 2.49432713e-01 4.21916097e-01 -9.67163518e-02
1.18855305e-01 8.03562284e-01 -2.65122056e-01 -3.31790894e-01
-4.87220079e-01 1.59084141e-01 4.10832047e-01 -2.98550930e-02
1.06894724e-01 -6.02981210e-01 -2.54841238e-01 4.67708945e-01
-5.57892621e-01 -3.07739191e-02 6.16117299e-01 8.45963597e-01
1.89380512e-01 5.51455200e-01 6.79641306e-01 -5.29160738e-01
6.57055855e-01 -4.74246532e-01 -1.30212677e+00 4.06703502e-01
-5.29205501e-01 4.40311670e-01 6.57149553e-01 -5.74986935e-01
-7.07151532e-01 -1.43996969e-01 5.99484205e-01 -7.87948668e-02
3.85057539e-01 5.85064292e-01 2.71347255e-01 -1.71923593e-01
3.13205421e-01 1.00276321e-01 2.91585654e-01 -2.60045290e-01
2.76182592e-01 3.06826830e-01 2.14139953e-01 -8.94941986e-01
1.27059209e+00 3.57134014e-01 7.16811121e-01 -7.96651542e-01
-8.49660277e-01 -4.40622382e-02 -2.56686538e-01 -1.84245184e-01
4.49401766e-01 -8.77531111e-01 -1.58071184e+00 4.33073074e-01
-6.93729699e-01 -6.04304731e-01 -2.93399185e-01 5.61173737e-01
-9.51737702e-01 4.33355600e-01 -6.08181655e-01 -1.43661416e+00
2.08465844e-01 -1.00724220e+00 3.04911047e-01 2.37349749e-01
3.68164003e-01 -9.29864705e-01 2.85816550e-01 1.85375869e-01
3.09696287e-01 1.43330395e-01 1.01159859e+00 -9.80098128e-01
-9.24263358e-01 -1.53654009e-01 -7.04573393e-02 -4.99183387e-02
-8.24957862e-02 -2.96839714e-01 -5.12191594e-01 -6.92934453e-01
1.68427542e-01 -1.46569654e-01 5.95920980e-01 3.05628181e-01
3.88731509e-01 -5.96440136e-01 -2.69883245e-01 1.41200840e-01
1.52965152e+00 3.48121494e-01 1.33877918e-02 4.32133406e-01
4.20955837e-01 5.43076992e-01 4.94320512e-01 5.59333920e-01
5.40907323e-01 6.16970301e-01 4.90424514e-01 4.65446562e-01
5.68918824e-01 -4.09164697e-01 6.01131082e-01 8.57683241e-01
-2.79617459e-01 -7.07986429e-02 -6.18723869e-01 2.46217951e-01
-2.39248610e+00 -1.04708242e+00 2.49396145e-01 2.47833037e+00
6.50134921e-01 -1.72525734e-01 3.97784114e-01 5.50028030e-03
6.70696378e-01 -2.15116307e-01 -7.20325232e-01 -2.67699838e-01
-1.08492121e-01 -1.51856259e-01 7.87120402e-01 7.69504607e-01
-9.02254105e-01 5.19688427e-01 6.36493587e+00 6.66496575e-01
-3.55480611e-01 2.91550308e-01 -1.37123270e-02 5.60661368e-02
7.88996741e-02 1.49293572e-01 -6.35254979e-01 3.28611545e-02
8.87242854e-01 -6.04540646e-01 1.14433897e+00 8.39256108e-01
2.59938806e-01 -5.91446459e-02 -1.37265587e+00 7.73039460e-01
-2.08961982e-02 -7.86652565e-01 -4.17880625e-01 4.84427303e-01
6.80611134e-01 7.39495009e-02 1.18517488e-01 4.55370367e-01
8.25548172e-01 -7.49119997e-01 7.44287968e-01 4.59500313e-01
-8.17909092e-02 -8.96268547e-01 7.67310739e-01 4.83601540e-01
-1.26716173e+00 -7.41970420e-01 -2.67618358e-01 -4.10362303e-01
-2.64229417e-01 -2.16705948e-02 -4.62670535e-01 7.14987934e-01
3.14576626e-01 6.66052759e-01 -5.51569819e-01 9.15060103e-01
-1.00250557e-01 3.68474126e-01 -7.12968886e-01 -2.98099071e-01
2.66513199e-01 -6.23454034e-01 7.56738961e-01 5.97689986e-01
9.24237892e-02 -9.62861106e-02 3.76160294e-01 5.65263689e-01
3.55898514e-02 -1.80681154e-01 -5.53588271e-01 -4.12560642e-01
5.01411378e-01 9.95428205e-01 -6.00525618e-01 1.78769082e-01
-4.30385679e-01 3.31669390e-01 6.04387522e-01 5.98381519e-01
-8.59899163e-01 -1.51451126e-01 5.49303889e-01 -1.99350819e-01
4.54467058e-01 -4.68773782e-01 2.51552433e-01 -1.03242552e+00
8.52179974e-02 -9.69549119e-01 4.33252722e-01 -2.02619672e-01
-1.60345185e+00 1.77992016e-01 1.96895704e-01 -1.01030171e+00
-4.39755470e-01 -6.25568926e-01 -8.02969187e-02 1.24215253e-01
-1.33667898e+00 -1.00425732e+00 2.24516660e-01 5.41072667e-01
1.85954422e-01 -4.62030113e-01 6.50401473e-01 9.91472602e-03
-6.78948879e-01 5.05163431e-01 5.38827717e-01 -1.45362094e-01
3.04732174e-01 -1.35110509e+00 -3.52851450e-01 8.39793205e-01
-9.17471051e-02 7.43245602e-01 9.39000189e-01 -3.28319520e-01
-2.09651303e+00 -8.92675579e-01 2.74019808e-01 -4.34266925e-01
1.44731557e+00 -4.15765047e-01 -3.26950788e-01 1.08284438e+00
5.82453702e-03 -2.75290340e-01 3.96641433e-01 2.68413633e-01
-4.35620248e-01 -3.58943909e-01 -8.79227638e-01 8.24606180e-01
9.91127908e-01 -4.60423887e-01 -4.98376101e-01 2.48010382e-01
6.08533919e-01 -1.08327940e-01 -9.06239748e-01 1.50695473e-01
6.74910963e-01 -6.01013899e-01 8.64884377e-01 -6.35344386e-01
-8.54458660e-02 -5.11515737e-01 -3.74523401e-01 -1.12674034e+00
-1.54211089e-01 -9.75894094e-01 -3.32139194e-01 9.18724239e-01
1.84363633e-01 -1.27339435e+00 2.51519412e-01 2.85132498e-01
3.56066376e-01 -5.70514202e-01 -1.31300330e+00 -1.17691469e+00
2.78966993e-01 -5.74325621e-02 1.62705824e-01 7.26167858e-01
2.55438358e-01 3.85735393e-01 -6.30661607e-01 5.20564198e-01
1.03909123e+00 1.16286092e-01 8.00287068e-01 -1.27946377e+00
-7.88089871e-01 -5.13914168e-01 -4.68703747e-01 -8.64589989e-01
3.97737414e-01 -6.93507850e-01 -5.42927347e-02 -1.10773301e+00
4.70506370e-01 -5.79724014e-01 -3.51317704e-01 4.19929504e-01
1.95363224e-01 -1.08343452e-01 4.94302839e-01 1.60990119e-01
-8.64226878e-01 6.33057535e-01 9.38261509e-01 -6.49571717e-02
-1.89792544e-01 2.12266654e-01 -5.83456337e-01 7.78160691e-01
7.09579468e-01 -5.99659979e-01 -7.35137522e-01 -5.14738820e-02
8.65303099e-01 5.59754716e-03 6.67345345e-01 -6.77890956e-01
4.45092291e-01 -2.67952353e-01 -2.63877809e-01 -3.74165773e-01
2.17312902e-01 -9.91705418e-01 1.76163092e-01 7.74379492e-01
-6.42098308e-01 9.21163335e-02 -2.14537069e-01 1.09153056e+00
4.02203202e-01 -3.20653975e-01 5.65493345e-01 1.90062121e-01
-1.31148815e-01 2.52786309e-01 -5.68895280e-01 8.28997120e-02
1.08369434e+00 2.31079623e-01 -5.08219719e-01 -6.73582196e-01
-5.39248407e-01 5.28603613e-01 8.00255090e-02 1.39268860e-01
2.31991932e-02 -1.05307114e+00 -5.90449214e-01 4.29330952e-02
-2.54179627e-01 -3.88500303e-01 1.67306051e-01 1.32979846e+00
-1.96641564e-01 3.80455434e-01 -1.53952399e-02 -2.83234864e-01
-9.36292231e-01 7.83875048e-01 6.96555734e-01 -4.50964987e-01
-4.96437967e-01 2.26071090e-01 2.53398299e-01 -4.26410228e-01
2.67994463e-01 -3.93901356e-02 8.95842463e-02 -2.07851484e-01
3.59265029e-01 7.21254230e-01 -4.89193678e-01 -2.66372770e-01
-3.05930257e-01 6.33146346e-01 -2.39973664e-01 -3.21572483e-01
1.52882862e+00 -4.65268105e-01 -1.77434117e-01 5.61966836e-01
5.85996866e-01 -1.68546051e-01 -1.41512918e+00 -3.19523960e-01
-4.09228764e-02 -8.41291621e-02 -2.78119624e-01 -4.34529752e-01
-9.01425600e-01 4.86075521e-01 4.34926152e-01 6.08419597e-01
8.60079169e-01 8.47634524e-02 -3.05493246e-03 1.10697746e+00
7.96786129e-01 -1.05201876e+00 -9.16253477e-02 6.37495875e-01
7.27540791e-01 -7.14932740e-01 2.56983023e-02 -1.96779937e-01
-2.28723079e-01 8.91532302e-01 3.88814509e-01 -4.25596744e-01
6.09215736e-01 3.23628426e-01 -4.19549286e-01 -2.48993095e-03
-1.02630389e+00 -8.73585269e-02 -3.13882351e-01 7.32745707e-01
-3.58325422e-01 1.54153276e-02 -4.21005011e-01 7.73657024e-01
-1.97518528e-01 -1.81165218e-01 6.90671086e-01 7.90261924e-01
-3.08258891e-01 -1.07041132e+00 -9.68918353e-02 1.19672783e-01
-2.04758331e-01 2.50300378e-01 -1.49145544e-01 1.10978329e+00
-4.63296622e-01 1.11126089e+00 -2.16453180e-01 1.23726107e-01
1.81625187e-01 -1.92790642e-01 8.50628436e-01 3.62964347e-02
-1.87018141e-01 5.67990392e-02 -1.30888477e-01 -4.14229304e-01
-7.04179645e-01 -6.34256601e-01 -9.24281418e-01 -4.89068151e-01
-4.48943257e-01 3.57075453e-01 2.92838901e-01 1.33547831e+00
1.79235131e-01 2.01389864e-01 9.35992002e-01 -4.59317863e-01
-1.25082862e+00 -7.45580018e-01 -8.28099132e-01 -3.04054290e-01
4.40383434e-01 -8.97651792e-01 -7.63891697e-01 -4.67328906e-01] | [3.784022808074951, 2.1405839920043945] |
66f35b3f-f970-48c2-99cc-4063171e10bc | real-time-3d-indoor-human-image-capturing | null | null | https://arxiv.org/abs/1812.07099 | https://arxiv.org/pdf/1812.07099.pdf | Real Time 3D Indoor Human Image Capturing Based on FMCW Radar | Compared to traditional camera-based computer vision and imaging, radio imaging based on wireless sensing does not require lighting and is friendly to privacy. This work proposes a deep learning radio imaging solution to visualize real-time user indoor activities. The proposed solution uses a low-power, MIMO Frequency Modulated Continuous Wave (FMCW) radar array to capture the reflected signals from human objects, and then constructs 3D human visualization through a serials of data analytics including: 1) a data preprocessing mechanism to remove background static reflection, 2) a signal processing mechanism to transfer received complex radar signals to a matrix containing spatial information, and 3) a deep learning scheme to filter abnormal frames resulted from rough surface of human body. This solution has been extensively evaluated in an indoor research lab. The constructed real-time human images are compared to the camera images captured at the same time. The results show that the proposed radio imaging solution can result in significantly high accuracy.
IEEE publication: "Real-Time Indoor 3D Human Imaging Based on MIMO Radar Sensing"
https://doi.org/10.1109/ICME.2019.00244 | ['Saeed ALI-AlQarni', 'Wenjun Shi', 'Shaoen Wu', 'Nan Zhang', 'Honggang Wang', 'Hangqing Guo'] | 2019-08-05 | null | null | null | 2019-ieee-international-conference-on-2 | ['rf-based-pose-estimation'] | ['computer-vision'] | [ 4.21456784e-01 -4.26309496e-01 8.37020636e-01 -2.21966848e-01
-3.26730102e-01 -4.23309244e-02 2.77215213e-01 -4.76420820e-01
-4.39666152e-01 4.83248800e-01 4.35100272e-02 -3.36026639e-01
-1.05533786e-01 -9.20211136e-01 -4.46652442e-01 -8.18314075e-01
-1.71426624e-01 3.13415155e-02 -3.27510536e-01 7.82784745e-02
-4.52200584e-02 4.54381973e-01 -1.38582087e+00 1.62007287e-01
5.31588554e-01 1.25550759e+00 3.46856534e-01 8.94951642e-01
2.99045026e-01 8.38825643e-01 -9.93553698e-01 5.78534186e-01
5.51279664e-01 -2.11917579e-01 2.58795589e-01 5.03473021e-02
3.22180659e-01 -7.09029675e-01 -6.13124907e-01 8.75213027e-01
8.34991693e-01 -5.98030575e-02 4.89484489e-01 -9.00305748e-01
-9.15481687e-01 -5.61086722e-02 -8.65724266e-01 2.75980026e-01
8.14932108e-01 -2.91552991e-02 -1.12940304e-01 -8.66852224e-01
9.82748568e-02 1.00438273e+00 9.25616741e-01 1.74507394e-01
-6.82312965e-01 -1.06981421e+00 -6.54073000e-01 3.29597145e-01
-1.38416624e+00 -8.48604366e-02 9.25011933e-01 -5.49190938e-01
5.90495884e-01 3.24889660e-01 1.02378654e+00 1.12263775e+00
8.88524294e-01 2.09744826e-01 1.45220578e+00 -6.91733181e-01
2.27244735e-01 3.33322138e-02 4.19193536e-01 9.96738970e-01
4.95167971e-01 4.02438074e-01 -2.69064039e-01 3.53106484e-02
8.73371840e-01 6.26489341e-01 -3.93346429e-01 -1.32020310e-01
-1.05595589e+00 6.56457245e-01 4.23923761e-01 4.52953011e-01
-6.24170244e-01 2.37396911e-01 -2.80618697e-01 3.01055104e-01
1.84305146e-01 1.73374377e-02 7.12993219e-02 4.84190822e-01
-8.02355826e-01 -1.36922255e-01 5.92256188e-01 4.56838429e-01
5.12397945e-01 4.59529370e-01 1.05714254e-01 4.80133116e-01
7.94992566e-01 1.30009055e+00 2.67336369e-01 -8.05936038e-01
1.00361660e-01 7.17308000e-02 2.54956752e-01 -1.38708448e+00
-6.62174702e-01 -6.88439190e-01 -1.06968653e+00 5.86489558e-01
-4.97563882e-03 -7.20147669e-01 -1.04929006e+00 1.09651625e+00
1.17845617e-01 3.65399480e-01 1.91483676e-01 1.11265910e+00
1.07993615e+00 1.07618499e+00 -3.92776549e-01 -1.74796402e-01
1.40244174e+00 -4.01693016e-01 -9.69960272e-01 -2.32734412e-01
-2.48006850e-01 -6.71848774e-01 5.84098220e-01 8.05069625e-01
-5.66761196e-01 -8.53949308e-01 -1.42143118e+00 3.60437065e-01
-2.06158087e-01 3.78915161e-01 5.06107092e-01 1.03179169e+00
-8.33550155e-01 -3.28231841e-01 -5.22277296e-01 -4.13860649e-01
8.96605253e-02 1.56523392e-01 -7.63039514e-02 -3.39149743e-01
-1.19468021e+00 7.61344910e-01 -1.21605493e-01 4.25777733e-01
-8.98856997e-01 -6.26669943e-01 -7.90798247e-01 -1.81986809e-01
9.12137423e-03 -7.91724920e-01 9.20272350e-01 -7.52164006e-01
-1.13465321e+00 6.51907682e-01 2.72151142e-01 -4.83773738e-01
2.39619240e-01 -4.71799284e-01 -7.55464494e-01 3.63811284e-01
-7.38264695e-02 1.44032780e-02 1.11664629e+00 -1.16994989e+00
-4.64046359e-01 -5.98652244e-01 -2.54590929e-01 1.11547939e-01
3.01035289e-02 -1.65018618e-01 -2.68951450e-02 -4.57321048e-01
1.11491404e-01 -6.65434659e-01 -8.55831355e-02 -3.86100747e-02
-1.02951102e-01 6.08279288e-01 1.17661941e+00 -9.94909227e-01
7.59012341e-01 -1.88450742e+00 -6.44617438e-01 5.00222564e-01
9.14672688e-02 3.23868662e-01 2.56685138e-01 1.67696193e-01
8.62940401e-02 -5.39308846e-01 -6.14446811e-02 1.15280144e-01
-3.96466583e-01 -3.86487335e-01 1.32543653e-01 8.02007854e-01
-5.36143124e-01 4.45109725e-01 -3.97706151e-01 -2.35648468e-01
6.01864994e-01 9.22382355e-01 1.22597963e-01 8.75969678e-02
4.87392515e-01 5.78289092e-01 -4.60981756e-01 9.22575176e-01
9.23614383e-01 1.31506413e-01 3.43991704e-02 -3.36235434e-01
-1.98720619e-01 -7.12536514e-01 -1.32834780e+00 1.23119223e+00
-5.87038398e-01 8.45439792e-01 6.23363316e-01 -1.00291264e+00
1.36553204e+00 3.34582001e-01 6.77278876e-01 -1.27920663e+00
3.87884319e-01 -2.81911135e-01 -4.43658322e-01 -8.19147766e-01
2.54371222e-02 -1.46323785e-01 -1.61063492e-01 3.86260778e-01
-3.68861616e-01 2.64590949e-01 -5.15951514e-01 -1.13197915e-01
1.45826864e+00 -3.31433192e-02 2.88740516e-01 -8.17442462e-02
6.39891148e-01 -1.74782460e-03 3.97062331e-01 7.76826441e-01
-5.39508574e-02 3.09934556e-01 -7.06036389e-01 -6.32947087e-01
-5.45406222e-01 -1.49391520e+00 1.17946699e-01 6.65057659e-01
3.64820421e-01 4.26764846e-01 -5.88816822e-01 -4.47917655e-02
-1.08274303e-01 5.87742209e-01 -3.53840202e-01 -8.95831920e-03
-4.36638743e-01 -7.21382558e-01 3.92332643e-01 1.74006209e-01
9.97599423e-01 -1.08225155e+00 -1.36932421e+00 5.60302064e-02
-1.43544689e-01 -9.03384447e-01 4.09313664e-02 6.04146644e-02
-5.41187465e-01 -9.86785352e-01 -7.43096471e-01 -7.80995011e-01
5.38532257e-01 8.74101043e-01 7.75617301e-01 -6.34236494e-03
-1.11666870e+00 1.09703565e+00 -3.21322918e-01 -9.45175946e-01
1.96295977e-01 -8.65277112e-01 1.95714951e-01 1.44778371e-01
5.73319554e-01 -4.31702048e-01 -6.68487310e-01 -7.07186311e-02
-5.38374066e-01 -1.19947284e-01 1.04020333e+00 5.66084862e-01
2.45786250e-01 4.18103933e-01 3.51107538e-01 -5.18278718e-01
5.53780556e-01 -3.16444933e-01 -7.97427952e-01 1.33196965e-01
-3.67128015e-01 -5.99317074e-01 4.56166983e-01 -2.40131274e-01
-1.45630002e+00 1.33692235e-01 3.30483556e-01 -3.13191861e-01
-4.09754843e-01 3.62232149e-01 -5.70204370e-02 -1.60446078e-01
7.86151409e-01 4.19094026e-01 2.20954254e-01 -2.50107765e-01
-1.08112497e-02 7.98876047e-01 8.43914270e-01 5.32071851e-02
1.09358287e+00 6.86876953e-01 3.55578810e-02 -1.56440687e+00
-5.05427063e-01 -4.61718231e-01 -3.18306088e-01 -7.39849210e-01
1.31172156e+00 -1.26040530e+00 -6.15316987e-01 5.10329485e-01
-9.14528847e-01 -5.00889979e-02 3.14631075e-01 9.67290521e-01
-2.19160259e-01 2.74449170e-01 -5.39866269e-01 -1.11921644e+00
-6.46525741e-01 -4.19842303e-01 7.08631575e-01 4.55538601e-01
6.53496012e-02 -7.26201653e-01 1.65631518e-01 8.00366998e-01
4.77315038e-01 7.09893048e-01 4.45502430e-01 2.76231691e-02
-7.01056778e-01 -3.80218208e-01 -7.55021647e-02 2.69495875e-01
1.29806608e-01 -4.55255419e-01 -9.94116664e-01 -4.02554840e-01
6.25263572e-01 2.17697993e-02 6.22956336e-01 8.56593668e-01
8.02296698e-01 -1.29359458e-02 -3.48365963e-01 7.73190081e-01
1.74727333e+00 8.18219662e-01 8.40298712e-01 2.33524546e-01
5.96514940e-01 3.98692489e-01 7.51719654e-01 5.34650445e-01
-1.01894811e-01 2.92801201e-01 4.43551928e-01 -4.42966729e-01
-3.98902483e-02 2.20521033e-01 6.34209812e-01 4.40954149e-01
-1.40878111e-01 -5.92121743e-02 -7.99182951e-01 -1.53029049e-02
-1.25151873e+00 -1.31113303e+00 -3.05511683e-01 1.87892199e+00
-6.89174831e-02 -1.67804554e-01 -2.39238590e-01 2.27995977e-01
6.83886230e-01 1.09667458e-01 -3.73542666e-01 -1.97186455e-01
2.59096175e-01 6.42334223e-01 6.65625989e-01 4.52516526e-01
-1.12682533e+00 1.33866474e-01 5.27268887e+00 -1.65377483e-02
-1.38321185e+00 1.87339410e-01 3.68743092e-02 -1.42866299e-01
-7.78988795e-03 -6.24968529e-01 -2.03086406e-01 3.07688206e-01
6.44554317e-01 1.62804589e-01 3.15351516e-01 7.31935441e-01
4.16111529e-01 -2.77916193e-01 -4.22269940e-01 1.47989190e+00
4.52463865e-01 -1.05979574e+00 -3.42901587e-01 2.56464928e-01
2.35319078e-01 -2.15751693e-01 1.71379820e-01 1.54707819e-01
1.20051410e-02 -1.11629689e+00 5.15370131e-01 1.02506924e+00
6.28328145e-01 -8.25301707e-01 7.48685360e-01 3.87855917e-01
-1.12145591e+00 -4.41564918e-01 -3.08918834e-01 -3.30648243e-01
4.08670120e-02 8.21551263e-01 -8.50581229e-01 4.63486284e-01
1.01613951e+00 3.23321939e-01 -3.26499701e-01 9.06313062e-01
1.35283008e-01 4.71127987e-01 -1.02147222e-01 -1.67039499e-01
-6.67811260e-02 -3.43262106e-01 4.44065303e-01 1.22941804e+00
9.61413980e-01 4.13145691e-01 2.10715026e-01 5.68929732e-01
4.60268378e-01 -3.38951230e-01 -1.10396171e+00 3.76385659e-01
4.19876188e-01 1.37162960e+00 -5.50906718e-01 -8.72200578e-02
-5.44903457e-01 8.03503633e-01 -6.48409545e-01 4.16035295e-01
-1.02549672e+00 -6.58329487e-01 2.14925066e-01 4.29083526e-01
3.59763801e-01 -5.49627006e-01 -1.02901727e-01 -7.20319390e-01
-9.52745676e-02 -6.46860957e-01 2.47777626e-01 -1.21986961e+00
-8.75205815e-01 3.14293563e-01 -1.61739185e-01 -1.16408503e+00
1.03003614e-01 -7.44776905e-01 -5.98370016e-01 7.26753712e-01
-1.00274563e+00 -1.01642466e+00 -1.02939403e+00 9.90714073e-01
3.91705811e-01 -5.14993310e-01 7.90825009e-01 5.10259569e-01
-2.90209681e-01 9.75598767e-02 6.65335953e-02 1.98679164e-01
4.75225657e-01 -8.49926889e-01 -2.09689096e-01 1.11927998e+00
2.72250138e-02 4.80018705e-01 7.32549727e-01 -7.58061886e-01
-1.92187417e+00 -1.00306129e+00 6.70351163e-02 6.44613132e-02
-6.99120685e-02 -4.02393460e-01 -4.30796713e-01 5.89258373e-01
4.65322971e-01 -8.39072913e-02 8.58960271e-01 -3.75236839e-01
3.93075794e-02 -5.81616044e-01 -1.67726421e+00 1.14327043e-01
5.33267498e-01 -7.01434091e-02 -8.14254642e-01 6.75470680e-02
2.60286897e-01 -1.06549516e-01 -4.67409581e-01 1.77426785e-01
9.12386060e-01 -1.10613644e+00 1.18553030e+00 3.74063432e-01
-1.65962502e-01 -5.66764712e-01 -3.78780305e-01 -9.16502357e-01
-6.66039705e-01 -3.36738944e-01 -9.13472194e-03 5.97886086e-01
-1.17123872e-01 -5.50417185e-01 7.37319827e-01 -6.51895478e-02
-6.70760572e-02 -1.69652939e-01 -6.24112427e-01 -4.64630991e-01
-8.39017272e-01 -4.59689468e-01 1.77303869e-02 7.18466938e-01
-6.12297118e-01 3.18172127e-01 -7.93939233e-01 8.05947781e-01
1.29374397e+00 -8.03285241e-02 7.79877365e-01 -1.55987203e+00
-3.46186399e-01 4.03622419e-01 -3.92564476e-01 -5.22772849e-01
-3.58966500e-01 -4.13899809e-01 1.11040883e-01 -1.91951442e+00
-1.12965718e-01 -9.35473442e-02 -2.98338532e-01 -5.62503561e-02
4.84991372e-01 3.92432600e-01 4.17010300e-02 -4.60279807e-02
-2.81880796e-01 1.14218615e-01 8.24593127e-01 -1.98540911e-01
2.66708191e-02 1.69393137e-01 -4.97560203e-01 8.41527998e-01
8.96808088e-01 -1.64832100e-01 -4.65696871e-01 -6.02243781e-01
-2.18212917e-01 4.41077858e-01 8.72250199e-01 -1.85964346e+00
4.34294313e-01 4.06479114e-04 1.36769187e+00 -8.33613753e-01
6.87394261e-01 -1.30493903e+00 5.04784644e-01 1.15442312e+00
2.27292344e-01 -1.77427346e-03 -6.12501688e-02 8.73583317e-01
2.28784755e-01 7.62940124e-02 8.97780836e-01 -4.65301007e-01
-9.58382070e-01 -1.76357776e-01 -9.81177330e-01 -7.28406250e-01
1.25692296e+00 -5.64461708e-01 -3.36204469e-01 -5.97729921e-01
-6.47367001e-01 -1.52282417e-01 -1.07233569e-01 9.29674134e-02
1.17006993e+00 -1.21245003e+00 -4.66762811e-01 5.08264303e-01
-2.90560007e-01 -5.01669228e-01 4.78630066e-01 4.73820537e-01
-1.03217471e+00 5.27605712e-01 -7.11826801e-01 -6.43593907e-01
-1.63636649e+00 4.03974235e-01 3.47296834e-01 1.46566302e-01
-8.88988316e-01 5.47712326e-01 -1.63207039e-01 -2.00362802e-01
4.15942609e-01 -3.02154440e-02 -2.66085684e-01 -1.74678966e-01
6.86338842e-01 6.16459250e-01 -1.48978457e-01 -4.04932797e-01
-5.74911833e-01 8.02304566e-01 3.10846031e-01 -4.02206510e-01
1.33928239e+00 -3.05422813e-01 1.40363723e-01 2.96001107e-01
8.90627861e-01 2.47652158e-01 -8.90737295e-01 -1.53739065e-01
-3.39054435e-01 -4.49269950e-01 2.29404047e-01 -1.12262034e+00
-1.08637393e+00 7.46074796e-01 1.58262968e+00 1.23427607e-01
1.54092884e+00 -4.83972371e-01 8.27533364e-01 6.26622200e-01
7.39607334e-01 -9.97072279e-01 3.28651398e-01 2.14027405e-01
7.15775073e-01 -8.59318852e-01 3.41782093e-01 -5.84096424e-02
-3.07023555e-01 1.15332973e+00 4.20945674e-01 -3.27444732e-01
8.37817669e-01 7.44879961e-01 5.22590756e-01 -3.22378427e-01
-1.61429495e-01 -3.47196087e-02 -1.27075642e-01 1.29598784e+00
1.75131768e-01 5.53370751e-02 7.92652741e-02 3.34546417e-01
-1.21900171e-01 2.69472510e-01 6.20133936e-01 1.09057403e+00
-9.54847515e-01 -2.37232804e-01 -1.26569045e+00 3.86532456e-01
-5.10817230e-01 3.67090255e-01 -1.97910413e-01 7.79198349e-01
4.23954278e-01 1.25442088e+00 9.87455323e-02 -5.32007337e-01
2.53430605e-01 -1.70247093e-01 6.40094817e-01 -3.69751900e-01
-2.00734362e-01 8.03290233e-02 -1.90676138e-01 -4.52273041e-01
-2.82773674e-01 -4.36873376e-01 -1.21733415e+00 -9.64553282e-02
1.72725618e-01 -4.17468324e-02 1.00492346e+00 6.72345340e-01
1.34245351e-01 7.98344553e-01 6.29076958e-01 -8.18157911e-01
-1.07427582e-01 -7.48031735e-01 -9.37697172e-01 -1.01857841e-01
5.04395127e-01 -6.45499527e-01 -3.09444338e-01 -2.29205191e-02] | [6.858942031860352, 0.5047256946563721] |
4c541186-dc51-4730-88be-e0a0ceb194c1 | masked-rpca-sparse-and-low-rank-decomposition | 1909.08049 | null | https://arxiv.org/abs/1909.08049v1 | https://arxiv.org/pdf/1909.08049v1.pdf | Masked-RPCA: Sparse and Low-rank Decomposition Under Overlaying Model and Application to Moving Object Detection | Foreground detection in a given video sequence is a pivotal step in many computer vision applications such as video surveillance system. Robust Principal Component Analysis (RPCA) performs low-rank and sparse decomposition and accomplishes such a task when the background is stationary and the foreground is dynamic and relatively small. A fundamental issue with RPCA is the assumption that the low-rank and sparse components are added at each element, whereas in reality, the moving foreground is overlaid on the background. We propose the representation via masked decomposition (i.e. an overlaying model) where each element either belongs to the low-rank or the sparse component, decided by a mask. We propose the Masked-RPCA algorithm to recover the mask and the low-rank components simultaneously, utilizing linearizing and alternating direction techniques. We further extend our formulation to be robust to dynamic changes in the background and enforce spatial connectivity in the foreground component. Our study shows significant improvement of the detected mask compared to post-processing on the sparse component obtained by other frameworks. | ['Amirhossein Khalilian-Gourtani', 'Yao Wang', 'Shervin Minaee'] | 2019-09-17 | null | null | null | null | ['moving-object-detection'] | ['computer-vision'] | [ 7.88918972e-01 -1.82802007e-01 1.99335262e-01 2.76036263e-01
-1.64558157e-01 -4.54104841e-01 4.78789002e-01 -5.15578032e-01
3.03962585e-02 5.87209821e-01 3.94617021e-01 3.98748219e-02
-1.28073990e-01 -4.13262099e-01 -5.06103754e-01 -1.28458512e+00
-1.71744600e-01 3.15568119e-01 5.90953350e-01 2.59191066e-01
-2.10154504e-02 4.93409187e-01 -1.54251206e+00 5.35534084e-01
6.49301291e-01 6.66298449e-01 2.60373205e-01 9.33929145e-01
-4.75015715e-02 1.01536334e+00 -4.18822199e-01 5.37043437e-02
6.48902476e-01 -4.11738932e-01 -2.88305581e-01 7.59933412e-01
6.42705798e-01 -1.73519522e-01 -4.37160105e-01 1.27238727e+00
-2.09061764e-02 9.30427760e-02 5.39057255e-01 -1.26428640e+00
-8.13643821e-03 2.52988279e-01 -1.39264405e+00 5.33096075e-01
2.93751031e-01 -2.27276161e-01 3.62345517e-01 -1.03444147e+00
7.49317288e-01 1.15044594e+00 4.59743470e-01 1.56377062e-01
-1.20373857e+00 -4.31519210e-01 6.18207037e-01 1.92900136e-01
-1.35624027e+00 -6.39359772e-01 1.15791798e+00 -6.52500093e-01
3.05615872e-01 5.36893010e-01 6.15902603e-01 6.86885476e-01
7.59183541e-02 7.38473773e-01 1.07608259e+00 -4.77030963e-01
1.69504732e-01 -2.32928008e-01 3.01186830e-01 6.74078822e-01
6.86247587e-01 -1.78148404e-01 -4.66191381e-01 -4.77004886e-01
8.24244380e-01 2.87793428e-01 -6.02506757e-01 -6.96046412e-01
-1.31816387e+00 3.96495402e-01 -2.56154448e-01 3.43586653e-01
-7.99752057e-01 -6.89546689e-02 -9.82007105e-03 -1.16107427e-01
2.94443697e-01 -3.11864793e-01 -2.91456580e-01 2.68856496e-01
-1.53704441e+00 1.83219381e-03 7.48265386e-01 8.28614175e-01
6.40961707e-01 3.02924037e-01 -2.96954103e-02 5.96494615e-01
4.71571952e-01 6.12342358e-01 2.23059967e-01 -1.17741215e+00
3.60878766e-01 5.42428732e-01 1.85364395e-01 -1.42209661e+00
-4.17624302e-02 -5.09949028e-01 -1.10011184e+00 3.60387385e-01
4.01812792e-01 -2.05709517e-01 -1.04409802e+00 1.50642347e+00
7.06708074e-01 1.00571167e+00 7.55657405e-02 9.34656143e-01
5.99753439e-01 8.85425091e-01 -2.10559636e-01 -9.82910514e-01
1.28309643e+00 -8.02868366e-01 -9.96534646e-01 -2.35561609e-01
-2.93475747e-01 -9.85441923e-01 1.67531997e-01 5.95269740e-01
-1.03961909e+00 -5.29121161e-01 -9.67546165e-01 2.40564600e-01
2.71169752e-01 3.31678897e-01 3.97359073e-01 5.33388019e-01
-9.99227524e-01 5.56380451e-02 -9.72241402e-01 -2.47229591e-01
6.18048720e-02 3.57297063e-01 -5.96145391e-01 -3.52485389e-01
-5.47317266e-01 4.80363399e-01 2.45748796e-02 4.75808531e-01
-1.08302140e+00 -3.68152201e-01 -6.11298978e-01 -1.80451855e-01
5.80632031e-01 -6.34142101e-01 4.98306215e-01 -1.38446474e+00
-1.14362061e+00 7.39107370e-01 -7.19544649e-01 -2.87493974e-01
6.05597734e-01 -1.78461343e-01 -3.20249230e-01 4.83183771e-01
1.26935080e-01 5.85448332e-02 1.55604208e+00 -1.68288302e+00
-6.81344986e-01 -3.65538031e-01 -2.65030801e-01 2.74775207e-01
2.33102992e-01 2.19410986e-01 -7.95391500e-01 -8.50352407e-01
8.50314379e-01 -1.02787805e+00 -4.44977373e-01 -1.17267057e-01
-1.35893434e-01 4.87569630e-01 1.40878046e+00 -1.11599851e+00
1.17492044e+00 -2.41701174e+00 5.70173085e-01 3.39327097e-01
5.12829840e-01 1.00245140e-01 6.26868755e-02 2.26338990e-02
-3.17801952e-01 -5.89281857e-01 -5.88718593e-01 -1.61977857e-01
-6.23847425e-01 3.25846791e-01 -1.67476505e-01 9.55107808e-01
1.16145186e-01 7.54566938e-02 -8.99104476e-01 -6.81508839e-01
2.45156810e-01 7.38529086e-01 -5.07010698e-01 1.30240843e-01
2.65858561e-01 5.34493804e-01 -3.03959489e-01 7.79974103e-01
1.17011452e+00 -1.44430533e-01 3.59474361e-01 -5.89819193e-01
-3.92007768e-01 -3.85885894e-01 -2.07931995e+00 1.27187777e+00
2.44590044e-01 6.80785596e-01 1.05902386e+00 -8.78943920e-01
5.05613267e-01 4.28189635e-01 9.89994884e-01 1.91409752e-01
1.26643684e-02 -2.84811109e-02 1.59896478e-01 -4.45360899e-01
5.17577708e-01 -3.66065130e-02 5.87724745e-01 2.37414036e-02
-1.12653099e-01 5.04554391e-01 4.17939484e-01 4.46245939e-01
1.11505556e+00 1.90462455e-01 3.68638992e-01 -4.37244445e-01
8.14581394e-01 2.47864313e-02 9.72114027e-01 5.05130112e-01
-2.56046981e-01 8.52041185e-01 4.07187045e-01 -2.55237609e-01
-6.70292199e-01 -8.16777050e-01 5.30839041e-02 8.32316697e-01
1.24550775e-01 -1.18527310e-02 -7.87151694e-01 -2.33247980e-01
-2.68426239e-01 7.64545426e-02 -6.55389011e-01 3.12170357e-01
-9.24486697e-01 -1.00575864e+00 -1.44272029e-01 7.02843517e-02
4.81914163e-01 -4.94372308e-01 -6.47699475e-01 2.63178647e-01
-3.53668511e-01 -1.22190464e+00 -4.90191460e-01 8.66290778e-02
-9.90036666e-01 -1.36301494e+00 -9.07809317e-01 -8.73680055e-01
1.12530160e+00 9.68731046e-01 8.74767661e-01 -1.73188299e-02
-1.48799300e-01 5.50550818e-01 -2.90223539e-01 6.42638952e-02
-5.70181049e-02 -7.29000986e-01 1.50181547e-01 9.60244954e-01
-1.06653990e-02 -6.21065736e-01 -4.14661229e-01 2.66281128e-01
-8.97607505e-01 1.73124373e-01 6.32529020e-01 5.22204220e-01
7.26390004e-01 5.87312877e-01 -3.23352456e-01 -9.41964090e-01
7.57593960e-02 -5.52934051e-01 -7.64887094e-01 2.08415881e-01
8.95393863e-02 -4.35332656e-01 1.29917979e-01 -6.83109164e-01
-1.18599200e+00 7.55417705e-01 4.99058276e-01 -9.27845836e-01
-1.91696305e-02 3.91791999e-01 -6.10202014e-01 -2.29007483e-01
2.38493070e-01 4.63447988e-01 -1.30671963e-01 -4.99915898e-01
2.99589247e-01 1.85097530e-01 6.74167335e-01 -4.42396611e-01
1.14968610e+00 1.12502408e+00 2.16779143e-01 -1.35292196e+00
-4.09588516e-01 -8.96573126e-01 -8.91132176e-01 -4.21214640e-01
7.92687595e-01 -1.12811518e+00 -2.31501192e-01 4.56242830e-01
-1.08526719e+00 7.88425505e-02 -1.44656792e-01 5.24035275e-01
-1.00640483e-01 8.87773454e-01 -4.37620014e-01 -1.09894562e+00
-1.48762437e-02 -1.05363321e+00 8.75174820e-01 1.27371699e-01
1.00586750e-01 -6.21602237e-01 8.37761760e-02 3.37296188e-01
1.44103453e-01 4.57205445e-01 4.24061894e-01 -1.91449627e-01
-8.23432565e-01 -2.25359827e-01 -8.02186877e-02 1.87687799e-01
2.13451058e-01 3.74628067e-01 -9.71451640e-01 -3.57453853e-01
5.64858794e-01 6.66120470e-01 8.84773672e-01 9.61706936e-01
4.22257513e-01 -4.37631696e-01 -4.20054704e-01 5.79013288e-01
1.33988297e+00 3.08862597e-01 5.85926235e-01 -4.50352281e-02
1.02987194e+00 7.66504526e-01 4.09584820e-01 3.37115586e-01
-4.80246022e-02 4.91373986e-01 1.48725495e-01 -3.06822926e-01
-3.85303527e-01 2.35139161e-01 5.17360866e-01 1.00612617e+00
-3.64676684e-01 5.80031909e-02 -7.86723733e-01 5.69520652e-01
-2.04952526e+00 -1.23011482e+00 -7.22983420e-01 2.18139434e+00
3.95939529e-01 -1.12247411e-02 4.14204337e-02 3.86407942e-01
1.02392805e+00 2.91639090e-01 -3.02316993e-01 3.23649913e-01
-5.43304145e-01 -1.45893112e-01 4.65143263e-01 8.20930064e-01
-1.29158378e+00 5.86822033e-01 6.64550734e+00 4.45344597e-01
-9.48975801e-01 1.35165825e-01 4.42363560e-01 -9.86503959e-02
1.45684838e-01 1.76198438e-01 -5.64875364e-01 5.21158338e-01
4.05698210e-01 6.46469817e-02 3.67875725e-01 6.94607079e-01
4.33522999e-01 -4.25622612e-01 -7.09134579e-01 9.49670851e-01
3.52157861e-01 -1.05917668e+00 9.10593793e-02 1.67872548e-01
1.00409055e+00 -3.49788547e-01 -8.16041678e-02 -2.75535762e-01
1.08556837e-01 -4.91089463e-01 6.51476860e-01 4.56527144e-01
4.10107464e-01 -4.31322247e-01 5.50809503e-01 2.48369515e-01
-1.55895638e+00 8.83630738e-02 -3.20644200e-01 -1.00259535e-01
2.74591625e-01 9.63113189e-01 -2.63867497e-01 4.55483258e-01
5.43746054e-01 7.85176694e-01 -1.73225507e-01 1.04763508e+00
1.46890106e-02 7.00141728e-01 -3.50083321e-01 8.48160684e-01
2.51092501e-02 -8.65737259e-01 1.05201590e+00 1.30686939e+00
1.22944504e-01 5.35493970e-01 5.74646056e-01 4.71567601e-01
3.10287923e-01 -5.05693518e-02 -4.56717998e-01 2.14615077e-01
9.07598510e-02 1.29982591e+00 -1.13832307e+00 -5.00536144e-01
-7.40133166e-01 1.08408451e+00 -3.85805130e-01 7.50835121e-01
-5.47775924e-01 2.90290326e-01 6.35852098e-01 2.41071045e-01
6.11874461e-01 -3.99890006e-01 -1.16304003e-01 -1.46308076e+00
1.41194895e-01 -1.18003643e+00 4.05319840e-01 -5.27619660e-01
-9.65405881e-01 3.73856694e-01 5.06123621e-03 -1.38734508e+00
1.50162041e-01 -5.13178349e-01 -5.75982928e-01 8.23278069e-01
-1.37458098e+00 -1.20599902e+00 -3.91686320e-01 9.51687336e-01
6.04220271e-01 -1.21873163e-01 2.23724499e-01 5.74569702e-01
-7.85324454e-01 -2.68497974e-01 1.23923466e-01 1.25846326e-01
2.86592752e-01 -8.00101459e-01 -3.76218259e-01 1.75566244e+00
2.07259610e-01 7.76902199e-01 9.55454648e-01 -1.06302416e+00
-1.60849047e+00 -9.05870199e-01 5.29811203e-01 -3.23248208e-01
6.79443240e-01 -1.22203350e-01 -9.74249482e-01 6.25863612e-01
1.32642910e-01 2.52399862e-01 6.93039298e-01 -3.16943049e-01
-1.38336690e-02 -4.88013104e-02 -1.04239261e+00 3.54518771e-01
7.62727976e-01 -2.26557821e-01 -5.63595116e-01 1.88989520e-01
2.78252751e-01 -3.35105866e-01 -3.39097202e-01 2.60000110e-01
4.69243705e-01 -9.36730146e-01 9.56563294e-01 -2.82668561e-01
4.19321060e-02 -1.21388590e+00 -4.37301338e-01 -6.25350416e-01
-7.18917608e-01 -7.78694510e-01 -6.78789377e-01 1.10084462e+00
2.63960883e-02 -1.47032708e-01 1.02783704e+00 5.10934651e-01
6.26114681e-02 -9.40852910e-02 -1.03569877e+00 -4.62384582e-01
-7.20407724e-01 -1.08705588e-01 -8.39009508e-02 1.12927377e+00
-5.14701366e-01 6.59415871e-02 -8.08766067e-01 8.87836814e-01
1.08538353e+00 1.73249438e-01 7.53548741e-01 -1.23835981e+00
-3.98288310e-01 -3.86143401e-02 -3.70795876e-01 -9.53262269e-01
-1.57009035e-01 -3.20791334e-01 1.43009335e-01 -1.39077950e+00
5.25362790e-01 -1.04183756e-01 -3.68923873e-01 4.73633520e-02
-3.08494598e-01 1.63774699e-01 3.20213437e-01 5.15719831e-01
-4.10957217e-01 1.09000333e-01 9.04665947e-01 -2.13181168e-01
-2.69580603e-01 2.57682372e-02 -4.88884956e-01 1.00948095e+00
2.04438761e-01 -5.29577434e-01 -3.89105618e-01 -3.32984746e-01
-1.52598113e-01 1.27195522e-01 1.52377099e-01 -9.44352984e-01
4.25943345e-01 -1.75744832e-01 5.61570764e-01 -6.73921645e-01
4.56249148e-01 -1.34165800e+00 8.19429159e-01 4.08367068e-01
2.76890159e-01 -8.03295232e-04 5.18132374e-02 9.38423038e-01
-2.88586229e-01 -1.41284138e-01 8.11691284e-01 -2.75334090e-01
-6.63426399e-01 1.44603044e-01 -6.96055830e-01 -2.80503899e-01
1.06910872e+00 -6.64433360e-01 9.53822955e-03 -2.54769206e-01
-8.64791036e-01 -1.31534755e-01 5.28547943e-01 -8.98532942e-02
6.86950922e-01 -9.84454155e-01 -9.26238894e-01 3.76819104e-01
-4.44973081e-01 1.24389296e-02 3.17149788e-01 1.12698531e+00
-7.13603377e-01 -5.18538356e-02 -1.40725166e-01 -7.24533498e-01
-1.68855262e+00 7.64091909e-01 9.79054049e-02 -2.09262863e-01
-7.71916211e-01 7.85144925e-01 5.77394485e-01 5.05517840e-01
3.55183184e-01 -1.92668423e-01 -5.10365784e-01 2.36659065e-01
8.76391411e-01 7.69022107e-01 -3.29489589e-01 -1.27869654e+00
-4.56671089e-01 8.73327434e-01 1.94014698e-01 -6.78789839e-02
1.28280580e+00 -2.03289345e-01 -6.49946511e-01 3.77790302e-01
7.84123540e-01 5.74146807e-01 -1.53654742e+00 -2.36208498e-01
-1.01416288e-02 -6.75230324e-01 2.75685430e-01 -1.97960734e-01
-1.43212509e+00 4.87680376e-01 6.84989035e-01 6.98368922e-02
1.31754935e+00 -3.86095881e-01 3.03898215e-01 7.58549795e-02
2.01116771e-01 -7.17115045e-01 -1.59516692e-01 1.92565188e-01
9.22405720e-01 -9.50318396e-01 5.53311527e-01 -8.67550015e-01
-5.87355196e-01 9.03038442e-01 2.13222444e-01 -3.71215761e-01
8.41504037e-01 5.04072428e-01 7.45985731e-02 -2.07523361e-01
-6.00335538e-01 -2.15088069e-01 4.29697365e-01 7.49909699e-01
2.88859516e-01 -5.58663718e-02 -7.52744600e-02 1.53037921e-01
4.52550322e-01 -2.02587157e-01 5.64796031e-01 1.00353396e+00
-5.18130660e-01 -6.93498075e-01 -1.16676259e+00 3.09444308e-01
-5.66867769e-01 -7.60371834e-02 -4.03081715e-01 3.91928643e-01
3.09219956e-01 1.14038861e+00 -1.64271444e-01 -1.04237385e-02
1.82548627e-01 -5.87582923e-02 3.71392965e-01 -5.67060232e-01
-2.68213183e-01 9.82185543e-01 -9.69429314e-02 -5.91293991e-01
-9.94081676e-01 -1.10736358e+00 -8.21826279e-01 1.05751054e-02
-2.77787089e-01 1.29602313e-01 3.77070785e-01 5.63574731e-01
7.80296996e-02 5.02041578e-01 3.66933376e-01 -1.12507606e+00
2.77409285e-01 -6.15935147e-01 -7.66353667e-01 2.95477629e-01
7.40618229e-01 -6.47644460e-01 -5.79824567e-01 8.31798732e-01] | [9.021259307861328, -0.8195126056671143] |
6210c00d-d9c9-4110-b2e1-c555f9231e97 | ai-imagery-and-the-overton-window | 2306.00080 | null | https://arxiv.org/abs/2306.00080v2 | https://arxiv.org/pdf/2306.00080v2.pdf | AI Imagery and the Overton Window | AI-based text-to-image generation has undergone a significant leap in the production of visually comprehensive and aesthetic imagery over the past year, to the point where differentiating between a man-made piece of art and an AI-generated image is becoming more difficult. Generative Models such as Stable Diffusion, Midjourney and others are expected to affect several major industries in technological and ethical aspects. Striking the balance between raising human standard of life and work vs exploiting one group of people to enrich another is a complex and crucial part of the discussion. Due to the rapid growth of this technology, the way in which its models operate, and gray area legalities, visual and artistic domains - including the video game industry, are at risk of being taken over from creators by AI infrastructure owners. This paper is a literature review examining the concerns facing both AI developers and users today, including identity theft, data laundering and more. It discusses legalization challenges and ethical concerns, and concludes with how AI generative models can be tremendously useful in streamlining the process of visual creativity in both static and interactive media given proper regulation. Keywords: AI text-to-image generation, Midjourney, Stable Diffusion, AI Ethics, Game Design, Digital Art, Data Laundering | ['Sarah K. Amer'] | 2023-05-31 | null | null | null | null | ['ethics'] | ['miscellaneous'] | [ 3.93932134e-01 5.00569999e-01 3.16223204e-01 1.74234867e-01
-4.45814542e-02 -8.90227795e-01 9.74009097e-01 -1.87670901e-01
-1.38825610e-01 5.24304748e-01 5.42375803e-01 -3.12283099e-01
-8.14476907e-02 -7.29787230e-01 -2.88128138e-01 -3.89242232e-01
4.50153977e-01 4.15051132e-01 -1.63772598e-01 -4.44055259e-01
4.70418364e-01 3.47377926e-01 -1.63958621e+00 1.10875398e-01
9.01339352e-01 9.50629830e-01 -1.79252490e-01 7.49642193e-01
-2.25366324e-01 1.21657538e+00 -7.43211985e-01 -1.13264430e+00
5.01649022e-01 -8.69382143e-01 -5.94243526e-01 1.82137653e-01
2.75286108e-01 -2.05537111e-01 2.14546263e-01 9.90045905e-01
5.18513918e-01 -2.91349590e-01 5.04376650e-01 -1.48338318e+00
-1.20019364e+00 4.02610093e-01 -6.71778619e-01 -2.69119978e-01
2.35837594e-01 6.59773052e-01 4.78016257e-01 -5.02647877e-01
1.21930003e+00 1.06301987e+00 4.93379980e-01 4.60046947e-01
-1.12238622e+00 -8.58634412e-01 -3.76567155e-01 -2.10123032e-01
-1.31234169e+00 -5.69762647e-01 6.07364178e-01 -1.11674225e+00
6.64948046e-01 4.92710203e-01 1.36033010e+00 8.66353452e-01
3.24288189e-01 2.56134272e-01 1.04550183e+00 -5.53616047e-01
5.39929792e-02 3.69166255e-01 -6.62928283e-01 3.11760336e-01
5.34210920e-01 -1.82013456e-02 -5.73605955e-01 1.97814987e-03
1.16454554e+00 -4.37503815e-01 9.56568494e-02 -2.21667707e-01
-1.08297563e+00 7.49585986e-01 1.08195725e-03 5.11486948e-01
-6.15127146e-01 3.85814697e-01 4.21203643e-01 8.94209296e-02
4.13572490e-01 7.33296573e-01 5.16484082e-01 -9.91835535e-01
-7.78368533e-01 3.23813766e-01 4.80357349e-01 6.53492570e-01
4.02003437e-01 3.05078924e-01 4.74263169e-02 5.24752378e-01
2.64163613e-01 3.58563364e-01 1.84777334e-01 -1.16103625e+00
-1.11606628e-01 7.80130863e-01 7.51120597e-02 -1.20576739e+00
3.14063400e-01 -1.78787723e-01 -5.80959499e-01 9.76734877e-01
1.49122834e-01 -4.53853369e-01 -8.91758502e-01 1.20128679e+00
2.25105107e-01 -6.27639234e-01 -2.20973611e-01 9.13041055e-01
5.61370015e-01 6.25419378e-01 3.64957631e-01 3.63612697e-02
1.22672844e+00 -4.03318018e-01 -8.66709232e-01 -5.04965603e-01
2.19390705e-01 -9.69515979e-01 1.07253850e+00 5.17328560e-01
-1.52919281e+00 -1.37508571e-01 -1.07215762e+00 -2.45549932e-01
-1.36307910e-01 -2.34724045e-01 5.38783193e-01 1.19573474e+00
-1.03316033e+00 2.60041863e-01 -2.73575872e-01 -6.44226432e-01
1.07731545e+00 -2.99446750e-02 -2.57875413e-01 9.45375040e-02
-5.81080854e-01 8.19718361e-01 1.67947769e-01 -4.12967876e-02
-2.54491061e-01 -5.73440611e-01 -4.77014422e-01 -3.36329281e-01
3.75644356e-01 -7.88162887e-01 9.49765980e-01 -1.79913843e+00
-1.28665400e+00 1.26013553e+00 5.03484726e-01 -2.87765741e-01
8.69389772e-01 -2.94101655e-01 -4.09023255e-01 -1.60826966e-01
3.62913698e-01 7.27696121e-01 8.43927801e-01 -1.39132440e+00
-6.50636017e-01 -2.00025216e-01 2.69193873e-02 2.85220265e-01
-1.10890009e-01 3.74493718e-01 -6.31129835e-03 -5.91149688e-01
-4.39106435e-01 -7.50668824e-01 -1.09655466e-02 1.62367314e-01
-1.59083456e-01 1.55816495e-01 9.30927157e-01 -6.31961882e-01
1.17620409e+00 -2.38963270e+00 -1.49318382e-01 3.12060416e-01
5.65209746e-01 2.53637731e-01 5.09085596e-01 7.81073689e-01
2.80755430e-01 6.57598019e-01 1.33284241e-01 1.77854836e-01
6.39622435e-02 -3.86993065e-02 -3.11818942e-02 2.83934977e-02
-5.17389812e-02 1.02249718e+00 -8.36790800e-01 -4.74751264e-01
-9.86325219e-02 6.49157405e-01 -3.06684196e-01 -2.06710324e-01
1.19617902e-01 4.54850405e-01 -1.40120730e-01 5.38180649e-01
4.04612571e-01 -2.82644504e-03 1.34499595e-01 2.78208822e-01
-6.48463309e-01 -4.14190352e-01 -8.27307463e-01 1.40237582e+00
-2.24263191e-01 1.13800204e+00 1.69107139e-01 -9.35516972e-03
8.92895699e-01 3.71213436e-01 3.04275662e-01 -9.53167856e-01
4.19948429e-01 1.12052873e-01 3.24526727e-01 -5.51083267e-01
8.38368833e-01 -5.95086336e-01 -3.60856801e-02 7.44720936e-01
-6.33769512e-01 -5.84290326e-01 -4.03879061e-02 3.06250423e-01
8.06135058e-01 4.88388598e-01 1.86309338e-01 4.66929469e-03
-1.50015727e-01 4.56622064e-01 3.71007949e-01 1.57508820e-01
-3.12333018e-01 5.62777638e-01 7.22865045e-01 -5.04367352e-01
-1.48066473e+00 -9.66412663e-01 3.39977324e-01 9.25548792e-01
7.09737316e-02 -4.19447303e-01 -1.03108704e+00 -1.71927825e-01
-1.86100960e-01 1.08100915e+00 -5.27059853e-01 -2.92053789e-01
-1.04496494e-01 -3.59070785e-02 6.46453261e-01 1.54549196e-01
6.30306721e-01 -1.31010890e+00 -1.44871473e+00 2.14948133e-01
7.89904222e-02 -8.56155038e-01 -3.46324801e-01 -6.27601027e-01
-3.27834547e-01 -7.43220091e-01 -6.20844603e-01 -4.69475001e-01
6.16263926e-01 -4.71327193e-02 1.06537366e+00 -8.35522041e-02
-6.08278215e-01 4.64098990e-01 -3.76332194e-01 -1.07869208e+00
-7.93703556e-01 -5.17577052e-01 -5.26802242e-01 1.35730002e-02
1.60142705e-01 -7.52681792e-01 -6.69067979e-01 1.39556363e-01
-1.06512237e+00 6.53784990e-01 3.94061029e-01 3.04348826e-01
1.45088986e-01 1.07518295e-02 6.16350293e-01 -9.20150578e-01
9.46629643e-01 -3.68185282e-01 -1.33975763e-02 2.70563737e-02
-6.50412083e-01 -4.45613801e-01 1.18372008e-01 -4.08956528e-01
-1.30536330e+00 -2.67100662e-01 3.78863364e-01 -9.52781886e-02
2.54499400e-03 3.02597374e-01 -7.05336928e-02 5.36497794e-02
1.08987951e+00 -2.98153698e-01 4.21174824e-01 6.75861239e-02
6.06360793e-01 7.21316874e-01 6.41870975e-01 -1.99823633e-01
8.68039191e-01 7.49689639e-01 -2.58226722e-01 -8.29821885e-01
1.64155081e-01 1.11322068e-01 -2.60486960e-01 -1.09702551e+00
1.15951812e+00 -5.80378175e-01 -4.75763947e-01 5.17847180e-01
-8.83136272e-01 -2.73272157e-01 -1.00662267e+00 3.14071798e-03
-4.88053769e-01 -6.81206733e-02 -6.07243851e-02 -1.00923538e+00
-4.43305701e-01 -8.59886885e-01 5.52383423e-01 4.69846249e-01
-6.65486872e-01 -5.58007240e-01 3.15257870e-02 8.13960850e-01
6.02488995e-01 1.04696953e+00 7.60026515e-01 1.07251823e-01
-5.78651667e-01 -4.59823847e-01 -3.19667876e-01 2.33230919e-01
3.13643306e-01 3.96916211e-01 -7.22982883e-01 3.33273947e-01
-8.30676705e-02 -3.62218738e-01 -1.63277462e-01 1.90784678e-01
4.33003381e-02 -5.18596351e-01 1.27397501e-03 1.42065376e-01
1.50310576e+00 8.58774364e-01 1.23267746e+00 5.14890015e-01
7.04484999e-01 6.88713491e-01 6.70486033e-01 6.32933617e-01
1.89481363e-01 1.02893777e-01 2.53759861e-01 -1.61426410e-01
-1.53532103e-01 -2.89019257e-01 5.30934036e-02 3.33725929e-01
-8.15182686e-01 4.85072359e-02 -1.22853422e+00 5.72415292e-01
-1.80308092e+00 -1.13450670e+00 -2.43641391e-01 2.09527755e+00
5.45949340e-01 3.51576984e-01 5.10054529e-01 3.44576910e-02
5.93514264e-01 -1.19325928e-01 -3.29347163e-01 -1.02278435e+00
-1.60232000e-02 1.35587588e-01 4.15978581e-01 -3.67604122e-02
-3.57292324e-01 9.17405367e-01 6.01108456e+00 7.00402856e-01
-1.07577646e+00 3.16688091e-01 8.76420319e-01 -1.58003896e-01
-5.55292249e-01 1.14193738e-01 8.10226649e-02 3.24866265e-01
4.59348470e-01 -7.54472971e-01 5.05869687e-01 5.93283653e-01
4.01205212e-01 -5.56880653e-01 -5.93245924e-01 1.06413531e+00
2.72744238e-01 -1.44881213e+00 -9.32867825e-02 3.75667483e-01
9.09215689e-01 -5.43423831e-01 4.27936256e-01 -4.37051117e-01
5.18797100e-01 -1.28223729e+00 1.41593766e+00 4.72698212e-01
1.04798198e+00 -9.09285486e-01 2.86430269e-01 -1.22191355e-01
-8.04820240e-01 -5.91307022e-02 1.47848397e-01 -5.49917281e-01
4.68086183e-01 2.58996248e-01 -6.61112845e-01 1.16023675e-01
6.68895006e-01 5.15484884e-02 -4.25724149e-01 9.12526786e-01
6.40473142e-02 2.44811013e-01 1.09577633e-01 5.11309803e-02
1.41949922e-01 -4.87317055e-01 6.64690733e-01 7.97510684e-01
3.66828173e-01 1.03429213e-01 -6.92992926e-01 1.12133133e+00
2.85230786e-01 1.80503443e-01 -9.74528790e-01 -8.41594040e-01
2.90996164e-01 1.14275587e+00 -1.01399386e+00 3.25748809e-02
-1.21212520e-01 1.09193814e+00 -3.60528290e-01 1.13243997e-01
-8.04663301e-01 -3.91839206e-01 7.19434738e-01 9.05758560e-01
-1.11297019e-01 -3.65790546e-01 -7.79088378e-01 -3.20371479e-01
-1.73243806e-01 -1.08482504e+00 -2.67824382e-01 -1.21978295e+00
-8.86132419e-01 3.93148750e-01 -1.11439742e-01 -9.53343391e-01
-3.08487535e-01 -1.02728918e-01 -6.25305951e-01 7.02985048e-01
-3.24892551e-01 -1.64295173e+00 -3.14280421e-01 -1.09922709e-02
2.80774176e-01 -1.41533807e-01 3.54458809e-01 8.17050412e-02
-8.83402452e-02 3.13444048e-01 1.01129208e-02 -1.40037239e-01
5.31410515e-01 -7.29268014e-01 5.91006160e-01 7.18482375e-01
-6.59505054e-02 3.68283689e-01 6.82405174e-01 -9.62775707e-01
-1.35520673e+00 -4.43327814e-01 5.86046278e-01 -4.83644634e-01
5.14360547e-01 -4.10880417e-01 -1.42193228e-01 5.63462317e-01
7.37962723e-01 -1.00842917e+00 6.73686266e-01 -3.65163505e-01
5.30254468e-02 1.25059023e-01 -1.27203965e+00 1.18725204e+00
1.27503967e+00 -4.69077766e-01 -1.50296912e-01 -1.37902185e-01
4.58327353e-01 -1.06617741e-01 -6.48073137e-01 -3.37729394e-01
9.85002875e-01 -1.22773218e+00 5.69471776e-01 -6.90122396e-02
7.19522297e-01 -1.30536139e-01 4.59188759e-01 -9.81739759e-01
-4.75280046e-01 -1.35219419e+00 7.85240471e-01 1.69954860e+00
3.31049621e-01 -5.14975846e-01 8.03568363e-01 1.33178806e+00
-1.15345046e-01 -2.73754179e-01 -7.19531059e-01 -4.87207979e-01
-1.38399035e-01 -5.43919921e-01 5.90567112e-01 1.02514303e+00
2.15011626e-01 3.63711655e-01 -4.65193361e-01 -5.50695062e-01
3.71720552e-01 -5.02319336e-01 1.01447451e+00 -1.32713568e+00
4.27092053e-02 -6.50184155e-01 -9.14869726e-01 5.01821898e-02
-8.64412606e-01 -4.44932193e-01 -2.48721033e-01 -2.03171062e+00
9.10945295e-04 -1.04290508e-01 6.75461113e-01 1.38302907e-01
3.64456207e-01 6.21340215e-01 7.75221646e-01 2.96895981e-01
-1.21210866e-01 -8.28665942e-02 1.65796781e+00 9.32148919e-02
-3.77842218e-01 -6.67161345e-01 -1.29160285e+00 6.39130890e-01
5.95997870e-01 -1.46987915e-01 -7.73031950e-01 -1.18358858e-01
1.12954295e+00 -3.61648977e-01 4.08949226e-01 -1.10173488e+00
6.38144836e-02 -4.18954849e-01 2.62628615e-01 6.33492768e-02
2.95574307e-01 -9.59980845e-01 1.29125440e+00 3.92321289e-01
1.55160844e-01 1.23002328e-01 1.61411881e-01 4.80119139e-02
1.86432227e-01 -1.55654803e-01 5.37402034e-01 -2.41171703e-01
-4.55292612e-01 -3.00118625e-01 -6.37928367e-01 1.77708060e-01
1.46203077e+00 -1.20555055e+00 -4.32315856e-01 -8.34228754e-01
-4.57654595e-01 -1.88450336e-01 1.09772754e+00 4.08119589e-01
4.81316745e-01 -1.31387091e+00 -6.72819793e-01 1.38598651e-01
-4.23829593e-02 -2.34668739e-02 3.89598966e-01 4.89045411e-01
-1.09313285e+00 -2.27661222e-01 -8.11469316e-01 -1.72602723e-03
-1.18469381e+00 3.40931952e-01 6.11100439e-03 2.75017291e-01
-6.80714130e-01 7.83709884e-01 4.96808976e-01 6.05185866e-01
-3.16789180e-01 9.31127518e-02 8.43991265e-02 2.28817105e-01
2.66538292e-01 7.48331428e-01 -6.09480560e-01 -1.02939439e+00
1.74810626e-02 4.39013302e-01 2.25759283e-01 -6.61229789e-01
1.06997669e+00 -2.06235930e-01 -1.35443434e-01 5.63160539e-01
4.59083706e-01 1.78603828e-01 -1.17968690e+00 6.97512448e-01
-3.28872621e-01 -9.40294802e-01 -1.79372773e-01 -1.17070484e+00
-9.65695143e-01 6.54749155e-01 6.00989699e-01 6.11248493e-01
1.09091103e+00 7.05830678e-02 7.84050643e-01 -6.36920929e-01
1.61218405e-01 -1.51713943e+00 1.67484820e-01 -2.45156467e-01
1.26617920e+00 -5.76807618e-01 7.04525113e-02 -3.59929472e-01
-1.41604924e+00 6.24195814e-01 5.62703431e-01 1.78374231e-01
3.77395093e-01 4.32825774e-01 1.98235750e-01 -4.54195440e-01
-2.92710811e-01 -6.87762350e-02 -3.90319377e-02 1.14635992e+00
4.29855257e-01 1.94321379e-01 -6.29480779e-01 4.01699573e-01
-4.97958481e-01 4.53357697e-01 7.53301024e-01 1.08372903e+00
-2.13835299e-01 -1.06092680e+00 -6.93271458e-01 3.64100099e-01
-7.01383173e-01 1.67991057e-01 -1.05939114e+00 9.61526930e-01
7.11203814e-01 7.79408872e-01 3.72807115e-01 -4.44510072e-01
1.56186417e-01 8.65535736e-02 2.90880740e-01 -2.23648831e-01
-1.06328797e+00 1.68930933e-01 6.23618484e-01 -5.53369932e-02
-2.26708144e-01 -7.76593447e-01 -1.11705768e+00 -7.33272076e-01
-4.31527980e-02 -2.20322043e-01 1.00174427e+00 4.13764209e-01
7.54473090e-01 3.55044216e-01 5.92757016e-03 -6.28646672e-01
4.97746497e-01 -5.46721339e-01 -7.47827947e-01 3.34731758e-01
-3.64887595e-01 -2.79632181e-01 1.34096786e-01 4.42346990e-01] | [9.326991081237793, 6.3547234535217285] |
61a8593e-17af-4a03-bd2d-6b9924973eb6 | mvtn-multi-view-transformation-network-for-3d | 2011.13244 | null | https://arxiv.org/abs/2011.13244v3 | https://arxiv.org/pdf/2011.13244v3.pdf | MVTN: Multi-View Transformation Network for 3D Shape Recognition | Multi-view projection methods have demonstrated their ability to reach state-of-the-art performance on 3D shape recognition. Those methods learn different ways to aggregate information from multiple views. However, the camera view-points for those views tend to be heuristically set and fixed for all shapes. To circumvent the lack of dynamism of current multi-view methods, we propose to learn those view-points. In particular, we introduce the Multi-View Transformation Network (MVTN) that regresses optimal view-points for 3D shape recognition, building upon advances in differentiable rendering. As a result, MVTN can be trained end-to-end along with any multi-view network for 3D shape classification. We integrate MVTN in a novel adaptive multi-view pipeline that can render either 3D meshes or point clouds. MVTN exhibits clear performance gains in the tasks of 3D shape classification and 3D shape retrieval without the need for extra training supervision. In these tasks, MVTN achieves state-of-the-art performance on ModelNet40, ShapeNet Core55, and the most recent and realistic ScanObjectNN dataset (up to 6% improvement). Interestingly, we also show that MVTN can provide network robustness against rotation and occlusion in the 3D domain. The code is available at https://github.com/ajhamdi/MVTN . | ['Bernard Ghanem', 'Silvio Giancola', 'Abdullah Hamdi'] | 2020-11-26 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Hamdi_MVTN_Multi-View_Transformation_Network_for_3D_Shape_Recognition_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Hamdi_MVTN_Multi-View_Transformation_Network_for_3D_Shape_Recognition_ICCV_2021_paper.pdf | iccv-2021-1 | ['3d-shape-retrieval', '3d-shape-recognition', '3d-classification', 'multi-view-3d-shape-retrieval', '3d-object-retrieval'] | ['computer-vision', 'computer-vision', 'computer-vision', 'computer-vision', 'computer-vision'] | [-2.74404794e-01 -2.72491395e-01 9.66356024e-02 -5.45843959e-01
-8.70797038e-01 -9.17977810e-01 8.42888415e-01 -4.17321950e-01
1.02770858e-01 -2.09268153e-01 1.19852237e-01 -2.96958297e-01
1.48581401e-01 -8.47366154e-01 -8.55382860e-01 -5.08150160e-01
2.37823054e-01 8.79298508e-01 7.00274110e-02 -1.87023386e-01
3.09204403e-02 1.15659690e+00 -1.44387639e+00 5.92574477e-01
1.96091905e-01 1.09089482e+00 -1.77525356e-02 7.18715787e-01
-1.78646132e-01 1.39378384e-01 5.60719669e-02 -4.08006221e-01
6.72859967e-01 4.48260039e-01 -5.63040912e-01 2.06494808e-01
1.26648188e+00 -5.50159037e-01 -4.31028217e-01 7.01076746e-01
8.96638811e-01 -1.56413585e-01 4.81951863e-01 -1.05783713e+00
-9.03552175e-01 4.71903533e-02 -6.72555745e-01 -3.50596616e-04
2.00425863e-01 2.82335490e-01 9.87789869e-01 -1.70603848e+00
6.32761240e-01 1.60439515e+00 6.64364576e-01 6.65612519e-01
-1.21939933e+00 -6.25312626e-01 3.40767443e-01 -2.23618433e-01
-1.14354146e+00 -6.67344093e-01 8.57512355e-01 -3.43501121e-01
1.30101001e+00 2.13902429e-01 8.18176568e-01 1.03949964e+00
3.00105866e-02 8.61912787e-01 9.03497934e-01 1.17576785e-01
-2.24569276e-01 -2.72905588e-01 -3.03700209e-01 7.75584936e-01
4.51140525e-03 8.47587511e-02 -5.42305291e-01 -3.52103025e-01
9.23908591e-01 1.03377581e-01 -1.90469138e-02 -1.10913229e+00
-1.00555658e+00 4.88767922e-01 3.96496773e-01 -1.20330215e-01
-2.61870414e-01 2.11904883e-01 3.81134540e-01 3.09353709e-01
1.00306106e+00 -4.22768891e-02 -7.65775621e-01 -4.53585107e-03
-4.98893052e-01 2.74999976e-01 8.13114583e-01 1.08643031e+00
5.53210497e-01 3.28163981e-01 2.04443321e-01 1.01251721e+00
4.98160034e-01 1.05035591e+00 -3.12259316e-01 -9.53893423e-01
7.61982322e-01 8.18279028e-01 -3.06414127e-01 -9.06988680e-01
-4.14523453e-01 -5.47345221e-01 -7.43527770e-01 4.09507692e-01
1.74732432e-01 1.60205901e-01 -1.11915171e+00 1.39552486e+00
6.11807585e-01 1.70093834e-01 -1.33615449e-01 8.90193820e-01
1.27355909e+00 3.17542523e-01 -4.84406054e-01 5.57372391e-01
1.12023532e+00 -8.87355149e-01 1.52959928e-01 -1.89543903e-01
4.48200107e-01 -8.49769294e-01 1.13840449e+00 3.40478808e-01
-1.49236345e+00 -3.41463298e-01 -8.35861921e-01 -2.30102703e-01
-2.05738455e-01 1.80489514e-02 5.18326283e-01 4.42003280e-01
-1.23503113e+00 5.48023343e-01 -1.15018928e+00 -2.57291496e-01
8.42943728e-01 3.78134042e-01 -4.85073566e-01 -3.44391286e-01
-3.21443111e-01 7.31310546e-01 -2.89899617e-01 1.31231785e-01
-9.37302649e-01 -1.26570666e+00 -7.14915156e-01 -9.13582444e-02
2.70693123e-01 -1.20565379e+00 1.13156974e+00 -4.67084914e-01
-1.40044117e+00 1.28192174e+00 -2.11210638e-01 4.84349281e-02
4.69369471e-01 -1.91315398e-01 -1.38052672e-01 1.79973647e-01
-1.79984778e-01 6.49200976e-01 7.97228396e-01 -1.53555346e+00
-2.46299669e-01 -8.07806313e-01 5.18789515e-02 4.21722949e-01
-2.26933956e-02 -1.18694372e-01 -7.32352674e-01 -2.73340553e-01
6.69300497e-01 -1.19130766e+00 4.66000587e-02 4.89249229e-01
-1.70092776e-01 -2.67427891e-01 1.12941504e+00 -3.36676836e-01
1.44044802e-01 -2.11155367e+00 2.13940337e-01 1.92147866e-01
4.08992469e-01 2.68880576e-01 -3.48379821e-01 2.73556739e-01
-1.01520352e-01 9.15123746e-02 -4.66932096e-02 -7.25335956e-01
-1.58018079e-02 3.52590680e-01 -3.68356019e-01 6.27143264e-01
1.11789227e-01 1.11872268e+00 -5.10081172e-01 1.26306340e-01
4.31676865e-01 9.51993585e-01 -6.52694404e-01 4.09258567e-02
-2.79212236e-01 5.34442902e-01 -3.69164437e-01 8.69676054e-01
9.79871750e-01 -7.06792176e-01 -1.42467245e-01 -4.24138606e-01
-1.99009385e-02 3.97375733e-01 -1.05675673e+00 2.12093449e+00
-6.60248578e-01 3.29323590e-01 2.66687691e-01 -5.78269601e-01
8.74571562e-01 3.51045758e-01 8.10133219e-01 -4.63626772e-01
-7.57591501e-02 2.71767944e-01 -1.87665805e-01 -1.11159906e-01
2.41302654e-01 6.72924966e-02 3.92346501e-01 5.63114583e-01
-6.04363950e-03 -5.77991784e-01 -3.70768934e-01 5.15678786e-02
7.43365586e-01 4.21566039e-01 -1.24966621e-01 -8.01201761e-02
3.87274683e-01 -3.21334779e-01 2.39343867e-01 3.13460141e-01
3.01409274e-01 9.67582881e-01 1.13029554e-01 -8.71437669e-01
-1.29572737e+00 -1.34596968e+00 -2.10789323e-01 1.06897950e+00
-1.55325219e-01 -2.28230134e-01 -2.22444728e-01 -7.21262038e-01
4.68029022e-01 3.57043654e-01 -2.86089569e-01 2.32092008e-01
-8.80822182e-01 -3.07844877e-01 5.87386250e-01 6.62730873e-01
3.91167372e-01 -6.90534532e-01 -4.14034992e-01 -1.09536834e-01
2.45859846e-01 -1.34888923e+00 -6.39115453e-01 -1.58869609e-01
-1.16824532e+00 -1.03712618e+00 -5.28674245e-01 -4.17748779e-01
7.69954741e-01 7.05847979e-01 1.45793724e+00 1.47710502e-01
6.28996864e-02 9.47362006e-01 2.24963557e-02 -4.38369602e-01
-3.12536210e-01 2.88436741e-01 -2.24026628e-02 -1.88169017e-01
1.54909700e-01 -1.18402815e+00 -7.03785598e-01 4.17695612e-01
-6.02359951e-01 1.32850438e-01 2.81509757e-01 5.37559867e-01
9.16401982e-01 -9.83130991e-01 1.31356761e-01 -7.51688361e-01
1.07781351e-01 -1.73972204e-01 -8.01717997e-01 1.54602513e-01
-5.68990886e-01 -1.50430560e-01 4.71498787e-01 -1.88312382e-01
-7.20225394e-01 1.85178265e-01 -3.94241214e-01 -1.15264535e+00
-3.02424788e-01 1.88649133e-01 -2.32176334e-01 -4.59751517e-01
4.27453935e-01 3.61792684e-01 1.95445091e-01 -7.95242906e-01
5.87439001e-01 3.14512461e-01 1.94652215e-01 -6.14811242e-01
9.58588421e-01 9.97741044e-01 2.25737214e-01 -7.40382969e-01
-9.38664615e-01 -3.95762533e-01 -8.10868204e-01 -1.39095873e-01
6.09709680e-01 -1.21552074e+00 -8.74164462e-01 5.36286831e-01
-1.17636633e+00 -2.39048451e-01 -1.38279736e-01 1.25675127e-01
-8.66304934e-01 2.55882949e-01 -5.13178587e-01 -3.94219875e-01
-7.60614395e-01 -1.17799878e+00 1.53569365e+00 -2.47899249e-01
2.59630740e-01 -1.01854396e+00 -2.16007382e-01 3.75803709e-01
4.23889816e-01 2.37609163e-01 9.60233450e-01 -5.94169676e-01
-9.73025799e-01 1.40662253e-01 -3.34805518e-01 2.71927297e-01
1.67092204e-01 -9.92154330e-03 -1.28483748e+00 -7.10481703e-01
5.06345518e-02 -3.07076067e-01 7.67978430e-01 2.69948363e-01
1.31601238e+00 -8.42247829e-02 -1.62473232e-01 1.38236713e+00
1.35553515e+00 -8.87449160e-02 2.72767395e-01 -1.84189267e-02
1.29525399e+00 2.38466948e-01 -1.06230541e-03 3.18657011e-01
6.37278616e-01 8.00574303e-01 7.94845462e-01 -1.39781073e-01
-3.53684753e-01 -2.35352069e-01 3.74757461e-02 1.15905547e+00
-2.14348808e-01 -7.68724754e-02 -1.19813466e+00 3.65833014e-01
-1.46131778e+00 -8.07519138e-01 -2.00170785e-01 2.03473520e+00
2.53109485e-01 1.03958830e-01 9.06900987e-02 -4.36390191e-01
1.62877738e-01 5.41082501e-01 -1.00790119e+00 -2.97862113e-01
-2.40109831e-01 2.73594242e-02 5.00077128e-01 3.85131896e-01
-9.97975707e-01 8.68333995e-01 5.45485449e+00 5.09262025e-01
-1.45599151e+00 1.22756310e-01 5.16264260e-01 -3.98043752e-01
-6.34211779e-01 -2.00500086e-01 -8.51749182e-01 -2.28049397e-01
4.79181230e-01 2.45316982e-01 5.63350141e-01 7.58444905e-01
-9.45506059e-03 4.85694528e-01 -1.26281679e+00 1.27039635e+00
2.90647209e-01 -1.51021183e+00 3.88401210e-01 2.96617597e-01
7.46551216e-01 1.02130175e+00 2.70582110e-01 -7.73564503e-02
3.05268914e-01 -9.12570715e-01 6.87123954e-01 4.18486893e-01
1.03798890e+00 -6.78821981e-01 1.12092100e-01 4.71129656e-01
-1.39667571e+00 2.69411385e-01 -4.64744747e-01 3.54683638e-01
2.03381225e-01 4.10036117e-01 -6.79607809e-01 7.50003576e-01
6.47003770e-01 1.08785760e+00 -4.79111046e-01 8.11467171e-01
8.50124285e-02 2.47665808e-01 -6.15210295e-01 4.36241716e-01
3.37172985e-01 -1.81688398e-01 9.56237495e-01 7.81151414e-01
4.41313773e-01 1.22730970e-01 3.26653838e-01 9.65732455e-01
-9.64089707e-02 -1.29730552e-01 -1.07750893e+00 2.87925690e-01
2.55935878e-01 1.22763050e+00 -5.55080533e-01 -2.10028052e-01
-6.60901368e-01 7.14693725e-01 4.25647169e-01 2.56072789e-01
-5.11273146e-01 4.41027939e-01 9.24599826e-01 3.17354411e-01
6.93987906e-01 -5.33572674e-01 -5.74561000e-01 -1.31226194e+00
2.34777689e-01 -8.69962633e-01 1.89207807e-01 -8.28604400e-01
-1.62667596e+00 5.95834136e-01 6.53223768e-02 -1.39341795e+00
4.82083037e-02 -8.46061945e-01 -3.07293713e-01 9.05596733e-01
-1.51243436e+00 -1.50218153e+00 -2.46333361e-01 5.92614591e-01
6.62754536e-01 -3.09412330e-01 8.19819808e-01 3.77351105e-01
-1.01032024e-02 5.23643374e-01 -1.14899218e-01 1.08453296e-01
5.34161747e-01 -1.23861635e+00 1.22655320e+00 3.79953504e-01
4.24092174e-01 4.61618811e-01 1.27693743e-01 -5.41824162e-01
-2.04035902e+00 -1.07320035e+00 4.35362220e-01 -9.41742480e-01
3.04926246e-01 -5.59214652e-01 -9.44878757e-01 8.10475111e-01
1.96208451e-02 6.37866735e-01 3.92267168e-01 1.44360960e-01
-7.99599171e-01 -1.25582218e-01 -9.37737644e-01 4.61275667e-01
1.61663270e+00 -7.23296344e-01 -1.60765633e-01 3.26785296e-01
6.53313577e-01 -1.21865189e+00 -1.01727486e+00 6.00199640e-01
7.59412050e-01 -1.08882797e+00 1.52303457e+00 -6.98227286e-01
4.46738303e-01 -9.78683457e-02 -4.29357857e-01 -1.36568296e+00
-2.49131456e-01 -2.60143101e-01 -2.62292176e-01 7.55040348e-01
3.97587061e-01 -7.28753686e-01 9.86248016e-01 5.80048561e-01
-5.18678486e-01 -1.28219008e+00 -1.04724932e+00 -7.26081192e-01
2.90195554e-01 -5.61410010e-01 9.07085061e-01 8.36459100e-01
-9.33683038e-01 3.77693355e-01 -5.13139628e-02 4.10210133e-01
7.23318875e-01 6.08800650e-01 1.07099414e+00 -1.33432484e+00
-2.28384867e-01 -4.74727452e-01 -2.06038937e-01 -1.44051266e+00
1.92274898e-01 -1.39830244e+00 -4.49470490e-01 -1.61492169e+00
7.23603964e-02 -6.01550758e-01 -1.63194090e-02 4.89459157e-01
3.57528776e-01 2.52475172e-01 5.41309595e-01 1.42103255e-01
-2.96109647e-01 5.65757215e-01 1.89333987e+00 -3.62583585e-02
-1.42088428e-01 2.36604691e-01 -4.01789844e-01 9.70696747e-01
6.72653496e-01 -3.55121851e-01 -4.88983810e-01 -1.08294809e+00
3.27273130e-01 1.62907213e-01 7.11146355e-01 -5.07772088e-01
1.69089034e-01 -5.89107201e-02 5.27829528e-01 -1.26124811e+00
8.34602058e-01 -9.33691978e-01 1.76384479e-01 8.73381943e-02
7.77952597e-02 5.77804148e-01 3.34266633e-01 4.64475691e-01
2.05381647e-01 3.36152375e-01 6.91066802e-01 -4.00693655e-01
-2.57980466e-01 8.77585113e-01 2.20349193e-01 1.38840795e-01
5.19264460e-01 -4.05066401e-01 -3.92251164e-01 -2.42843300e-01
-7.42668211e-01 1.86384007e-01 6.21810734e-01 6.60421073e-01
8.56666386e-01 -1.55899715e+00 -8.16917181e-01 4.67681140e-01
2.50924267e-02 4.49114144e-01 2.53058106e-01 8.00154030e-01
-2.95865089e-01 2.67799258e-01 -4.96504121e-02 -1.16164100e+00
-1.37994099e+00 2.50997871e-01 6.16445005e-01 -2.88338941e-02
-1.07504666e+00 7.32895970e-01 2.82741338e-01 -1.05498278e+00
7.75393695e-02 -4.57029313e-01 1.60739824e-01 -1.85604006e-01
1.47575647e-01 3.07040960e-01 4.38300312e-01 -6.84486985e-01
-3.68177533e-01 1.01511729e+00 1.19721154e-02 7.76298568e-02
1.84457731e+00 9.93442759e-02 -1.75773829e-01 5.38050115e-01
1.41506672e+00 7.95655102e-02 -1.41684055e+00 -2.65835643e-01
-4.30234969e-01 -6.00585103e-01 -6.14882037e-02 -6.62623644e-01
-1.53148377e+00 9.95256960e-01 5.50830305e-01 -1.04446106e-01
7.40773797e-01 1.26562953e-01 9.36838508e-01 5.44743955e-01
4.18058366e-01 -5.05376518e-01 4.41765971e-03 9.37949479e-01
1.21946657e+00 -1.16366446e+00 1.40470684e-01 -5.31011581e-01
-2.43727013e-01 1.15821052e+00 7.50250518e-01 -3.95634413e-01
9.43115473e-01 3.34163219e-01 2.03095108e-01 -7.52068222e-01
-1.06530678e+00 1.51585028e-01 6.84350848e-01 4.12210315e-01
3.07165682e-01 1.28202625e-02 5.52910984e-01 -8.20054859e-02
-2.37645283e-01 -4.57152128e-01 1.31206289e-01 6.08504415e-01
4.30567004e-03 -1.08764195e+00 -1.56907931e-01 5.37654996e-01
-2.41636410e-01 -9.84357391e-03 -4.33151245e-01 6.58226013e-01
-2.44533896e-01 3.01253885e-01 1.44116774e-01 -3.49551946e-01
7.74600327e-01 1.39264733e-01 7.44815469e-01 -6.64483011e-01
-7.25958645e-01 1.87692329e-01 5.12354868e-03 -8.43321800e-01
-4.69419390e-01 -8.76492739e-01 -9.26207006e-01 -4.94709015e-01
-1.11585148e-01 -6.76349282e-01 6.34634793e-01 7.50719249e-01
8.34641874e-01 1.68176785e-01 6.48258090e-01 -1.17097008e+00
-7.64508724e-01 -4.92400467e-01 -1.07576884e-01 3.22202533e-01
3.06458801e-01 -5.69614649e-01 -2.04739198e-01 -2.87890255e-01] | [8.271098136901855, -3.581665515899658] |
cc23d227-9366-4b18-894e-bef3f3168144 | forecasting-west-nile-virus-with-graph-neural | 2212.11367 | null | https://arxiv.org/abs/2212.11367v1 | https://arxiv.org/pdf/2212.11367v1.pdf | Forecasting West Nile Virus with Graph Neural Networks: Harnessing Spatial Dependence in Irregularly Sampled Geospatial Data | Machine learning methods have seen increased application to geospatial environmental problems, such as precipitation nowcasting, haze forecasting, and crop yield prediction. However, many of the machine learning methods applied to mosquito population and disease forecasting do not inherently take into account the underlying spatial structure of the given data. In our work, we apply a spatially aware graph neural network model consisting of GraphSAGE layers to forecast the presence of West Nile virus in Illinois, to aid mosquito surveillance and abatement efforts within the state. More generally, we show that graph neural networks applied to irregularly sampled geospatial data can exceed the performance of a range of baseline methods including logistic regression, XGBoost, and fully-connected neural networks. | ['Rebecca Smith', 'William Brown', 'Bo Li', 'Trevor Harris', 'Adam Tonks'] | 2022-12-21 | null | null | null | null | ['crop-yield-prediction', 'crop-yield-prediction'] | ['computer-vision', 'miscellaneous'] | [ 1.59593955e-01 8.86314660e-02 -3.13752919e-01 -2.21175432e-01
3.95961940e-01 -2.37509310e-01 6.37154043e-01 7.25964904e-01
-1.04010655e-02 7.55679965e-01 4.34135526e-01 -1.26180995e+00
-3.01256835e-01 -1.52204835e+00 -5.78889906e-01 -5.49123824e-01
-6.69888139e-01 3.05509865e-01 -2.70601481e-01 -5.20406127e-01
-4.01105136e-02 6.69910848e-01 -8.71306539e-01 -2.07809016e-01
9.33087409e-01 5.55709481e-01 -6.77593276e-02 7.03631818e-01
-5.65663874e-02 5.93602657e-01 -4.17742521e-01 4.37000543e-02
-1.31919473e-01 -3.60507518e-02 -3.09890658e-01 -2.71937460e-01
2.75271714e-01 -3.96658152e-01 -3.01726818e-01 9.34866190e-01
1.82280988e-01 1.90967135e-02 9.54811931e-01 -1.19013262e+00
-8.12919796e-01 4.79600698e-01 -9.51426923e-01 4.66332614e-01
8.06857347e-02 -5.36034163e-03 5.93272507e-01 -4.23593938e-01
2.85815448e-01 1.33528483e+00 9.79823768e-01 2.26423498e-02
-1.15929222e+00 -6.10510528e-01 3.42003047e-01 -2.11219117e-01
-1.21048594e+00 -3.38679627e-02 5.35455644e-01 -6.99606121e-01
1.14847624e+00 1.72459990e-01 7.55733013e-01 5.04264593e-01
5.62543213e-01 2.54842907e-01 8.52301598e-01 -2.38805801e-01
4.19272661e-01 -5.75850189e-01 7.87853897e-02 8.49303901e-01
6.79154038e-01 3.15312803e-01 -2.12706208e-01 -5.59010863e-01
6.22007549e-01 5.14975965e-01 -9.09435451e-02 1.82465643e-01
-5.44961035e-01 1.27889740e+00 9.59266067e-01 1.62848800e-01
-6.67358756e-01 1.93229571e-01 -1.91123746e-02 -9.40026343e-02
1.44205511e+00 6.52684212e-01 -3.40766996e-01 4.56895053e-01
-1.19862330e+00 2.12583050e-01 7.18935966e-01 2.05358326e-01
7.99192667e-01 4.65212524e-01 3.53136033e-01 5.16460657e-01
4.89876926e-01 9.80741680e-01 -2.66158193e-01 -1.17475703e-01
3.65600258e-01 8.73714149e-01 -8.05123076e-02 -1.68200445e+00
-7.51706898e-01 1.14732176e-01 -1.01183546e+00 1.18297175e-01
1.15608111e-01 -8.07529449e-01 -1.34669626e+00 1.26071370e+00
3.67500514e-01 5.01387596e-01 -1.14399314e-01 6.16025865e-01
7.09950447e-01 1.10802627e+00 5.46764970e-01 -4.92727682e-02
7.36213207e-01 -3.08889568e-01 -6.21949971e-01 -2.55970120e-01
4.45321560e-01 -2.00926110e-01 2.53945231e-01 -1.38074279e-01
-3.72057825e-01 -8.17916468e-02 -7.75318444e-01 5.98840833e-01
-1.14530838e+00 -2.88676292e-01 1.18302119e+00 4.98900384e-01
-1.39187109e+00 5.91582119e-01 -1.09936976e+00 -7.90665030e-01
5.07926166e-01 2.97906399e-01 -2.49557242e-01 -4.49458025e-02
-1.20939851e+00 1.14245272e+00 2.92051584e-01 6.18303001e-01
-5.40688276e-01 -4.91034210e-01 -1.13343954e+00 2.22576067e-01
-8.19117352e-02 -2.54763156e-01 2.17890203e-01 -3.71352077e-01
-8.53670180e-01 3.42725635e-01 -1.26681805e-01 -3.05897534e-01
-1.88995525e-01 -7.84908384e-02 -4.91748869e-01 -3.52914453e-01
2.00057015e-01 4.98221636e-01 4.08658475e-01 -8.84015560e-01
-4.60271627e-01 -8.04332614e-01 -2.27534756e-01 2.31251433e-01
-4.23804462e-01 4.26566601e-02 1.95319414e-01 -9.06377017e-01
-1.02678560e-01 -7.68563271e-01 -7.99101055e-01 -2.11793795e-01
-4.70229894e-01 -6.59550205e-02 9.22020376e-01 -1.06556046e+00
1.18118906e+00 -1.90984774e+00 2.81822030e-02 5.44977784e-01
3.02724630e-01 4.11287695e-01 -4.14399832e-01 6.68188155e-01
2.23568991e-01 3.48905772e-01 -3.73406827e-01 5.02598226e-01
-5.51942885e-01 2.96108514e-01 -4.14960027e-01 5.00739157e-01
5.34902096e-01 7.96876729e-01 -1.23691368e+00 -5.45125641e-02
3.72191727e-01 7.62024879e-01 -1.32620305e-01 1.15045190e-01
-5.77015996e-01 1.50385141e-01 -7.01366246e-01 6.74148142e-01
7.61227727e-01 -4.43561345e-01 4.32671934e-01 4.90730375e-01
7.90039357e-03 2.61250194e-02 -5.36755562e-01 7.71042585e-01
-1.29284188e-01 7.77512610e-01 -1.46787271e-01 -8.27678204e-01
9.83724535e-01 1.12702392e-01 6.10731721e-01 -3.51774424e-01
-4.39972699e-01 -2.52788633e-01 -2.55248100e-01 -2.58040965e-01
5.03519297e-01 3.05194438e-01 3.83288205e-01 6.70950115e-01
-4.21874702e-01 3.17570828e-02 -1.01935513e-01 -4.97886240e-02
9.90481198e-01 -1.48354669e-03 4.90924984e-01 -5.03628373e-01
-1.56932816e-01 4.74617481e-01 1.21977545e-01 5.03083825e-01
1.75290674e-01 1.25977442e-01 4.72387880e-01 -1.23821998e+00
-7.99417496e-01 -8.38413417e-01 -1.71051413e-01 1.33083498e+00
-1.07398577e-01 -1.37725726e-01 -5.69772840e-01 -4.31869000e-01
5.59552372e-01 6.08813643e-01 -8.90171349e-01 2.17268020e-01
-3.18026900e-01 -1.74649620e+00 4.59399462e-01 4.74268258e-01
1.88612327e-01 -1.01967382e+00 -3.35688114e-01 4.42614436e-01
2.00059608e-01 -6.49947464e-01 1.29157066e-01 2.25823358e-01
-9.96397495e-01 -1.18871069e+00 -5.01013160e-01 -3.94146442e-01
8.28018486e-01 2.65702218e-01 9.72801626e-01 2.20950097e-01
-3.24182808e-01 1.95297785e-02 -2.17180084e-02 -5.74830651e-01
-1.46654546e-01 1.90368071e-01 -4.21131738e-02 -4.12917048e-01
7.89420426e-01 -4.13860559e-01 -6.52217269e-01 6.20708577e-02
-7.41464257e-01 2.67267320e-02 1.76114231e-01 5.15110850e-01
1.78260073e-01 3.76104236e-01 5.10276198e-01 -8.48672986e-01
8.09810519e-01 -1.11007333e+00 -8.12616825e-01 6.24231517e-01
-7.87505507e-01 -3.67201604e-02 2.97696471e-01 -2.09001005e-01
-6.70374691e-01 1.45919904e-01 3.40650082e-01 1.92663983e-01
-2.24890381e-01 1.32381487e+00 3.00266981e-01 -1.40619248e-01
6.43013954e-01 -3.91129926e-02 5.88576868e-02 -1.29674107e-01
3.32235008e-01 6.39716506e-01 1.34238049e-01 -8.87994766e-02
7.03405023e-01 3.89287442e-01 3.39871824e-01 -1.17340457e+00
-3.15427721e-01 -2.71167189e-01 -3.95951569e-01 -2.62877464e-01
9.38270271e-01 -9.70124722e-01 -4.07433599e-01 5.47987878e-01
-1.04033709e+00 -6.28520787e-01 3.97199988e-01 4.57670778e-01
1.13494590e-01 -1.47257879e-01 -4.70748693e-01 -8.90199900e-01
-4.57408905e-01 -6.52098060e-01 8.76212776e-01 2.91181713e-01
9.71724861e-04 -1.65296960e+00 5.26178718e-01 -9.98070687e-02
9.02658403e-01 9.53920722e-01 1.07989562e+00 -4.57514465e-01
-2.66131043e-01 -1.29882187e-01 -3.76125634e-01 -3.25448424e-01
9.32352126e-01 4.14958477e-01 -8.14065754e-01 -3.08142513e-01
-5.60639262e-01 1.16854459e-01 9.80981231e-01 1.07754183e+00
6.69693947e-01 -8.89088929e-01 -7.16681421e-01 6.35123551e-01
1.42613840e+00 2.18415022e-01 1.78412393e-01 1.62726894e-01
8.40658426e-01 8.05400968e-01 -2.68886588e-03 2.60306031e-01
4.79332954e-01 2.60465652e-01 6.69071376e-01 -7.93817461e-01
3.43274415e-01 -4.05896902e-01 1.06566492e-02 1.00161150e-01
-4.23419684e-01 -6.29608274e-01 -1.64601839e+00 6.08854890e-01
-1.88142431e+00 -9.96935666e-01 -4.97232005e-02 2.11359286e+00
3.03165287e-01 -3.86277586e-01 7.63870627e-02 -4.16142553e-01
8.36932182e-01 6.47434294e-01 -5.31809270e-01 -3.98815840e-01
-2.19045535e-01 1.86068848e-01 1.09328544e+00 6.57313704e-01
-1.45493460e+00 1.29215741e+00 7.28075027e+00 1.80815190e-01
-1.49843478e+00 -3.66339386e-01 1.13968945e+00 1.23796694e-01
-2.41712749e-01 -2.59643346e-01 -5.01826584e-01 2.57364392e-01
1.06213915e+00 1.53906737e-02 8.23643208e-01 6.30893588e-01
5.53889811e-01 -1.85795754e-01 -6.43480122e-01 1.71431124e-01
-3.02006043e-02 -1.56118596e+00 1.57782450e-01 3.65501255e-01
9.68683183e-01 5.53743303e-01 9.21821371e-02 -2.23561540e-01
1.20770061e+00 -1.53602052e+00 -4.33543533e-01 4.55578268e-01
5.52540302e-01 -6.00980401e-01 5.70878685e-01 4.69846465e-02
-1.23310053e+00 1.13341786e-01 -3.84843528e-01 -5.30938447e-01
1.13592550e-01 5.67498982e-01 -1.25455427e+00 1.16242923e-01
7.55068123e-01 8.97131026e-01 -6.49159372e-01 8.75050724e-01
-2.55124837e-01 9.79747057e-01 -5.91853499e-01 -3.44370157e-01
5.10392249e-01 -2.26699397e-01 1.88072279e-01 9.99414921e-01
2.58989990e-01 3.93376261e-01 8.03406984e-02 5.15488625e-01
1.74584895e-01 8.60249698e-02 -1.28116834e+00 -2.37402275e-01
2.63446838e-01 6.81783617e-01 -6.70421183e-01 -4.19556675e-03
-3.24059963e-01 2.99042732e-01 3.02832305e-01 8.73145878e-01
-3.10495853e-01 -2.61085153e-01 7.06183434e-01 3.53930265e-01
7.23075122e-02 -1.70050323e-01 -2.60088474e-01 -9.30019379e-01
-5.03239274e-01 -3.98616374e-01 4.32373405e-01 -7.43501365e-01
-1.41263378e+00 5.05452216e-01 9.55590755e-02 -4.27366555e-01
-6.55325472e-01 -6.50730789e-01 -1.23402083e+00 1.12212098e+00
-1.75556040e+00 -1.34222341e+00 -2.17019498e-01 3.56248438e-01
-1.18234672e-01 -2.76227951e-01 9.39080119e-01 -2.87147552e-01
-4.90980864e-01 -3.23333591e-01 4.54827905e-01 3.42591822e-01
5.16745001e-02 -1.07497394e+00 1.13104022e+00 8.95299315e-01
-3.33043151e-02 4.35741514e-01 4.97298777e-01 -1.27532697e+00
-1.26146078e+00 -1.70931637e+00 1.14458382e+00 -1.26459626e-02
9.32768762e-01 -6.19431585e-02 -9.20641124e-01 7.99495280e-01
1.51223645e-01 3.61011736e-02 5.92108250e-01 6.22026026e-01
-2.17021376e-01 -8.94739851e-02 -1.14556527e+00 4.59531486e-01
3.79075706e-01 -3.88236225e-01 -1.11095920e-01 8.29592407e-01
4.48467314e-01 3.91912088e-02 -7.57466197e-01 5.48814416e-01
3.99569482e-01 -1.66570455e-01 8.08487535e-01 -1.13431585e+00
4.66985554e-01 -1.05183691e-01 -1.47381395e-01 -1.65587246e+00
-4.95960534e-01 -2.67319024e-01 3.18593197e-02 5.86795688e-01
5.56744099e-01 -6.15969419e-01 1.04938459e+00 6.42087698e-01
5.17143071e-01 -4.16915178e-01 -7.49812365e-01 -3.69210154e-01
3.69365633e-01 6.95009083e-02 1.08337760e+00 1.26262236e+00
1.78693891e-01 1.81694523e-01 -5.89919925e-01 7.24498153e-01
7.72521317e-01 2.83464557e-03 4.85883236e-01 -1.42177439e+00
1.84600696e-01 -5.10875344e-01 -6.42250657e-01 -3.24085891e-01
1.13661677e-01 -4.38572228e-01 -2.77539998e-01 -1.90715110e+00
-8.44432041e-02 -6.20575249e-01 -2.47787192e-01 1.02606773e+00
-5.85055709e-01 7.73069859e-02 -2.61286050e-01 -7.18358904e-02
-4.38371859e-03 3.77470106e-01 5.81504941e-01 -5.51779985e-01
-5.69513500e-01 3.34075242e-01 -4.97220784e-01 5.31683743e-01
9.67515707e-01 -4.52924997e-01 -2.55827665e-01 -9.71557617e-01
4.81511235e-01 3.75381172e-01 3.40271473e-01 -6.20380580e-01
2.49752253e-01 -8.29532087e-01 5.82624197e-01 -4.59850371e-01
-9.45687443e-02 -7.71896064e-01 3.93538952e-01 5.80858409e-01
-1.59591034e-01 4.20547158e-01 5.96606791e-01 6.02994442e-01
-5.12367375e-02 5.08694112e-01 1.04266845e-01 1.13595344e-01
-5.49292803e-01 7.39111841e-01 -6.76072299e-01 -4.27318722e-01
8.81964386e-01 -9.50651318e-02 -8.44703257e-01 -4.17788386e-01
-5.76847434e-01 5.09623647e-01 2.97430545e-01 4.74085689e-01
6.45938635e-01 -9.69882667e-01 -1.06338167e+00 2.04422072e-01
-1.56079726e-02 -1.39259383e-01 -2.18388274e-01 6.89673200e-02
-9.57252324e-01 5.39603770e-01 -2.72727668e-01 -3.38982165e-01
-8.64722013e-01 3.78224283e-01 2.79560655e-01 -1.98318437e-01
-1.08987831e-01 4.95098531e-01 2.19455305e-02 -6.68585598e-01
-5.56510314e-02 -3.28618050e-01 -5.01900733e-01 -1.63012460e-01
4.97106194e-01 3.61706257e-01 -2.17592940e-01 -6.54308617e-01
-3.45966637e-01 2.50960439e-01 3.29069346e-01 1.07736275e-01
1.94529748e+00 3.49343121e-01 -3.80607665e-01 8.32811594e-02
6.40240788e-01 -3.96897495e-01 -1.28326333e+00 -2.04692427e-02
3.42800498e-01 -2.35178143e-01 4.76344347e-01 -8.76660109e-01
-1.25891685e+00 8.87734294e-01 5.37757099e-01 4.79935884e-01
9.46366370e-01 -3.51117164e-01 1.62622944e-01 5.07144928e-01
7.01745152e-02 -4.77850109e-01 -6.59780502e-01 2.05638498e-01
4.75214362e-01 -1.48848939e+00 4.21254396e-01 -8.82200897e-02
-2.44429201e-01 1.13233685e+00 5.23807108e-01 3.42453308e-02
1.27788413e+00 3.60027969e-01 -7.09321797e-02 -4.66273427e-01
-6.58467770e-01 -1.21204004e-01 2.34980255e-01 1.03881013e+00
3.12180132e-01 7.42255628e-01 3.64966273e-01 -3.93067122e-01
1.87396154e-01 1.74856052e-01 2.51701921e-01 7.51751542e-01
-7.46970415e-01 -5.02471209e-01 -3.79322290e-01 9.58738983e-01
-3.31678510e-01 -8.66119564e-01 -4.09036219e-01 4.97083724e-01
-2.86046863e-01 1.01815879e+00 3.88415158e-01 -2.33772904e-01
-1.28907084e-01 -4.60704029e-01 -1.34102419e-01 -5.12594402e-01
-5.49512804e-01 -2.29040533e-01 -1.91099077e-01 -2.01338664e-01
-4.69522566e-01 -4.20029134e-01 -7.11231053e-01 -8.48291218e-01
-3.75286877e-01 1.96632788e-01 9.76831675e-01 8.72260928e-01
3.56450617e-01 -1.44442935e-02 5.97203672e-01 -8.09343219e-01
1.28139928e-01 -1.10214138e+00 -6.70904100e-01 -1.03885040e-01
4.95532274e-01 -3.42128485e-01 -1.23146467e-01 -2.15055138e-01] | [6.708925247192383, 2.725522518157959] |
f64de075-589d-4af7-9057-95344f8c8568 | reduction-of-subjective-listening-effort-for | 2111.01914 | null | https://arxiv.org/abs/2111.01914v1 | https://arxiv.org/pdf/2111.01914v1.pdf | Reduction of Subjective Listening Effort for TV Broadcast Signals with Recurrent Neural Networks | Listening to the audio of TV broadcast signals can be challenging for hearing-impaired as well as normal-hearing listeners, especially when background sounds are prominent or too loud compared to the speech signal. This can result in a reduced satisfaction and increased listening effort of the listeners. Since the broadcast sound is usually premixed, we perform a subjective evaluation for quantifying the potential of speech enhancement systems based on audio source separation and recurrent neural networks (RNN). Recently, RNNs have shown promising results in the context of sound source separation and real-time signal processing. In this paper, we separate the speech from the background signals and remix the separated sounds at a higher signal-to-noise ratio. This differs from classic speech enhancement, where usually only the extracted speech signal is exploited. The subjective evaluation with 20 normal-hearing subjects on real TV-broadcast material shows that our proposed enhancement system is able to reduce the listening effort by around 2 points on a 13-point listening effort rating scale and increases the perceived sound quality compared to the original mixture. | ['Bernd T. Meyer', 'Jan Rennies', 'Ragini Sinha', 'Hannah Baumgartner', 'Rainer Huber', 'Nils L. Westhausen'] | 2021-11-02 | null | null | null | null | ['audio-source-separation'] | ['audio'] | [ 3.99373084e-01 5.10699004e-02 5.55429101e-01 -6.91891760e-02
-1.16558218e+00 -1.60310000e-01 4.90983669e-03 2.61125714e-01
-6.13073826e-01 5.36677659e-01 6.55491531e-01 -6.48617595e-02
-2.51322299e-01 -4.72765148e-01 -2.05334023e-01 -8.03534627e-01
-1.39995907e-02 -3.02699149e-01 2.48179853e-01 -5.65080702e-01
-1.16902195e-01 3.00627440e-01 -2.20808482e+00 2.50098735e-01
7.63134658e-01 9.27891791e-01 6.16480291e-01 1.09820771e+00
8.74646679e-02 5.73934555e-01 -1.19136846e+00 -8.52202922e-02
-2.48752348e-02 -4.88574058e-01 -3.66391838e-01 2.05479413e-02
2.07190230e-01 -1.32124618e-01 -1.13466211e-01 1.25990009e+00
1.20925725e+00 4.43853945e-01 1.05118051e-01 -7.17364371e-01
-8.77860561e-02 6.70027614e-01 -7.09228963e-02 2.77889013e-01
5.84415734e-01 -1.90936193e-01 7.29608476e-01 -7.89277911e-01
7.48031400e-03 9.92308855e-01 6.38754606e-01 3.28346789e-01
-1.27877939e+00 -5.53435504e-01 -1.70101359e-01 4.26357061e-01
-1.15648985e+00 -9.23121572e-01 1.09967637e+00 -1.15575798e-01
9.53848183e-01 6.53704166e-01 6.53507113e-01 9.00269508e-01
-1.60919100e-01 3.81311208e-01 9.23979819e-01 -8.07745039e-01
3.89988482e-01 2.88305640e-01 1.34472728e-01 -4.38799739e-01
-2.43714243e-01 2.04556614e-01 -4.16881859e-01 1.25743777e-01
2.90292740e-01 -2.91788399e-01 -6.39285922e-01 4.26035225e-01
-7.22482562e-01 4.11337495e-01 2.38368616e-01 9.63443696e-01
-7.90755749e-01 -1.19282313e-01 3.42782974e-01 5.00213325e-01
5.63368559e-01 4.60216880e-01 -2.15915278e-01 -4.12770748e-01
-1.13069260e+00 8.34743902e-02 6.97788477e-01 1.28113240e-01
4.44952846e-02 5.07329285e-01 -1.98240817e-01 1.55898917e+00
1.62358344e-01 3.91432852e-01 5.69632351e-01 -8.65297437e-01
1.01393335e-01 -1.61903918e-01 2.10553929e-01 -1.08784568e+00
-3.84107262e-01 -9.32441115e-01 -8.34253550e-01 6.06088459e-01
2.17218012e-01 -3.15513998e-01 -7.60930002e-01 1.82506728e+00
2.69096166e-01 -6.36805147e-02 2.16249481e-01 9.22316730e-01
8.01551878e-01 9.94701564e-01 -1.19903013e-01 -7.03409851e-01
1.34161043e+00 -8.19704175e-01 -1.42690051e+00 -2.60980338e-01
-1.94334924e-01 -1.07083118e+00 1.24392283e+00 7.59877861e-01
-1.57104826e+00 -9.02645111e-01 -1.05234778e+00 2.72623122e-01
-1.22603931e-01 -5.60759418e-02 -2.80017704e-01 1.23863065e+00
-1.37138009e+00 3.83461654e-01 -4.39671636e-01 6.83736503e-02
-2.25106627e-01 3.18031341e-01 -1.07580498e-01 1.66246429e-01
-1.40130520e+00 6.68216705e-01 -3.00263651e-02 3.15723628e-01
-6.41391397e-01 -7.51652658e-01 -7.00584233e-01 4.33094114e-01
2.07420781e-01 -2.28770331e-01 1.54968846e+00 -9.75655198e-01
-1.83865368e+00 2.97424525e-01 -1.67829841e-01 -4.02239531e-01
4.09911096e-01 -4.55697745e-01 -1.12728906e+00 2.60825127e-01
-2.06173211e-01 3.80920223e-03 1.10911012e+00 -1.03624821e+00
-6.06124878e-01 -1.61742657e-01 -2.70962417e-01 2.07215801e-01
-4.99801338e-01 4.51272875e-01 -9.25868601e-02 -1.03447771e+00
1.91658422e-01 -4.99330670e-01 -1.36299983e-01 -5.05027890e-01
-1.00322373e-01 2.54863828e-01 6.62947118e-01 -1.22155941e+00
1.16962123e+00 -2.69404244e+00 -6.21443838e-02 1.67349398e-01
-3.22690494e-02 4.91487235e-01 -7.16294721e-02 3.82537171e-02
-4.60476518e-01 -2.14757636e-01 6.72534630e-02 -4.65816468e-01
-1.54301718e-01 -2.34346315e-01 1.21113397e-02 1.83533117e-01
-2.85898685e-01 -1.76243205e-02 -7.24458039e-01 -1.24387436e-01
2.24434555e-01 1.01836848e+00 -4.74883676e-01 4.78418708e-01
2.74043173e-01 2.89863557e-01 3.79532069e-01 2.50989437e-01
7.98815548e-01 6.46194935e-01 -2.15172678e-01 -2.38894239e-01
-2.00078264e-01 4.61704701e-01 -1.48758066e+00 1.24689102e+00
-7.90728211e-01 8.42467546e-01 8.55308235e-01 -5.88420987e-01
1.00298548e+00 9.59677100e-01 2.64829487e-01 -8.28767776e-01
1.96116731e-01 3.43043685e-01 3.19877803e-01 -6.61006987e-01
6.47048056e-01 -5.17323434e-01 3.72605085e-01 1.45536751e-01
-8.02563652e-02 -2.27668360e-01 7.69987404e-02 -2.05896273e-01
8.89551222e-01 -7.04711497e-01 2.00578928e-01 -9.17178951e-03
6.25474811e-01 -9.54181373e-01 1.70573950e-01 5.59778571e-01
-2.43229568e-01 4.46447223e-01 1.17395245e-01 5.49179912e-01
-6.34795368e-01 -1.13612509e+00 2.77337171e-02 1.55491924e+00
-2.68425047e-01 -2.62261182e-01 -8.66217613e-01 2.74345785e-01
-6.24327898e-01 1.00279415e+00 -1.05007298e-01 -2.54824042e-01
-2.81327039e-01 -3.42262149e-01 5.72093129e-01 2.96353281e-01
1.62176847e-01 -1.27392316e+00 -4.02209371e-01 5.24949014e-01
-5.47076523e-01 -7.81709790e-01 -3.17808747e-01 3.23792011e-01
-4.39861774e-01 -1.16226085e-01 -1.15576732e+00 -9.79145944e-01
2.43382365e-01 2.21946239e-01 7.01349974e-01 -2.80750185e-01
-1.28167734e-01 3.98297429e-01 -4.98608530e-01 -6.75156891e-01
-7.65480340e-01 -4.16465580e-01 1.90251097e-01 1.76699162e-01
-1.23757079e-01 -9.52981651e-01 -4.64626431e-01 5.19378185e-01
-8.42992187e-01 -3.15726489e-01 3.53154391e-01 5.42416990e-01
1.92229614e-01 9.76783991e-01 8.69268417e-01 5.39808013e-02
1.40298736e+00 -5.75675070e-02 -3.04347038e-01 -3.28142583e-01
-1.49369672e-01 -5.88863075e-01 4.81202811e-01 -7.44424820e-01
-1.48224175e+00 -3.46994340e-01 -8.46877933e-01 1.44369984e-02
-3.91402900e-01 4.43385601e-01 -5.87228179e-01 1.57082707e-01
8.48264456e-01 6.54386654e-02 3.19645181e-03 -7.53274918e-01
-5.05731329e-02 1.13050544e+00 8.19250762e-01 2.64021158e-02
4.33298081e-01 -2.86342166e-02 -4.14789677e-01 -1.31862986e+00
-2.85334468e-01 -6.15535200e-01 -6.06794618e-02 -5.98444998e-01
6.16611183e-01 -5.37442744e-01 -5.52137852e-01 7.26426542e-01
-1.11388028e+00 -1.44885287e-01 -5.01601279e-01 9.51499224e-01
-3.17197978e-01 1.60145879e-01 -6.06702447e-01 -1.43655658e+00
-4.32720959e-01 -1.44990635e+00 5.58627188e-01 2.23346010e-01
-3.36585343e-01 -5.83573043e-01 3.83441150e-02 3.04377526e-01
9.19846833e-01 -2.50533938e-01 6.67322457e-01 -4.24364567e-01
4.14392918e-01 -3.35002601e-01 2.87768364e-01 7.83264458e-01
5.43852866e-01 -3.87579829e-01 -1.48996222e+00 -1.45894483e-01
7.29924619e-01 6.57852367e-02 4.76752579e-01 9.83660936e-01
7.34427810e-01 -2.35616893e-01 2.97722220e-01 -1.56534985e-02
9.56001580e-01 5.83538771e-01 9.88453507e-01 1.94612816e-02
1.49539916e-03 1.01203835e+00 4.07264173e-01 2.63088524e-01
-4.75790679e-01 6.05199158e-01 3.24445039e-01 -5.74220777e-01
-4.23893750e-01 8.75596702e-02 5.92795968e-01 9.61582541e-01
-1.52815834e-01 -2.87138611e-01 -5.01523554e-01 5.34033954e-01
-1.01878953e+00 -9.61848021e-01 -1.27164006e-01 2.45310712e+00
9.32601035e-01 3.15131128e-01 4.08889532e-01 1.14668357e+00
1.11118007e+00 1.55653059e-02 -1.00264542e-01 -5.99137247e-01
-1.06075980e-01 5.19175291e-01 2.47824173e-02 8.35547626e-01
-7.83755898e-01 3.44293535e-01 6.45969725e+00 9.81910169e-01
-1.37641108e+00 2.86687672e-01 2.11445928e-01 -5.03381312e-01
8.09777156e-02 -6.13197088e-01 -2.14296922e-01 3.46077651e-01
1.36589158e+00 -1.06334575e-01 4.29867357e-01 5.98749757e-01
7.29172349e-01 9.47511569e-03 -7.63278604e-01 1.12678373e+00
1.02564536e-01 -6.36861503e-01 -4.76379275e-01 -4.31977995e-02
-1.45531376e-03 -2.88103759e-01 3.39816600e-01 2.26916775e-01
-4.64231133e-01 -9.22856212e-01 8.70270073e-01 3.68763626e-01
5.89429557e-01 -9.19004321e-01 8.01259279e-01 2.12339506e-01
-9.30919707e-01 -1.20907158e-01 -8.00107867e-02 -1.01711258e-01
4.96631354e-01 8.06041121e-01 -8.37905526e-01 2.26820484e-01
9.00447309e-01 -8.85970816e-02 -2.06440702e-01 1.41335309e+00
-2.60705799e-01 8.86019886e-01 -3.47771853e-01 9.95694548e-02
-2.62227029e-01 1.67424962e-01 1.17406511e+00 1.32446420e+00
5.18022954e-01 -6.39370531e-02 -4.03846443e-01 5.55786490e-01
1.10251963e-01 4.37707156e-01 -1.37931600e-01 9.71342176e-02
1.04735352e-01 1.05025494e+00 -5.60620368e-01 -1.02811873e-01
-4.78563532e-02 6.72420144e-01 -6.52659416e-01 4.35858548e-01
-5.56288779e-01 -9.57953155e-01 4.17474747e-01 2.40850151e-01
1.57692477e-01 1.05932295e-01 1.48148581e-01 -3.58496100e-01
2.09217086e-01 -1.13023365e+00 2.38958150e-02 -1.16417718e+00
-8.35522115e-01 1.06819832e+00 -2.16394886e-01 -1.31766653e+00
-2.65577376e-01 -4.12328243e-01 -6.07725441e-01 1.04813147e+00
-1.04453146e+00 -3.85194749e-01 -1.26867160e-01 5.67140996e-01
7.49758184e-01 6.70869872e-02 8.12651634e-01 8.11159909e-01
-2.58114070e-01 6.27290785e-01 -3.21766399e-02 -3.62722725e-01
5.92565000e-01 -9.95782554e-01 -1.89993352e-01 8.82612646e-01
-1.81045935e-01 4.47544217e-01 1.38267398e+00 -2.46817812e-01
-6.51837051e-01 -6.79360211e-01 7.32909381e-01 3.55628908e-01
4.69877630e-01 -3.94021273e-01 -1.01428294e+00 -1.81106821e-01
4.38133478e-01 -4.62475926e-01 8.90627921e-01 1.77990422e-02
1.45656526e-01 -4.63912249e-01 -1.24352145e+00 6.41629398e-01
5.11284173e-01 -5.96033096e-01 -7.67607331e-01 -1.27479639e-02
9.37375546e-01 2.85709482e-02 -7.30130255e-01 2.50200152e-01
5.53114951e-01 -1.12436104e+00 1.06818378e+00 -4.55813035e-02
1.03282712e-01 -2.59212017e-01 -3.51314902e-01 -1.77026141e+00
-2.47147650e-01 -8.97099435e-01 3.01184058e-01 1.59852374e+00
6.21197879e-01 -5.15188575e-01 3.17407906e-01 3.99226487e-01
-2.11016461e-02 -2.58059651e-02 -9.23602223e-01 -8.25476885e-01
-4.17634517e-01 -9.49232101e-01 2.85063207e-01 3.58712584e-01
2.31191158e-01 4.44955975e-01 -5.66616058e-01 3.55221212e-01
3.50382358e-01 -6.60975039e-01 3.93384457e-01 -1.17931974e+00
-6.07942998e-01 -5.69325864e-01 -5.44189930e-01 -7.09614694e-01
-2.17283264e-01 -3.22678685e-01 4.19562399e-01 -1.56277394e+00
-5.16308844e-01 2.14277834e-01 -5.49587011e-01 1.65390372e-02
-8.46620053e-02 5.25997765e-02 1.19050823e-01 -3.73786718e-01
4.91166823e-02 5.63982248e-01 1.03340364e+00 -3.64765823e-01
-6.99290216e-01 7.02860057e-01 -6.99168861e-01 8.54627788e-01
5.55512369e-01 -5.32969415e-01 -5.07236660e-01 -2.68232767e-02
8.14375505e-02 4.24521774e-01 1.70627028e-01 -1.13748419e+00
2.54148453e-01 3.80926818e-01 -9.45781693e-02 -6.16764426e-01
8.19162428e-01 -1.17534232e+00 3.10151547e-01 3.61765414e-01
-5.29366434e-01 -4.96867836e-01 5.16530335e-01 3.98612231e-01
-5.22918344e-01 -4.20131087e-01 9.89084840e-01 1.37231752e-01
-1.27669170e-01 -5.78608334e-01 -9.74587560e-01 -4.87401277e-01
4.23901200e-01 -2.62263864e-01 4.06165645e-02 -1.10771322e+00
-1.21931422e+00 -5.70308268e-01 -3.18582118e-01 1.15414120e-01
6.51889026e-01 -1.02465093e+00 -7.96562135e-01 2.34176308e-01
-3.21171105e-01 -4.08454657e-01 7.29804277e-01 1.09280252e+00
-9.72221345e-02 2.27062479e-02 -2.43215293e-01 -3.88039768e-01
-1.78143930e+00 4.60698038e-01 6.28069818e-01 1.23720758e-01
-5.40024042e-01 8.92030597e-01 5.39427437e-02 2.09086850e-01
7.05236495e-01 -4.91572678e-01 -6.11870766e-01 2.49313608e-01
9.11975145e-01 9.78060484e-01 5.13889492e-01 -5.32981455e-01
-1.78880785e-02 2.76405126e-01 7.26849809e-02 -7.12824106e-01
1.37033904e+00 -2.56845117e-01 -1.84896573e-01 5.45824170e-01
1.20668864e+00 7.66470671e-01 -4.78954732e-01 -2.65180226e-02
-3.05149138e-01 -5.10687411e-01 6.58540726e-01 -8.71327460e-01
-9.98956680e-01 8.90962720e-01 1.13204098e+00 8.49326491e-01
1.77775669e+00 -2.10067511e-01 6.62967324e-01 1.48174450e-01
5.06020524e-02 -1.18525147e+00 2.18537420e-01 1.01584867e-01
1.27836120e+00 -6.17122710e-01 -5.22509277e-01 -3.48301739e-01
-2.82177895e-01 8.30480516e-01 1.50570288e-01 2.85295546e-01
7.83555567e-01 5.69728851e-01 3.23948443e-01 2.05937862e-01
-1.76178083e-01 -3.02016616e-01 2.46585876e-01 7.29698300e-01
5.74884832e-01 -1.22801512e-02 -1.78707484e-02 9.05092478e-01
-7.29080141e-01 -2.64974147e-01 5.75586498e-01 6.77076578e-01
-7.53506184e-01 -8.64758670e-01 -1.17738509e+00 1.39893502e-01
-8.58811140e-01 -1.20024741e-01 -1.88735113e-01 1.18417300e-01
-8.32300559e-02 1.85715783e+00 -8.47964641e-03 -3.70115817e-01
8.51615608e-01 1.56628072e-01 1.09149568e-01 -3.24726701e-01
-6.11253083e-01 1.07894707e+00 8.96180421e-02 -1.05144083e-01
-4.41618353e-01 -4.39847976e-01 -9.26707566e-01 1.21875219e-01
-5.67206860e-01 4.46007520e-01 9.89251733e-01 6.16071463e-01
7.11992756e-02 1.34370542e+00 6.26797915e-01 -1.04594386e+00
-2.14849606e-01 -1.47202897e+00 -9.70255613e-01 1.65630400e-01
7.82884836e-01 -3.42954725e-01 -8.39153171e-01 3.65784392e-02] | [15.097996711730957, 5.780060768127441] |
5108f6af-b307-4c0b-b175-fd89b82d83d8 | rethinking-image-restoration-for-object | null | null | https://openreview.net/forum?id=se2oxj-6Nz | https://openreview.net/pdf?id=se2oxj-6Nz | Rethinking Image Restoration for Object Detection | Although image restoration has achieved significant progress, its potential to assist object detectors in adverse imaging conditions lacks enough attention. It is reported that the existing image restoration methods cannot improve the object detector performance and sometimes even reduce the detection performance. To address the issue, we propose a targeted adversarial attack in the restoration procedure to boost object detection performance after restoration. Specifically, we present an ADAM-like adversarial attack to generate pseudo ground truth for restoration training. Resultant restored images are close to original sharp images, and at the same time, lead to better results of object detection. We conduct extensive experiments in image dehazing and low light enhancement and show the superiority of our method over conventional training and other domain adaptation and multi-task methods. The proposed pipeline can be applied to all restoration methods and detectors in both one- and two-stage. | ['Xiaochun Cao', 'Tao Wang', 'Wenqi Ren', 'Shangquan Sun'] | 2022-11-01 | null | null | null | nips-2022-11 | ['image-dehazing'] | ['computer-vision'] | [ 7.02804387e-01 -2.72599667e-01 3.33663732e-01 -4.16189246e-02
-1.11563611e+00 -2.85810798e-01 5.52955031e-01 -2.29544535e-01
-2.91273206e-01 5.86429060e-01 1.32152811e-01 -1.42939687e-01
2.16709659e-01 -6.53444886e-01 -7.15387166e-01 -9.44904685e-01
3.97706330e-01 1.61922332e-02 5.93623221e-01 -2.62745410e-01
3.44149947e-01 4.70724463e-01 -1.27283907e+00 5.08947492e-01
9.23798680e-01 8.18198860e-01 2.91201651e-01 7.80034065e-01
3.15382004e-01 8.10749829e-01 -8.59364629e-01 -4.10126358e-01
6.14134312e-01 -2.47117132e-01 -5.48141599e-01 4.93577868e-01
6.34378672e-01 -7.35342860e-01 -3.90230000e-01 1.40905774e+00
8.38166475e-01 1.89530164e-01 4.93973225e-01 -9.95338559e-01
-1.04429615e+00 2.33106807e-01 -9.94583607e-01 6.31783307e-01
3.36611360e-01 3.82654130e-01 4.35972005e-01 -1.28046191e+00
2.28340551e-01 1.46731603e+00 7.15686977e-01 6.50139153e-01
-1.26672387e+00 -5.99315464e-01 -4.87190522e-02 3.21883351e-01
-1.18147540e+00 -4.84855562e-01 8.23755562e-01 -1.69970229e-01
4.97881144e-01 4.20522168e-02 2.18445882e-01 1.18054307e+00
2.82301635e-01 9.67273235e-01 1.45186949e+00 -2.98569709e-01
-2.66374070e-02 2.14932989e-02 -2.46932983e-01 4.44193363e-01
2.98048079e-01 4.69167054e-01 -3.23995650e-01 8.60448107e-02
7.90284097e-01 8.55671167e-02 -4.44538087e-01 2.02253357e-01
-1.03009009e+00 5.77910662e-01 7.12377787e-01 -1.98014826e-02
-5.63094199e-01 6.65980801e-02 2.30215758e-01 9.86395627e-02
4.31510836e-01 1.47711426e-01 6.57683685e-02 6.72491312e-01
-8.80656421e-01 2.89663494e-01 2.55410403e-01 4.57690269e-01
5.45759499e-01 5.20409644e-01 -4.70464885e-01 7.62469590e-01
2.35640302e-01 5.61937749e-01 3.52954745e-01 -7.45912254e-01
1.45931631e-01 3.82603347e-01 2.68726379e-01 -9.76855040e-01
-6.41184151e-02 -7.07652569e-01 -1.04122949e+00 7.85483837e-01
2.82382131e-01 1.14168920e-01 -1.20208180e+00 1.15038681e+00
4.31058794e-01 5.38595140e-01 2.28263766e-01 1.27907193e+00
9.49681997e-01 5.90780020e-01 1.76519886e-01 1.18200034e-02
1.40038121e+00 -1.05134428e+00 -7.64884114e-01 -6.43757999e-01
-2.22086892e-01 -1.22483695e+00 9.37598050e-01 4.91017997e-01
-1.12405801e+00 -8.99016857e-01 -1.18085277e+00 4.81399857e-02
7.27711841e-02 6.42229319e-02 5.79863250e-01 8.62990618e-01
-8.97606730e-01 4.21901137e-01 -7.12018430e-01 -5.21656536e-02
8.51835728e-01 1.54072464e-01 -2.58823276e-01 -5.04350603e-01
-9.37774479e-01 1.16600752e+00 3.31707388e-01 1.85057655e-01
-1.56264353e+00 -7.58344948e-01 -6.44797981e-01 -1.71616048e-01
2.20690504e-01 -7.43319154e-01 9.82783377e-01 -8.99940550e-01
-1.27081919e+00 8.46939921e-01 1.13007650e-01 -7.51326144e-01
6.26661837e-01 -2.94263303e-01 -5.73500216e-01 2.62077600e-01
1.01031296e-01 5.58632374e-01 1.35104537e+00 -1.53044295e+00
-6.48545563e-01 -1.80253267e-01 4.76081818e-02 1.65068105e-01
-1.98399112e-01 3.29715610e-01 -2.70548314e-01 -1.02041137e+00
4.96678650e-02 -5.00227332e-01 -4.37859893e-01 1.63924187e-01
-5.18817246e-01 1.91407084e-01 1.05231547e+00 -8.96220028e-01
4.82989669e-01 -2.16733646e+00 -2.14777201e-01 -3.04992139e-01
-4.50833049e-03 6.38193011e-01 -4.60878313e-01 -3.38265449e-02
-9.04094800e-02 -1.27914965e-01 -4.80226606e-01 -3.71029735e-01
-4.59003210e-01 1.68586284e-01 -5.52330136e-01 8.55606556e-01
4.51392174e-01 7.15244949e-01 -7.54999161e-01 -4.50134456e-01
6.03826284e-01 6.86425507e-01 -4.33917522e-01 1.90079510e-01
-2.03403700e-02 6.58317506e-01 -2.80501783e-01 9.58458006e-01
1.25475001e+00 -1.05326533e-01 -3.38804811e-01 -5.67672133e-01
2.71632493e-01 -2.18861297e-01 -1.39278269e+00 1.20020485e+00
-3.68602723e-01 6.64970458e-01 2.63792872e-01 -1.07507527e+00
7.70873547e-01 1.84772611e-01 2.39527643e-01 -7.69357383e-01
3.66243050e-02 1.22750551e-01 2.64274329e-02 -3.06434274e-01
4.77093250e-01 -3.47292066e-01 4.75340068e-01 1.57529309e-01
-7.67626017e-02 -1.54353529e-01 -9.48779508e-02 4.12324741e-02
1.04648280e+00 9.33327898e-02 2.83883810e-01 2.53575649e-02
7.86839128e-01 -6.83800280e-02 5.47250986e-01 9.14990425e-01
-2.95592159e-01 9.20877039e-01 -2.54817635e-01 -3.46617132e-01
-1.01537800e+00 -1.39679801e+00 -3.80271226e-01 9.46057498e-01
5.91054916e-01 3.31782699e-01 -6.64544880e-01 -5.55735230e-01
-6.26521334e-02 4.92954582e-01 -3.23425710e-01 -3.35761935e-01
-6.21608198e-01 -1.27365470e+00 7.00210452e-01 4.01785403e-01
9.11564827e-01 -1.13543594e+00 -2.65926898e-01 2.01713920e-01
-4.14042830e-01 -1.47128427e+00 -4.70906824e-01 -9.87701640e-02
-7.89860845e-01 -1.14547932e+00 -8.63321722e-01 -8.61606061e-01
7.94910848e-01 5.91875732e-01 8.20395112e-01 2.95075238e-01
-5.45258820e-01 1.94186464e-01 -2.98365563e-01 -3.43549520e-01
-8.01007509e-01 -7.18767226e-01 1.37628227e-01 2.66581953e-01
-1.52974740e-01 -5.17486036e-01 -9.41694617e-01 5.82821190e-01
-1.26447034e+00 -1.93047091e-01 8.74092519e-01 9.73744571e-01
4.68509912e-01 2.86806315e-01 7.32954919e-01 -5.55498302e-01
6.18456483e-01 -1.88011769e-02 -6.04346037e-01 1.94447741e-01
-7.74254620e-01 -1.10678844e-01 5.34855425e-01 -7.21150219e-01
-1.39685690e+00 2.09499091e-01 -3.29731584e-01 -5.01257479e-01
-2.23239735e-01 2.22905651e-02 -5.71068525e-02 -7.54547417e-01
1.02261770e+00 4.17914122e-01 -4.27216254e-02 -5.49149334e-01
3.65263432e-01 5.22024751e-01 1.02476895e+00 -1.61018983e-01
1.27817237e+00 8.32501829e-01 3.74980122e-02 -5.88152349e-01
-9.46949422e-01 -4.00241464e-01 -2.14791387e-01 -3.04199576e-01
6.47213995e-01 -1.13818181e+00 -6.03081584e-01 6.83415413e-01
-1.21291828e+00 -1.63676500e-01 -1.79253206e-01 2.98343420e-01
-2.59440422e-01 7.68389642e-01 -7.32878268e-01 -6.91348076e-01
-6.40503943e-01 -1.24594176e+00 9.87443566e-01 4.36707139e-01
6.51825309e-01 -7.73014605e-01 -3.98713976e-01 6.14461482e-01
6.58295572e-01 1.82376146e-01 4.71171260e-01 -4.58309650e-01
-8.32383215e-01 -3.63294303e-01 -5.12677073e-01 7.93255508e-01
1.70456804e-02 -4.30598527e-01 -1.08548784e+00 -4.97826248e-01
2.51022667e-01 -1.48419470e-01 1.20622456e+00 5.17665148e-01
1.01409185e+00 -2.26404533e-01 -1.94940969e-01 6.48772657e-01
1.50804496e+00 -1.38730064e-01 1.12531722e+00 5.45114636e-01
5.21545112e-01 2.36046746e-01 6.39028430e-01 2.11688176e-01
-1.71754912e-01 7.69455254e-01 7.70050585e-01 -6.04256511e-01
-8.99168730e-01 -7.65089542e-02 7.03653753e-01 -7.19961477e-03
-9.16841161e-03 -2.14128047e-01 -6.12969458e-01 5.94977915e-01
-1.41579592e+00 -1.06922269e+00 -3.42974454e-01 1.97514248e+00
7.65107274e-01 1.35203123e-01 -1.60033822e-01 9.04354677e-02
1.00586009e+00 1.13938190e-01 -6.43231392e-01 1.82244271e-01
-4.78726506e-01 2.82246321e-01 7.20352650e-01 5.21134079e-01
-1.21908402e+00 1.06345379e+00 6.95555401e+00 1.03590715e+00
-8.08316946e-01 4.10231739e-01 5.94148040e-01 1.60585612e-01
9.05162618e-02 5.00843339e-02 -6.37073517e-01 2.17442617e-01
2.33960479e-01 -1.07245721e-01 4.64923263e-01 8.39008570e-01
1.92873120e-01 -8.23434442e-02 -7.45731533e-01 9.40419734e-01
3.87222618e-01 -1.26507425e+00 1.63706288e-01 -1.53604060e-01
9.09652233e-01 -8.31460953e-02 3.05447817e-01 9.96640921e-02
3.24139923e-01 -8.51698995e-01 6.41217470e-01 2.21784085e-01
4.77839738e-01 -5.83492458e-01 7.77240634e-01 2.98066169e-01
-9.52979445e-01 -1.86450601e-01 -5.78931093e-01 1.56648353e-01
5.47621667e-01 6.51844263e-01 -9.19660389e-01 4.87386048e-01
6.80513203e-01 4.66424346e-01 -7.46668875e-01 1.43344700e+00
-4.67085898e-01 5.16035378e-01 9.22519043e-02 5.60445130e-01
-1.21050499e-01 -4.36063968e-02 9.81970429e-01 1.11420405e+00
5.60046025e-02 9.35547426e-02 3.75366271e-01 8.61794889e-01
-6.13005050e-02 -3.04820210e-01 -3.04884434e-01 4.71754909e-01
9.24662724e-02 1.35808158e+00 -7.16898382e-01 -2.43644804e-01
-3.87357533e-01 1.39510810e+00 -3.35242629e-01 5.27518511e-01
-9.03424442e-01 -6.55122567e-03 6.63963795e-01 1.45952865e-01
1.90393507e-01 3.11178789e-02 -2.37767503e-01 -1.00839627e+00
3.35661322e-02 -1.10021770e+00 3.31176698e-01 -8.15220952e-01
-1.49112666e+00 6.16684556e-01 -3.46645266e-01 -1.24247921e+00
2.06520781e-01 -7.18609631e-01 -6.98331118e-01 8.13826859e-01
-1.82438159e+00 -1.33194959e+00 -4.83366877e-01 8.51533413e-01
7.83929229e-01 -2.91315824e-01 4.16192532e-01 6.50110364e-01
-5.29294968e-01 5.05291939e-01 4.37813886e-02 2.16576934e-01
7.89397299e-01 -1.01384127e+00 4.78595525e-01 1.66962969e+00
8.05850029e-02 3.40409487e-01 1.01667976e+00 -7.76694417e-01
-1.39994025e+00 -1.26674426e+00 -1.19305514e-01 -2.67927796e-01
3.96518946e-01 1.33234421e-02 -1.01826274e+00 4.24666822e-01
2.45402321e-01 3.72854203e-01 -1.51853729e-03 -3.88556749e-01
-2.78123796e-01 -2.05918893e-01 -1.41606009e+00 5.24888217e-01
7.07789600e-01 -4.74617243e-01 -6.31070018e-01 6.60483181e-01
6.34795368e-01 -6.28218055e-01 -4.77854788e-01 5.51225483e-01
1.93841606e-01 -9.27088559e-01 1.82062709e+00 -2.42953509e-01
4.51891661e-01 -6.33383334e-01 -2.21252844e-01 -1.13089490e+00
-2.15075865e-01 -4.20834422e-01 -7.93506429e-02 1.18639231e+00
8.80038440e-02 -5.20919442e-01 6.08505845e-01 7.18326196e-02
-2.98957109e-01 -1.91025078e-01 -8.91403854e-01 -8.26716661e-01
-2.99232125e-01 -2.65076607e-01 1.68547362e-01 5.90663970e-01
-7.50173390e-01 1.26721516e-01 -7.97799945e-01 7.83196270e-01
1.25740004e+00 6.34854585e-02 7.25280285e-01 -8.81766438e-01
-3.17133814e-01 -2.03605607e-01 -4.37551200e-01 -9.64772165e-01
-4.59484644e-02 -7.32818782e-01 3.48248512e-01 -1.64489281e+00
3.57532144e-01 -1.16120636e-01 -3.45398992e-01 3.48127276e-01
-5.51888347e-01 8.81978750e-01 4.04006690e-02 2.51090497e-01
-3.94250184e-01 4.37055767e-01 1.44949579e+00 -3.43153358e-01
1.61276817e-01 1.65907621e-01 -7.81571984e-01 6.58155382e-01
5.32821357e-01 -6.67500734e-01 8.40053800e-03 -4.15372312e-01
-2.43686557e-01 -2.65984442e-02 1.09064150e+00 -1.14765739e+00
3.67651820e-01 -4.17238958e-02 6.61345899e-01 -4.69282627e-01
2.40176156e-01 -7.09016085e-01 -2.98126247e-02 6.40516043e-01
-1.33973509e-02 -3.24424326e-01 2.67389238e-01 8.61511350e-01
-3.76477629e-01 -1.60060138e-01 1.49338901e+00 -1.44834682e-01
-9.05141532e-01 2.11477384e-01 -2.82118112e-01 -2.58674562e-01
9.25041795e-01 -3.65345359e-01 -4.59860623e-01 -2.34480724e-01
-5.19841135e-01 -1.05851442e-01 2.50011832e-01 4.89113748e-01
1.08333611e+00 -1.15106070e+00 -1.15635777e+00 1.60736561e-01
-1.26582757e-01 -2.90453196e-01 4.98693824e-01 8.04754555e-01
-4.54284906e-01 -3.50546479e-01 -4.12663013e-01 -5.92013597e-01
-1.30403268e+00 8.05465043e-01 6.82785332e-01 -3.84682082e-02
-1.02693820e+00 7.41988182e-01 3.68026763e-01 -4.01761942e-02
1.75472841e-01 6.91674501e-02 -9.80932266e-02 -3.66070509e-01
9.84675109e-01 4.38740700e-01 1.69706538e-01 -5.02116740e-01
-3.90544385e-01 3.74598265e-01 -2.55510271e-01 6.30842969e-02
1.33236217e+00 -1.57475948e-01 6.83540851e-02 -4.39721197e-01
7.12706864e-01 9.64824297e-03 -1.42199039e+00 -3.18631768e-01
-2.85876900e-01 -9.24762607e-01 4.17508811e-01 -9.20926809e-01
-1.24252534e+00 7.80148029e-01 1.24304736e+00 2.83473786e-02
1.50022054e+00 -1.72810838e-01 7.68862545e-01 2.50573196e-02
1.49051547e-01 -6.94420457e-01 6.62212014e-01 1.40830511e-02
1.03093147e+00 -1.54488361e+00 3.01477343e-01 -5.81270933e-01
-6.19725883e-01 9.11054075e-01 6.59584165e-01 -3.81799906e-01
2.09956214e-01 3.91299903e-01 2.03258350e-01 -3.57059166e-02
-2.16045588e-01 -3.48579824e-01 3.18409383e-01 9.73755360e-01
7.77033567e-02 -3.02965015e-01 -7.08749741e-02 3.13751191e-01
4.01811689e-01 -1.28946140e-01 6.96084380e-01 3.78639460e-01
-5.99213004e-01 -1.08782530e+00 -9.42964971e-01 4.20250222e-02
-8.08315277e-01 -2.77828455e-01 3.77856866e-02 6.06844842e-01
-1.37411775e-02 1.34858108e+00 -3.83255631e-01 -2.05405444e-01
4.33901757e-01 -4.79025394e-01 4.22879606e-01 -3.45148236e-01
-6.48245096e-01 1.00938067e-01 -2.95979202e-01 -3.82146537e-01
-4.20555800e-01 -4.23499197e-01 -1.04277861e+00 -1.82795554e-01
-5.70690811e-01 -3.05028498e-01 4.96776611e-01 8.35960448e-01
1.52609730e-02 7.37258613e-01 6.17498159e-01 -1.08418584e+00
-8.04968476e-01 -1.02489054e+00 -4.94284421e-01 5.95428944e-01
5.89974225e-01 -6.42566562e-01 -4.68171567e-01 2.58575380e-01] | [10.781844139099121, -2.530362844467163] |
c6af1d88-8fbf-4d04-aac4-d281527f3c9d | beyond-detection-visual-realism-assessment-of | 2306.05985 | null | https://arxiv.org/abs/2306.05985v1 | https://arxiv.org/pdf/2306.05985v1.pdf | Beyond Detection: Visual Realism Assessment of Deepfakes | In the era of rapid digitalization and artificial intelligence advancements, the development of DeepFake technology has posed significant security and privacy concerns. This paper presents an effective measure to assess the visual realism of DeepFake videos. We utilize an ensemble of two Convolutional Neural Network (CNN) models: Eva and ConvNext. These models have been trained on the DeepFake Game Competition (DFGC) 2022 dataset and aim to predict Mean Opinion Scores (MOS) from DeepFake videos based on features extracted from sequences of frames. Our method secured the third place in the recent DFGC on Visual Realism Assessment held in conjunction with the 2023 International Joint Conference on Biometrics (IJCB 2023). We provide an over\-view of the models, data preprocessing, and training procedures. We also report the performance of our models against the competition's baseline model and discuss the implications of our findings. | ['Borut Batagelj', 'Vitomir Štruc', 'Peter Peer', 'Luka Dragar'] | 2023-06-09 | null | null | null | null | ['face-swapping'] | ['computer-vision'] | [-2.38232315e-01 -2.64839083e-02 2.24105213e-02 -5.14501333e-01
-3.89060974e-01 -3.75255883e-01 5.72734773e-01 -4.69450355e-01
-7.64741361e-01 3.70758384e-01 3.39544922e-01 -3.89088780e-01
2.25728552e-04 -4.61199522e-01 -5.57376504e-01 -9.80351046e-02
-3.00164878e-01 -5.17135561e-01 -3.60449851e-01 -1.79296225e-01
2.92106062e-01 5.60418546e-01 -1.40293801e+00 5.21332383e-01
2.18926445e-01 1.71189082e+00 -6.20314837e-01 1.11525047e+00
6.08573318e-01 9.46884811e-01 -7.02827454e-01 -1.15013027e+00
6.80015326e-01 -2.01308757e-01 -8.49028051e-01 -2.54838765e-01
8.51199090e-01 -9.24032450e-01 -7.95353889e-01 8.94568264e-01
9.10225987e-01 -1.33816823e-02 3.34860861e-01 -1.59017348e+00
-4.73539203e-01 -6.28607050e-02 -4.11397666e-01 4.48811769e-01
2.45077550e-01 3.24354053e-01 6.04835927e-01 -8.94119263e-01
6.04182601e-01 1.04916859e+00 1.10698271e+00 8.14127207e-01
-8.58515084e-01 -9.24156427e-01 -4.36350673e-01 4.77669746e-01
-1.54921532e+00 -8.54399145e-01 4.77186888e-01 -4.10681933e-01
1.04774070e+00 1.50670558e-01 8.57138097e-01 1.12479305e+00
4.54306781e-01 6.28224671e-01 9.69525516e-01 -1.66118994e-01
2.25019470e-01 6.85065007e-03 -1.06890507e-01 7.44198442e-01
-3.99515182e-02 4.73150969e-01 -6.34411156e-01 1.33635458e-02
8.25252056e-01 -8.56574595e-01 -1.12040691e-01 -3.45104009e-01
-6.64369166e-01 1.04748356e+00 2.97449201e-01 -2.23718703e-01
-3.77173722e-01 3.76870483e-01 7.79876888e-01 3.82486641e-01
3.30276370e-01 2.98222452e-01 -4.77757275e-01 -5.00568986e-01
-9.14779186e-01 2.86867827e-01 7.21751571e-01 3.27349842e-01
1.96688697e-01 1.02194428e-01 -1.13060817e-01 6.44003391e-01
3.77834350e-01 -4.20273934e-03 2.03414172e-01 -1.44072735e+00
2.60300606e-01 -4.67884243e-02 1.69811830e-01 -1.60787046e+00
-1.55515581e-01 6.79627359e-02 -6.25867724e-01 6.92907512e-01
3.97862494e-01 -4.45880324e-01 -7.75212944e-01 1.52758777e+00
-1.68572858e-01 4.50681001e-02 -5.42707294e-02 8.51371527e-01
1.13384318e+00 2.51506865e-01 1.10510305e-01 1.71122894e-01
9.92610455e-01 -4.73389745e-01 -7.40630448e-01 3.58186781e-01
4.47516322e-01 -5.22203028e-01 8.04711282e-01 1.00527596e+00
-1.24494064e+00 -7.89636374e-01 -1.36464667e+00 -4.05095406e-02
-4.75635201e-01 2.01723248e-01 6.43214345e-01 1.45782936e+00
-1.53990471e+00 6.62406206e-01 -4.73599821e-01 -3.17648858e-01
8.09166789e-01 8.98277104e-01 -6.27343237e-01 1.57541677e-01
-1.25705183e+00 9.86609817e-01 9.06158537e-02 2.73218900e-01
-8.40007305e-01 -2.29900971e-01 -8.29055130e-01 -7.08084255e-02
1.00605316e-01 -1.01969928e-01 1.31730962e+00 -1.32972360e+00
-1.66794884e+00 1.08370459e+00 3.57854992e-01 -6.23393297e-01
5.97149670e-01 -1.88077718e-01 -6.38641179e-01 1.18260436e-01
-4.88837421e-01 1.24861360e+00 4.36961502e-01 -7.57073164e-01
-4.58653003e-01 -5.23708500e-02 5.26913524e-01 6.02533780e-02
-2.53794461e-01 2.78075933e-01 -4.65796679e-01 -3.55591685e-01
-7.34450161e-01 -9.07252014e-01 7.93789513e-03 3.30399543e-01
-7.91459158e-02 -4.70051542e-02 8.40919435e-01 -8.44006717e-01
1.27002466e+00 -2.36082697e+00 -3.12263399e-01 3.79997045e-01
5.72540700e-01 7.79138565e-01 -3.28780532e-01 -2.50168052e-02
-2.75095820e-01 3.23828161e-01 6.02071285e-01 -2.15971336e-01
-1.01969957e-01 -2.45664507e-01 2.49472260e-01 3.63362312e-01
-3.68593186e-02 1.07951713e+00 -4.61911649e-01 -1.46535963e-01
4.84238535e-01 5.61085463e-01 -6.94304168e-01 -3.82005572e-02
4.53705162e-01 1.08635500e-02 1.70035213e-01 5.67677259e-01
9.01074827e-01 -4.01504487e-02 1.64015979e-01 -5.07590413e-01
1.13390468e-01 -1.80617385e-02 -9.06185865e-01 1.62344694e+00
-1.29022375e-01 1.28884661e+00 -8.41703191e-02 -6.04874671e-01
7.08742619e-01 4.41755891e-01 4.58133578e-01 -1.08795702e+00
7.18654931e-01 -1.17201366e-01 2.00353116e-01 -6.42337024e-01
7.61825383e-01 2.97126144e-01 1.09971091e-01 9.66345891e-02
2.56481797e-01 1.84220061e-01 -3.97198908e-02 2.69072264e-01
8.98124039e-01 -7.07349107e-02 2.52201170e-01 -2.43751243e-01
2.89029062e-01 -4.79103088e-01 4.74946797e-01 4.58016336e-01
-8.78799915e-01 5.90016842e-01 6.49423242e-01 -9.12921369e-01
-1.08643758e+00 -8.96507263e-01 2.31301486e-01 5.41585803e-01
-1.08818226e-02 -7.72386670e-01 -1.08578527e+00 -8.41000140e-01
-2.34564260e-01 1.12579301e-01 -1.05753529e+00 -2.81232536e-01
-2.55276468e-02 -3.97665858e-01 1.07442844e+00 5.35813868e-01
9.83498693e-01 -1.18152177e+00 -8.03205490e-01 -2.16405973e-01
-3.75913471e-01 -1.24853396e+00 -3.55668455e-01 -2.23847792e-01
-2.21049890e-01 -1.28573382e+00 -3.92307639e-01 -3.34611595e-01
1.03096791e-01 -6.20860755e-02 1.03715813e+00 6.25965446e-02
-3.32253724e-01 4.99514699e-01 -1.96332172e-01 -6.72077715e-01
-1.33063927e-01 -2.76742101e-01 2.13951975e-01 -1.61394384e-02
9.04202044e-01 -8.59755576e-02 -8.32730353e-01 3.35210621e-01
-8.77894223e-01 8.85248706e-02 1.77478626e-01 6.48189068e-01
-8.73786118e-03 -1.09558724e-01 5.43625534e-01 -3.67613763e-01
8.46947134e-01 -1.59076944e-01 -5.33522248e-01 1.69376627e-01
-5.08993030e-01 -5.18279672e-01 -2.07557194e-02 -3.58718097e-01
-7.55266726e-01 -2.31326022e-03 -2.26954475e-01 -4.85826463e-01
-3.08652744e-02 5.97990811e-01 -3.12764257e-01 -7.03376770e-01
7.27087617e-01 -5.25897980e-01 1.20214432e-01 1.89908475e-01
4.41751741e-02 9.19692576e-01 8.60967636e-01 -9.13213938e-02
1.86875910e-01 1.80520266e-01 -9.90312770e-02 -7.29667068e-01
-4.64453697e-02 4.69852276e-02 -2.05516934e-01 -8.03920865e-01
8.48585546e-01 -9.64704394e-01 -1.20285344e+00 9.97367501e-01
-1.05271053e+00 -3.09960485e-01 -4.43659239e-02 7.04233646e-01
-5.02744794e-01 4.57871705e-01 -7.00370312e-01 -7.40692496e-01
-2.37205505e-01 -1.11102259e+00 4.47351813e-01 3.37621599e-01
-4.30675149e-01 -7.46427000e-01 1.52640328e-01 5.22604704e-01
4.76689577e-01 3.31597567e-01 3.21476519e-01 -3.85064065e-01
-2.37667039e-01 -4.08136010e-01 -4.60493118e-01 9.30327713e-01
-1.22300275e-01 1.87927037e-01 -1.26959395e+00 -2.73034334e-01
-4.01560843e-01 -8.68332148e-01 4.59188819e-01 7.67242789e-01
1.44408691e+00 -6.96291998e-02 4.28778738e-01 7.90268362e-01
1.23846233e+00 5.27516186e-01 1.44758141e+00 3.50017846e-01
2.80132204e-01 5.80003798e-01 3.14352214e-01 5.11875629e-01
3.49238247e-01 6.03256524e-01 5.89427471e-01 -2.08880275e-01
1.09702066e-01 -1.09242927e-02 2.61165053e-01 2.25943193e-01
-4.72822547e-01 -4.84697759e-01 -7.84183145e-01 3.29368979e-01
-1.48538351e+00 -9.24880326e-01 1.75280854e-01 1.87799001e+00
1.51206970e-01 -1.49788698e-02 3.53640974e-01 2.66822308e-01
5.09181380e-01 4.61009741e-02 -2.35994995e-01 -1.04617536e+00
-7.88365826e-02 4.73503858e-01 5.24806619e-01 1.24496296e-01
-1.16089201e+00 7.56679893e-01 7.24604273e+00 8.10126781e-01
-1.31060755e+00 -2.79646635e-01 1.25070727e+00 -7.83104822e-02
2.58516699e-01 -5.23356497e-01 -1.75971180e-01 2.93113798e-01
1.32244992e+00 -1.77280873e-01 5.45373738e-01 6.14251971e-01
2.73741454e-01 -2.10356742e-01 -9.54219520e-01 1.31785619e+00
2.95097113e-01 -1.59667313e+00 -2.26327911e-01 6.56403244e-01
5.93996227e-01 4.78022080e-03 6.90231144e-01 3.10549170e-01
2.69675881e-01 -1.62902570e+00 4.02501881e-01 4.43640828e-01
1.26510012e+00 -1.11758542e+00 9.94125843e-01 -3.20812672e-01
-9.97141838e-01 -2.00549960e-01 -1.05586760e-01 -2.55385816e-01
-2.45545626e-01 -2.14677304e-01 -4.93446916e-01 2.14378148e-01
1.07917511e+00 4.33516949e-01 -5.41033566e-01 1.01540291e+00
4.50851142e-01 3.42773318e-01 1.03627786e-01 6.07950874e-02
9.13542435e-02 3.56481642e-01 -1.03982814e-01 1.07297361e+00
2.81283081e-01 2.20521659e-01 -4.88017619e-01 2.36237183e-01
-4.13612485e-01 2.38688178e-02 -5.10962009e-01 -3.09742808e-01
1.48677275e-01 1.29908657e+00 -5.74591458e-01 -2.67214835e-01
-6.78815484e-01 9.63739514e-01 -7.45080337e-02 2.21217126e-01
-8.51376295e-01 -4.47782993e-01 9.47179973e-01 -5.04871272e-02
1.46484017e-01 4.52436469e-02 -3.90454769e-01 -1.18812442e+00
-2.33271852e-01 -1.38132048e+00 1.56696901e-01 -1.09561718e+00
-1.12297547e+00 6.70746684e-01 -3.54428515e-02 -1.23311996e+00
-4.17779356e-01 -9.37241316e-01 -2.85878211e-01 7.35946834e-01
-1.01136422e+00 -1.01716328e+00 -5.41389585e-01 8.18217695e-01
2.24645138e-01 -5.40043890e-01 1.00464046e+00 4.60649639e-01
-1.90778300e-01 1.16577411e+00 -1.60925344e-01 6.26551449e-01
6.84150696e-01 -7.20119655e-01 7.44798720e-01 7.33129680e-01
-4.64644693e-02 3.38590652e-01 3.08498621e-01 -3.76905859e-01
-8.56190860e-01 -7.34287739e-01 7.68302858e-01 -4.94689107e-01
3.37148756e-01 -2.58176416e-01 -2.32712775e-01 5.79941392e-01
4.90422487e-01 5.74272983e-02 9.62022007e-01 -8.04788992e-02
-3.20253015e-01 -3.13119516e-02 -1.57215190e+00 4.83391702e-01
6.89732432e-01 -7.53169835e-01 8.24490190e-02 -5.46335101e-01
-5.02787977e-02 -4.47346359e-01 -7.21494615e-01 5.03655136e-01
1.29296732e+00 -1.41785586e+00 8.52731347e-01 -6.20946109e-01
7.91761816e-01 9.98113975e-02 -4.46382761e-01 -1.08616209e+00
-1.14443943e-01 -6.15082562e-01 -7.61944354e-02 7.60663688e-01
4.14973907e-02 -5.88826761e-02 1.23296189e+00 9.92219508e-01
3.79605889e-01 -6.17895126e-01 -9.22989249e-01 -5.19607067e-01
-4.86547723e-02 -7.27188587e-01 4.54248726e-01 9.29125369e-01
-1.03832046e-02 -7.60148913e-02 -1.02077460e+00 -2.28820309e-01
2.82571882e-01 -1.02613127e+00 1.08869159e+00 -1.16170514e+00
-6.10995404e-02 -2.93571055e-01 -1.35669327e+00 -6.46647096e-01
-1.80156693e-01 -1.43916562e-01 -1.97056249e-01 -1.03076553e+00
3.58088434e-01 3.50806653e-01 -5.21681905e-01 3.50375146e-01
3.15049171e-01 8.63087952e-01 4.67835218e-01 -5.43895423e-01
-4.59689885e-01 2.15103269e-01 8.26938450e-01 -2.08311882e-02
1.15894482e-01 -3.03716868e-01 -7.20844567e-01 6.62538350e-01
8.26538563e-01 1.71222866e-01 -5.86622894e-01 -3.73071253e-01
3.46383303e-01 1.59106553e-01 6.85065150e-01 -1.26005650e+00
1.19461432e-01 -8.47370550e-03 9.25399423e-01 -3.71365875e-01
6.58329368e-01 -7.64666736e-01 2.72763252e-01 4.25966322e-01
-5.43308377e-01 2.84510314e-01 6.24568939e-01 8.84617418e-02
-1.23509072e-01 2.44456261e-01 8.45929861e-01 1.22333087e-01
-9.68616366e-01 4.32817608e-01 -7.64207244e-01 -3.41284484e-01
8.03865075e-01 -6.60406351e-01 -2.63566047e-01 -1.01101136e+00
-8.37155163e-01 -5.35514690e-02 3.49871308e-01 5.63080072e-01
9.55093980e-01 -1.51733923e+00 -4.83492851e-01 3.42026711e-01
-4.41318797e-03 -9.16153312e-01 5.99791467e-01 4.29637581e-01
-7.69281209e-01 3.58015090e-01 -1.00611615e+00 -7.14711398e-02
-1.91380584e+00 2.19496086e-01 6.20058000e-01 -8.76686722e-02
3.91429923e-02 9.50299084e-01 1.34049460e-01 -2.13905454e-01
4.03169155e-01 7.70049766e-02 -4.38319623e-01 -3.06063592e-01
7.92564154e-01 5.26136279e-01 8.32612514e-02 -9.23563719e-01
-3.39428365e-01 7.44843334e-02 -6.49412870e-02 -4.40257937e-01
1.04231513e+00 -1.87520340e-01 4.67249602e-01 -5.80598116e-02
1.41679692e+00 -5.23797274e-02 -1.27002978e+00 3.59938264e-01
-1.98477551e-01 -7.51724064e-01 3.87312651e-01 -1.12909293e+00
-1.33540285e+00 9.87461567e-01 1.30656135e+00 -1.17977858e-01
1.34287214e+00 -6.53143585e-01 7.45197237e-01 2.84645170e-01
2.57750124e-01 -1.20468915e+00 2.49283046e-01 3.79063427e-01
4.77603704e-01 -1.21453929e+00 -2.03916773e-01 -1.41587153e-01
-7.78802037e-01 1.18587589e+00 8.94621909e-01 -8.08331147e-02
7.92309582e-01 1.63585469e-01 4.47054982e-01 -1.99511975e-01
-5.96464813e-01 3.04785371e-01 5.34504533e-01 9.95257556e-01
3.60057086e-01 -6.35004565e-02 -2.09638536e-01 7.17485189e-01
-1.75180376e-01 8.32915604e-01 7.09981263e-01 6.86698556e-01
2.67073393e-01 -7.91582823e-01 6.36520088e-02 3.66828501e-01
-1.03410196e+00 -2.43194122e-02 -8.57187092e-01 9.60494995e-01
4.13977593e-01 1.26293147e+00 9.17787999e-02 -1.33522570e+00
1.19883046e-01 -3.04185718e-01 4.80577499e-01 6.67851120e-02
-8.07998300e-01 -2.53638446e-01 4.03869539e-01 -9.98690724e-01
-5.73156774e-01 -3.51715505e-01 -5.35044611e-01 -1.08520186e+00
-1.76916286e-01 1.40851855e-01 9.60610569e-01 8.18798542e-01
3.06190073e-01 2.68624216e-01 5.52504599e-01 -9.26054239e-01
9.93329752e-03 -7.22384572e-01 -6.79540873e-01 8.28412101e-02
2.81629205e-01 -3.99492234e-01 2.86386702e-02 -1.11015060e-03] | [13.355147361755371, 1.6094558238983154] |
12f9e452-9414-4b4d-a7d0-c69400a27ec9 | spoken-language-identification-with | null | null | https://www.researchgate.net/publication/301828095_Spoken_Language_Identification_with_Phonotactics_Methods_on_Minangkabau_Sundanese_and_Javanese_Languages | https://www.researchgate.net/publication/301828095_Spoken_Language_Identification_with_Phonotactics_Methods_on_Minangkabau_Sundanese_and_Javanese_Languages | Spoken Language Identification with Phonotactics Methods on Minangkabau, Sundanese, and Javanese Languages | Research in the field of spoken language identification (spoken LID) on local languages helps to extend the outreach of
technology to local language speakers. This research also contributes to the preservation of local languages. In this paper, we
report our work on identifying spoken data in three local Indonesian languages: Minangkabau, Sundanese and Javanese.
Statistical phonotactics models are created to map the speech signals into the language used by the speaker. We use two
phonotactics methods, namely Phone Recognition followed by Language Modelling (PRLM) and Parallel Phone Recognition
followed by Language Modelling (PPRLM). PRLM method shows the highest accuracy using the phone recognizer trained for
English and Russian with the average of 77.42% and 75.94% respectively. | ['and Mirna Adriani', 'Amalia Zahra', 'Nur Endah Safitri'] | 2016-10-01 | null | null | null | null | ['spoken-language-identification'] | ['speech'] | [-5.56753129e-02 -1.94282621e-01 -1.73196241e-01 -5.92530429e-01
-8.62012088e-01 -7.90602565e-01 8.31249774e-01 -3.50476891e-01
-5.51660359e-01 8.70152533e-01 5.59123456e-01 -7.29831338e-01
3.17418247e-01 -3.92879486e-01 -3.03314537e-01 -5.23627698e-01
2.75611430e-01 7.39547789e-01 5.01086637e-02 -2.63637006e-01
5.11852145e-01 6.90162539e-01 -1.24263191e+00 3.86380792e-01
7.29845285e-01 5.04144616e-02 5.94362676e-01 1.06764960e+00
-4.88081723e-01 8.00512373e-01 -4.91339624e-01 1.28103137e-01
3.43157426e-02 -5.40011525e-01 -1.22707307e+00 1.73392579e-01
3.38836491e-01 -1.74222946e-01 -1.21403318e-02 8.87770057e-01
3.29311490e-01 1.96593970e-01 7.64454007e-01 -7.13358223e-01
-6.34090543e-01 9.32779729e-01 -2.45153278e-01 2.78722942e-01
4.43967164e-01 -3.49016011e-01 5.57144165e-01 -1.10088766e+00
4.38391536e-01 1.63762188e+00 5.11285305e-01 5.91810167e-01
-6.60905600e-01 -6.02059364e-01 -1.07973412e-01 -2.59847432e-01
-1.67011833e+00 -7.47282147e-01 4.10257161e-01 -3.31318140e-01
1.27477825e+00 2.42778391e-01 1.20419875e-01 4.97970134e-01
-1.69370726e-01 4.47854817e-01 1.65176725e+00 -1.08120215e+00
-1.03981055e-01 7.36259699e-01 5.05887568e-01 8.04285705e-01
-2.18700636e-02 -1.01589978e-01 -5.43453276e-01 2.22282130e-02
5.85412800e-01 -4.88292694e-01 1.15295641e-01 7.48102903e-01
-8.70763481e-01 9.83486891e-01 -4.94685888e-01 6.60755277e-01
-3.97262692e-01 -3.25762153e-01 1.51354253e-01 2.13788986e-01
2.50421166e-01 -1.24477722e-01 -3.95350873e-01 -3.57662767e-01
-9.42281306e-01 -2.53170639e-01 9.93426859e-01 8.29292774e-01
7.00284183e-01 5.23479879e-01 5.17039120e-01 1.43426430e+00
8.64029527e-01 8.71935129e-01 8.73508513e-01 -5.36593974e-01
3.10945511e-01 3.02620262e-01 -1.62546765e-02 -3.93647194e-01
-8.92982110e-02 2.34553978e-01 -1.28996953e-01 -1.65219724e-01
2.99906403e-01 -4.12542611e-01 -1.24521005e+00 1.30667317e+00
7.80892074e-02 -1.86766639e-01 8.44591379e-01 4.34233308e-01
8.33293676e-01 1.36693347e+00 -5.44941868e-04 -4.98913266e-02
1.49014974e+00 -1.05992556e+00 -6.43033445e-01 -3.34681600e-01
6.44216359e-01 -1.33908784e+00 1.17620194e+00 3.53620350e-01
-7.33986199e-01 -4.79786873e-01 -5.98589659e-01 1.64004862e-01
-6.95512414e-01 3.91361296e-01 3.21320921e-01 1.24154329e+00
-1.32053626e+00 -3.70093763e-01 -7.04366863e-01 -9.93008196e-01
-3.06872845e-01 7.21214294e-01 -4.38252568e-01 -5.19709177e-02
-1.08131075e+00 7.20721841e-01 4.30658549e-01 -4.95187379e-03
-7.35917807e-01 1.32017164e-02 -8.13721836e-01 -4.13423330e-01
-3.78332078e-01 3.44954103e-01 1.13639438e+00 -9.43303764e-01
-1.90288436e+00 1.27926517e+00 -5.37725329e-01 -1.72980130e-01
1.40750725e-02 -1.96186658e-02 -7.35157847e-01 -1.40931144e-01
-4.00418825e-02 5.63125253e-01 2.93305010e-01 -1.25858450e+00
-1.06591308e+00 -1.58529118e-01 -6.51470482e-01 2.93661803e-01
-2.68310934e-01 7.79253781e-01 -2.92360008e-01 -4.21407580e-01
2.38251209e-01 -9.42364454e-01 3.63878131e-01 -1.17664623e+00
-2.41420403e-01 -4.32117850e-01 9.85173881e-01 -1.36869884e+00
1.33083415e+00 -1.91977715e+00 -4.23825532e-01 3.66263032e-01
-8.62483919e-01 4.14719641e-01 5.71818650e-02 8.96178842e-01
2.15671420e-01 1.03391916e-01 -1.62015632e-01 -5.32869101e-01
-1.78139910e-01 8.26681376e-01 -4.09667015e-01 2.74854779e-01
-9.26609188e-02 6.12769127e-01 -3.87875974e-01 -4.62547868e-01
2.18124345e-01 4.22491908e-01 -1.42998844e-02 8.87628496e-02
3.77144545e-01 3.02545995e-01 -3.64747308e-02 9.07350361e-01
6.15360737e-01 6.59256458e-01 2.93237895e-01 5.09068370e-01
-9.42232728e-01 7.99258173e-01 -1.21955848e+00 9.53169882e-01
-6.39870167e-01 4.50564176e-01 2.01021463e-01 -6.92428410e-01
1.25205100e+00 5.10276020e-01 -2.26303101e-01 -3.72096419e-01
-2.68771648e-02 8.64214957e-01 1.93422273e-01 -5.87697923e-01
6.92672849e-01 -2.26991624e-01 -2.19204769e-01 6.42152607e-01
1.13355756e-01 8.47572386e-02 -1.72326520e-01 -1.44644633e-01
9.95921269e-02 -1.96855202e-01 4.81298953e-01 -8.67054045e-01
9.43166912e-01 1.21243551e-01 2.03585282e-01 8.06403041e-01
-8.57080892e-02 2.35393763e-01 -1.48159429e-01 -1.44528300e-01
-9.28314567e-01 -8.24388504e-01 -1.37946039e-01 1.50678802e+00
-4.70649570e-01 1.52892143e-01 -9.85752523e-01 -3.42261642e-01
-4.52589601e-01 7.42576241e-01 1.31989956e-01 1.87724933e-01
-9.11752939e-01 -5.66742897e-01 1.10477078e+00 2.11546496e-01
7.14059591e-01 -1.45641673e+00 1.41396254e-01 1.95520356e-01
-4.21627387e-02 -1.08944654e+00 -7.33570397e-01 1.28221795e-01
-6.07227623e-01 -2.24018008e-01 -7.67958343e-01 -1.84281588e+00
6.11287057e-01 7.07400218e-02 5.77508330e-01 -2.24387452e-01
1.80773959e-01 3.92941207e-01 -3.86457831e-01 -3.73821884e-01
-9.21857715e-01 3.19729626e-01 5.09478509e-01 1.12330042e-01
6.84431612e-01 -8.60305354e-02 2.74273217e-01 1.00030549e-01
-5.38923264e-01 -3.66991729e-01 4.18203712e-01 6.04620993e-01
3.67787153e-01 -7.36570079e-03 4.13259953e-01 -1.02707410e+00
5.96341074e-01 -3.59227568e-01 -6.27838433e-01 5.72747946e-01
-2.77604550e-01 -7.84053206e-02 4.88549739e-01 -4.19865370e-01
-1.19100082e+00 3.42905015e-01 -5.20523190e-01 4.28627312e-01
-5.56775808e-01 5.16287804e-01 -2.47997910e-01 -3.76893640e-01
1.22920431e-01 9.88920748e-01 -1.82606861e-01 -6.80449903e-01
-6.80361018e-02 1.51927531e+00 7.43484616e-01 -4.09964293e-01
3.58140349e-01 1.20257037e-02 -6.77117646e-01 -1.76258886e+00
-7.73535371e-02 -8.42235506e-01 -6.89580262e-01 -1.96484938e-01
9.21737313e-01 -8.22426915e-01 -4.82006788e-01 1.08797455e+00
-1.10425377e+00 -2.77564913e-01 2.33405292e-01 8.42118680e-01
1.95038375e-02 3.99056882e-01 -6.15669608e-01 -1.58231282e+00
-3.47207516e-01 -9.62202430e-01 8.88953567e-01 3.20797473e-01
-4.10062969e-01 -1.42671633e+00 2.64115959e-01 2.87124544e-01
2.97772855e-01 -6.09374821e-01 8.61925721e-01 -9.26586986e-01
-6.61703646e-02 3.06278728e-02 2.39251539e-01 5.98910570e-01
5.55932462e-01 1.68168962e-01 -1.15643263e+00 -1.01393528e-01
-3.79062556e-02 -2.60926127e-01 4.70722765e-01 4.25651073e-01
3.54121514e-02 -3.99383962e-01 1.02688111e-01 2.55386680e-01
1.37398458e+00 8.09123516e-01 5.81543326e-01 1.75746694e-01
6.58584893e-01 8.12424481e-01 2.92128593e-01 5.96021414e-02
6.03788376e-01 2.39853680e-01 -5.86287975e-01 1.74324997e-02
-1.31325498e-01 -5.15260935e-01 1.03141344e+00 1.45617628e+00
2.67296314e-01 -4.89159107e-01 -1.44548774e+00 6.39740646e-01
-1.35304511e+00 -5.08472562e-01 -3.38198274e-01 2.28184390e+00
8.05472970e-01 -2.20534310e-01 3.37966233e-01 1.23554274e-01
8.56985092e-01 -1.74303055e-01 3.65844339e-01 -1.16849744e+00
-3.71637672e-01 4.28550035e-01 7.57239223e-01 1.62238050e+00
-8.82920444e-01 1.74016440e+00 6.53461838e+00 7.09741771e-01
-1.42735040e+00 3.94518301e-03 4.22023594e-01 7.88478971e-01
-2.30489418e-01 1.46337628e-01 -1.68475878e+00 4.30909753e-01
1.57246244e+00 2.80549943e-01 6.37075186e-01 4.48364437e-01
4.90019292e-01 -3.43144506e-01 -5.00162244e-01 9.61161256e-01
3.29065740e-01 -8.91065180e-01 4.19381201e-01 1.37011468e-01
6.98129773e-01 3.07939768e-01 -1.82895027e-02 9.64128301e-02
4.78296101e-01 -1.07756829e+00 6.74643099e-01 4.36359853e-01
5.57937980e-01 -8.37163031e-01 8.65870237e-01 6.82147324e-01
-1.10004842e+00 2.49030828e-01 -3.40618938e-01 -1.45840615e-01
1.70948252e-01 -1.90167487e-01 -1.43672001e+00 -5.03559783e-02
4.62126285e-01 1.31475419e-01 -2.32380345e-01 4.02325898e-01
-6.45174459e-02 1.40683961e+00 -7.91436017e-01 -3.33510637e-01
7.04620063e-01 -3.64462405e-01 3.40306610e-01 1.59120643e+00
3.91651481e-01 6.63623437e-02 4.54945385e-01 1.96323738e-01
1.90263167e-01 5.64616263e-01 -6.15995944e-01 -4.49481398e-01
6.60679758e-01 7.44840980e-01 -7.74872720e-01 -3.48513752e-01
-3.03951710e-01 1.05275857e+00 -3.73645709e-03 2.21047565e-01
-1.06600843e-01 -3.47396016e-01 2.88180798e-01 -1.31607011e-01
-9.02988669e-03 -5.47510207e-01 -1.19122997e-01 -5.46239138e-01
-2.79135346e-01 -1.05220997e+00 3.41203958e-01 -2.88470834e-01
-1.03424251e+00 8.98924053e-01 4.95475195e-02 -4.16930765e-01
-6.97723627e-01 -8.54911923e-01 -6.39461577e-01 1.48112237e+00
-1.44205451e+00 -1.47323406e+00 2.46744230e-01 4.08379227e-01
8.66358101e-01 -8.56613338e-01 1.19236243e+00 1.04673214e-01
-5.20922601e-01 5.88997483e-01 2.64838248e-01 4.24346536e-01
4.12875980e-01 -1.05560517e+00 2.26707742e-01 8.76921833e-01
4.33872283e-01 8.24576557e-01 4.43578303e-01 -6.16002083e-01
-1.15841699e+00 -8.27657938e-01 2.01340222e+00 -2.72342503e-01
4.17894304e-01 -6.30198359e-01 -7.89825082e-01 8.09911489e-01
5.84717691e-01 -5.84629536e-01 8.72545004e-01 -1.37192547e-01
-1.08045563e-02 4.09221239e-02 -1.10143518e+00 2.95677304e-01
1.90296605e-01 -9.70227778e-01 -3.62600505e-01 3.93993378e-01
3.14974636e-01 4.50748242e-02 -6.50721848e-01 -3.05118114e-01
6.39320850e-01 -5.14470518e-01 4.82048810e-01 -4.46836889e-01
-4.16617393e-01 -2.33679369e-01 -5.92273951e-01 -8.20394218e-01
-4.32676896e-02 -7.69707739e-01 6.95815444e-01 1.83282876e+00
6.69943810e-01 -1.21000850e+00 6.85078382e-01 3.43103439e-01
-5.39472327e-02 -7.60911703e-02 -7.88726389e-01 -7.21234739e-01
2.33782157e-01 -2.96087831e-01 2.81057090e-01 9.15138125e-01
-2.82729924e-01 4.27630126e-01 -4.60463077e-01 4.60497022e-01
1.93572566e-01 -3.19912225e-01 3.44122440e-01 -8.39899004e-01
-1.83086306e-01 -2.14979295e-02 -1.27603441e-01 -1.04659748e+00
3.39982033e-01 -7.09499776e-01 2.96913892e-01 -1.22400701e+00
-1.12703674e-01 -4.36191112e-01 7.89698155e-04 7.06883371e-01
3.07459533e-01 2.98385054e-01 -2.83547174e-02 2.82117605e-01
1.79047108e-01 -4.23324630e-02 5.00488937e-01 3.07635039e-01
-6.10870004e-01 1.09506026e-01 -4.37898993e-01 6.90954208e-01
1.15386522e+00 -4.93245006e-01 -2.46289149e-01 -5.49887836e-01
-4.73160416e-01 -1.51305988e-01 -1.44387767e-01 -8.13566267e-01
2.97614962e-01 -1.30261794e-01 -1.44644901e-01 -7.27067709e-01
1.78189069e-01 -5.93166590e-01 5.91915241e-03 6.68192506e-01
-2.04792753e-01 5.58407843e-01 3.20977211e-01 -2.76203603e-01
-4.77205604e-01 -6.03305340e-01 1.07094371e+00 -2.46882334e-01
-1.22226822e+00 -2.96936959e-01 -1.18607354e+00 -2.87801325e-01
7.89358675e-01 -6.37318671e-01 2.44424380e-02 -5.20979702e-01
-4.27942693e-01 1.19927332e-01 1.06756113e-01 3.02341372e-01
5.90865612e-01 -1.03020501e+00 -8.69568348e-01 7.39185274e-01
-1.77807540e-01 -7.11969018e-01 -1.34877920e-01 4.92934108e-01
-1.09973538e+00 1.02080715e+00 -1.08743906e-01 -2.97232896e-01
-1.68181181e+00 -2.29607746e-01 3.91864449e-01 1.02017865e-01
1.33949921e-01 1.15517390e+00 -2.52453864e-01 -1.09248376e+00
2.20271781e-01 -1.29000425e-01 -5.64070523e-01 -1.85647786e-01
2.76880056e-01 3.78595322e-01 -1.06113434e-01 -1.42706692e+00
-7.27567852e-01 6.92029238e-01 -2.60455996e-01 -6.97723746e-01
7.68290818e-01 -3.18404317e-01 -5.01928568e-01 9.28046584e-01
1.24376953e+00 6.71905875e-01 -4.07165706e-01 2.48257220e-02
3.38158876e-01 7.72510022e-02 7.53151253e-02 -8.82181287e-01
-3.33843857e-01 7.56497443e-01 7.15157330e-01 -1.13784790e-01
8.89968514e-01 -9.49000269e-02 6.78165019e-01 4.35684770e-01
2.55276769e-01 -1.32899404e+00 -6.67645037e-01 1.16689289e+00
3.43075126e-01 -1.25313199e+00 -4.75055993e-01 -3.93308491e-01
-7.87613571e-01 1.12970710e+00 2.56642610e-01 -3.07398569e-02
8.81323040e-01 3.35287541e-01 6.32504225e-01 2.38814831e-01
-3.72765929e-01 -1.20849706e-01 2.13948354e-01 7.44593441e-01
8.39854777e-01 2.78549135e-01 -4.44561005e-01 2.84099013e-01
-5.72099864e-01 -3.98351938e-01 4.52196777e-01 9.15823936e-01
-9.20140505e-01 -1.45482612e+00 -8.53901327e-01 7.84606934e-02
-5.70282757e-01 -5.31898677e-01 -8.47207367e-01 8.65149796e-01
2.17819259e-01 1.13886678e+00 1.98350757e-01 -4.04699147e-01
-1.29366815e-01 4.68916357e-01 5.62006384e-02 -9.38866138e-01
-6.63727164e-01 5.86082101e-01 4.71046388e-01 4.40750033e-01
-4.34798509e-01 -8.73313546e-01 -1.59472644e+00 -3.48541886e-01
-1.38692409e-01 7.64403164e-01 8.79074275e-01 9.29970443e-01
-2.30121955e-01 -3.22270781e-01 7.26578414e-01 -1.93571433e-01
-3.29770356e-01 -1.11838734e+00 -9.24677610e-01 -3.94292831e-01
1.98148310e-01 1.02569364e-01 -1.80156991e-01 2.74473041e-01] | [14.243820190429688, 6.84208869934082] |
0b4b8e3c-b369-427a-bcf0-72c83c16a166 | predicting-gaze-in-egocentric-video-by | 1803.09125 | null | http://arxiv.org/abs/1803.09125v3 | http://arxiv.org/pdf/1803.09125v3.pdf | Predicting Gaze in Egocentric Video by Learning Task-dependent Attention Transition | We present a new computational model for gaze prediction in egocentric videos
by exploring patterns in temporal shift of gaze fixations (attention
transition) that are dependent on egocentric manipulation tasks. Our assumption
is that the high-level context of how a task is completed in a certain way has
a strong influence on attention transition and should be modeled for gaze
prediction in natural dynamic scenes. Specifically, we propose a hybrid model
based on deep neural networks which integrates task-dependent attention
transition with bottom-up saliency prediction. In particular, the
task-dependent attention transition is learned with a recurrent neural network
to exploit the temporal context of gaze fixations, e.g. looking at a cup after
moving gaze away from a grasped bottle. Experiments on public egocentric
activity datasets show that our model significantly outperforms
state-of-the-art gaze prediction methods and is able to learn meaningful
transition of human attention. | ['Zhenqiang Li', 'Minjie Cai', 'Yifei Huang', 'Yoichi Sato'] | 2018-03-24 | predicting-gaze-in-egocentric-video-by-1 | http://openaccess.thecvf.com/content_ECCV_2018/html/Huang_Predicting_Gaze_in_ECCV_2018_paper.html | http://openaccess.thecvf.com/content_ECCV_2018/papers/Huang_Predicting_Gaze_in_ECCV_2018_paper.pdf | eccv-2018-9 | ['eye-tracking'] | ['computer-vision'] | [ 2.91249037e-01 9.98048782e-02 -1.84883028e-01 -3.40336651e-01
1.11799031e-01 -3.08134053e-02 3.65282029e-01 -2.63411552e-01
-1.78844005e-01 3.97075444e-01 4.66821462e-01 2.81291828e-02
-1.44511327e-01 -1.04925089e-01 -8.90685380e-01 -6.87194228e-01
-1.12849995e-01 -2.14520663e-01 3.24937195e-01 -2.43613735e-01
9.47127879e-01 7.54591078e-02 -2.02313066e+00 1.89057842e-01
4.99652535e-01 8.99673045e-01 7.34407902e-01 7.06733763e-01
3.64934504e-01 1.09515047e+00 -2.40068242e-01 9.65770483e-02
-2.13223353e-01 -6.33940518e-01 -1.08863032e+00 2.20309868e-02
5.07112741e-01 -1.94509417e-01 -2.33042352e-02 9.34146643e-01
2.14814752e-01 5.99114239e-01 6.96863651e-01 -1.19481587e+00
-8.49751592e-01 3.12235057e-01 -8.08045983e-01 1.02830744e+00
5.52648187e-01 4.61123198e-01 9.10217464e-01 -7.62659550e-01
8.08486581e-01 1.04784942e+00 8.44521821e-02 7.82834530e-01
-1.03288782e+00 -4.12655205e-01 7.06596136e-01 8.14369380e-01
-9.83409047e-01 -3.51790905e-01 1.04266953e+00 -8.09110165e-01
9.19594228e-01 1.13460369e-01 6.79899931e-01 1.39133418e+00
7.71775961e-01 9.40722823e-01 6.90993249e-01 -5.10078311e-01
1.00231975e-01 -2.95420319e-01 1.39584705e-01 6.33762121e-01
-1.95314541e-01 -1.77021012e-01 -1.14126778e+00 3.03883612e-01
4.78472978e-01 3.08737218e-01 -4.93188888e-01 -6.25283241e-01
-1.41147590e+00 5.32141984e-01 8.69692564e-01 2.39028826e-01
-7.35400558e-01 2.78545856e-01 1.76672891e-01 -1.05446309e-01
6.91484392e-01 5.37755728e-01 -3.59480172e-01 -2.59621352e-01
-7.49412715e-01 1.09551638e-01 2.03307614e-01 9.92585897e-01
9.25045609e-01 -1.78575709e-01 -6.97522581e-01 2.38322511e-01
2.88450271e-01 4.22098339e-01 6.08757257e-01 -8.97686958e-01
3.37238461e-01 5.66210151e-01 3.01060051e-01 -8.87039840e-01
-5.79939723e-01 5.24489991e-02 -1.05911098e-01 2.08223939e-01
4.70231026e-01 -1.92528367e-01 -7.35391259e-01 2.01971102e+00
1.85704321e-01 3.59690219e-01 -2.71040887e-01 1.17688680e+00
5.24248660e-01 2.55270690e-01 4.21272278e-01 -4.33519095e-01
1.61996901e+00 -1.15397751e+00 -1.03092873e+00 -3.31610709e-01
6.11759603e-01 -4.38117474e-01 1.31644249e+00 2.07376897e-01
-1.34686720e+00 -7.61058986e-01 -6.05134189e-01 -1.72859892e-01
-2.12883160e-01 -9.13969651e-02 4.11709517e-01 5.83008602e-02
-1.00432634e+00 4.28639144e-01 -7.91663885e-01 -8.17532122e-01
5.09317696e-01 3.18988472e-01 -7.91942403e-02 4.04817253e-01
-9.42905366e-01 8.54567528e-01 9.07497108e-02 1.81997716e-01
-1.07258987e+00 -6.51586294e-01 -9.04208302e-01 4.77776945e-01
5.94585955e-01 -9.57914114e-01 1.42814243e+00 -1.43097174e+00
-1.58700895e+00 8.11453521e-01 -1.04745209e+00 -4.21261549e-01
-7.99633283e-03 -7.09932089e-01 -2.32878491e-01 3.36060971e-01
2.21576430e-02 8.58258367e-01 1.47864568e+00 -8.06395411e-01
-7.24843740e-01 -5.47154546e-01 2.42619216e-01 4.49731201e-01
-2.80250460e-01 4.07274038e-01 -1.82035312e-01 -1.67331457e-01
-2.13682622e-01 -1.13730454e+00 2.38444716e-01 -1.91670954e-01
-2.04195276e-01 -7.80574858e-01 9.50296879e-01 -3.23797166e-01
1.72924089e+00 -1.95216203e+00 6.28202677e-01 -3.68854344e-01
3.99744332e-01 -3.04888673e-02 2.44256437e-01 1.71678096e-01
-4.04079169e-01 -1.00399546e-01 2.21107438e-01 -4.33586657e-01
-3.57908964e-01 -4.21469271e-01 -4.00720209e-01 3.03023160e-01
3.38091612e-01 1.06142378e+00 -1.31610608e+00 -3.83303165e-01
3.01098023e-02 3.27628583e-01 -7.66367912e-01 3.58709455e-01
-3.07484239e-01 7.87690401e-01 -5.06420732e-01 5.53667009e-01
1.63138524e-01 -6.57324195e-01 -4.28376757e-02 -1.72843935e-03
-3.36134732e-01 3.95151943e-01 -2.07098261e-01 1.88672435e+00
-1.53381914e-01 9.93236721e-01 -4.72158045e-01 -4.67260063e-01
4.42030847e-01 1.46364748e-01 1.86001331e-01 -7.40473747e-01
3.38849008e-01 -4.64550614e-01 1.75084829e-01 -1.09742355e+00
7.36152232e-01 3.50327313e-01 1.79778039e-01 8.43742788e-01
1.42574444e-01 4.83544290e-01 1.67014316e-01 -1.64915901e-02
6.87070012e-01 7.89889693e-01 3.38630825e-01 -4.58398849e-01
7.16253161e-01 -2.21754462e-01 3.34943593e-01 4.54481423e-01
-7.20537126e-01 5.52913427e-01 8.59302461e-01 -5.29131293e-01
-6.53931975e-01 -4.15169179e-01 3.96755487e-01 2.04643846e+00
4.58817899e-01 -2.03506812e-01 -1.12585235e+00 -7.78203487e-01
-5.17370343e-01 1.00042808e+00 -1.53534341e+00 -5.63262343e-01
-7.87308753e-01 -2.63662905e-01 -4.27953929e-01 5.98324418e-01
1.25878066e-01 -1.83720636e+00 -1.26501477e+00 -3.24591011e-01
-1.61992788e-01 -5.80706000e-01 -8.90020311e-01 1.76534103e-03
-8.46169531e-01 -1.28276873e+00 -7.83043563e-01 -5.81930161e-01
8.17730486e-01 6.66594446e-01 1.06666255e+00 -1.12356067e-01
9.83116329e-02 6.04720891e-01 -2.20225334e-01 -7.23845065e-01
4.23891455e-01 3.62674206e-01 2.36033097e-01 3.31338346e-01
1.07242596e+00 -3.18621218e-01 -1.05297554e+00 3.34272772e-01
-3.00745100e-01 2.94436403e-02 9.87870768e-02 6.80610895e-01
2.39706397e-01 -8.36923718e-01 2.66229659e-01 -7.05772221e-01
6.02422357e-01 -8.23149562e-01 -3.85944873e-01 2.99564838e-01
-3.86121958e-01 6.12351857e-02 -6.02740329e-03 -4.38652307e-01
-1.35867143e+00 -1.65034756e-01 7.58084118e-01 -9.56845164e-01
-1.80879131e-01 2.74206132e-01 4.04968351e-01 2.97840446e-01
7.64511943e-01 4.99402061e-02 -7.17774406e-02 -2.25702971e-01
2.71319393e-02 1.87831491e-01 1.75621226e-01 -1.51562914e-01
2.60249019e-01 5.20982385e-01 5.91746345e-02 -5.08943021e-01
-1.23712599e+00 -4.97878850e-01 -1.10647476e+00 -6.62958980e-01
1.32690573e+00 -7.22589433e-01 -1.47858763e+00 4.33603257e-01
-1.02806747e+00 -5.92339933e-01 -6.55500442e-02 4.82357115e-01
-8.70601058e-01 -1.15208970e-02 -1.04565322e-01 -8.54318500e-01
-2.84724802e-01 -1.08634198e+00 1.25692427e+00 8.13958883e-01
-4.51710552e-01 -1.13207042e+00 6.47996888e-02 1.58578366e-01
2.72763133e-01 2.58556474e-02 5.67615449e-01 -3.22351366e-01
-8.89063179e-01 4.20208484e-01 -1.36255622e-01 -2.76091397e-01
4.13963825e-01 1.19947284e-01 -1.09270465e+00 -1.67507917e-01
3.09144240e-02 -1.22787356e-01 7.57030547e-01 8.43275189e-01
1.30846655e+00 -1.13403238e-02 -7.19533205e-01 5.72809517e-01
8.19791675e-01 9.26475525e-02 6.02869034e-01 2.98892587e-01
7.46395350e-01 8.30421984e-01 1.11731935e+00 3.21058691e-01
5.84635317e-01 6.58321619e-01 8.45992863e-01 4.40936208e-01
1.88743904e-01 -7.50529915e-02 3.36384296e-01 2.11728081e-01
-6.19869292e-01 -2.35819876e-01 -8.36160302e-01 7.90124357e-01
-1.98277617e+00 -1.23597276e+00 -7.76965767e-02 2.09397173e+00
3.66196513e-01 1.15579948e-01 2.21598431e-01 -4.61077273e-01
8.88702810e-01 3.43728095e-01 -8.80623221e-01 -6.25632405e-01
5.43697238e-01 -1.84770897e-01 -2.91717350e-02 2.53433436e-01
-1.06361103e+00 1.19189620e+00 6.55508947e+00 -2.81060617e-02
-1.38765919e+00 2.02481627e-01 2.61544555e-01 -4.58409041e-01
1.41296044e-01 -2.03003839e-01 -9.84150589e-01 6.94960952e-01
9.05637622e-01 -1.19649842e-01 3.30174983e-01 8.93543243e-01
4.39242005e-01 -4.75322455e-01 -1.31608474e+00 8.67574751e-01
5.19263387e-01 -9.63541508e-01 -2.11064756e-01 9.75909010e-02
6.18845403e-01 -6.24467321e-02 4.46622729e-01 1.86858177e-01
-6.52959272e-02 -7.77658701e-01 7.64853776e-01 1.36645281e+00
4.41493005e-01 -4.60183471e-01 3.25949728e-01 4.03554291e-01
-8.80747736e-01 -5.03491104e-01 -2.10780635e-01 -3.36140215e-01
2.14352190e-01 -4.06720728e-01 -7.96334326e-01 7.99588859e-02
1.13497126e+00 1.16970122e+00 -7.70424247e-01 7.82896340e-01
-6.00032151e-01 4.63657558e-01 2.94292867e-01 -2.21135676e-01
2.50578493e-01 7.38227665e-02 6.93429172e-01 7.30993986e-01
3.41364563e-01 1.76714912e-01 -6.16964281e-01 1.01819062e+00
1.36461526e-01 -1.99275970e-01 -6.72517359e-01 2.81928807e-01
4.04422134e-02 1.11055338e+00 -3.63500983e-01 -4.18804176e-02
-2.85132468e-01 8.32387865e-01 6.04485571e-01 6.22868240e-01
-9.06006873e-01 -1.66617304e-01 9.02058840e-01 2.46423274e-01
8.07615757e-01 5.28524332e-02 1.00132696e-01 -1.09865201e+00
-1.06159292e-01 -1.44371375e-01 3.71491075e-01 -1.72918332e+00
-7.81257093e-01 5.54383695e-01 -2.07205825e-02 -1.38565135e+00
-5.78904033e-01 -3.65418166e-01 -1.00739861e+00 1.15685880e+00
-1.74724698e+00 -1.11470735e+00 -7.94635713e-01 7.80803740e-01
9.30072963e-01 -1.24164954e-01 5.02593040e-01 -4.06652480e-01
-7.71389782e-01 3.83023143e-01 -5.73910534e-01 -3.43379140e-01
9.05611515e-01 -1.14823139e+00 1.98518038e-01 8.80460441e-01
-2.99457788e-01 9.92775798e-01 7.77508557e-01 -5.54075778e-01
-1.08222663e+00 -7.56333470e-01 1.06415927e+00 -1.01320899e+00
5.71877539e-01 -2.50663608e-01 -8.95048976e-01 1.16367900e+00
6.91098094e-01 -5.21050952e-03 6.70760691e-01 4.35864180e-01
7.06865266e-03 1.49649933e-01 -5.60416102e-01 8.03616524e-01
1.21035790e+00 -6.00317657e-01 -1.08844507e+00 1.88131630e-01
6.19405687e-01 -4.40340638e-01 -3.11353147e-01 1.50810525e-01
5.62243462e-01 -1.08795261e+00 6.56918943e-01 -1.13181639e+00
6.51519775e-01 -2.51028150e-01 4.61418480e-01 -1.14610124e+00
-8.30087364e-01 -9.14908171e-01 -7.28627503e-01 7.38738656e-01
-8.39047953e-02 -2.91627049e-01 5.57967484e-01 5.32933831e-01
-1.90917969e-01 -4.89834011e-01 -5.56778312e-01 -1.13920532e-01
-5.02316236e-01 8.44375938e-02 5.21831870e-01 7.52482712e-01
5.22655427e-01 5.30480742e-01 -4.82361436e-01 -1.68476850e-02
1.83107510e-01 1.12699702e-01 9.24812257e-01 -1.38588798e+00
4.96960610e-01 -4.46636230e-01 -2.38248438e-01 -1.17780423e+00
4.62327242e-01 -2.87503392e-01 -4.65446055e-06 -1.03912926e+00
3.03710610e-01 2.60946810e-01 -8.12561035e-01 3.80504102e-01
-6.45774901e-01 1.91305548e-01 2.21200898e-01 4.19527531e-01
-1.01350069e+00 5.67686319e-01 1.57186413e+00 2.57898837e-01
-5.25660455e-01 1.54424012e-01 -8.42159688e-01 7.96697974e-01
4.64850843e-01 -3.64582896e-01 -5.13272166e-01 -3.80338401e-01
6.22585356e-01 -2.09778436e-02 2.56315202e-01 -8.71050417e-01
5.08583009e-01 -3.54907453e-01 6.96956292e-02 -7.30111361e-01
3.91137689e-01 -6.98966146e-01 -5.39019346e-01 2.50189096e-01
-6.14194751e-01 9.43956338e-03 1.07178308e-01 8.52328718e-01
5.83329201e-02 -2.09821567e-01 4.70427245e-01 3.48896310e-02
-8.40342045e-01 1.91758394e-01 -4.24815148e-01 -1.48998424e-01
1.18060577e+00 -4.34027433e-01 -4.26841021e-01 -3.95368248e-01
-1.13433969e+00 2.92373359e-01 3.86958569e-01 9.20493066e-01
2.44719356e-01 -1.05263078e+00 -2.63103813e-01 2.87230372e-01
4.49131221e-01 -2.36147135e-01 4.32239622e-01 1.28179300e+00
-1.23931225e-02 6.82800949e-01 -5.58453381e-01 -1.02982080e+00
-1.32976162e+00 1.18134236e+00 3.02571952e-01 7.09082037e-02
-4.25103664e-01 1.04614544e+00 1.05017662e+00 4.77127045e-01
-1.55615686e-02 -6.00675881e-01 -1.08635700e+00 2.39411160e-01
7.62736082e-01 3.01476151e-01 -3.37924689e-01 -9.38404679e-01
-2.61880994e-01 7.30343640e-01 -2.07338423e-01 5.17506659e-01
1.28356504e+00 -6.55491412e-01 -2.70901527e-02 9.28483427e-01
6.29372716e-01 -3.75988543e-01 -1.92004502e+00 -2.70277202e-01
-9.90460515e-02 -5.24861038e-01 -1.07918531e-02 -5.11641502e-01
-9.82049584e-01 1.17882252e+00 5.66652954e-01 2.39691600e-01
1.24793339e+00 1.20237708e-01 3.71427506e-01 4.82694298e-01
2.31870100e-01 -1.00499618e+00 6.28076196e-01 7.16325223e-01
1.01177287e+00 -1.31055582e+00 -1.33740846e-02 -9.09654200e-02
-9.73118782e-01 9.89058495e-01 1.06053174e+00 -2.35505670e-01
8.22728217e-01 -7.30788410e-01 -2.86963612e-01 -5.98718703e-01
-1.00120795e+00 -2.96100408e-01 6.69480920e-01 4.63969886e-01
3.27507854e-01 -2.80351579e-01 1.20481171e-01 3.77330005e-01
3.54472883e-02 2.50103593e-01 4.63137299e-01 8.66410732e-01
-4.85446841e-01 -2.82123536e-01 -2.09193423e-01 1.46153495e-01
-4.17997152e-01 -1.19531333e-01 -1.38113841e-01 6.01946175e-01
-6.53380807e-03 5.53957820e-01 5.32476664e-01 -1.21982776e-01
1.48139939e-01 5.19183397e-01 4.79975879e-01 -8.81311178e-01
-6.43870533e-01 4.08379221e-03 -4.73433614e-01 -9.52548742e-01
-1.01183581e+00 -1.16520238e+00 -7.55048275e-01 -1.49381785e-02
-3.49969983e-01 -2.66611814e-01 3.80926162e-01 1.12238991e+00
6.93309128e-01 8.82467330e-01 5.60782850e-01 -1.35212517e+00
1.53617665e-01 -1.37717021e+00 -5.23248732e-01 3.66476178e-01
7.16937602e-01 -1.21613526e+00 -4.12306100e-01 6.00904703e-01] | [13.975506782531738, 0.04987473413348198] |
844a1811-5df0-4632-b58c-76151c6cb079 | sentence-level-sign-language-recognition | 2211.14447 | null | https://arxiv.org/abs/2211.14447v1 | https://arxiv.org/pdf/2211.14447v1.pdf | Sentence-Level Sign Language Recognition Framework | We present two solutions to sentence-level SLR. Sentence-level SLR required mapping videos of sign language sentences to sequences of gloss labels. Connectionist Temporal Classification (CTC) has been used as the classifier level of both models. CTC is used to avoid pre-segmenting the sentences into individual words. The first model is an LRCN-based model, and the second model is a Multi-Cue Network. LRCN is a model in which a CNN as a feature extractor is applied to each frame before feeding them into an LSTM. In the first approach, no prior knowledge has been leveraged. Raw frames are fed into an 18-layer LRCN with a CTC on top. In the second approach, three main characteristics (hand shape, hand position, and hand movement information) associated with each sign have been extracted using Mediapipe. 2D landmarks of hand shape have been used to create the skeleton of the hands and then are fed to a CONV-LSTM model. Hand locations and hand positions as relative distance to head are fed to separate LSTMs. All three sources of information have been then integrated into a Multi-Cue network with a CTC classification layer. We evaluated the performance of proposed models on RWTH-PHOENIX-Weather. After performing an excessive search on model hyper-parameters such as the number of feature maps, input size, batch size, sequence length, LSTM memory cell, regularization, and dropout, we were able to achieve 35 Word Error Rate (WER). | ['Atra Akandeh'] | 2022-11-13 | null | null | null | null | ['sign-language-recognition'] | ['computer-vision'] | [ 4.86657560e-01 -5.60391508e-02 -1.21301927e-01 -2.93747187e-01
-7.48807669e-01 -4.71098483e-01 4.61915433e-01 -4.90064561e-01
-9.52974677e-01 4.92087245e-01 4.15088058e-01 -9.06844065e-02
2.73369700e-01 -1.98547825e-01 -5.92035830e-01 -7.42495000e-01
-8.70331470e-03 2.42665812e-01 6.16339445e-01 6.88844686e-03
3.74482840e-01 5.83393574e-01 -1.44039667e+00 5.83453357e-01
3.09608966e-01 1.05162871e+00 4.90651757e-01 1.00564110e+00
-3.78936417e-02 9.78864253e-01 -6.80263162e-01 3.50727476e-02
2.08521426e-01 -5.20022273e-01 -6.79858267e-01 -7.18389377e-02
6.26125813e-01 -3.81065875e-01 -6.07190967e-01 7.02687860e-01
8.73472750e-01 2.91187853e-01 6.22355759e-01 -7.69988298e-01
-7.12176487e-02 3.96483988e-01 -3.82535189e-01 1.63043395e-01
3.12421590e-01 4.89707589e-01 8.39028060e-01 -6.96954429e-01
1.01798093e+00 1.26817465e+00 5.48874915e-01 7.37466753e-01
-8.64197671e-01 -4.23098326e-01 1.63366631e-01 2.35155270e-01
-9.19533908e-01 -3.07527304e-01 5.60649395e-01 -4.10473168e-01
1.33153999e+00 1.28543019e-01 9.13612366e-01 1.28954530e+00
1.64332241e-01 1.07373130e+00 9.85643804e-01 -7.01926053e-01
2.72193085e-02 -1.99527238e-02 1.84459999e-01 6.89158618e-01
-3.60235870e-01 -1.50619755e-02 -8.48524451e-01 2.98476368e-01
1.02862525e+00 -1.46274477e-01 -2.35995933e-01 1.95317626e-01
-1.21456456e+00 6.24424875e-01 5.33144951e-01 6.28540456e-01
-3.66725206e-01 4.37951744e-01 6.88475847e-01 2.64421016e-01
5.77714965e-02 1.18296884e-01 -4.92738098e-01 -4.38643321e-02
-9.98790860e-01 -9.43325758e-02 4.56169844e-01 9.48568821e-01
1.28586426e-01 1.37274653e-01 -6.84496343e-01 6.09758794e-01
3.74066085e-01 5.17752945e-01 9.53596711e-01 -7.60212243e-01
7.91362822e-01 4.84021097e-01 -1.70174763e-01 -7.95435905e-01
-5.08364379e-01 -1.98032096e-01 -3.68111640e-01 3.05212766e-01
6.78000689e-01 -2.77987570e-01 -1.41642308e+00 1.51958728e+00
-2.23716214e-01 1.48522690e-01 4.41918895e-02 1.18602562e+00
8.52666557e-01 5.32910228e-01 1.23590030e-01 -4.76723537e-02
1.39572072e+00 -9.91116345e-01 -6.08585000e-01 -2.37780839e-01
6.28684819e-01 -6.42923772e-01 1.11241901e+00 4.24239725e-01
-1.06617498e+00 -6.73776507e-01 -8.36629450e-01 -2.76257843e-01
-4.52809095e-01 6.50258005e-01 2.46313542e-01 3.14041525e-01
-1.26438284e+00 5.41325331e-01 -8.53606105e-01 -5.67458272e-01
1.79323062e-01 5.53101838e-01 -3.74346256e-01 1.17783800e-01
-1.06230223e+00 1.14730918e+00 2.95728564e-01 7.18452990e-01
-7.55163193e-01 -1.02095030e-01 -8.51066113e-01 -3.00673693e-01
5.14030382e-02 -4.02743995e-01 9.65541005e-01 -1.13633502e+00
-1.85754073e+00 7.92028248e-01 -4.22922760e-01 -3.92212719e-01
6.18648529e-01 -1.18064247e-01 -1.88500285e-01 3.75547498e-01
-2.52807081e-01 9.27797556e-01 1.26993310e+00 -9.89853323e-01
-7.37185359e-01 -1.90837339e-01 -2.29002714e-01 3.03957939e-01
-2.22729221e-02 3.82635236e-01 -4.52953309e-01 -6.47915840e-01
3.35304886e-01 -1.00317919e+00 7.27163032e-02 -7.34434053e-02
-2.54589945e-01 -2.18503878e-01 9.01650429e-01 -1.22988069e+00
1.00679660e+00 -2.12018967e+00 5.29065073e-01 1.73280045e-01
-1.30719155e-01 5.07457674e-01 -4.16304797e-01 3.17118838e-02
4.47037928e-02 -1.40832111e-01 3.51210572e-02 -4.99858856e-01
-2.59307802e-01 1.46002784e-01 -1.38610572e-01 4.88788784e-01
9.72465798e-02 9.14211154e-01 -6.96237385e-01 -5.37628353e-01
4.68895137e-01 8.36930275e-01 -2.92872518e-01 2.45179832e-02
-3.72420043e-01 5.43565750e-01 -2.30794758e-01 4.72387820e-01
1.78084612e-01 1.23800009e-01 1.65877610e-01 -4.57176387e-01
-1.49054870e-01 3.57258081e-01 -1.02435279e+00 1.90475714e+00
-6.15569115e-01 1.07519603e+00 -2.52251476e-01 -5.73740125e-01
7.00502396e-01 6.78518057e-01 2.05592737e-01 -5.62608898e-01
5.26123166e-01 3.04141998e-01 2.98551042e-02 -8.44266355e-01
1.43230259e-01 1.78214461e-01 1.08596593e-01 2.72396356e-01
2.39836365e-01 1.27800092e-01 2.39664957e-01 -1.72870725e-01
8.98041487e-01 6.01126492e-01 -1.80705622e-01 2.22386450e-01
6.46629572e-01 -1.13348730e-01 3.54995936e-01 5.13985276e-01
-6.99981526e-02 8.91107142e-01 4.03394789e-01 -3.70958328e-01
-8.68568182e-01 -7.06475616e-01 3.11503917e-01 1.07064497e+00
-5.25286674e-01 4.05620188e-02 -7.55036473e-01 -6.13320529e-01
-3.29564840e-01 6.41390204e-01 -5.83815575e-01 1.30080834e-01
-1.04391539e+00 -2.72738457e-01 6.95526123e-01 7.55623698e-01
4.39443409e-01 -1.78411567e+00 -1.15206265e+00 1.71129733e-01
-1.63385540e-01 -1.10520232e+00 -6.52256131e-01 4.28989947e-01
-9.46196139e-01 -9.06497419e-01 -1.01276779e+00 -1.11103773e+00
7.99588740e-01 -3.60254914e-01 3.12236935e-01 -1.50657028e-01
-3.78704518e-01 1.75857514e-01 -4.38800305e-01 -1.39518389e-02
-1.41010344e-01 8.58454406e-02 3.46251316e-02 1.61366120e-01
3.36570382e-01 -3.35096985e-01 -5.26512682e-01 3.58803533e-02
-5.85935891e-01 -5.79983741e-02 7.28847146e-01 6.91390634e-01
4.00904685e-01 -5.11392474e-01 1.07045442e-01 -2.91874796e-01
4.46483612e-01 1.48402512e-01 -5.59600115e-01 2.63442189e-01
1.40974596e-01 1.67995170e-01 4.77725238e-01 -8.15426886e-01
-9.53201413e-01 5.86493969e-01 -9.45723727e-02 -5.45351982e-01
-3.19927871e-01 2.86396652e-01 -8.35995562e-03 3.58910002e-02
3.01960856e-01 5.35249114e-01 5.74182868e-02 -4.08855349e-01
3.34553540e-01 8.19347382e-01 5.34801424e-01 -4.40719314e-02
3.78331691e-01 2.33579472e-01 -1.78132430e-01 -1.04555035e+00
-6.65101707e-01 -3.21824908e-01 -1.26665533e+00 -4.97608483e-01
1.26034868e+00 -5.66361368e-01 -8.25649977e-01 8.50761712e-01
-1.42961740e+00 -6.47032797e-01 -1.92715883e-01 7.31523335e-01
-7.27409482e-01 1.20910607e-01 -6.91947937e-01 -8.56841326e-01
-4.36449170e-01 -1.07785237e+00 8.48240912e-01 3.01513374e-01
-3.81621808e-01 -9.43544447e-01 -2.93141633e-01 3.08823824e-01
3.04104000e-01 2.40390226e-01 7.14118242e-01 -6.39272332e-01
-2.39273325e-01 -3.15235138e-01 -1.65398210e-01 4.92660016e-01
8.37187096e-02 -1.16951801e-01 -1.10266626e+00 -3.28005791e-01
4.10092808e-02 -3.56909633e-01 8.51127267e-01 8.37514162e-01
7.15377927e-01 -2.47898310e-01 -2.39015505e-01 5.44022143e-01
1.18783128e+00 4.12656367e-01 6.65487647e-01 2.40280107e-01
7.38978505e-01 6.21841967e-01 3.83227110e-01 -1.03203975e-01
4.10010479e-02 6.98388875e-01 -3.67848650e-02 -6.92825913e-02
-6.38317287e-01 -7.61689097e-02 7.76451349e-01 8.26009810e-01
-3.61999512e-01 -1.52723268e-01 -7.01684237e-01 3.16055834e-01
-1.73671174e+00 -9.86251891e-01 -1.67184800e-01 2.03937411e+00
6.39140725e-01 3.55839640e-01 2.78547525e-01 2.49879092e-01
5.44514596e-01 1.40070617e-01 -5.79303145e-01 -4.06067133e-01
-1.59988940e-01 3.06309443e-02 6.61473274e-01 6.53816223e-01
-8.80938768e-01 1.40853739e+00 6.01523399e+00 4.39918697e-01
-1.68741286e+00 9.59669501e-02 6.85864612e-02 -4.82852221e-01
3.53789777e-01 -2.93511242e-01 -1.02161288e+00 5.37989080e-01
9.03131783e-01 7.39374042e-01 4.77865756e-01 3.42397004e-01
6.55393660e-01 -3.51223677e-01 -1.09228551e+00 8.88678491e-01
4.28851336e-01 -9.32284176e-01 -3.92537788e-02 -1.12881437e-01
3.35866034e-01 2.86957711e-01 -8.98338258e-02 5.57106845e-02
-2.75867358e-02 -1.00387919e+00 9.26400304e-01 8.59303474e-01
1.11693275e+00 -4.08803731e-01 8.03046882e-01 1.64929688e-01
-1.40632403e+00 -2.80192554e-01 -8.33637919e-03 1.29588515e-01
3.08902651e-01 -2.69086629e-01 -7.31833935e-01 2.04826966e-01
5.54249644e-01 6.06596291e-01 -4.81437802e-01 9.91511762e-01
-6.28789246e-01 6.29278302e-01 -5.74476480e-01 -2.27852687e-01
4.45909411e-01 1.63670272e-01 5.07796466e-01 1.54115307e+00
2.10871801e-01 -5.39667942e-02 -1.25428110e-01 4.78697747e-01
3.24610502e-01 -1.34768546e-01 -3.99266720e-01 1.46077022e-01
1.30819872e-01 9.79191959e-01 -8.48087430e-01 -2.41428167e-01
-3.75904799e-01 1.33495975e+00 -5.55956783e-03 7.02394664e-01
-5.64674973e-01 -5.57179272e-01 5.41658849e-02 2.64109392e-02
6.08845115e-01 -3.94089103e-01 -2.22755462e-01 -9.96785104e-01
2.03351304e-01 -7.82629251e-01 4.31977153e-01 -8.48024607e-01
-8.14063609e-01 8.69571805e-01 -2.14905322e-01 -1.14516509e+00
-5.63274443e-01 -9.42618072e-01 -4.46476370e-01 1.06907654e+00
-1.43133485e+00 -1.37207305e+00 -1.65060118e-01 8.07355881e-01
8.54946494e-01 -3.86555791e-01 7.21596003e-01 1.57445490e-01
-5.28577864e-01 5.19021571e-01 -2.83362180e-01 5.30761003e-01
6.24116004e-01 -1.04688478e+00 2.57856339e-01 8.10794830e-01
9.69547555e-02 4.73596066e-01 4.56277639e-01 -7.12756634e-01
-1.20675874e+00 -8.49717021e-01 1.12312829e+00 -4.41525161e-01
5.36844194e-01 -5.88587582e-01 -4.71949697e-01 8.67963791e-01
5.75479381e-02 -9.92373284e-03 1.66186675e-01 -5.20761251e-01
-1.43503830e-01 -4.27708812e-02 -9.80617404e-01 5.22364557e-01
9.01202261e-01 -7.10390210e-01 -8.18844318e-01 2.45791778e-01
4.00324374e-01 -4.76865888e-01 -5.76736569e-01 -1.08301550e-01
7.36643434e-01 -5.26790857e-01 6.13314271e-01 -5.71609676e-01
5.67499436e-02 -3.39939415e-01 -1.43025845e-01 -1.11161852e+00
2.28257049e-02 -5.43122828e-01 -1.95082687e-02 9.15656924e-01
5.18135130e-01 -4.45190221e-01 6.85894132e-01 4.67212856e-01
-6.36990145e-02 -5.06021857e-01 -1.13782227e+00 -7.82493591e-01
-1.74855813e-01 -5.03333211e-01 -1.25551507e-01 3.59971255e-01
-1.07965343e-01 4.16867405e-01 -3.06156576e-01 3.61212157e-02
3.93029779e-01 -3.25007915e-01 3.14116478e-01 -9.64839399e-01
-8.71756598e-02 -5.38349271e-01 -3.18454295e-01 -1.15692890e+00
1.98177859e-01 -7.73098171e-01 2.57924139e-01 -1.80151045e+00
-8.79082233e-02 -7.13482723e-02 -1.68533921e-01 8.34238112e-01
3.70623231e-01 3.92140150e-02 4.73478347e-01 1.07628979e-01
-9.67337936e-02 1.78102702e-01 1.10159910e+00 -2.37529445e-02
-5.24375141e-01 1.26079887e-01 3.88071239e-01 8.27850521e-01
6.14646554e-01 -3.89182180e-01 -1.31178543e-01 -7.05004692e-01
-2.34075651e-01 2.51823068e-01 5.12142718e-01 -1.20759320e+00
5.95711052e-01 2.06849366e-01 4.23430115e-01 -8.19771647e-01
5.38189352e-01 -9.10720885e-01 -2.83581376e-01 6.73813581e-01
-5.37610173e-01 -9.86579284e-02 1.83154553e-01 2.05431134e-01
-1.00913070e-01 -3.56438905e-01 7.74328709e-01 -2.23864958e-01
-8.00941408e-01 -7.06456974e-02 -6.29500151e-01 -2.50229567e-01
7.20827341e-01 -5.37226558e-01 1.07780889e-01 -2.00280637e-01
-1.08866549e+00 1.47449031e-01 -5.51036298e-02 5.63432574e-01
7.34088600e-01 -1.01629758e+00 -6.22751594e-01 3.32000643e-01
-3.53537649e-01 -1.24775670e-01 4.06778827e-02 9.74490285e-01
-5.74955344e-01 8.40497136e-01 -3.69577348e-01 -6.00212693e-01
-1.46363258e+00 2.38374412e-01 6.31362557e-01 -2.64794622e-02
-8.49202693e-01 9.68248069e-01 -4.62200075e-01 -2.24432781e-01
8.98382068e-01 -6.28578782e-01 -4.16121304e-01 4.34840858e-01
5.19160867e-01 2.59575397e-01 -6.55284598e-02 -7.90941238e-01
-4.59313244e-01 1.04694295e+00 3.67844738e-02 -6.23183370e-01
1.35054231e+00 1.28799543e-01 1.73400998e-01 6.81405485e-01
1.20166278e+00 -3.73896241e-01 -1.56327260e+00 -6.37948886e-02
1.61499903e-01 -5.02449349e-02 1.27606109e-01 -9.61625755e-01
-1.23798907e+00 1.12952149e+00 8.02645981e-01 -4.18556839e-01
1.05615938e+00 -2.87187248e-01 7.99368083e-01 3.13539922e-01
2.84697652e-01 -1.30584061e+00 1.41400754e-01 7.33418643e-01
1.08637166e+00 -9.98064637e-01 -3.98042291e-01 1.73790649e-01
-7.94816136e-01 1.38250196e+00 5.12652874e-01 -1.32864997e-01
3.48120838e-01 3.24909687e-01 3.20647299e-01 -6.22835122e-02
-4.32219535e-01 -3.26813430e-01 4.96649891e-01 4.99273986e-01
4.83267814e-01 -3.06320131e-01 -1.81862250e-01 3.84967417e-01
-7.26837367e-02 3.09910893e-01 1.58245400e-01 9.26990092e-01
-2.65337318e-01 -7.92749465e-01 -1.84466287e-01 2.75604725e-01
-2.02331588e-01 -1.54028907e-01 -2.96258777e-01 7.45651186e-01
2.08852082e-01 6.73755705e-01 1.59864128e-01 -4.88281459e-01
4.77645814e-01 3.36321056e-01 6.71789765e-01 -6.38249934e-01
-9.21008646e-01 1.99291885e-01 7.74606392e-02 -5.31817257e-01
-2.47570395e-01 -8.31158698e-01 -1.48171341e+00 3.72742355e-01
-8.77087936e-02 -4.13512439e-01 7.93585420e-01 1.28560317e+00
-5.71971014e-02 4.76479739e-01 3.44763212e-02 -1.19376254e+00
-4.64510441e-01 -1.32145321e+00 -2.84307450e-01 4.70023528e-02
5.06141841e-01 -3.99030060e-01 -3.75172526e-01 2.66355246e-01] | [9.17330551147461, -6.480647087097168] |
dbb7e8c5-cd09-4662-8673-d78770ce14eb | gradual-test-time-adaptation-by-self-training | 2208.07736 | null | https://arxiv.org/abs/2208.07736v2 | https://arxiv.org/pdf/2208.07736v2.pdf | Introducing Intermediate Domains for Effective Self-Training during Test-Time | Experiencing domain shifts during test-time is nearly inevitable in practice and likely results in a severe performance degradation. To overcome this issue, test-time adaptation continues to update the initial source model during deployment. A promising direction are methods based on self-training which have been shown to be well suited for gradual domain adaptation, since reliable pseudo-labels can be provided. In this work, we address two problems that exist when applying self-training in the setting of test-time adaptation. First, adapting a model to long test sequences that contain multiple domains can lead to error accumulation. Second, naturally, not all shifts are gradual in practice. To tackle these challenges, we introduce GTTA. By creating artificial intermediate domains that divide the current domain shift into a more gradual one, effective self-training through high quality pseudo-labels can be performed. To create the intermediate domains, we propose two independent variations: mixup and light-weight style transfer. We demonstrate the effectiveness of our approach on the continual and gradual corruption benchmarks, as well as ImageNet-R. To further investigate gradual shifts in the context of urban scene segmentation, we publish a new benchmark: CarlaTTA. It enables the exploration of several non-stationary domain shifts. | ['Bin Yang', 'Mario Döbler', 'Robert A. Marsden'] | 2022-08-16 | null | null | null | null | ['scene-segmentation'] | ['computer-vision'] | [ 3.99441957e-01 -3.28776181e-01 -1.90327555e-01 -4.72063243e-01
-8.89018178e-01 -7.22482920e-01 5.09496033e-01 6.37069941e-02
-5.65244973e-01 1.06833887e+00 -1.23863205e-01 -2.46676639e-01
2.43706405e-01 -5.80794930e-01 -8.53043616e-01 -5.85418642e-01
-6.38652146e-02 6.62642837e-01 8.19813848e-01 -3.47484082e-01
4.39481363e-02 3.44894439e-01 -1.55325437e+00 2.46063128e-01
1.29929173e+00 7.34917223e-01 1.01374879e-01 6.17533505e-01
-2.40802884e-01 5.84223747e-01 -9.41616237e-01 -2.39520118e-01
3.68431449e-01 -5.03682554e-01 -8.78759682e-01 4.36965525e-01
5.20776629e-01 -2.65331179e-01 5.26812151e-02 9.34113562e-01
6.18629992e-01 2.74228185e-01 4.99867976e-01 -1.12787616e+00
-2.43575111e-01 4.42496717e-01 -5.59637666e-01 3.32016677e-01
2.36712351e-01 4.96675342e-01 4.27770674e-01 -7.35335350e-01
7.75695741e-01 1.03677440e+00 7.20950425e-01 6.00484312e-01
-1.34093046e+00 -6.82672977e-01 4.15068418e-01 4.38791633e-01
-1.26107669e+00 -3.88775915e-01 8.45824957e-01 -3.23660642e-01
6.00628614e-01 5.77060468e-02 4.08302248e-01 1.36192632e+00
-2.28193417e-01 9.50463176e-01 1.44316244e+00 -3.65601778e-01
4.54422235e-01 2.12631136e-01 -8.12093839e-02 8.12008753e-02
6.28255075e-03 1.20284475e-01 -2.55576223e-01 7.49826133e-02
4.05444711e-01 -4.14835840e-01 -2.31057778e-01 -5.68093657e-01
-1.06072307e+00 4.68509167e-01 3.15416634e-01 2.65933216e-01
-2.09326312e-01 -1.09883472e-01 6.61283672e-01 6.28047228e-01
6.52684808e-01 5.08564591e-01 -5.76156378e-01 -5.01323462e-01
-1.15603769e+00 2.32603237e-01 6.89138412e-01 8.94729257e-01
8.92221332e-01 2.06921771e-01 -2.11768895e-01 1.05066741e+00
-1.44097641e-01 3.01315844e-01 5.58297575e-01 -7.94324875e-01
5.69264710e-01 1.94799900e-01 2.18182936e-01 -4.03342903e-01
-1.44051746e-01 -6.85754538e-01 -6.92277491e-01 2.42029324e-01
6.97555840e-01 -2.09028974e-01 -1.34718919e+00 1.81079853e+00
5.48777521e-01 6.83590353e-01 2.75489967e-02 7.24434376e-01
2.68519282e-01 4.44899768e-01 1.32982820e-01 -1.64000943e-01
6.98159277e-01 -1.01425660e+00 -4.38351691e-01 -4.67588454e-01
5.55403054e-01 -6.72732770e-01 1.41750741e+00 5.15715480e-01
-9.86841798e-01 -7.30949461e-01 -1.07970631e+00 5.15539050e-01
-3.17222595e-01 -3.90752286e-01 -6.29515480e-03 4.92701322e-01
-9.31556106e-01 6.84836686e-01 -8.43491256e-01 -4.14323419e-01
4.55272257e-01 1.42569229e-01 -1.38006508e-01 -3.58097047e-01
-1.21421361e+00 8.74567807e-01 4.60179687e-01 -1.79404348e-01
-1.00339723e+00 -6.67034864e-01 -6.89749122e-01 -2.98216730e-01
6.27455294e-01 -3.89553100e-01 1.38252735e+00 -1.32404375e+00
-1.62998652e+00 8.45240533e-01 -4.85893935e-02 -4.96500701e-01
9.35130477e-01 -2.88264334e-01 -6.00144804e-01 -1.18012324e-01
1.16525359e-01 6.27228737e-01 1.07295883e+00 -1.49263930e+00
-6.59903288e-01 -7.53987879e-02 -2.28157938e-02 2.90261000e-01
-4.50858265e-01 -2.25781307e-01 -5.40852427e-01 -6.79762959e-01
-2.04466224e-01 -1.06747866e+00 -3.17680597e-01 -3.27841431e-01
-1.92014799e-01 1.19226702e-01 8.45359802e-01 -4.38054025e-01
1.25859356e+00 -2.27465391e+00 8.82644206e-02 2.37943023e-01
-1.14728913e-01 6.03620052e-01 -3.35295230e-01 4.51235939e-03
-7.79590756e-02 7.50638545e-02 -7.17498183e-01 -3.85317445e-01
-2.13322058e-01 5.73788583e-01 -1.39839575e-01 2.41994411e-01
4.55857128e-01 6.23231053e-01 -1.11997354e+00 -3.89974266e-01
4.28112336e-02 1.48968890e-01 -6.74778640e-01 1.35637924e-01
-4.21095192e-01 8.86479199e-01 -1.71226040e-01 5.61223984e-01
7.91818738e-01 -1.35412380e-01 -1.34549560e-02 2.56846488e-01
5.43700391e-03 2.38422573e-01 -1.23320568e+00 1.67704463e+00
-6.28762186e-01 5.98119617e-01 -1.20270066e-01 -1.09980893e+00
7.32689798e-01 9.08057243e-02 3.43088806e-01 -8.54461670e-01
-1.95572764e-01 3.92549425e-01 9.98418406e-02 -4.04953778e-01
5.47899306e-01 -2.02530578e-01 2.15138122e-02 9.26194638e-02
-3.21526341e-02 -3.55075538e-01 4.77830738e-01 -1.71596900e-01
1.24444199e+00 4.21944022e-01 9.17251557e-02 3.36431991e-03
4.89179999e-01 2.21152097e-01 7.04284430e-01 7.25811958e-01
-4.38990027e-01 7.34899282e-01 2.67986953e-01 -3.25131983e-01
-1.14870894e+00 -1.10031819e+00 -1.24019049e-01 1.03650963e+00
2.64519572e-01 -3.25953364e-02 -7.68682361e-01 -1.00058353e+00
-1.21963561e-01 7.92686820e-01 -4.08117324e-01 -2.84084857e-01
-8.88226151e-01 -8.00653934e-01 5.34574747e-01 5.07852793e-01
8.23641598e-01 -1.06901574e+00 -7.10894763e-01 5.06229460e-01
-2.90202230e-01 -1.39270842e+00 -4.31328803e-01 4.17719960e-01
-9.34637189e-01 -8.21449995e-01 -9.99450684e-01 -6.54664576e-01
5.27764857e-01 1.15285188e-01 1.53750014e+00 -9.10093859e-02
5.16459346e-02 2.35375211e-01 -5.87942302e-01 -1.73358247e-01
-8.44479859e-01 3.27151805e-01 -4.33954075e-02 -2.28661038e-02
1.08750217e-01 -7.24862456e-01 -5.52666128e-01 6.52039170e-01
-1.13985229e+00 -1.61795944e-01 5.69513917e-01 9.81200695e-01
7.46516228e-01 1.41833350e-01 6.38754308e-01 -1.21169031e+00
5.93156636e-01 -6.47774875e-01 -5.36422849e-01 2.08283320e-01
-6.34185791e-01 1.29875168e-01 9.97145414e-01 -9.20362294e-01
-1.12085032e+00 -1.76049247e-02 -3.63923788e-01 -4.58012611e-01
-5.32797754e-01 5.15454888e-01 -1.60383675e-02 -3.86219891e-03
1.09308255e+00 2.96319336e-01 -3.32571357e-01 -3.57658178e-01
3.83187145e-01 5.17085671e-01 5.70264697e-01 -7.93050528e-01
1.09755123e+00 3.04727912e-01 -4.19801027e-01 -8.03845823e-01
-7.06916809e-01 -5.79447031e-01 -4.39146131e-01 -2.39145204e-01
4.89081115e-01 -8.65275860e-01 3.43283862e-01 7.17059553e-01
-7.88992703e-01 -1.18938482e+00 -6.89463019e-01 1.02891825e-01
-6.19884968e-01 4.76400077e-01 -2.65700608e-01 -3.17742348e-01
7.72226229e-02 -1.21397626e+00 7.60273278e-01 9.79044065e-02
-1.26128957e-01 -9.83319104e-01 2.10268646e-01 -8.24706256e-02
5.87361813e-01 4.08625752e-01 6.02743030e-01 -6.72871947e-01
-2.78061152e-01 1.08425140e-01 3.21258046e-02 6.70327902e-01
1.36237517e-01 -1.15001589e-01 -7.80115008e-01 -5.10248780e-01
-1.78735450e-01 -5.18098116e-01 8.29957247e-01 1.59616042e-02
1.14009941e+00 7.00272247e-02 -1.35812059e-01 6.94517255e-01
1.25646472e+00 2.47296780e-01 8.04522574e-01 6.84279323e-01
4.71773565e-01 2.83196390e-01 9.92716968e-01 3.41915637e-01
3.81381750e-01 9.13440883e-01 3.33664536e-01 -2.50997007e-01
-4.65126812e-01 -6.02302216e-02 4.66208011e-01 6.81661844e-01
8.09451044e-02 -3.51074815e-01 -1.16895020e+00 8.46354544e-01
-1.64836133e+00 -7.88684428e-01 5.38375638e-02 2.42704248e+00
1.14451802e+00 6.06304824e-01 3.55015993e-01 1.92926839e-01
6.00956798e-01 3.11528202e-02 -8.19238424e-01 -2.32426360e-01
-1.37067899e-01 3.55195284e-01 6.13118470e-01 3.94429177e-01
-1.17357540e+00 1.13125324e+00 6.08002949e+00 9.04278755e-01
-1.35697997e+00 3.60053360e-01 6.00771189e-01 5.92928380e-02
-2.15650722e-01 -4.01702896e-02 -6.20509505e-01 7.48257339e-01
1.06310868e+00 1.36611760e-01 2.61731535e-01 8.86056840e-01
5.04081920e-02 -2.33603671e-01 -9.36082363e-01 7.28197992e-01
-1.03153236e-01 -8.26865852e-01 -2.71628290e-01 -2.32520714e-01
1.13599706e+00 2.93350965e-01 1.94195956e-01 6.14455342e-01
5.45802116e-01 -5.25773823e-01 7.32520401e-01 3.48084345e-02
1.12940037e+00 -6.60609484e-01 5.93008041e-01 4.13621545e-01
-1.05530083e+00 -1.29951224e-01 -2.56037861e-01 1.94854304e-01
1.63019672e-01 6.80657744e-01 -1.04575694e+00 5.07164061e-01
6.80455685e-01 6.65557861e-01 -6.92947268e-01 1.37946212e+00
-3.87371808e-01 9.68644798e-01 -4.35585320e-01 3.84211183e-01
2.21374720e-01 3.16700041e-02 6.54479325e-01 1.46339703e+00
3.08532000e-01 -3.33447635e-01 3.07004243e-01 5.99196970e-01
3.75219919e-02 -1.51069984e-01 -5.16034842e-01 4.31071818e-01
4.50702727e-01 7.51314461e-01 -8.05272341e-01 -5.06431639e-01
-2.57377833e-01 1.33254194e+00 2.14031041e-01 5.67309439e-01
-1.07158911e+00 -2.00459678e-02 6.21958613e-01 3.45897019e-01
3.81795853e-01 -3.68460506e-01 -1.61006987e-01 -1.19825482e+00
1.29806980e-01 -1.21222675e+00 2.51780182e-01 -3.36259514e-01
-1.14135790e+00 7.79070139e-01 2.98534632e-01 -1.54287684e+00
-3.54925007e-01 -1.46833658e-01 -4.20961112e-01 5.31015277e-01
-2.03608131e+00 -8.38100195e-01 -5.74757934e-01 6.15873516e-01
1.06064892e+00 -1.60104467e-03 5.35363138e-01 6.32829547e-01
-4.19917971e-01 7.93884754e-01 2.53186762e-01 -1.76167369e-01
1.11109316e+00 -1.41560543e+00 7.28095233e-01 1.06443644e+00
1.20390214e-01 9.69283730e-02 9.16490734e-01 -5.49853206e-01
-6.85837269e-01 -1.32281125e+00 2.61586189e-01 -3.41344267e-01
5.03234565e-01 -2.91249156e-01 -1.21483219e+00 4.47364748e-01
-2.93286908e-02 2.18116000e-01 3.97122562e-01 -9.57513303e-02
-2.72091061e-01 -2.03436509e-01 -1.07035673e+00 5.86834252e-01
1.15465641e+00 -1.70926318e-01 -3.44033509e-01 1.89657047e-01
6.69398010e-01 -6.83354378e-01 -6.25768185e-01 6.33219898e-01
1.33149490e-01 -9.93750989e-01 9.28818822e-01 -2.38642424e-01
2.98455000e-01 -3.72253209e-01 1.80661932e-01 -1.75308084e+00
3.38844443e-03 -6.39333487e-01 -5.74667342e-02 1.32646096e+00
4.34259534e-01 -6.23179555e-01 9.17797625e-01 2.24085778e-01
-2.45386600e-01 -5.34698665e-01 -1.01170969e+00 -1.26441014e+00
3.15465748e-01 -5.00528693e-01 6.28290236e-01 9.64713037e-01
-4.56305772e-01 3.61784957e-02 -3.15478504e-01 5.98266604e-04
3.95793945e-01 -2.10520938e-01 9.96207476e-01 -1.00705719e+00
-4.23009962e-01 -2.83684671e-01 -2.40431309e-01 -1.36295664e+00
1.18075125e-01 -5.21068990e-01 4.46854383e-01 -1.08834815e+00
-2.30860129e-01 -8.54111850e-01 -3.72155309e-01 3.08155149e-01
-3.23795199e-01 2.73629278e-01 1.51781127e-01 2.06590250e-01
-7.82616436e-01 4.18829650e-01 1.26129735e+00 -1.20987825e-01
-5.19203365e-01 2.43598446e-01 -3.62072855e-01 5.23812234e-01
9.80879188e-01 -5.53711057e-01 -5.80495059e-01 -5.21634638e-01
6.80745840e-02 -1.09138995e-01 7.23589733e-02 -1.40081799e+00
-1.02095697e-02 -1.57131895e-01 -4.43270765e-02 -3.37390393e-01
8.41242746e-02 -9.22451913e-01 2.99693998e-02 2.17048213e-01
-1.48416504e-01 6.22398444e-02 4.47548181e-01 4.89008516e-01
-4.66742128e-01 -1.64319828e-01 1.11618626e+00 -3.38874124e-02
-1.07363307e+00 1.17394790e-01 -2.50003427e-01 3.89696777e-01
1.10266924e+00 -5.59048176e-01 2.23857118e-03 -3.16168636e-01
-8.04772556e-01 4.34153795e-01 6.96344376e-01 4.07594353e-01
5.37019789e-01 -1.14838326e+00 -7.00330853e-01 1.68840364e-01
2.75308043e-01 4.41755056e-01 1.47223487e-01 7.83564866e-01
-4.96105969e-01 -2.61613250e-01 -2.21378848e-01 -8.55585217e-01
-9.82455432e-01 4.28793937e-01 3.58949244e-01 -6.89360559e-01
-3.97152722e-01 8.14593852e-01 3.64540704e-02 -4.07054573e-01
1.60961419e-01 -3.47341299e-01 -8.75875726e-02 -1.45805823e-02
2.71448821e-01 1.42408326e-01 2.38884613e-01 -4.56344128e-01
-1.96256965e-01 5.39677501e-01 -1.79894745e-01 -2.17364654e-01
1.25753152e+00 -2.41312191e-01 4.16967660e-01 4.35616314e-01
8.98926735e-01 -8.91888738e-02 -1.75249493e+00 -4.16911811e-01
2.91034639e-01 -4.90684688e-01 -4.18687075e-01 -1.02454567e+00
-8.91405523e-01 8.13736379e-01 8.17074835e-01 1.24356553e-01
1.44436061e+00 -2.94108689e-01 8.31880391e-01 6.69147298e-02
4.31615174e-01 -1.36628413e+00 3.68021727e-01 6.90059543e-01
6.45312369e-01 -1.41263211e+00 -3.16698611e-01 -2.64381170e-01
-6.85595810e-01 8.17778468e-01 8.67830634e-01 -4.34092395e-02
2.97417372e-01 2.94562757e-01 1.59405842e-01 2.93625236e-01
-6.42528594e-01 -4.53603804e-01 8.66923598e-04 8.21632206e-01
9.75161791e-02 -8.73217136e-02 -1.88923746e-01 -7.10699102e-03
1.27143543e-02 1.42557979e-01 5.47324836e-01 1.02244234e+00
-2.35985741e-01 -1.44371891e+00 -5.38214266e-01 1.58886462e-01
-3.23604792e-01 7.75158405e-02 -1.20051377e-01 8.07386577e-01
3.42446864e-01 7.29439557e-01 -6.42336607e-02 -3.04790229e-01
6.87741399e-01 1.54743299e-01 3.84245247e-01 -7.64995635e-01
-4.34149832e-01 1.11126512e-01 1.06861129e-01 -4.57876444e-01
-4.52640831e-01 -7.42673934e-01 -1.04470479e+00 -6.09369799e-02
-2.29422763e-01 1.79550238e-02 5.01548827e-01 9.00075912e-01
1.54087007e-01 7.99452901e-01 5.71106672e-01 -7.25368261e-01
-7.39286482e-01 -8.51351082e-01 -2.08961159e-01 8.33387971e-01
3.73694211e-01 -7.24578977e-01 -2.80509531e-01 1.45295650e-01] | [9.70718765258789, 1.5189100503921509] |
a3519b25-43e7-4166-bb37-4c94e9305cf3 | smile-and-laugh-expressions-detection-based | 2101.01874 | null | https://arxiv.org/abs/2101.01874v1 | https://arxiv.org/pdf/2101.01874v1.pdf | Smile and Laugh Expressions Detection Based on Local Minimum Key Points | In this paper, a smile and laugh facial expression is presented based on dimension reduction and description process of the key points. The paper has two main objectives; the first is to extract the local critical points in terms of their apparent features, and the second is to reduce the system's dependence on training inputs. To achieve these objectives, three different scenarios on extracting the features are proposed. First of all, the discrete parts of a face are detected by local binary pattern method that is used to extract a set of global feature vectors for texture classification considering various regions of an input-image face. Then, in the first scenario and with respect to the correlation changes of adjacent pixels on the texture of a mouth area, a set of local key points are extracted using the Harris corner detector. In the second scenario, the dimension reduction of the extracted points of first scenario provided by principal component analysis algorithm leading to reduction in computational costs and overall complexity without loss of performance and flexibility, etc. | ['Majid Harouni', 'Mina Mohammadi Dashti'] | 2021-01-06 | null | null | null | null | ['texture-classification'] | ['computer-vision'] | [ 2.86203891e-01 8.72619301e-02 6.80763274e-02 -1.37771279e-01
-1.83300421e-01 -1.71536624e-01 4.44927186e-01 1.54091239e-01
-6.85157031e-02 2.65924603e-01 -1.91119611e-02 3.08300883e-01
-4.64039594e-01 -6.17228627e-01 -6.51279911e-02 -9.88946795e-01
1.00315206e-01 6.49886802e-02 1.06660753e-01 -2.33824924e-01
6.35919869e-01 1.15839911e+00 -1.80908966e+00 1.42898113e-01
1.46048591e-01 1.23663878e+00 -1.83497697e-01 4.01435286e-01
-8.29386860e-02 4.31058109e-01 -1.42373949e-01 -7.83933848e-02
2.42820680e-01 -4.32630688e-01 -4.34336662e-01 6.05484307e-01
9.23437253e-02 -1.96730345e-01 3.26480180e-01 1.23102164e+00
4.43973839e-01 -3.94621678e-02 1.04432070e+00 -9.78960216e-01
1.20249636e-01 -2.63133973e-01 -9.50284004e-01 -4.54871021e-02
3.56781155e-01 -6.33032480e-03 7.03427255e-01 -1.25828779e+00
8.39190781e-01 1.33037746e+00 4.30452436e-01 3.25251132e-01
-1.09763932e+00 -3.08903992e-01 -4.62483943e-01 3.56022090e-01
-1.43709588e+00 -6.76391602e-01 1.34993875e+00 -4.88719761e-01
4.03771222e-01 3.41527194e-01 5.76249480e-01 4.37814951e-01
3.02081287e-01 4.32126641e-01 1.26009810e+00 -7.97535121e-01
2.71704763e-01 4.39867646e-01 3.00138384e-01 9.27898943e-01
1.56279668e-01 -1.53205603e-01 -4.06220526e-01 -2.63077289e-01
5.29849172e-01 5.70075661e-02 6.46026060e-03 -5.63428819e-01
-4.26518112e-01 7.31437802e-01 -2.56760836e-01 7.08280563e-01
-4.93749827e-01 -4.85592991e-01 3.25084984e-01 4.67458479e-02
1.19188912e-01 -1.18787088e-01 -1.65448517e-01 2.05421215e-03
-1.06364059e+00 -9.83973444e-02 6.69529140e-01 3.73518169e-01
9.40910935e-01 -1.99581429e-01 2.94467419e-01 4.75403786e-01
5.10119796e-01 4.59519655e-01 5.60023010e-01 -3.66678208e-01
2.99020201e-01 1.13226593e+00 -1.26405761e-01 -1.75563467e+00
-5.76285481e-01 -1.65873393e-01 -6.72250569e-01 6.47531271e-01
2.42675200e-01 -1.05591565e-01 -4.30308461e-01 1.08710444e+00
7.14159071e-01 -2.37617835e-01 9.37404260e-02 5.72295249e-01
6.05859518e-01 4.99410748e-01 -3.52488995e-01 -6.13785565e-01
1.54767919e+00 -3.03365082e-01 -7.16997385e-01 2.36409232e-01
2.15207264e-01 -1.02275431e+00 7.75852084e-01 3.28394055e-01
-7.31017351e-01 -5.96826017e-01 -1.13051343e+00 3.93618852e-01
-2.59323359e-01 6.86266720e-01 1.28439024e-01 5.59991956e-01
-8.37603629e-01 4.61259574e-01 -5.45934856e-01 -4.93422925e-01
7.38278329e-02 4.32742387e-01 -4.86705601e-01 3.91614825e-01
-4.80598986e-01 6.88828528e-01 4.33368497e-02 4.12695378e-01
-1.52745575e-01 -9.20972601e-02 -6.79655015e-01 2.48337179e-01
-1.32857516e-01 6.55295774e-02 4.76873666e-01 -1.31681788e+00
-1.65983438e+00 7.37328172e-01 -4.49738979e-01 2.11260766e-01
5.56937099e-01 2.56873161e-01 -3.92587215e-01 6.69752598e-01
-1.52622938e-01 2.57670343e-01 1.33745301e+00 -1.04636121e+00
-5.21588802e-01 -8.31616104e-01 -7.58507133e-01 3.35965186e-01
-4.60626006e-01 2.44914964e-01 -2.97341675e-01 -1.98033184e-01
4.90277380e-01 -8.34175169e-01 1.18995070e-01 -8.10894445e-02
-1.96117014e-01 -2.72654384e-01 1.36391091e+00 -7.84841359e-01
1.04522097e+00 -2.51049685e+00 1.73475631e-02 7.99399197e-01
-6.71689808e-02 2.07215965e-01 1.12719893e-01 3.12306046e-01
-1.65664732e-01 -1.58828512e-01 -2.60773227e-02 -5.73886260e-02
-3.50864768e-01 -3.42533886e-01 -1.51345483e-03 7.52556801e-01
4.05044645e-01 1.31271690e-01 -2.80240268e-01 -8.36557746e-01
4.52509344e-01 5.82218826e-01 -3.93093944e-01 -1.93052992e-01
3.72700602e-01 3.88241231e-01 -6.65852249e-01 6.80138588e-01
9.78535712e-01 4.57862228e-01 2.74391621e-01 -4.61364865e-01
-1.07003249e-01 -3.61196101e-01 -1.76629901e+00 7.10886538e-01
-1.83696061e-01 5.03656149e-01 1.54623076e-01 -9.50491965e-01
1.37734759e+00 4.30392653e-01 7.74961531e-01 -5.41259527e-01
4.09322619e-01 2.94458359e-01 -2.10112542e-01 -5.37480235e-01
3.38883214e-02 8.59850124e-02 2.68175334e-01 2.16999918e-01
4.45130393e-02 5.86354136e-02 -4.05931249e-02 -4.02976274e-01
3.31649393e-01 1.72017798e-01 7.31915593e-01 -4.07877386e-01
1.11974275e+00 -4.03135359e-01 3.17728847e-01 -2.17154711e-01
-2.70592988e-01 1.23372301e-01 7.56903112e-01 -5.93253791e-01
-8.38939726e-01 -3.88968050e-01 -3.23758394e-01 3.01220328e-01
3.21218595e-02 3.36907012e-03 -9.34168279e-01 -4.43581253e-01
-1.69777006e-01 5.26295491e-02 -6.55470371e-01 5.77285513e-02
-5.02016783e-01 -6.44688725e-01 9.56520438e-03 -1.94490731e-01
6.75450027e-01 -1.16456461e+00 -1.28081477e+00 -7.57592395e-02
3.38469386e-01 -6.19459391e-01 2.34298618e-03 -1.13707602e-01
-1.04119515e+00 -1.25367057e+00 -4.95969713e-01 -7.70863891e-01
9.09779906e-01 6.99432865e-02 3.52067024e-01 2.05075052e-02
-5.92159212e-01 2.00502053e-01 -2.86451936e-01 -3.03825200e-01
-1.59982428e-01 -2.86541939e-01 2.33697314e-02 6.98108613e-01
6.82826221e-01 -4.38404977e-01 -4.80332196e-01 3.40327442e-01
-6.47669673e-01 -4.33981508e-01 5.82010806e-01 6.81863844e-01
4.52298611e-01 2.06073135e-01 2.49992445e-01 -2.78120160e-01
4.51818645e-01 -9.93309096e-02 -7.10202336e-01 1.38273031e-01
-3.21745455e-01 1.25396371e-01 4.59569722e-01 -1.61102414e-01
-1.13579834e+00 4.56090987e-01 8.18353966e-02 1.94589626e-02
-2.54015625e-01 1.73235431e-01 -2.97867984e-01 -3.42959374e-01
5.40240288e-01 5.24842203e-01 3.41705978e-01 -3.86462808e-01
-5.23179322e-02 7.87174463e-01 -2.22265106e-02 -9.98398885e-02
6.71855092e-01 6.58675611e-01 6.80441856e-01 -1.40588212e+00
-5.33663929e-02 -6.23962998e-01 -9.68314767e-01 -4.17916864e-01
7.20058858e-01 -2.00823143e-01 -1.04512572e+00 4.05286938e-01
-1.04825675e+00 4.90136832e-01 -3.58480439e-02 3.86571407e-01
-6.26820922e-01 7.42010832e-01 -2.13098109e-01 -1.11602414e+00
-5.91846168e-01 -1.24590814e+00 9.27833974e-01 4.67378467e-01
-1.70802996e-01 -6.86235726e-01 -1.19837016e-01 -5.42614423e-02
-1.02528399e-02 4.55613405e-01 8.08218181e-01 -3.45027834e-01
-2.03225136e-01 -5.25352538e-01 -3.39192078e-02 4.28701848e-01
3.54400247e-01 5.14083087e-01 -7.59253561e-01 -1.66474998e-01
6.37145936e-01 1.24763586e-01 2.27947980e-01 4.40191299e-01
5.48096001e-01 -5.22522867e-01 -4.17236507e-01 4.51517612e-01
1.53408182e+00 7.59412587e-01 6.00987196e-01 3.06942463e-01
1.38025492e-01 8.96771908e-01 8.69429648e-01 5.98102152e-01
-3.58726472e-01 7.36807823e-01 3.48027945e-01 -3.04253668e-01
-6.59078956e-02 1.00119375e-01 4.13450003e-01 4.01033878e-01
-3.77770998e-02 4.78583723e-01 -5.49631596e-01 4.48532224e-01
-1.50839865e+00 -9.80079353e-01 3.21412738e-03 2.20560026e+00
2.42797658e-01 -1.43224731e-01 2.14480743e-01 6.39036298e-01
9.41817522e-01 -7.59417471e-03 -2.93161124e-01 -6.71781838e-01
-3.64171155e-02 1.41487107e-01 1.73624381e-01 5.38820148e-01
-9.22284663e-01 6.22653365e-01 5.29866314e+00 8.32031488e-01
-1.52661383e+00 -3.00048351e-01 6.97950065e-01 3.83389533e-01
2.89701968e-01 6.93944225e-04 -9.85845745e-01 4.60848391e-01
3.26518536e-01 3.14710945e-01 4.07412052e-02 9.20713961e-01
2.91396469e-01 -4.63123798e-01 -5.33271313e-01 1.23150408e+00
2.41722569e-01 -6.90578878e-01 1.61277622e-01 2.20702305e-01
4.68305647e-01 -8.02422702e-01 2.28122696e-01 -4.69429970e-01
-7.09737420e-01 -7.40460336e-01 5.05680203e-01 7.97743440e-01
4.86286789e-01 -1.16227686e+00 8.23989272e-01 1.26660928e-01
-1.14299560e+00 -1.80089593e-01 -3.80577922e-01 2.16263812e-02
-3.00292403e-01 4.80308682e-01 -9.85654652e-01 4.09453511e-01
4.17193323e-01 1.65284529e-01 -4.51505423e-01 8.31200242e-01
-3.51327881e-02 3.74796718e-01 -4.41791952e-01 -4.43916291e-01
1.37585565e-01 -5.19593835e-01 7.42729664e-01 1.07256067e+00
4.35772002e-01 -4.03379500e-02 -3.46904665e-01 5.22843838e-01
5.45418262e-01 6.93201900e-01 -6.72242761e-01 2.95717120e-01
4.15085912e-01 1.66854227e+00 -8.88404906e-01 -1.45583481e-01
-1.41994819e-01 8.81123900e-01 -1.60268441e-01 7.92987794e-02
-3.11319500e-01 -6.98572993e-01 1.43185064e-01 3.79303068e-01
2.68530548e-01 -1.94208637e-01 -3.56048167e-01 -7.03023553e-01
1.70886114e-01 -8.77908230e-01 2.25318506e-01 -2.77474523e-01
-5.44791281e-01 5.46979904e-01 -2.06744462e-01 -1.30013299e+00
-4.10469174e-01 -6.72574282e-01 -5.10263741e-01 9.23117340e-01
-1.16538799e+00 -1.20860219e+00 -3.01363379e-01 8.45422626e-01
4.28092569e-01 -3.47555280e-01 6.14263058e-01 4.31339070e-02
-4.05749381e-01 4.84018385e-01 2.31116712e-01 5.56761585e-02
4.23913360e-01 -5.69503486e-01 -3.52672487e-01 8.63344491e-01
-2.28314534e-01 4.92339551e-01 7.00116515e-01 -5.11932075e-01
-1.16152942e+00 -3.25928360e-01 9.90309656e-01 3.58897783e-02
2.46719137e-01 -7.56177902e-02 -2.73318231e-01 2.32084110e-01
-2.34465167e-01 -1.97314009e-01 2.92677790e-01 -3.40028048e-01
1.82128593e-01 -3.85894686e-01 -1.51433897e+00 5.46643913e-01
2.45762214e-01 -3.79911363e-01 -4.20022935e-01 1.39224619e-01
-4.13605511e-01 8.94009843e-02 -6.76210642e-01 3.00375640e-01
7.82094538e-01 -1.15252542e+00 6.97141111e-01 -1.47570491e-01
1.27579883e-01 -4.10454363e-01 -9.03841630e-02 -7.21150637e-01
-2.69927830e-01 -5.65553963e-01 3.62790585e-01 1.34417892e+00
1.65623739e-01 -5.42129457e-01 8.61053169e-01 1.15347281e-01
4.61650103e-01 -6.25281096e-01 -1.16474903e+00 -3.24160755e-01
-5.63503861e-01 3.00726175e-01 3.78554374e-01 6.41747534e-01
2.78025985e-01 2.83624172e-01 -2.47729659e-01 5.63122556e-02
6.09817326e-01 1.89990774e-01 7.95513928e-01 -1.30859768e+00
1.48051664e-01 -2.95897722e-01 -9.91585255e-01 -4.51687217e-01
-2.64730573e-01 -3.66613805e-01 -2.86247432e-01 -1.00664902e+00
8.65897834e-02 -4.67975959e-02 -3.08082569e-02 2.80534983e-01
2.58154929e-01 1.32108554e-01 1.38887927e-01 3.13189775e-01
1.51299730e-01 3.22914839e-01 1.02902293e+00 1.38135552e-01
-4.78835762e-01 1.83697239e-01 -2.43543506e-01 9.07841682e-01
5.58047354e-01 -1.04119517e-01 -2.49633238e-01 3.48568857e-01
-1.28003448e-01 2.79053926e-01 2.06484079e-01 -1.22093451e+00
1.39496149e-02 -1.72258481e-01 5.81589222e-01 -6.26333892e-01
3.94794255e-01 -1.01670635e+00 6.21297248e-02 7.33409166e-01
2.01381490e-01 1.22229472e-01 6.04661927e-02 1.56921044e-01
-4.04756516e-01 -5.53468168e-01 1.25994539e+00 6.76715970e-02
-7.09585190e-01 -2.52797991e-01 -3.32815200e-01 -8.43350530e-01
1.43182492e+00 -7.80734718e-01 1.84797898e-01 -2.37834603e-01
-7.02480316e-01 -6.01705909e-01 3.24451864e-01 1.69838294e-02
5.96148789e-01 -1.07847428e+00 -4.11043167e-01 5.60591280e-01
-1.59115911e-01 -4.31610227e-01 5.06194055e-01 9.66768324e-01
-8.60385895e-01 4.30583775e-01 -7.79026330e-01 -5.95159709e-01
-1.82063460e+00 6.15934134e-01 3.22684288e-01 -6.46806434e-02
-3.65847409e-01 3.72541010e-01 -1.31582171e-01 1.93594456e-01
-9.44892541e-02 -1.83323026e-01 -8.50956142e-01 5.60992777e-01
4.32163805e-01 5.45563877e-01 2.36894384e-01 -1.14441895e+00
-5.28089762e-01 1.29212403e+00 3.14504504e-01 -1.15285866e-01
1.19945288e+00 -2.04046622e-01 -3.71209234e-01 7.21060932e-02
1.46458519e+00 6.53478265e-01 -1.15324163e+00 2.62777954e-02
2.24224374e-01 -4.85408932e-01 -9.04414654e-02 -4.05988961e-01
-8.07119071e-01 9.01896894e-01 1.07798517e+00 1.95157353e-03
1.48620510e+00 -4.89337713e-01 1.48345202e-01 1.86733827e-01
1.07853830e-01 -1.19985068e+00 -3.01085003e-02 1.53934926e-01
9.17559683e-01 -8.95744145e-01 -3.64321023e-02 -5.24948597e-01
-4.73166257e-01 1.78993571e+00 9.33467373e-02 -2.25174367e-01
8.31779122e-01 4.86558005e-02 -1.02763727e-01 -4.54068482e-01
-6.49861470e-02 -3.75771672e-02 4.31476295e-01 4.44429398e-01
1.57184914e-01 -1.29092321e-01 -9.49405134e-01 3.17258000e-01
-1.29863814e-01 9.54796895e-02 3.21674049e-01 7.93001115e-01
-8.39937449e-01 -7.16593862e-01 -7.38815367e-01 1.69177670e-02
-4.12823349e-01 1.32771865e-01 -3.64025861e-01 8.56306970e-01
3.48426938e-01 1.01783752e+00 -8.57952237e-02 -3.54464114e-01
2.45755509e-01 1.48834005e-01 3.43550414e-01 -1.21429406e-01
-1.34742513e-01 2.48869687e-01 -2.37675533e-01 -3.30305517e-01
-2.88794547e-01 -8.89806092e-01 -9.14510429e-01 5.55971339e-02
-1.61263540e-01 2.66865730e-01 1.10872185e+00 7.28992164e-01
2.83446968e-01 -8.52188542e-02 9.89923239e-01 -9.09651875e-01
-6.63759649e-01 -9.41144645e-01 -8.11497033e-01 4.85373735e-01
2.62922734e-01 -6.91729128e-01 -5.02918541e-01 -3.10667418e-02] | [13.13843822479248, 0.6967126727104187] |
517c3b6b-6597-4e96-92c3-e6902137815a | mime-mutual-information-minimisation | 2001.05636 | null | https://arxiv.org/abs/2001.05636v1 | https://arxiv.org/pdf/2001.05636v1.pdf | MIME: Mutual Information Minimisation Exploration | We show that reinforcement learning agents that learn by surprise (surprisal) get stuck at abrupt environmental transition boundaries because these transitions are difficult to learn. We propose a counter-intuitive solution that we call Mutual Information Minimising Exploration (MIME) where an agent learns a latent representation of the environment without trying to predict the future states. We show that our agent performs significantly better over sharp transition boundaries while matching the performance of surprisal driven agents elsewhere. In particular, we show state-of-the-art performance on difficult learning games such as Gravitar, Montezuma's Revenge and Doom. | ['Craig Atkinson', 'Lech Szymanski', 'Brendan McCane', 'Haitao Xu'] | 2020-01-16 | null | null | null | null | ['montezumas-revenge'] | ['playing-games'] | [-1.56778827e-01 5.94443202e-01 2.38305889e-02 -1.72527730e-01
-5.18457532e-01 -4.80755240e-01 9.45063591e-01 5.14088273e-02
-6.81629956e-01 1.18612266e+00 -4.76680025e-02 -1.95068657e-01
-4.10463363e-01 -7.18861222e-01 -8.94372582e-01 -8.35682392e-01
-8.98718297e-01 6.80827677e-01 9.42029059e-02 -6.72251582e-01
2.96837717e-01 -1.90805361e-01 -1.30113840e+00 -2.12152526e-01
7.91603208e-01 5.09945095e-01 2.32092544e-01 9.39807177e-01
3.73008370e-01 1.20956707e+00 -2.61268854e-01 3.17100523e-04
5.06951749e-01 -8.16197991e-01 -8.79738331e-01 -9.28230658e-02
-3.28947932e-01 -2.70917296e-01 -3.98828954e-01 1.31269205e+00
7.33577237e-02 4.30216461e-01 5.78433216e-01 -1.23051059e+00
-3.12163264e-01 1.13521731e+00 -5.28566062e-01 1.56111166e-01
2.89587528e-01 8.59125614e-01 1.15943682e+00 -9.03267190e-02
6.26051486e-01 1.30297256e+00 4.28220659e-01 8.84637713e-01
-1.18531144e+00 -4.24219608e-01 4.08019602e-01 1.32935360e-01
-6.83301747e-01 -1.69252872e-01 4.88665760e-01 -1.21558048e-01
1.30521190e+00 3.89836840e-02 9.36880648e-01 1.22222757e+00
7.03530371e-01 9.73334372e-01 1.52710927e+00 -1.92303732e-01
9.26835656e-01 -1.26395121e-01 -4.10194039e-01 8.96913111e-01
6.55195937e-02 9.82394993e-01 -8.04303348e-01 -1.05683178e-01
6.15740716e-01 -3.13779473e-01 -1.21158801e-01 -8.18478167e-01
-1.13559425e+00 8.75084341e-01 4.64880675e-01 -6.53817281e-02
-6.14959717e-01 6.86650872e-01 5.02000712e-02 1.05969238e+00
-1.40475675e-01 1.23039198e+00 -6.39868677e-01 -4.93378937e-01
-3.16024929e-01 5.34294426e-01 9.94443834e-01 5.13870895e-01
8.32256794e-01 2.48660520e-01 4.24857855e-01 8.63888487e-02
6.15405023e-01 2.94307232e-01 7.89103270e-01 -1.46594346e+00
1.65180534e-01 3.70279923e-02 4.96921003e-01 -3.44622016e-01
-6.51318908e-01 -5.52261949e-01 -3.62980723e-01 8.58953357e-01
2.88218558e-01 -8.56032670e-01 -7.53645122e-01 2.20283985e+00
8.95347595e-02 7.62529671e-02 9.21881199e-01 7.46125221e-01
-1.65350270e-02 7.14259088e-01 -1.33733526e-01 -4.72687364e-01
7.31220067e-01 -1.04758334e+00 -4.82447922e-01 -1.06926394e+00
5.86783051e-01 1.39121756e-01 8.90619397e-01 6.11619711e-01
-1.28872728e+00 5.43014221e-02 -1.28733873e+00 6.79637313e-01
-1.04072437e-01 -9.74489808e-01 1.08439565e+00 2.50278175e-01
-1.15348113e+00 1.10641801e+00 -1.24965179e+00 -4.67084646e-01
2.36412853e-01 4.53828871e-01 -2.80261844e-01 6.82505429e-01
-1.26879656e+00 1.25424540e+00 7.14226484e-01 -2.89131910e-01
-1.78230715e+00 -2.57887870e-01 -7.37834096e-01 -8.86395350e-02
7.30548441e-01 -5.78659892e-01 1.76919746e+00 -9.46192741e-01
-2.07020998e+00 6.01127684e-01 5.02017438e-01 -1.23558593e+00
6.61333084e-01 -3.35180551e-01 2.52859350e-02 -1.94352120e-01
-1.14950612e-01 8.66535425e-01 7.64196038e-01 -1.18113244e+00
-5.08367002e-01 -1.58536717e-01 4.46532965e-02 6.98058486e-01
9.09410566e-02 -3.55562121e-01 6.18344188e-01 1.81312393e-02
6.62671626e-02 -1.10085022e+00 -8.80813181e-01 -5.06950557e-01
-3.17304015e-01 -3.26532304e-01 4.13342535e-01 1.09426267e-01
4.67097402e-01 -1.85809791e+00 4.12349612e-01 5.36975870e-03
1.80587769e-01 -2.63686717e-01 -3.39601010e-01 6.98103964e-01
8.24131668e-02 -2.41497047e-02 -1.69474855e-01 -1.93450581e-02
4.11504567e-01 5.04702747e-01 -4.31785285e-01 3.97045463e-01
-7.13791326e-02 1.11376715e+00 -1.38631058e+00 1.02112377e-02
-1.25625938e-01 -3.92395794e-01 -7.96890378e-01 3.65374804e-01
-8.31310153e-01 4.26057130e-01 -5.88747025e-01 1.66035235e-01
1.28375828e-01 -3.51646066e-01 5.12504160e-01 9.27147329e-01
-2.06492215e-01 5.78361928e-01 -9.48703229e-01 1.75686085e+00
-3.13630700e-02 5.18891692e-01 -6.59401789e-02 -8.61849427e-01
7.35943139e-01 1.37247920e-01 2.14626372e-01 -6.50083423e-01
1.92894787e-01 1.74399689e-01 3.32399964e-01 -2.57502377e-01
3.46764147e-01 -4.95282590e-01 -3.61530751e-01 8.06254029e-01
1.63458943e-01 -6.65178716e-01 1.20845828e-02 3.34985614e-01
1.28568900e+00 3.19908261e-01 5.19255996e-01 -5.46579361e-01
-1.77760780e-01 8.21332261e-02 6.58712864e-01 1.38774753e+00
-4.18662280e-01 -9.09232944e-02 9.08408701e-01 -6.06177390e-01
-7.44305789e-01 -1.49899971e+00 4.86875027e-01 1.01096094e+00
2.85221934e-01 -7.00965643e-01 -7.84876585e-01 -6.65450037e-01
-1.71522066e-01 9.59422708e-01 -9.62106109e-01 -4.57894117e-01
-2.47466788e-01 -7.44088352e-01 1.50667667e-01 3.85132991e-02
7.35707760e-01 -1.38654852e+00 -1.35720897e+00 2.35988617e-01
2.86531359e-01 -2.75894165e-01 -5.42607717e-02 9.67853248e-01
-9.59687352e-01 -8.03949952e-01 -1.14899330e-01 -3.57686549e-01
2.52303898e-01 -2.42816687e-01 1.01593113e+00 -3.53437066e-01
-2.92343907e-02 5.10689914e-01 -9.91295464e-03 -5.00455320e-01
-5.15077412e-01 -2.55782660e-02 6.85894787e-01 -7.45209038e-01
2.55890816e-01 -8.90236795e-01 -5.49679756e-01 4.55313995e-02
-6.52445674e-01 1.33613065e-01 5.20780146e-01 1.04278839e+00
1.56983405e-01 2.71610498e-01 6.25601530e-01 -5.03749311e-01
1.01101804e+00 -6.88239157e-01 -1.00364196e+00 -3.41730975e-02
-9.11401451e-01 7.16169059e-01 5.11557400e-01 -4.80797946e-01
-1.20565295e+00 -2.09913291e-02 3.39149147e-01 1.48821235e-01
-5.80500066e-03 3.19847345e-01 4.80935365e-01 -7.51271993e-02
1.12358212e+00 4.16888118e-01 3.51821780e-01 -2.16501728e-01
4.12566096e-01 2.30409220e-01 4.28117871e-01 -7.45033145e-01
8.46466243e-01 1.43189117e-01 -6.09594062e-02 -3.59156162e-01
-7.62764692e-01 2.26864457e-01 -4.41379808e-02 -3.13497275e-01
6.73393309e-01 -8.19978774e-01 -1.27646422e+00 4.72524732e-01
-5.61625004e-01 -1.05347681e+00 -8.95238519e-01 4.85106051e-01
-1.38093805e+00 9.70468596e-02 -6.97965443e-01 -1.03077567e+00
1.58880837e-02 -9.05311763e-01 2.68667549e-01 7.75293112e-01
-1.46934167e-01 -8.55390489e-01 8.13408911e-01 -1.05321102e-01
6.43859386e-01 5.82769327e-02 6.55843318e-01 -7.15753138e-01
-8.47884655e-01 4.87461835e-01 4.64004278e-01 -5.79480976e-02
-3.19669256e-04 -3.20209831e-01 -5.56631565e-01 -5.64314067e-01
2.51049906e-01 -9.93260503e-01 1.05876172e+00 1.95412785e-01
3.44393849e-01 -6.64172113e-01 -1.35606632e-01 3.65998596e-01
1.17668116e+00 3.95149320e-01 4.11611199e-01 8.87581170e-01
-1.22076139e-01 6.19399369e-01 6.43446684e-01 9.42201376e-01
3.27462286e-01 2.09346399e-01 1.02439678e+00 7.10619748e-01
5.99580705e-01 -4.85946596e-01 8.17252457e-01 4.36892301e-01
9.89265442e-02 5.98024763e-02 -8.00458372e-01 5.12143075e-01
-2.30845213e+00 -9.92841363e-01 6.93028748e-01 1.96941626e+00
1.33291137e+00 6.00184083e-01 1.70481741e-01 -5.76200962e-01
7.39031360e-02 3.28127861e-01 -1.18673348e+00 -6.07101321e-01
-1.81317981e-02 -8.88644755e-02 3.98472995e-01 8.68949473e-01
-7.97503769e-01 1.40605497e+00 7.73435211e+00 3.15313518e-01
-6.49566054e-01 -1.88344508e-01 5.36404908e-01 -4.28659409e-01
-3.03927213e-01 2.74462581e-01 -3.41561586e-01 1.49740145e-01
1.10656917e+00 -4.07631099e-01 1.07783437e+00 1.06316197e+00
-7.29833841e-02 -5.55210352e-01 -1.17037058e+00 6.35791481e-01
-4.97209728e-01 -1.24945211e+00 -5.54273307e-01 1.55200168e-01
1.09297657e+00 5.20825982e-01 4.01932329e-01 7.18412101e-01
1.52339947e+00 -1.06518996e+00 6.25780463e-01 4.86511707e-01
-7.91766644e-02 -8.97544086e-01 3.38745773e-01 7.07341492e-01
-3.93673539e-01 -3.21877003e-01 -3.94982338e-01 -7.93297410e-01
-1.34968653e-01 8.59183259e-03 -1.05422425e+00 -1.65119067e-01
4.07948941e-01 4.02819574e-01 -2.34934852e-01 8.00894380e-01
-4.94433939e-01 3.85044307e-01 -5.19738615e-01 -5.94036281e-01
9.67461824e-01 -2.71113306e-01 7.18506396e-01 6.39779031e-01
1.51759610e-01 1.19230440e-02 1.03580192e-01 1.10113096e+00
-1.42602110e-02 -4.77989018e-01 -8.44596565e-01 -1.46071017e-01
9.37337652e-02 7.28532255e-01 -4.65011001e-01 -1.65040061e-01
2.90378630e-01 9.08750594e-01 6.38705492e-01 3.16969246e-01
-5.59609592e-01 -9.97445285e-02 1.13024867e+00 -6.46609247e-01
2.21468166e-01 -2.27463961e-01 1.42293215e-01 -1.25776756e+00
-4.05962914e-01 -9.60823953e-01 3.36564213e-01 -6.21613562e-01
-1.02824759e+00 3.99809152e-01 -1.55906320e-01 -7.94346333e-01
-9.88031268e-01 -3.57158214e-01 -7.43204236e-01 1.87972784e-01
-1.32941985e+00 -1.63241982e-01 2.71759897e-01 3.60844761e-01
5.87515175e-01 -4.07340556e-01 8.04315805e-01 -1.00343108e+00
-3.85756969e-01 1.48581803e-01 4.57756162e-01 -3.69804859e-01
3.07598114e-01 -1.80132043e+00 6.91880822e-01 6.67124808e-01
2.45987386e-01 4.75329101e-01 1.33817828e+00 -6.50336087e-01
-1.59055829e+00 -5.64665258e-01 1.19429089e-01 -2.81006724e-01
1.02522373e+00 -1.85483083e-01 -5.45552254e-01 1.13761199e+00
7.21453011e-01 -1.83472753e-01 2.45744064e-01 1.66872829e-01
-3.96090806e-01 2.15545595e-01 -1.10890138e+00 1.07527900e+00
8.65333259e-01 -3.24478835e-01 -1.03062105e+00 5.02560973e-01
6.59995675e-01 -3.81573796e-01 -4.10747766e-01 5.93246296e-02
3.92858267e-01 -1.19283068e+00 5.43925643e-01 -1.17532158e+00
3.38014156e-01 -1.25109509e-01 4.78473864e-02 -1.92514229e+00
-2.64815062e-01 -1.68041384e+00 -3.22726995e-01 2.72250146e-01
4.45155770e-01 -8.47126245e-01 1.14185226e+00 4.86855537e-01
3.08859378e-01 -6.74938679e-01 -1.07222712e+00 -1.12041414e+00
3.12919676e-01 -1.78491920e-01 3.15091580e-01 8.98338199e-01
1.06749249e+00 2.75013298e-01 -6.64545774e-01 -8.44478458e-02
1.01511395e+00 1.20823488e-01 4.85559821e-01 -6.99989736e-01
-6.98255956e-01 -4.91601616e-01 -1.66905373e-01 -1.13429642e+00
1.11637779e-01 -5.84430158e-01 5.95275998e-01 -1.14130592e+00
3.02869141e-01 -2.05173984e-01 -6.04377508e-01 4.89507705e-01
1.48016572e-01 -2.97660887e-01 3.83465409e-01 7.70524442e-02
-1.40227175e+00 1.00274110e+00 1.25512230e+00 -1.04339495e-02
-3.83076310e-01 -4.62136082e-02 -9.20391142e-01 9.31801438e-01
1.20742655e+00 -6.54648185e-01 -4.27492380e-01 -7.36023709e-02
8.10071647e-01 3.94543886e-01 5.68111353e-02 -1.13609326e+00
5.33975899e-01 -6.95863187e-01 -2.52323616e-02 -4.29598801e-02
3.60947818e-01 -2.37531468e-01 -7.46605247e-02 1.14747024e+00
-9.58719373e-01 -1.58014446e-02 -1.24657582e-02 8.27808499e-01
2.45939299e-01 -3.34969819e-01 8.74973238e-01 -6.31989062e-01
-7.29478538e-01 5.09579517e-02 -1.11667502e+00 3.15438598e-01
1.12584054e+00 3.26299369e-01 -6.22304678e-01 -8.46440554e-01
-7.64892340e-01 8.01470935e-01 5.48891962e-01 2.59061217e-01
6.88907325e-01 -9.81102288e-01 -6.25216305e-01 -6.95880502e-02
-8.10851902e-02 -1.93193883e-01 2.44229082e-02 4.46047366e-01
-3.66122425e-01 2.42361814e-01 -6.56043470e-01 -2.09662139e-01
-6.04727685e-01 4.45187151e-01 8.28933716e-01 -5.36067605e-01
-5.79673111e-01 1.07001412e+00 2.91626096e-01 -6.76598728e-01
2.37164974e-01 -1.88711822e-01 1.19790711e-01 -2.77577400e-01
4.84934986e-01 1.35811791e-01 -5.21237075e-01 2.60944337e-01
-1.93958208e-01 -1.73062216e-02 -5.06106496e-01 -6.02001846e-01
1.39684141e+00 -2.31749304e-02 3.57120484e-02 5.16568899e-01
6.61245763e-01 -5.99022329e-01 -2.02923632e+00 -3.08962941e-01
2.72376508e-01 -2.04746246e-01 5.94007187e-02 -9.75950301e-01
-4.81376588e-01 4.66848910e-01 5.44906497e-01 5.37449718e-01
6.35538340e-01 1.07340716e-01 4.86598253e-01 1.29782367e+00
7.21274793e-01 -1.42683327e+00 5.01167476e-01 8.85800898e-01
6.90921843e-01 -1.34004807e+00 -2.43890569e-01 5.70198357e-01
-8.14304948e-01 9.54936624e-01 7.22049296e-01 -4.54018325e-01
3.31366211e-01 3.50610405e-01 -1.44082487e-01 -3.86776000e-01
-1.69888318e+00 -1.83429375e-01 -5.05392551e-01 7.17916608e-01
-3.39283645e-01 2.02507362e-01 1.60503045e-01 2.23088995e-01
-5.37850320e-01 -3.28288376e-01 9.47681248e-01 1.16821706e+00
-1.17422163e+00 -8.36167812e-01 -1.56985205e-02 3.07447791e-01
-2.96959847e-01 -3.79030965e-02 -4.80725467e-01 6.13678336e-01
-5.50243556e-01 8.00375462e-01 -7.93799311e-02 -3.66545558e-01
-2.43432283e-01 1.20134349e-03 7.26234138e-01 -4.01964068e-01
-5.05679786e-01 -3.22401077e-01 -5.31743504e-02 -9.79130328e-01
-3.51481438e-02 -9.74867165e-01 -1.44078219e+00 -2.02940255e-01
3.20755914e-02 6.16293252e-01 5.42203903e-01 1.04716873e+00
1.50856271e-01 4.10672724e-01 7.24013746e-01 -7.33771622e-01
-1.23442435e+00 -7.67468035e-01 -6.50059342e-01 2.87857912e-02
7.48110116e-01 -5.25750160e-01 -6.88269019e-01 -5.55441976e-01] | [3.9274983406066895, 1.6894569396972656] |
d328dab2-4802-4c56-a517-5375ba672e55 | deep-kalman-filters | 1511.05121 | null | http://arxiv.org/abs/1511.05121v2 | http://arxiv.org/pdf/1511.05121v2.pdf | Deep Kalman Filters | Kalman Filters are one of the most influential models of time-varying
phenomena. They admit an intuitive probabilistic interpretation, have a simple
functional form, and enjoy widespread adoption in a variety of disciplines.
Motivated by recent variational methods for learning deep generative models, we
introduce a unified algorithm to efficiently learn a broad spectrum of Kalman
filters. Of particular interest is the use of temporal generative models for
counterfactual inference. We investigate the efficacy of such models for
counterfactual inference, and to that end we introduce the "Healing MNIST"
dataset where long-term structure, noise and actions are applied to sequences
of digits. We show the efficacy of our method for modeling this dataset. We
further show how our model can be used for counterfactual inference for
patients, based on electronic health record data of 8,000 patients over 4.5
years. | ['Rahul G. Krishnan', 'Uri Shalit', 'David Sontag'] | 2015-11-16 | null | null | null | null | ['counterfactual-inference'] | ['miscellaneous'] | [ 4.89140674e-02 1.40277073e-01 -2.35037014e-01 -2.51366645e-01
-5.31438828e-01 -2.59416342e-01 8.83190215e-01 -2.70478010e-01
-4.17819738e-01 1.19162309e+00 8.31701458e-01 -4.37142491e-01
-5.48997700e-01 -8.29045534e-01 -8.46444368e-01 -7.62318850e-01
-3.58033657e-01 3.79751056e-01 -3.48118275e-01 4.36408781e-02
-5.58835268e-02 1.91989496e-01 -1.11925709e+00 1.35646477e-01
5.94571114e-01 5.30141115e-01 1.11313514e-01 5.34540355e-01
4.22071397e-01 7.05716312e-01 -3.27976823e-01 -6.46654427e-01
-2.94889454e-02 -4.00240183e-01 -5.73116243e-01 4.73363977e-03
-2.48835176e-01 -6.49023950e-01 -8.74226093e-01 8.16635668e-01
6.49014771e-01 1.04031302e-01 8.62833083e-01 -1.16031301e+00
-6.86449766e-01 8.32340539e-01 -7.47050121e-02 2.05586955e-01
1.03413954e-01 3.22085828e-01 8.57147753e-01 -1.90856680e-01
5.46927750e-01 1.37686098e+00 7.34695911e-01 6.70355558e-01
-1.50046694e+00 -4.70579743e-01 1.97452813e-01 1.50954291e-01
-1.11110163e+00 -3.90911609e-01 6.75873995e-01 -4.98577356e-01
7.07353592e-01 1.06581293e-01 9.94884193e-01 1.71463227e+00
1.13089919e+00 1.08498001e+00 1.22416437e+00 -8.06530938e-02
6.16324961e-01 -5.03865421e-01 -3.86246555e-02 4.02850330e-01
2.81697035e-01 8.44993711e-01 -4.23005491e-01 -6.35928452e-01
1.04965842e+00 5.67605078e-01 -3.33495826e-01 -3.76791388e-01
-1.10491061e+00 1.09872496e+00 1.18974514e-01 3.25365760e-03
-7.08111823e-01 7.84145415e-01 9.34097245e-02 -3.37305591e-02
5.98432124e-01 1.28110826e-01 -5.58544099e-01 -1.46684960e-01
-9.04445171e-01 6.77658319e-01 8.31686676e-01 4.48296219e-01
-4.39052768e-02 -6.65287301e-02 -6.06679201e-01 4.82497394e-01
6.70371473e-01 7.96826541e-01 3.72016966e-01 -1.05272734e+00
-1.67128682e-01 -9.52410772e-02 3.37428391e-01 -3.76173884e-01
-4.83880162e-01 -5.97101331e-01 -9.95080054e-01 -4.70921248e-02
4.29700881e-01 -3.64143997e-01 -1.13233244e+00 2.07436228e+00
4.05237600e-02 8.91383886e-01 6.22937977e-02 4.03151929e-01
1.33005366e-01 2.50661522e-01 1.97670430e-01 -7.61790633e-01
1.24900913e+00 -9.69494656e-02 -1.06529570e+00 8.49816948e-02
4.40683961e-01 -4.24192637e-01 4.88250881e-01 4.41718519e-01
-1.01933289e+00 1.23519875e-01 -4.49360222e-01 3.74099642e-01
-1.20533435e-02 -3.42772394e-01 9.95499134e-01 7.45912075e-01
-9.27117109e-01 9.14615571e-01 -1.32124817e+00 -2.10154906e-01
7.45233655e-01 1.09830819e-01 3.06943983e-01 -1.24470405e-01
-1.35252285e+00 7.45279133e-01 1.55381322e-01 6.08611889e-02
-1.38562393e+00 -1.12206113e+00 -5.57967067e-01 5.33921197e-02
3.17253530e-01 -1.52642894e+00 1.60524654e+00 -4.21767682e-01
-1.55434930e+00 4.87750679e-01 -1.92600816e-01 -1.02589393e+00
8.66284549e-01 -4.05493438e-01 -4.73958105e-01 -2.03583747e-01
-1.70203869e-03 1.49819002e-01 8.10956359e-01 -6.81814671e-01
-3.64543587e-01 -4.56383884e-01 -4.24409211e-02 -2.85330098e-02
2.82287091e-01 -3.69130403e-01 -1.29365651e-02 -1.01156127e+00
-2.31116310e-01 -1.12920403e+00 -6.32457852e-01 -2.70793766e-01
-6.32323384e-01 -1.54864773e-01 2.47653201e-01 -5.72452307e-01
1.44006228e+00 -1.89329553e+00 -5.71004488e-02 -7.59100541e-02
1.19639926e-01 -4.02469814e-01 1.16592988e-01 4.38873291e-01
-1.02317901e-02 9.94856805e-02 -5.17362177e-01 -2.59098202e-01
6.13901019e-02 4.50147063e-01 -6.19101644e-01 5.66079557e-01
-2.51850456e-01 1.26631653e+00 -1.02344275e+00 -1.20271392e-01
3.73839021e-01 4.85372961e-01 -7.69474506e-01 -6.08129725e-02
-3.70006800e-01 4.47293550e-01 -5.47155797e-01 1.95062995e-01
4.52311754e-01 -3.91244471e-01 4.72172141e-01 1.37465224e-01
1.42234325e-01 2.98580229e-01 -9.92259324e-01 1.69000459e+00
-2.46700779e-01 1.39906272e-01 -4.07991827e-01 -9.11832571e-01
6.32393807e-02 6.55068099e-01 7.35538900e-01 -2.42471442e-01
3.11192900e-01 -4.75918092e-02 2.57363021e-01 -4.20866251e-01
-6.20562285e-02 -8.59406412e-01 -1.79254189e-01 6.10386193e-01
-6.03878349e-02 -5.85940704e-02 -2.50725243e-02 1.87010691e-01
1.17420959e+00 6.40242994e-02 6.13429070e-01 -3.05511147e-01
-2.92105496e-01 -2.69430876e-01 7.09415138e-01 1.25518143e+00
-1.14077702e-01 4.25840050e-01 4.68934625e-01 -4.28011179e-01
-8.07398975e-01 -1.56753039e+00 -4.11046237e-01 4.08712357e-01
-3.21525633e-01 -8.77948403e-02 -4.35370088e-01 -3.98274183e-01
3.25939029e-01 1.08662510e+00 -9.67358232e-01 -3.56033057e-01
-2.05376700e-01 -1.46676314e+00 3.73178989e-01 6.34545982e-01
1.81523204e-01 -9.44689929e-01 -6.30310297e-01 5.49566448e-01
-2.37356111e-01 -5.29004693e-01 -1.61455870e-01 -1.59636199e-01
-1.13593447e+00 -1.12698019e+00 -8.94006968e-01 2.63642937e-01
3.77480686e-02 -2.74324208e-01 1.27707505e+00 -4.23076868e-01
-2.00266421e-01 7.13366687e-01 2.75651127e-01 -7.67399490e-01
-5.02549946e-01 -4.27345812e-01 5.06320715e-01 -1.15234241e-01
6.41872287e-02 -7.23811805e-01 -9.64045882e-01 -1.65074915e-01
-9.39903796e-01 5.76207228e-02 4.86065537e-01 1.01227272e+00
5.26739120e-01 -1.07984692e-01 5.93250036e-01 -1.08919108e+00
7.89736211e-01 -7.38828301e-01 -6.48703814e-01 2.78234690e-01
-7.11556375e-01 3.28697562e-01 1.28472149e-01 -4.57717359e-01
-1.40399551e+00 -1.74801663e-01 -2.04175338e-01 -4.41336960e-01
-2.72885039e-02 7.14064002e-01 -4.50283661e-03 8.27008128e-01
4.02859449e-01 1.99938253e-01 2.80986261e-03 -5.54121494e-01
7.23683536e-01 2.27657184e-01 5.64198673e-01 -3.99637341e-01
3.19338858e-01 1.04112363e+00 1.03109047e-01 -5.03879964e-01
-8.75854969e-01 1.01337500e-01 -1.52496740e-01 1.70509275e-02
8.14483464e-01 -1.04343176e+00 -1.23271382e+00 6.17178082e-01
-9.82203662e-01 -5.32573819e-01 -4.29600388e-01 8.45476747e-01
-1.21224236e+00 5.05818315e-02 -7.86271751e-01 -1.21353900e+00
-4.59879413e-02 -7.81052291e-01 8.94743085e-01 8.20670202e-02
-3.00631106e-01 -1.50156534e+00 5.41905940e-01 -4.00812216e-02
2.39963129e-01 2.85659254e-01 8.38414073e-01 -2.37167925e-01
-7.33890653e-01 8.58427510e-02 2.27669463e-01 -2.43221894e-02
3.12719345e-01 -8.21380317e-02 -8.66138160e-01 -3.05525571e-01
1.56875715e-01 2.00321645e-01 1.08537972e+00 1.49572015e+00
1.19173789e+00 -3.94120604e-01 -8.47984016e-01 4.62404281e-01
1.37996697e+00 3.99066687e-01 7.25114286e-01 -7.76801705e-02
2.79245883e-01 1.91532671e-01 2.83184588e-01 7.63596952e-01
3.45270365e-01 4.63523686e-01 4.92301255e-01 4.26341861e-01
2.76228458e-01 -5.06955385e-01 2.39129320e-01 3.88316393e-01
-1.03110574e-01 -3.78869772e-01 -7.42140710e-01 8.46544147e-01
-2.10858488e+00 -1.36342180e+00 3.33347470e-02 2.32705355e+00
7.08356977e-01 1.65075734e-01 2.15200216e-01 -2.98811793e-01
4.47574675e-01 1.40531674e-01 -8.38584065e-01 7.84668103e-02
3.65134627e-02 2.74872988e-01 6.12608850e-01 2.62781382e-01
-1.05571306e+00 5.51958799e-01 8.08151531e+00 6.85915053e-01
-7.22381473e-01 2.94556171e-01 7.25258589e-01 -4.38355923e-01
-5.35855114e-01 2.13442110e-02 -5.29398620e-01 8.58496428e-01
1.33823419e+00 -4.84830171e-01 2.02389076e-01 4.04236913e-01
7.74420142e-01 -1.34257123e-01 -1.30021751e+00 9.38816845e-01
-4.15343076e-01 -1.85343039e+00 -1.63085103e-01 2.66911626e-01
9.46497381e-01 2.27440238e-01 4.00982052e-01 1.75603345e-01
1.14546466e+00 -1.12532413e+00 6.69560254e-01 1.07466221e+00
5.09255469e-01 -6.43723071e-01 6.40876114e-01 4.57027942e-01
-3.67275834e-01 -2.35985428e-01 -1.57427773e-01 -2.94888318e-01
7.44990110e-01 9.14749026e-01 -6.77269280e-01 5.51844120e-01
5.07614613e-01 7.64181793e-01 2.50613868e-01 1.22393656e+00
-3.31078708e-01 9.18647468e-01 -4.83817697e-01 1.03592359e-01
-5.88226132e-02 -1.00760087e-01 5.59759080e-01 8.01708102e-01
4.16640967e-01 1.82782337e-01 -3.09972882e-01 9.58375752e-01
-6.37766114e-03 -5.68545580e-01 -7.95149326e-01 -1.13372959e-01
4.19241458e-01 5.12087107e-01 -5.25326371e-01 -3.62837285e-01
-2.93038875e-01 7.90386975e-01 -1.01144969e-01 6.00966632e-01
-9.29844677e-01 3.30484331e-01 1.10805309e+00 1.76873468e-02
2.68612266e-01 6.71009943e-02 -1.57313585e-01 -1.32688797e+00
-3.80501777e-01 -6.53015614e-01 7.10493445e-01 -7.03610837e-01
-1.61346185e+00 -1.01268902e-01 1.76154390e-01 -9.91750360e-01
-7.92257309e-01 -5.97389102e-01 -5.64101279e-01 6.91561699e-01
-1.11973548e+00 -7.83917785e-01 4.78925020e-01 5.70518017e-01
3.81296843e-01 3.54857743e-01 7.45122969e-01 -1.83363836e-02
-4.95565832e-01 7.14385211e-02 5.41398942e-01 -1.03329634e-02
2.89983153e-01 -1.26194298e+00 5.29371381e-01 7.93843150e-01
3.30693245e-01 9.35817420e-01 1.17702258e+00 -9.98346150e-01
-1.15526462e+00 -1.02249753e+00 5.77192903e-01 -8.72187376e-01
7.40962505e-01 -1.09011233e-01 -5.45415282e-01 1.12484550e+00
5.67131676e-02 -3.25772464e-01 7.97580600e-01 2.28713766e-01
6.55781999e-02 3.09269935e-01 -1.11081100e+00 8.66329253e-01
1.17253780e+00 -4.64451700e-01 -8.91409457e-01 3.76447827e-01
6.62922084e-01 -3.21833521e-01 -7.77989388e-01 2.95823097e-01
8.27530622e-01 -9.37548280e-01 1.08497298e+00 -1.08426857e+00
5.44571698e-01 9.82719287e-02 -8.58277008e-02 -1.60150182e+00
-4.31503117e-01 -9.03164923e-01 -4.73760456e-01 6.19918346e-01
1.87376186e-01 -8.37409675e-01 6.24652147e-01 6.01307034e-01
2.33461335e-01 -8.32210839e-01 -1.10717404e+00 -7.72601604e-01
3.55711997e-01 -8.72668266e-01 6.31847918e-01 7.50440300e-01
-1.05680756e-01 3.08162235e-02 -7.63589203e-01 7.44598582e-02
9.07549858e-01 1.92765087e-01 3.01081479e-01 -1.30347526e+00
-6.03801429e-01 -2.56279916e-01 -8.14765468e-02 -1.06889462e+00
4.83757816e-02 -7.03472078e-01 -1.48057148e-01 -1.62132263e+00
6.27554834e-01 -4.74881232e-02 -4.74745631e-01 1.20779462e-01
-2.64779806e-01 -1.13235645e-01 -2.00966448e-02 -7.67387152e-02
-2.32190251e-01 6.32644534e-01 9.78213787e-01 3.90239060e-02
-2.98967976e-02 4.83639270e-01 -7.89268196e-01 8.19241405e-01
5.50292134e-01 -5.31502783e-01 -6.04084074e-01 -2.25406624e-02
2.93871880e-01 4.13778156e-01 7.40368426e-01 -3.49585354e-01
-1.55714467e-01 -4.26736146e-01 4.24272597e-01 -6.09891891e-01
2.62201399e-01 -5.43338597e-01 7.77309120e-01 9.62751508e-01
-2.26888835e-01 -2.20992684e-01 1.70196548e-01 1.18708181e+00
1.90973327e-01 1.73645183e-01 4.35172260e-01 -3.38222176e-01
-3.72367650e-01 5.66756129e-01 -6.51654303e-01 -3.65891727e-03
7.55659401e-01 2.63237804e-01 -1.82952970e-01 -6.73586488e-01
-1.15646148e+00 1.70859456e-01 1.85209647e-01 3.70537579e-01
5.76506734e-01 -1.36290121e+00 -7.78527379e-01 -2.49201253e-01
-1.61171407e-01 -4.73787010e-01 7.92813838e-01 9.42173064e-01
1.47023620e-02 4.66015995e-01 6.78284168e-02 -5.07279813e-01
-7.05444992e-01 8.15259695e-01 4.16967154e-01 -6.85747445e-01
-6.69682205e-01 5.02159536e-01 5.84084630e-01 -3.08386069e-02
-3.36940706e-01 -4.29610312e-01 2.15791821e-01 -1.57236665e-01
5.88589489e-01 4.24775511e-01 -4.17498320e-01 -5.17979451e-02
-4.10819590e-01 -6.88061863e-02 -5.38738593e-02 -4.53520566e-01
1.52676880e+00 -3.56013387e-01 1.95229203e-01 7.73627043e-01
7.90660739e-01 -3.47726315e-01 -1.56354904e+00 -1.66443154e-01
-6.30817143e-03 -1.42767176e-01 -1.09638311e-02 -8.63215208e-01
-7.29380667e-01 6.40154481e-01 7.52829254e-01 2.25360602e-01
9.72603679e-01 2.00634450e-01 5.33853352e-01 -5.22550754e-02
5.12402236e-01 -7.72787035e-01 -2.80686140e-01 1.94235817e-01
6.90795660e-01 -9.62676346e-01 4.24218476e-02 -6.10047840e-02
-2.01437607e-01 4.84616756e-01 -3.70348483e-01 -2.94360459e-01
1.00958669e+00 4.30855423e-01 -1.25361130e-01 -1.84357122e-01
-1.16657352e+00 -1.53649122e-01 1.15486078e-01 5.63284099e-01
4.84216839e-01 5.79926789e-01 -4.63283032e-01 7.62350261e-01
-2.94013828e-01 7.21378267e-01 5.97900212e-01 5.97542465e-01
1.51010334e-01 -8.98912430e-01 -2.01665431e-01 8.21816146e-01
-8.86583209e-01 -2.47950166e-01 3.34716082e-01 7.25955129e-01
-3.10028851e-01 7.62318850e-01 1.49026200e-01 1.69943422e-01
1.16901532e-01 2.72204936e-01 7.65104830e-01 -3.27078819e-01
5.65778948e-02 1.89266965e-01 -1.65640548e-01 -7.04878390e-01
-4.86088842e-01 -1.14013314e+00 -7.95553327e-01 -5.39069295e-01
7.48256221e-03 -1.07121445e-01 3.15446913e-01 1.16845834e+00
4.48607057e-01 8.10266137e-01 2.49212667e-01 -3.53913575e-01
-9.62232530e-01 -1.01850188e+00 -7.60793328e-01 3.58387977e-01
4.37363267e-01 -8.33383024e-01 -3.43258649e-01 1.50623530e-01] | [8.04527759552002, 5.439292907714844] |
fd636e57-cb86-44de-bbf5-73ef58a73e2e | study-on-the-tea-market-in-india | 2304.07851 | null | https://arxiv.org/abs/2304.07851v1 | https://arxiv.org/pdf/2304.07851v1.pdf | Study on the tea market in India | India's tea business has a long history and plays a significant role in the economics of the nation. India is the world's second-largest producer of tea, with Assam and Darjeeling being the most well-known tea-growing regions. Since the British introduced tea cultivation to India in the 1820s, the nation has produced tea. Millions of people are employed in the tea sector today, and it contributes significantly to the Indian economy in terms of revenue. The production of tea has changed significantly in India over the years, moving more and more towards organic and sustainable practices. The industry has also had to deal with difficulties like competition from other nations that produce tea, varying tea prices, and labor-related problems. Despite these obstacles, the Indian tea business is still growing and produces a wide variety of teas, such as black tea, green tea, and chai tea. Additionally, the sector encourages travel through "tea tourism," which allows tourists to see how tea is made and discover its origins in India. Overall, India's tea business continues to play a significant role in its history, culture, and economy. | ['Sreayans Jain', 'Harshita Sachdev', 'Devansh Khandelwal', 'Adarsh Damani', 'Adit Vinod Nair'] | 2023-04-16 | null | null | null | null | ['culture'] | ['speech'] | [-4.06936586e-01 -3.14013034e-01 -6.01200938e-01 3.16221505e-01
-5.22079766e-01 -9.00440753e-01 4.78688538e-01 4.39581633e-01
-2.68020686e-02 8.12058985e-01 1.34664133e-01 -4.49179202e-01
1.65515229e-01 -1.07600427e+00 -1.41894743e-01 -1.01611745e+00
1.65681019e-01 1.39103800e-01 -1.15016922e-01 -6.04545295e-01
5.51464140e-01 5.93950093e-01 -9.11460042e-01 -5.30122280e-01
1.19177032e+00 5.59541285e-01 7.67611682e-01 8.94618109e-02
-3.73306602e-01 6.27186239e-01 -3.21263105e-01 -2.80820519e-01
9.31662694e-02 -8.49723458e-01 -7.60843813e-01 -1.87352709e-02
-4.00937021e-01 -2.60265768e-01 3.54754888e-02 8.07478786e-01
-2.73490757e-01 -6.66613579e-02 2.15157107e-01 -9.22120035e-01
-1.11986053e+00 5.27036428e-01 -8.08649540e-01 -4.35761482e-01
-1.91923916e-01 9.64010507e-02 8.98584843e-01 -6.90025866e-01
7.05202162e-01 9.08966601e-01 1.99311584e-01 1.32439192e-03
-1.17540431e+00 -9.16610837e-01 -1.32099167e-01 1.18780518e-02
-1.10262930e+00 -1.93433642e-01 4.62356269e-01 -1.10853747e-01
5.01284540e-01 3.31502229e-01 1.07324040e+00 1.97831020e-01
5.52539170e-01 4.29409862e-01 1.58178449e+00 -5.52883267e-01
1.22611254e-01 3.36569369e-01 -3.52841496e-01 3.12213063e-01
4.97680455e-01 -3.58135819e-01 -1.66423656e-02 2.00464144e-01
8.62443745e-01 1.46571547e-01 2.07251012e-02 -1.13503240e-01
-1.24839711e+00 1.09151542e+00 9.67649639e-01 7.63315558e-01
-5.71298361e-01 6.97012618e-02 3.30344468e-01 1.92658111e-01
2.33722493e-01 4.35152054e-01 -2.92123109e-01 -4.78306860e-01
-4.05061513e-01 5.98373003e-02 7.89237499e-01 5.22022128e-01
5.13841093e-01 1.65153503e-01 6.94143772e-01 1.07988894e+00
2.56614000e-01 8.61553073e-01 -8.94332826e-02 -1.22717071e+00
8.39890242e-02 7.62696624e-01 2.82650501e-01 -8.02056313e-01
1.82583734e-01 -2.07631543e-01 -5.40087879e-01 2.60811418e-01
8.71929646e-01 -1.69700049e-02 -8.86912704e-01 1.38189721e+00
1.08002856e-01 -7.63430357e-01 7.12425485e-02 6.24894977e-01
1.96232021e-01 9.26198542e-01 4.77785110e-01 -2.52258509e-01
1.30321264e+00 -8.15968990e-01 -8.30021322e-01 6.87915683e-02
3.53451222e-01 -9.79505658e-01 7.95668185e-01 3.51878226e-01
-1.16465485e+00 7.87379919e-04 -7.11606681e-01 1.77173316e-01
-6.19174242e-01 -4.21084881e-01 9.33027804e-01 8.67221415e-01
-5.81908882e-01 4.69224781e-01 -7.74849951e-01 -8.94834578e-01
6.08855367e-01 1.14017226e-01 -3.04165512e-01 -3.83516252e-01
-6.75257564e-01 1.19227242e+00 4.34800327e-01 5.23374975e-01
-2.85170138e-01 -4.69774306e-01 -5.47717214e-01 1.62319783e-02
4.45959330e-01 -1.24172568e-01 9.11717892e-01 -8.29057932e-01
-1.20374322e+00 6.71299100e-01 -3.25757235e-01 7.00977668e-02
1.13654859e-01 -6.03166781e-02 -8.16376150e-01 -1.48248943e-02
6.08158231e-01 5.11319399e-01 -2.16030404e-01 -1.26107752e+00
-1.01904905e+00 -4.79463100e-01 2.19390821e-02 -1.69153750e-01
9.52017605e-02 4.57326591e-01 2.33781226e-02 -5.07450700e-01
4.04834718e-01 -1.24215734e+00 -3.22711676e-01 -2.04484195e-01
-1.33878320e-01 -1.16880760e-01 1.14531040e+00 -8.45912397e-01
6.45300031e-01 -2.14391065e+00 -3.46487641e-01 2.49478713e-01
7.38417357e-02 -2.43997350e-01 2.88899094e-01 9.62202847e-01
3.87567818e-01 3.98488462e-01 -9.84842032e-02 6.82641208e-01
-2.05883503e-01 6.50353849e-01 5.32279573e-02 -2.05738447e-03
3.65069598e-01 6.70740068e-01 -1.49641061e+00 -2.22285554e-01
1.92211881e-01 4.65947896e-01 1.80014178e-01 -5.36775887e-01
2.51055270e-01 7.29022384e-01 -6.29670560e-01 1.43072414e+00
9.31572974e-01 -3.56968284e-01 7.72272825e-01 2.92580724e-01
-9.81368303e-01 1.43350229e-01 -6.81354702e-01 1.30179334e+00
-2.91973561e-01 9.25784945e-01 2.54791051e-01 -5.31081676e-01
8.66489947e-01 2.69245863e-01 5.21214247e-01 -1.13044453e+00
-1.47306537e-02 8.37275565e-01 2.43873045e-01 -3.36380601e-01
1.07465565e+00 -2.09653392e-01 1.45827457e-01 2.00066403e-01
-6.97616756e-01 -1.49705950e-02 6.34009778e-01 2.33694628e-01
2.98276007e-01 6.82067454e-01 2.61765748e-01 -5.31316221e-01
1.91039681e-01 7.76540279e-01 6.90251172e-01 -2.56895036e-01
-1.24353923e-01 -3.07987928e-01 3.10920656e-01 -8.08921754e-02
-1.08435488e+00 -1.31281424e+00 -5.24193048e-01 1.05473197e+00
2.72335678e-01 2.31093973e-01 -1.76415130e-01 4.88149971e-02
-3.91958505e-02 6.97309971e-01 -3.63488227e-01 5.27048647e-01
-6.73499584e-01 -5.45838296e-01 1.67303309e-02 3.88675332e-01
1.11465573e+00 -1.21624732e+00 -5.03580391e-01 5.53297997e-01
-3.85700345e-01 -4.88777608e-01 -2.37249002e-01 5.18811822e-01
-1.00170445e+00 -9.66574490e-01 -1.04114139e+00 -9.48176980e-01
4.81443346e-01 5.79467893e-01 6.98713303e-01 -3.34997296e-01
-1.82934970e-01 -2.70895362e-01 -3.37928027e-01 -8.12254429e-01
-7.72149980e-01 -1.64671257e-01 -3.78844678e-01 -6.09906971e-01
3.34475219e-01 -4.81301129e-01 -6.30471289e-01 1.63904764e-02
-4.17023689e-01 -3.33907120e-02 4.60386544e-01 5.61026454e-01
4.90959913e-01 3.05413425e-01 1.13417065e+00 -1.03701818e+00
1.00042909e-01 -7.11358607e-01 -3.96647781e-01 2.79696465e-01
-5.80119431e-01 -5.55574298e-01 2.29738876e-01 -1.41449243e-01
-1.04830873e+00 -2.56909251e-01 4.36611861e-01 5.77382863e-01
1.76338464e-01 6.00157380e-01 -1.64466068e-01 1.06562432e-02
1.55122653e-01 -8.02247673e-02 -2.41007265e-02 -4.54843819e-01
3.00049424e-01 5.72633147e-01 6.31448627e-01 -2.44220704e-01
7.80822635e-01 2.46794254e-01 1.10442922e-01 -1.16912925e+00
-3.28264356e-01 -3.09572339e-01 -5.09502590e-01 -2.72972018e-01
1.06024003e+00 -7.77156949e-01 -1.03461289e+00 3.98096263e-01
-4.95924205e-01 -2.51563907e-01 -3.98236141e-02 5.87323546e-01
1.66786864e-01 -1.50516436e-01 -5.61704755e-01 -1.14708662e+00
2.99270879e-02 -7.63373375e-01 4.18504328e-02 8.79129648e-01
-4.01107222e-03 -1.02593315e+00 -1.34415954e-01 3.45506996e-01
7.31291294e-01 1.00989056e+00 1.20648742e+00 2.03202754e-01
-1.04053938e+00 2.73283362e-01 -6.46868289e-01 1.57669559e-01
9.66286182e-01 -2.74132658e-03 -6.80934250e-01 -1.11033380e-01
-3.71702999e-01 -1.10432841e-01 5.10339320e-01 5.03345788e-01
1.92607507e-01 2.69310568e-02 -4.67200428e-01 -4.14898656e-02
1.85665596e+00 1.20144367e+00 6.56210244e-01 9.02031481e-01
4.98170793e-01 3.05096805e-01 8.47941041e-01 -5.23114242e-02
4.64425415e-01 -2.27037326e-01 5.14203131e-01 -4.50099766e-01
3.60749751e-01 -1.28023848e-01 6.58514053e-02 9.96874213e-01
-7.18757331e-01 1.73340023e-01 -6.51990771e-01 8.98200214e-01
-1.37616539e+00 -1.02242100e+00 -5.94470620e-01 2.22588563e+00
6.54707015e-01 -4.82888855e-02 3.26248884e-01 6.96526766e-02
2.88138539e-01 -2.72068828e-01 -5.77565253e-01 -7.17130601e-01
-1.30798131e-01 5.46903729e-01 9.66912925e-01 -7.68413246e-02
-6.81867599e-01 9.72831488e-01 6.48261356e+00 7.07943290e-02
-1.49646115e+00 -1.28037944e-01 4.68126893e-01 5.64737737e-01
-3.65294665e-01 4.64468390e-01 -1.59398213e-01 2.23494858e-01
5.83094358e-01 -5.15618145e-01 5.27945280e-01 4.80936587e-01
7.14154661e-01 -7.46717751e-01 -4.48559672e-01 3.37509751e-01
-6.08296096e-01 -1.49950945e+00 -4.71672684e-01 5.60908377e-01
8.78409445e-01 1.19307876e-01 -6.31623641e-02 6.52251020e-02
5.18979967e-01 -1.02878916e+00 5.54406106e-01 2.27582455e-01
5.62056601e-01 -1.21344304e+00 2.35028312e-01 9.46069658e-02
-1.42221951e+00 -1.88938096e-01 -1.86495021e-01 -2.44391918e-01
2.81958789e-01 2.89758861e-01 -5.41154027e-01 5.77440500e-01
1.00863087e+00 3.97373915e-01 -1.51264265e-01 8.08584630e-01
3.49728018e-02 7.13184237e-01 -1.88042343e-01 -6.48385435e-02
7.36381948e-01 -1.08949232e+00 1.89414129e-01 6.06062531e-01
3.34153801e-01 1.57414168e-01 1.84649333e-01 8.98635805e-01
-1.32431850e-01 3.70635927e-01 -4.47280467e-01 -9.24432576e-01
7.36022532e-01 8.88036788e-01 -1.28210127e+00 -1.22071318e-01
-3.19358677e-01 6.54076457e-01 -3.69318157e-01 3.25367481e-01
-5.28191447e-01 -6.67357981e-01 5.33677757e-01 -1.15354080e-02
1.45411387e-01 -4.70530093e-01 -9.42875594e-02 -3.17998171e-01
-3.39205563e-01 -7.61513114e-01 5.73577322e-02 -4.92960393e-01
-6.78914070e-01 1.87741712e-01 -3.25943679e-01 -7.92329073e-01
1.23526566e-01 -1.12389073e-01 -5.53625464e-01 1.18609118e+00
-1.45937943e+00 -1.57043374e+00 -6.33598790e-02 3.61221330e-03
6.87698960e-01 1.58597201e-01 1.00988054e+00 1.32417500e-01
-3.66685808e-01 6.73349723e-02 9.69704747e-01 2.84981937e-03
6.47647560e-01 -9.51055408e-01 1.47674561e-01 4.03242320e-01
-2.06276536e-01 6.58494174e-01 2.56776601e-01 -6.34317577e-01
-1.52336860e+00 -6.99696958e-01 1.24254894e+00 -9.03779566e-02
9.32243764e-01 1.37019502e-02 -5.60591936e-01 7.61888742e-01
5.26980996e-01 -9.97651994e-01 5.87009430e-01 2.46555537e-01
-5.09257168e-02 -1.34783998e-01 -1.15422606e+00 5.83699346e-01
2.56813556e-01 -3.17983091e-01 2.93176491e-02 4.53086495e-01
3.85314673e-01 -7.84115940e-02 -1.20403576e+00 -4.13051665e-01
1.13181973e+00 -5.31318724e-01 6.51381850e-01 -1.32763907e-01
4.59448725e-01 -2.12653384e-01 -3.54511850e-02 -1.09171283e+00
-6.37510359e-01 -6.71569884e-01 9.09221232e-01 1.56622720e+00
5.09394526e-01 -1.00643313e+00 6.03516042e-01 5.59056818e-01
-5.61341308e-02 -4.73745048e-01 -3.08710784e-01 -8.30771387e-01
3.03916633e-01 3.95638496e-01 6.88531339e-01 1.39067113e+00
-1.08574681e-01 3.22463103e-02 -3.75959605e-01 -1.91033497e-01
7.30662704e-01 5.30535281e-01 7.22193658e-01 -1.27130473e+00
2.98131078e-01 -4.06947315e-01 2.24376246e-01 -4.63286251e-01
-8.07510793e-01 -8.08311164e-01 -3.89246047e-01 -2.00263596e+00
3.49042177e-01 -8.78218889e-01 -9.65295732e-02 5.08450329e-01
1.50388300e-01 4.35171694e-01 7.57442594e-01 4.04497266e-01
4.87870991e-01 -3.63980353e-01 1.85856545e+00 -3.70355807e-02
-7.85397291e-01 -3.20753485e-01 -1.26697433e+00 2.99252987e-01
1.15055490e+00 -2.49747291e-01 -1.69445962e-01 -3.22229028e-01
2.14081690e-01 1.46303385e-01 -5.45360968e-02 -2.63378173e-01
-3.15243959e-01 -9.89209473e-01 2.95039922e-01 -8.38372767e-01
1.61695927e-01 -9.83311892e-01 8.99066985e-01 8.85155916e-01
3.03795546e-01 3.45781893e-02 6.85337856e-02 6.34499267e-02
7.42639005e-02 2.06207052e-01 8.53281200e-01 -2.57962227e-01
-8.16657007e-01 -2.41158560e-01 -5.45363843e-01 -3.23967129e-01
1.04317474e+00 -8.60055923e-01 -4.55214173e-01 -1.17695674e-01
-7.79440105e-01 5.16340256e-01 8.29988122e-01 3.50506634e-01
-1.44235089e-01 -1.29186356e+00 -7.54338920e-01 -2.57643938e-01
1.61176901e-02 -3.83474380e-01 2.05370516e-01 1.16095269e+00
-1.11451972e+00 1.12304881e-01 -7.83551335e-01 -4.15485680e-01
-8.02995920e-01 3.47941667e-01 -3.01908761e-01 -7.45865926e-02
-9.68627274e-01 3.14309180e-01 2.10227907e-01 3.26994270e-01
-3.48501265e-01 -4.14431065e-01 -8.35050270e-02 1.57980267e-02
2.36359373e-01 7.49828815e-01 -5.67877114e-01 -6.80630922e-01
-3.97828728e-01 3.39545816e-01 1.32258251e-01 2.32399151e-01
1.43992460e+00 -1.09655723e-01 -4.80456948e-01 8.40205133e-01
7.21329570e-01 4.60440278e-01 -8.10984313e-01 3.96013767e-01
-3.50153595e-02 -8.10567260e-01 -1.43936470e-01 -1.21865976e+00
-1.07819831e+00 4.41345900e-01 5.07442892e-01 3.22449803e-01
1.04090667e+00 -1.18997663e-01 1.18607318e+00 6.17051013e-02
4.00888801e-01 -1.15974581e+00 -3.83402228e-01 1.96854964e-01
5.98864734e-01 -1.17376876e+00 -6.50395155e-02 -2.35426903e-01
-5.42963624e-01 9.09533262e-01 -1.82304695e-01 -4.13244143e-02
5.84082186e-01 -1.34662256e-01 4.41606224e-01 1.36057707e-02
-9.71763805e-02 -1.81352183e-01 -5.34719229e-01 6.05855107e-01
9.98854876e-01 5.77663720e-01 -9.23246562e-01 -3.18374902e-01
-2.67086983e-01 2.00421706e-01 4.79421318e-01 1.13954711e+00
-7.80612648e-01 -1.54867852e+00 -6.45097852e-01 4.72325057e-01
-4.33279008e-01 1.93397831e-02 -2.64459550e-01 1.58941698e+00
3.10262471e-01 1.05472076e+00 -1.25783548e-01 4.08630133e-01
4.91268821e-02 3.61512229e-02 7.01318502e-01 1.09315395e-01
-8.22788775e-01 6.36217475e-01 2.84757882e-01 5.43172695e-02
-4.89509791e-01 -6.74386740e-01 -1.65144122e+00 -1.24400759e+00
-5.80209613e-01 5.29095888e-01 1.21076095e+00 6.56570613e-01
-3.43887627e-01 2.00457498e-01 9.66556847e-01 -3.78348053e-01
6.86294958e-02 -7.58424819e-01 -1.20768189e+00 -1.86497107e-01
-2.96040058e-01 -2.62550384e-01 3.97173822e-01 2.19422832e-01] | [9.09794807434082, 6.205521583557129] |
be6aa59f-04a8-4639-adf9-d199ff6679e4 | detecting-approximate-reflection-symmetry-in | 1706.08801 | null | http://arxiv.org/abs/1706.08801v6 | http://arxiv.org/pdf/1706.08801v6.pdf | Detecting Approximate Reflection Symmetry in a Point Set using Optimization on Manifold | We propose an algorithm to detect approximate reflection symmetry present in
a set of volumetrically distributed points belonging to $\mathbb{R}^d$
containing a distorted reflection symmetry pattern. We pose the problem of
detecting approximate reflection symmetry as the problem of establishing
correspondences between the points which are reflections of each other and we
determine the reflection symmetry transformation. We formulate an optimization
framework in which the problem of establishing the correspondences amounts to
solving a linear assignment problem and the problem of determining the
reflection symmetry transformation amounts to solving an optimization problem
on a smooth Riemannian product manifold. The proposed approach estimates the
symmetry from the geometry of the points and is descriptor independent. We
evaluate the performance of the proposed approach on the standard benchmark
dataset and achieve the state-of-the-art performance. We further show the
robustness of our approach by varying the amount of distortion in a perfect
reflection symmetry pattern where we perturb each point by a different amount
of perturbation. We demonstrate the effectiveness of the method by applying it
to the problem of 2-D and 3-D reflection symmetry detection along with
comparisons. | ['Rajendra Nagar', 'Shanmuganathan Raman'] | 2017-06-27 | null | null | null | null | ['symmetry-detection'] | ['computer-vision'] | [ 2.48614773e-01 2.87613720e-02 3.34112912e-01 -1.87033296e-01
-7.87015259e-01 -4.48290139e-01 7.62118459e-01 -9.25450176e-02
-2.63668180e-01 -1.43776059e-01 2.99018063e-02 3.41793551e-04
-3.01282555e-01 -5.62370539e-01 -7.26866961e-01 -5.60952604e-01
-3.74490879e-02 5.46871185e-01 6.93777055e-02 -2.33976111e-01
7.88862765e-01 8.07009101e-01 -1.34642422e+00 -1.23899564e-01
4.44778830e-01 9.32418168e-01 -7.41481707e-02 5.14673233e-01
3.90836507e-01 1.60512328e-01 -2.74958014e-01 -6.73367605e-02
8.37772667e-01 -3.77862841e-01 -1.16355026e+00 6.65799737e-01
6.52651787e-01 9.61524695e-02 -1.78717464e-01 1.05527830e+00
2.60269463e-01 3.62421751e-01 9.68054712e-01 -1.21881413e+00
-3.33899140e-01 -3.88224900e-01 -8.71434808e-01 -2.76244618e-03
6.64623380e-01 -5.63667238e-01 9.84322071e-01 -1.27551818e+00
8.60170305e-01 7.77246356e-01 5.14812529e-01 3.15010607e-01
-1.20004332e+00 -1.77585334e-01 -3.15129936e-01 4.20589000e-02
-1.84527528e+00 -6.03455305e-01 9.04942989e-01 -6.34213030e-01
6.88036978e-01 3.42058837e-01 1.93092152e-01 1.23904936e-01
6.62022978e-02 1.12660855e-01 8.10728192e-01 -4.70875740e-01
2.91416287e-01 -1.16684660e-01 -5.60676716e-02 8.83610189e-01
7.33925104e-02 -2.20707595e-01 -2.87194252e-01 -1.28370181e-01
7.93442667e-01 1.82218403e-01 -3.13762069e-01 -8.29883397e-01
-1.18224120e+00 7.86628306e-01 3.38758677e-01 9.26556438e-02
-2.47710451e-01 -3.33208650e-01 -9.44199115e-02 2.17025578e-01
3.62423718e-01 5.24488032e-01 2.05199376e-01 2.05157116e-01
-7.90076673e-01 2.46011853e-01 7.63618648e-01 1.03436339e+00
7.30168045e-01 -2.48166516e-01 3.24022174e-01 7.04267085e-01
2.34745771e-01 3.90458792e-01 9.83632430e-02 -1.01440263e+00
5.14597893e-01 7.48303354e-01 1.55943021e-01 -1.18618500e+00
-4.61681098e-01 -1.38956338e-01 -7.01241076e-01 3.12778145e-01
5.31443954e-01 2.45707408e-01 -4.11342233e-01 1.35248089e+00
3.89134884e-01 1.61237419e-01 -2.58053876e-02 9.64675248e-01
4.49217111e-01 4.40839201e-01 -7.86340594e-01 3.25026214e-02
1.20600259e+00 -5.73619306e-01 -1.56916052e-01 2.88225740e-01
4.87591058e-01 -1.01368999e+00 8.02901208e-01 3.34897608e-01
-1.13686204e+00 -2.44310111e-01 -1.13353431e+00 7.22305924e-02
2.85419449e-02 2.44694099e-01 -9.81403962e-02 3.89064699e-01
-8.94150138e-01 7.46721089e-01 -7.02320814e-01 -4.53998864e-01
1.78360477e-01 5.04184544e-01 -8.31786990e-01 -9.88956890e-04
-3.76223534e-01 8.13396990e-01 -3.46757233e-01 1.71733707e-01
-4.23169076e-01 -6.70134485e-01 -9.15084124e-01 -1.82517111e-01
5.67790195e-02 -4.51226383e-01 7.48584807e-01 -6.43010139e-01
-1.22297418e+00 1.41082335e+00 -3.30828846e-01 1.20143946e-02
6.83483601e-01 1.99336916e-01 -2.21658394e-01 2.35056311e-01
1.73999518e-01 1.58437788e-01 7.87270606e-01 -1.23941398e+00
-3.72148156e-01 -8.08517218e-01 1.06017869e-02 3.67596895e-01
1.12082981e-01 1.10248007e-01 -3.00213337e-01 -1.69360742e-01
9.59990799e-01 -1.09101748e+00 -7.66793191e-02 2.67669074e-02
-5.59461653e-01 -3.25320438e-02 9.68982935e-01 -6.28966153e-01
7.05466211e-01 -2.23491979e+00 3.33054632e-01 7.56828964e-01
3.26688766e-01 -2.44983897e-01 1.55982645e-02 4.01899993e-01
-4.24689472e-01 -2.74409652e-02 -2.60986626e-01 -4.20865506e-01
8.29742402e-02 -2.25619927e-01 -2.15042859e-01 1.45049226e+00
1.18373998e-01 3.65591913e-01 -5.29773533e-01 -1.43104851e-01
1.91171452e-01 5.36473811e-01 -4.84763205e-01 2.62546420e-01
3.73435020e-01 3.96703035e-01 -3.69734466e-01 3.25867474e-01
9.23175752e-01 -5.27050272e-02 -1.17190309e-01 -3.01470369e-01
-3.65017980e-01 2.21341759e-01 -1.80678821e+00 1.53661227e+00
-9.83538404e-02 3.84867549e-01 -3.38824838e-02 -1.10855556e+00
1.26966095e+00 2.81771496e-02 7.77922988e-01 -5.87617218e-01
2.70951450e-01 1.82692423e-01 -1.52102783e-02 -1.17655687e-01
4.96269792e-01 -2.32121989e-01 3.47489230e-02 7.23012149e-01
-2.54605323e-01 -4.15245950e-01 -8.61187801e-02 -2.05032885e-01
1.15297270e+00 -1.40059635e-01 2.07998887e-01 -5.78811288e-01
7.39312291e-01 -3.88974816e-01 1.43869579e-01 2.82124728e-01
1.44022793e-01 9.08628881e-01 4.63167012e-01 -4.50187325e-01
-1.40156472e+00 -1.31673586e+00 -4.13027316e-01 1.74353734e-01
3.68599623e-01 -3.69896382e-01 -8.39245319e-01 -3.10503781e-01
-3.30196768e-02 3.37016404e-01 -6.42161012e-01 -1.45742521e-01
-4.76371169e-01 -6.52280152e-01 1.04037151e-01 5.53966910e-02
7.53153682e-01 -3.68595839e-01 -6.04716241e-01 -3.63235593e-01
-2.31135771e-01 -1.02378118e+00 -7.23887622e-01 -4.34512407e-01
-8.58586907e-01 -1.23494613e+00 -7.23953366e-01 -9.54394341e-01
1.28056324e+00 4.28601772e-01 7.68843889e-01 5.21755591e-02
-5.00138521e-01 6.52532935e-01 4.46174527e-03 1.20265540e-02
-2.18456998e-01 -2.59657472e-01 1.73275724e-01 4.16820467e-01
1.10285290e-01 -5.40113807e-01 -6.16594493e-01 9.35563922e-01
-8.29691947e-01 -3.50960940e-01 1.04026318e-01 2.81200439e-01
8.62584710e-01 2.38861084e-01 -1.53348058e-01 -2.77403474e-01
5.31169355e-01 -2.32996032e-01 -7.11752653e-01 9.01909471e-02
-2.17131898e-01 4.29195791e-01 2.81720072e-01 1.02710007e-02
-4.04287070e-01 4.32494372e-01 2.79547602e-01 -5.05556464e-01
1.72675431e-01 8.25361684e-02 -1.66745543e-01 -6.41764343e-01
5.33879578e-01 3.66755933e-01 1.91091806e-01 -4.61745650e-01
1.15772173e-01 5.58267474e-01 4.12904888e-01 -6.46111786e-01
9.28919375e-01 1.08192468e+00 6.69735312e-01 -1.14843369e+00
-3.43488008e-01 -8.03007066e-01 -8.91652644e-01 -2.06251666e-01
6.80275798e-01 -4.36810166e-01 -8.36222827e-01 4.02445018e-01
-1.08203757e+00 2.01941952e-01 -3.17171305e-01 4.05964702e-01
-1.08470654e+00 8.38892400e-01 -2.54796265e-04 -5.68707585e-01
-1.39981732e-01 -1.16012847e+00 1.06753957e+00 -2.93688536e-01
-1.87995598e-01 -8.30902398e-01 4.71516937e-01 2.72446454e-01
1.05899032e-02 3.30466777e-01 7.51725733e-01 -5.24447978e-01
-5.64663470e-01 -5.05934954e-01 -5.69131114e-02 2.55212098e-01
2.75138646e-01 -1.97902732e-02 -5.52457392e-01 -2.51944363e-01
3.95423710e-01 2.35104516e-01 3.66816789e-01 1.67980313e-01
8.40521812e-01 -8.88748169e-02 -5.36340922e-02 6.74267709e-01
1.45847142e+00 -9.38867107e-02 7.60512948e-01 3.21860462e-01
5.13201416e-01 8.06519210e-01 4.77721244e-01 3.84786695e-01
2.50270784e-01 9.60439265e-01 4.00022894e-01 -1.14616059e-01
1.25905082e-01 -1.43963620e-01 -5.04421256e-03 4.77231979e-01
-4.96713102e-01 2.02823341e-01 -9.64504659e-01 3.39001894e-01
-1.64492524e+00 -1.01518166e+00 -5.19068502e-02 2.86268783e+00
1.04638264e-01 -2.60935545e-01 2.75867522e-01 2.79745668e-01
8.32327783e-01 -1.27786502e-01 -1.00391500e-01 -5.27794421e-01
9.81405303e-02 2.13805377e-01 5.30914009e-01 7.67908633e-01
-9.93083894e-01 4.73263681e-01 6.21750879e+00 2.42217481e-01
-1.25639307e+00 -2.48050094e-01 2.73697019e-01 1.54489130e-01
-4.33535166e-02 5.86732924e-02 -7.22631633e-01 7.58452117e-02
4.86491442e-01 -9.08174589e-02 4.23361450e-01 6.14893854e-01
2.85498589e-01 -1.84259340e-01 -1.35447347e+00 1.16642869e+00
5.51714122e-01 -1.15604699e+00 6.77829757e-02 4.45459127e-01
7.16937661e-01 -3.34140211e-02 2.69508600e-01 -4.34575707e-01
-1.90882996e-01 -1.00453603e+00 4.81351405e-01 4.82088268e-01
6.62795782e-01 -6.20177209e-01 3.13434273e-01 2.00144991e-01
-1.26107240e+00 3.68979335e-01 -4.15031433e-01 -5.53454906e-02
5.22414893e-02 1.09858848e-01 -9.12311554e-01 5.97628117e-01
3.74012381e-01 5.51152766e-01 -3.56461197e-01 1.25277758e+00
7.03804269e-02 -2.28731126e-01 -5.55723846e-01 3.03092510e-01
-3.21290120e-02 -8.10384512e-01 7.08708584e-01 9.54573333e-01
5.35708010e-01 1.18465200e-01 -7.44527876e-02 8.95880282e-01
-3.57782282e-03 3.83365482e-01 -7.75901139e-01 5.18120706e-01
3.98544483e-02 1.05898356e+00 -9.22796369e-01 1.77034318e-01
-3.22409838e-01 1.19914806e+00 3.40896755e-01 1.65055171e-01
-4.07832623e-01 -4.40250486e-01 8.06284368e-01 5.71927845e-01
4.13239151e-01 -4.88292038e-01 -4.13508452e-02 -1.19864476e+00
5.04825950e-01 -4.88589048e-01 2.44735867e-01 -7.72897124e-01
-9.02086318e-01 5.21757662e-01 -4.10061330e-02 -1.52894998e+00
-7.14603364e-02 -7.27993429e-01 -8.77467215e-01 1.04578352e+00
-1.03975964e+00 -5.88440418e-01 -3.80289078e-01 8.85961056e-01
4.59245779e-02 -6.42403448e-03 7.25231767e-01 7.33395144e-02
-9.91425291e-02 3.03569347e-01 1.50146455e-01 2.91439265e-01
3.00782114e-01 -1.08713222e+00 2.70242125e-01 9.44412470e-01
7.28749707e-02 6.87900722e-01 6.63398623e-01 -2.99972147e-01
-1.57369220e+00 -8.23683977e-01 1.11581707e+00 -5.72189093e-01
5.18798053e-01 -6.18626773e-01 -5.45325220e-01 7.34917462e-01
-4.65997487e-01 9.53651220e-02 4.51386064e-01 -2.72307932e-01
-4.80146527e-01 -7.68364668e-02 -1.53700793e+00 5.29652774e-01
1.12694061e+00 -6.11278236e-01 -7.54417479e-01 4.52179462e-01
-2.49919314e-02 -5.06266534e-01 -8.93682063e-01 2.71817982e-01
3.77661198e-01 -8.42653632e-01 1.00246739e+00 -5.68478525e-01
3.09660017e-01 -6.09341621e-01 -4.32684749e-01 -1.13600147e+00
-7.14900494e-02 -6.49130881e-01 6.08605504e-01 7.00277567e-01
2.24662185e-01 -5.65625131e-01 9.88526642e-01 5.96487224e-01
-5.48788458e-02 -6.91569507e-01 -1.37261367e+00 -9.13201034e-01
-9.71699059e-02 -1.52719155e-01 3.10327291e-01 8.00729752e-01
2.87112951e-01 5.05545847e-02 -1.92186274e-02 4.89082277e-01
8.34297240e-01 2.70058483e-01 8.41857016e-01 -1.08493769e+00
-1.53334931e-01 -2.65551537e-01 -1.17525399e+00 -1.04440320e+00
1.80612817e-01 -1.00056422e+00 -2.77832542e-02 -1.19982338e+00
1.51957422e-01 -2.80050755e-01 5.65657057e-02 -1.42457813e-01
5.79235971e-01 2.26247713e-01 4.13105190e-02 3.53792638e-01
-5.67845881e-01 5.13885915e-01 1.08341229e+00 3.92622612e-02
-1.55876607e-01 1.81201905e-01 -5.19709051e-01 7.24276781e-01
4.96176720e-01 -2.75638103e-01 -1.11156143e-01 -1.78690776e-01
2.23754361e-01 -6.94312453e-02 5.26968539e-01 -9.83515620e-01
1.75574109e-01 2.09004298e-01 9.08253193e-02 -2.87278920e-01
4.49846357e-01 -8.64297628e-01 1.98993802e-01 3.36935073e-01
3.95567436e-03 4.38727051e-01 -7.74010411e-03 2.48914003e-01
-6.90784603e-02 -2.22334817e-01 9.93644297e-01 5.95063902e-02
-2.21333072e-01 4.88843918e-01 1.04137950e-01 1.34205341e-01
1.16075158e+00 -4.84277695e-01 1.53171560e-02 -3.69930178e-01
-6.88304901e-01 -1.72319546e-01 8.89533222e-01 3.18617851e-01
9.40610051e-01 -1.24532330e+00 -8.37802947e-01 5.58819890e-01
3.10194641e-01 -6.82925805e-02 -1.11960799e-01 9.48601723e-01
-7.98558176e-01 3.51444840e-01 -2.40000382e-01 -8.39086115e-01
-1.53416085e+00 3.60482126e-01 8.39500010e-01 1.83889776e-01
-6.05723739e-01 4.91140753e-01 1.82369992e-01 -5.26413858e-01
-2.56814733e-02 -3.19261491e-01 -1.46168798e-01 -4.17326450e-01
4.01369810e-01 7.14201748e-01 3.98062676e-01 -1.12720561e+00
-4.67649728e-01 1.26671398e+00 1.31780192e-01 -2.41927326e-01
1.32824504e+00 3.72180045e-02 -2.65602082e-01 1.44140869e-01
1.68555140e+00 2.34190121e-01 -9.55587268e-01 -1.70819297e-01
-2.07647264e-01 -5.94177306e-01 -2.07478151e-01 -1.56587996e-02
-9.62928653e-01 6.19104445e-01 5.46310902e-01 -9.84135550e-03
8.18055928e-01 3.01448703e-01 1.52551129e-01 3.59468281e-01
3.01067501e-01 -8.87537599e-01 -3.00226286e-02 3.75608891e-01
1.11921036e+00 -9.72017288e-01 6.44089207e-02 -6.75484002e-01
-1.82104707e-01 1.28254271e+00 5.62397875e-02 -7.97874272e-01
9.77796972e-01 -1.54490739e-01 -1.44229069e-01 -5.25674522e-01
1.14315655e-02 1.69467777e-01 7.12195277e-01 3.43300670e-01
1.44007757e-01 -8.50894377e-02 -3.84249806e-01 -1.09813720e-01
-5.16592503e-01 -2.61501700e-01 5.51102161e-01 7.20647633e-01
-2.27434844e-01 -1.02299035e+00 -5.40679634e-01 1.03426166e-01
-2.64901936e-01 1.91752151e-01 -6.26583934e-01 8.08766007e-01
-1.06110647e-01 7.72515118e-01 5.55681050e-01 -3.21987212e-01
7.87021995e-01 -1.75938308e-01 8.40871572e-01 -4.81979221e-01
-9.63916332e-02 -2.38550216e-01 -2.14728296e-01 -7.28027999e-01
-3.72331828e-01 -1.01804566e+00 -1.27528560e+00 -4.35965061e-01
2.58323085e-02 2.40231082e-01 1.01854837e+00 8.10797095e-01
3.91639084e-01 -3.29241693e-01 1.11806405e+00 -6.89516008e-01
-7.86672115e-01 -4.42769825e-01 -9.24854815e-01 7.92692542e-01
2.48931885e-01 -6.81619048e-01 -7.16082633e-01 -2.36839414e-01] | [8.05241584777832, -2.3514676094055176] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.