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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
166dea06-e99e-4441-ad79-ad6a5aa659d7 | deep-hdr-hallucination-for-inverse-tone | 2106.09486 | null | https://arxiv.org/abs/2106.09486v1 | https://arxiv.org/pdf/2106.09486v1.pdf | Deep HDR Hallucination for Inverse Tone Mapping | Inverse Tone Mapping (ITM) methods attempt to reconstruct High Dynamic Range (HDR) information from Low Dynamic Range (LDR) image content. The dynamic range of well-exposed areas must be expanded and any missing information due to over/under-exposure must be recovered (hallucinated). The majority of methods focus on the former and are relatively successful, while most attempts on the latter are not of sufficient quality, even ones based on Convolutional Neural Networks (CNNs). A major factor for the reduced inpainting quality in some works is the choice of loss function. Work based on Generative Adversarial Networks (GANs) shows promising results for image synthesis and LDR inpainting, suggesting that GAN losses can improve inverse tone mapping results. This work presents a GAN-based method that hallucinates missing information from badly exposed areas in LDR images and compares its efficacy with alternative variations. The proposed method is quantitatively competitive with state-of-the-art inverse tone mapping methods, providing good dynamic range expansion for well-exposed areas and plausible hallucinations for saturated and under-exposed areas. A density-based normalisation method, targeted for HDR content, is also proposed, as well as an HDR data augmentation method targeted for HDR hallucination. | ['Kurt Debattista', 'Thomas Bashford-Rogers', 'Demetris Marnerides'] | 2021-06-17 | null | null | null | null | ['tone-mapping', 'inverse-tone-mapping'] | ['computer-vision', 'computer-vision'] | [ 5.04027605e-01 3.49132895e-01 7.42170541e-03 -2.52631638e-04
-6.23476923e-01 -1.45298868e-01 6.72326446e-01 -6.18630946e-01
-7.30910525e-02 1.03422165e+00 4.19024140e-01 2.00743705e-01
7.83049539e-02 -1.14003384e+00 -6.95983410e-01 -6.92880154e-01
2.54301220e-01 5.10765135e-01 3.49196374e-01 -8.00397098e-01
-5.79467975e-03 5.88133812e-01 -1.44573843e+00 3.06932718e-01
7.71663129e-01 8.18010747e-01 3.98438692e-01 5.68135023e-01
-1.40943930e-01 1.04374599e+00 -8.36439908e-01 -2.33551502e-01
4.22335744e-01 -9.72931743e-01 -5.83723128e-01 -6.90171123e-02
3.53235722e-01 -4.89698559e-01 -7.35111952e-01 7.79685974e-01
6.73611641e-01 -9.53602344e-02 7.02124000e-01 -9.28304315e-01
-9.61152971e-01 1.69989482e-01 -6.32531106e-01 7.19799772e-02
4.05360162e-01 1.12942338e-01 4.29020971e-01 -7.98478484e-01
9.13889289e-01 1.09251940e+00 7.16302693e-01 6.98215067e-01
-1.49908018e+00 -5.08563399e-01 -6.63011253e-01 2.56297171e-01
-1.15877724e+00 -2.83116430e-01 1.09368634e+00 -1.86604597e-02
8.05386841e-01 5.58175802e-01 6.64151549e-01 9.74890292e-01
3.38443875e-01 3.89112115e-01 1.75754786e+00 -6.08784974e-01
1.44817814e-01 2.29821220e-01 -9.12013650e-01 1.74604431e-01
-3.59337479e-01 4.60787803e-01 -5.57995737e-01 2.95331270e-01
1.14879584e+00 -1.61595941e-01 -5.82157075e-01 -1.52709782e-01
-1.07575023e+00 7.29855120e-01 5.51714599e-01 4.05474693e-01
-5.03666759e-01 1.21941473e-02 2.59308010e-01 7.21975327e-01
7.78853476e-01 5.84773183e-01 1.00359768e-01 -8.30841810e-02
-1.38203073e+00 2.10817218e-01 3.72370660e-01 6.92216218e-01
8.73238146e-01 5.33540010e-01 -3.66595417e-01 1.17956901e+00
9.11766198e-03 6.04930997e-01 4.53128904e-01 -9.72995937e-01
1.55883789e-01 2.50560910e-01 -2.64951158e-02 -7.58824587e-01
8.07857886e-02 -3.73368770e-01 -1.04947126e+00 7.87003875e-01
1.33597940e-01 1.33420765e-01 -1.25566494e+00 1.65898812e+00
3.86386700e-02 -6.96437135e-02 7.59507567e-02 1.07450700e+00
6.20501816e-01 9.75132167e-01 -1.97373673e-01 -2.27732852e-01
1.01694655e+00 -7.47907639e-01 -1.11247563e+00 -2.47631758e-01
-2.09323555e-01 -1.02057266e+00 1.17850494e+00 5.42067468e-01
-1.54682279e+00 -7.39948630e-01 -9.99800026e-01 -3.29912305e-01
-2.58269101e-01 -3.44183952e-01 1.09251946e-01 7.16484964e-01
-1.44714594e+00 6.79105639e-01 -2.25650713e-01 -1.89939231e-01
2.58769482e-01 1.66097552e-01 -3.35886002e-01 -2.99052298e-01
-1.52865493e+00 1.24642456e+00 1.30539641e-01 -6.17025457e-02
-8.20154786e-01 -9.56152201e-01 -7.29615510e-01 -1.23777032e-01
-3.58119123e-02 -6.16255701e-01 6.03327990e-01 -1.20615387e+00
-1.73034644e+00 9.49929357e-01 4.44340378e-01 -6.49221182e-01
9.37100351e-01 -3.71836498e-02 -5.41395485e-01 4.40633088e-01
-1.79754749e-01 9.50072408e-01 1.21608388e+00 -1.49923193e+00
1.53631821e-01 -1.38557896e-01 -3.82514805e-01 2.45545417e-01
-3.82576510e-02 -2.06691802e-01 -1.04824558e-01 -1.24948668e+00
-5.15570305e-03 -6.65105641e-01 6.15538843e-02 1.26558095e-01
-2.29838401e-01 6.55619383e-01 1.38699663e+00 -1.26331210e+00
1.01454079e+00 -1.81932187e+00 1.08801872e-01 -3.57960723e-03
-2.71880589e-02 5.04970312e-01 -1.99233577e-01 7.52097070e-01
-4.46357340e-01 -7.55577907e-02 -5.22530258e-01 -4.16315258e-01
-2.15443805e-01 2.49090567e-01 -6.63838744e-01 5.73556066e-01
1.32298827e-01 1.02856064e+00 -4.55175757e-01 -2.38720432e-01
8.39260161e-01 1.10057843e+00 -2.97797889e-01 3.83385807e-01
-3.08680713e-01 8.05553317e-01 2.09694847e-01 4.33277726e-01
9.90596473e-01 3.90580982e-01 -4.24898595e-01 -4.13909793e-01
-1.47992492e-01 -2.20032394e-01 -7.76913702e-01 1.65493488e+00
-8.73637974e-01 8.82531583e-01 -1.79447681e-02 -7.39777565e-01
1.42116845e+00 3.67313117e-01 5.32472014e-01 -1.29177368e+00
-1.18976235e-01 4.82939959e-01 -3.04867774e-01 -3.23724478e-01
5.66043258e-01 -8.03635657e-01 2.96503395e-01 2.86335975e-01
-9.88680869e-02 -7.00160563e-01 -3.01851779e-01 -1.04790241e-01
8.78554881e-01 3.01237673e-01 1.06974818e-01 -1.54254017e-02
4.95733321e-01 -2.23876044e-01 9.60275680e-02 4.50663954e-01
1.48094654e-01 1.52542186e+00 4.30017829e-01 -3.90426099e-01
-1.82447445e+00 -1.01662683e+00 -5.39149940e-02 3.79884303e-01
1.25076249e-01 2.00793877e-01 -9.17451859e-01 -1.72330797e-01
-5.35489500e-01 8.85310113e-01 -7.62266636e-01 -3.51876467e-01
-6.96492553e-01 -4.93437707e-01 5.46422064e-01 2.03015566e-01
1.00727344e+00 -1.57897735e+00 -4.78750527e-01 3.05056721e-01
-3.87578130e-01 -8.86579454e-01 -2.13288680e-01 1.09179251e-01
-8.06053936e-01 -5.11905491e-01 -1.65085530e+00 -4.70187634e-01
3.96441936e-01 -4.38692756e-02 1.28526258e+00 -2.01333389e-01
-4.57681596e-01 8.45154449e-02 -5.83517849e-01 -9.23349485e-02
-1.08220518e+00 -3.48151207e-01 -4.24221486e-01 1.50208429e-01
-3.18866283e-01 -7.63847530e-01 -8.15256000e-01 5.00388861e-01
-1.37973249e+00 1.93754897e-01 8.07158351e-01 1.00671995e+00
6.71219409e-01 1.39564157e-01 5.75302422e-01 -8.80552888e-01
4.31683779e-01 -2.05412880e-01 -1.90700814e-01 -3.15954797e-02
-6.23289585e-01 -6.73257485e-02 8.53727400e-01 -4.70479071e-01
-1.25499928e+00 -2.97359675e-01 -5.93286395e-01 -8.97843897e-01
-2.08036825e-01 -2.12177530e-01 -2.89226651e-01 -3.69985074e-01
8.79611671e-01 6.59446895e-01 2.81077653e-01 -4.15728301e-01
3.22918862e-01 4.14180577e-01 5.95952809e-01 -9.76658836e-02
9.28070903e-01 7.25941360e-01 2.62038913e-02 -8.51205587e-01
-1.82574153e-01 6.33432120e-02 -1.31701186e-01 -3.52985859e-01
8.89841974e-01 -8.20560038e-01 1.63973551e-02 6.37340903e-01
-8.25829327e-01 -8.09596062e-01 -1.01879978e+00 1.86630756e-01
-1.00421405e+00 2.25146726e-01 -6.54267848e-01 -6.46333158e-01
-2.85961092e-01 -9.99036849e-01 1.19142723e+00 1.01433337e-01
-1.45569164e-02 -9.62384641e-01 2.91064352e-01 3.46756905e-01
1.10709703e+00 6.04362011e-01 8.98203671e-01 1.10385083e-01
-5.06010294e-01 -3.31102982e-02 -1.10878050e-01 6.38624847e-01
6.65496886e-02 -4.82796997e-01 -1.10850966e+00 -3.25160623e-01
2.27450520e-01 -3.51431578e-01 1.06574285e+00 6.96757197e-01
9.11034167e-01 -3.30198199e-01 1.35360867e-01 6.04396284e-01
1.63465798e+00 4.59128432e-02 1.87676394e+00 3.78552616e-01
3.67784649e-01 6.19806290e-01 6.91196799e-01 2.34934151e-01
-3.06103617e-01 1.06598127e+00 6.09241366e-01 -8.94792736e-01
-1.04478729e+00 -3.34619403e-01 3.80745679e-01 1.17095634e-01
-1.35247067e-01 -5.23221493e-01 -2.97915041e-01 5.33121526e-01
-1.18696105e+00 -1.11597645e+00 -3.96738462e-02 2.24631500e+00
7.58254647e-01 -1.76556110e-02 1.11766592e-01 5.07864535e-01
8.64408851e-01 3.32903206e-01 -4.68216449e-01 -6.55259192e-01
-5.58487713e-01 6.32475019e-01 4.64303315e-01 6.06861055e-01
-5.94488442e-01 7.55505025e-01 6.29698753e+00 1.25288987e+00
-1.31588399e+00 3.51028323e-01 1.00583255e+00 1.11321323e-01
-6.53516173e-01 -1.88325778e-01 -6.62239417e-02 5.63846409e-01
7.97574103e-01 3.91719282e-01 6.69658184e-01 4.40774083e-01
3.64578873e-01 -2.41692334e-01 -3.70266348e-01 1.05490410e+00
2.91964144e-01 -1.21886337e+00 5.90410084e-03 2.53604621e-01
9.79936182e-01 -2.47298494e-01 5.67442596e-01 7.98558071e-02
-3.04236740e-01 -1.19579160e+00 8.05467188e-01 7.72755563e-01
1.54840529e+00 -9.91287529e-01 6.48753166e-01 3.95945180e-03
-8.10445547e-01 1.43064603e-01 -4.42573518e-01 3.04018199e-01
4.26707625e-01 8.37315977e-01 -6.07230783e-01 4.53671008e-01
6.48198783e-01 3.57460409e-01 -5.28437912e-01 6.01568282e-01
-2.19088107e-01 3.47217858e-01 -1.30699471e-01 6.07217371e-01
-7.45658353e-02 -2.94919252e-01 7.34577000e-01 1.05362511e+00
6.01073265e-01 7.86743220e-03 -5.54362893e-01 1.37018192e+00
9.11431462e-02 -2.84216274e-02 -8.47415209e-01 2.02803969e-01
8.13573003e-02 1.08928692e+00 -7.01142669e-01 -1.85314760e-01
-1.61915526e-01 1.52252555e+00 -3.07547063e-01 3.50636542e-01
-9.67746079e-01 -3.28144491e-01 1.52251169e-01 6.92589819e-01
2.17797726e-01 1.07608974e-01 -2.06717476e-01 -7.36574590e-01
-7.69821703e-02 -1.04543304e+00 1.05850011e-01 -1.27450359e+00
-1.08235288e+00 9.38250065e-01 -1.07044600e-01 -1.56698668e+00
-3.44036847e-01 -9.22683179e-02 -4.67354894e-01 1.13797581e+00
-1.61852205e+00 -1.36237657e+00 -4.22256470e-01 7.16445148e-01
6.82950318e-01 -3.06417376e-01 6.57470524e-01 4.88725573e-01
1.75577700e-01 3.25634211e-01 -1.33842062e-02 -4.53555554e-01
9.12746251e-01 -1.19800091e+00 3.40087146e-01 8.12659800e-01
-1.50148377e-01 2.18007192e-02 1.09391761e+00 -6.90161288e-01
-1.03902388e+00 -1.15768909e+00 5.19624352e-01 9.87950191e-02
-8.99952054e-02 -2.48612151e-01 -1.05210853e+00 1.75307170e-01
5.22116661e-01 9.02433395e-02 9.06862393e-02 -7.61151612e-01
-1.22122332e-01 -1.33324146e-01 -1.82401836e+00 4.68963921e-01
5.19325256e-01 -4.66672182e-01 -2.60490209e-01 -2.91458815e-02
4.81652200e-01 -4.67398107e-01 -8.96825135e-01 4.19069082e-01
2.23107606e-01 -1.66249490e+00 1.24780619e+00 3.18782061e-01
8.32212687e-01 -3.71177942e-01 1.43604837e-02 -1.47389913e+00
-5.05493954e-02 -7.27643132e-01 -1.35364518e-01 1.23887038e+00
7.08127022e-02 -5.09037077e-01 8.22734177e-01 2.27743253e-01
-1.82863921e-01 -4.46713835e-01 -8.43610704e-01 -8.32080245e-01
8.92327651e-02 1.01316376e-02 3.94165426e-01 9.21655476e-01
-6.19409561e-01 2.58923713e-02 -1.03565109e+00 -3.50124598e-01
4.91315335e-01 -3.91520970e-02 5.10086298e-01 -6.24203742e-01
-3.81370574e-01 -1.81783751e-01 -4.37287182e-01 -4.99288559e-01
-6.31636828e-02 -5.33407986e-01 1.79473236e-02 -1.59742403e+00
-1.39720440e-01 -5.58930159e-01 1.81771845e-01 2.94999570e-01
2.20644802e-01 1.18099284e+00 1.34945154e-01 2.60217130e-01
2.94208199e-01 8.00509691e-01 1.74417531e+00 4.48703431e-02
-2.58017987e-01 -2.60519892e-01 -2.42689714e-01 2.44050309e-01
7.64177263e-01 -2.67189890e-01 -4.82818961e-01 -1.02989100e-01
1.90527722e-01 4.15434599e-01 7.14939535e-01 -1.29228652e+00
-2.45036438e-01 2.40237236e-01 6.75793767e-01 -4.27471071e-01
8.06123435e-01 -9.02823567e-01 7.66892374e-01 5.60309350e-01
-2.94789165e-01 -1.67314395e-01 1.42112404e-01 3.78402472e-01
-5.15919745e-01 -2.82105966e-03 1.49085367e+00 -2.99287289e-01
-6.31771624e-01 1.46831289e-01 -2.26698846e-01 -7.95429647e-02
8.46281409e-01 -5.84466517e-01 -9.57513303e-02 -9.36296165e-01
-6.83495700e-01 -7.93114126e-01 8.54722202e-01 3.38700056e-01
8.35222542e-01 -1.35316467e+00 -9.04616058e-01 4.17707562e-01
-2.94946164e-01 -3.66116226e-01 6.74494326e-01 8.14897478e-01
-9.04040754e-01 3.54007632e-02 -7.46070325e-01 -2.69294173e-01
-9.19176996e-01 7.33570099e-01 4.80538905e-01 -3.33496571e-01
-1.02391088e+00 4.40210372e-01 1.96717188e-01 -7.93618485e-02
-1.46902576e-01 1.86462760e-01 -1.73013151e-01 -1.54047817e-01
5.25495708e-01 4.09597427e-01 2.25615978e-01 -7.27611303e-01
7.16801211e-02 7.71535039e-01 2.35151574e-01 -4.97486025e-01
1.40885794e+00 -3.63266349e-01 -1.15050711e-01 9.89087448e-02
1.16668069e+00 1.33453608e-01 -1.29399848e+00 1.09111607e-01
-8.24607015e-01 -7.36551285e-01 3.28868210e-01 -1.04667091e+00
-1.44182897e+00 1.01164329e+00 1.05269504e+00 1.84508100e-01
1.72888255e+00 -1.68710172e-01 9.01626050e-01 -4.31752294e-01
3.75522733e-01 -9.32115853e-01 3.18039536e-01 1.31893707e-02
1.34668243e+00 -8.75927806e-01 -8.21614340e-02 -5.42724013e-01
-7.57120430e-01 1.11529517e+00 2.89291918e-01 -3.80666375e-01
2.69749939e-01 4.54428256e-01 6.00516126e-02 -7.00673908e-02
-2.54129648e-01 -7.61391893e-02 1.50925949e-01 9.45796907e-01
2.54878581e-01 -3.74074936e-01 -2.44420856e-01 -2.29177043e-01
-1.60964832e-01 -8.13933089e-02 6.74841881e-01 4.36683834e-01
-1.61798313e-01 -1.08986270e+00 -7.86882520e-01 1.37084395e-01
-3.78972411e-01 -1.04655206e-01 -3.18184972e-01 9.58005965e-01
2.94213623e-01 7.29298592e-01 1.00091891e-02 -1.69706747e-01
2.81428993e-01 -2.77278543e-01 7.09220529e-01 -1.82539299e-01
-5.59452713e-01 2.97853798e-01 -9.42047015e-02 -6.44602776e-01
-4.05821621e-01 -2.31010616e-01 -8.51633906e-01 -4.10872608e-01
-9.34694335e-02 -3.18049520e-01 6.11273408e-01 5.04755199e-01
4.55924943e-02 7.29710817e-01 7.41207063e-01 -8.72848213e-01
7.58926868e-02 -9.70225096e-01 -9.74302292e-01 5.49784958e-01
4.81560647e-01 -5.36687374e-01 -4.03693527e-01 9.90901217e-02] | [11.034914016723633, -2.0511703491210938] |
4bdaa8b0-5558-4b2f-820e-c60f756b6a03 | neural-network-based-general-method-for | 1906.10935 | null | https://arxiv.org/abs/1906.10935v2 | https://arxiv.org/pdf/1906.10935v2.pdf | Solving Statistical Mechanics on Sparse Graphs with Feedback Set Variational Autoregressive Networks | We propose a method for solving statistical mechanics problems defined on sparse graphs. It extracts a small Feedback Vertex Set (FVS) from the sparse graph, converting the sparse system to a much smaller system with many-body and dense interactions with an effective energy on every configuration of the FVS, then learns a variational distribution parameterized using neural networks to approximate the original Boltzmann distribution. The method is able to estimate free energy, compute observables, and generate unbiased samples via direct sampling without auto-correlation. Extensive experiments show that our approach is more accurate than existing approaches for sparse spin glasses. On random graphs and real-world networks, our approach significantly outperforms the standard methods for sparse systems such as the belief-propagation algorithm; on structured sparse systems such as two-dimensional lattices our approach is significantly faster and more accurate than recently proposed variational autoregressive networks using convolution neural networks. | ['Hai-Jun Zhou', 'Feng Pan', 'Pan Zhang', 'Pengfei Zhou'] | 2019-06-26 | null | null | null | null | ['feedback-vertex-set-fvs'] | ['graphs'] | [ 1.03620857e-01 5.15460372e-01 -4.66810875e-02 -1.39607593e-01
-5.85213125e-01 -1.95180625e-01 7.84220099e-01 -2.14690939e-01
-6.67158291e-02 1.04300737e+00 3.65223765e-01 -1.40941087e-02
4.09158766e-02 -1.16558254e+00 -1.10301745e+00 -1.19902956e+00
-3.68792653e-01 1.41539419e+00 1.46752849e-01 -2.55099505e-01
2.67137647e-01 2.10024402e-01 -1.06739128e+00 -1.31597504e-01
5.75690150e-01 7.71198571e-01 3.65648568e-02 6.96415007e-01
1.54708862e-01 9.38489914e-01 1.18329957e-01 5.37373126e-02
5.97630963e-02 -3.83837461e-01 -9.13888693e-01 -3.28683615e-01
6.06589735e-01 -3.47605199e-02 -1.12155676e+00 1.35968065e+00
1.99173942e-01 7.11273909e-01 9.31068182e-01 -7.03595877e-01
-1.13458514e+00 1.13087428e+00 -2.75345325e-01 4.37352099e-02
7.01303706e-02 2.88789213e-01 1.15219271e+00 -6.60702944e-01
8.64508271e-01 1.48205280e+00 6.38483584e-01 5.85940480e-01
-1.87471533e+00 -3.14778864e-01 -8.28049928e-02 -1.40788406e-02
-1.33825195e+00 -5.14770508e-01 7.89980292e-01 -3.23906302e-01
1.31012881e+00 3.21635231e-02 9.14747059e-01 1.43240345e+00
5.75137258e-01 2.56505579e-01 1.29105759e+00 -3.47276986e-01
5.88360965e-01 -3.04671198e-01 5.39012432e-01 1.20332336e+00
6.25276089e-01 2.06283137e-01 -5.23899794e-01 -7.52805889e-01
9.17253315e-01 -1.58010274e-02 -2.64960796e-01 -6.15425706e-01
-1.20711505e+00 9.38093662e-01 5.01266897e-01 1.88442454e-01
-4.44390118e-01 9.96744812e-01 1.99387282e-01 2.94421129e-02
4.26329076e-01 2.30903402e-01 5.43235466e-02 2.91830003e-01
-1.10057819e+00 6.09360456e-01 1.44278765e+00 6.36080921e-01
1.11163247e+00 4.19470042e-01 -7.04938099e-02 2.97612369e-01
6.03721559e-01 1.02867711e+00 2.99265496e-02 -1.16356337e+00
-1.10511996e-01 -1.08430207e-01 9.59151685e-02 -8.01411510e-01
-2.71365494e-01 -1.81869447e-01 -1.34532547e+00 -5.54719232e-02
2.99164653e-01 1.09204061e-01 -1.23547924e+00 1.81178355e+00
2.21309051e-01 3.82709950e-01 -2.17547685e-01 9.21800613e-01
7.90040493e-01 8.86723995e-01 -1.44969136e-01 -3.10417295e-01
9.13170755e-01 -8.00551951e-01 -5.05031526e-01 -2.23813787e-01
-1.57525260e-02 -2.92175226e-02 5.01656651e-01 3.20308119e-01
-1.31005871e+00 -1.19288824e-01 -8.01850438e-01 -1.77309424e-01
-5.28718643e-02 -4.31452096e-01 1.16396415e+00 5.85345030e-01
-1.43221164e+00 1.13834810e+00 -1.41018271e+00 -2.93078065e-01
1.62891939e-01 6.48179770e-01 -3.64319146e-01 -4.67243157e-02
-1.22422016e+00 6.70607984e-01 1.04099460e-01 9.73635018e-02
-1.38888252e+00 -4.81446713e-01 -1.01356542e+00 -3.63688078e-03
2.33284041e-01 -1.24538589e+00 8.66291285e-01 -5.26425183e-01
-1.89776957e+00 3.41698349e-01 -3.60864758e-01 -5.86717486e-01
-3.13355446e-01 3.78793746e-01 -6.00383757e-03 -1.32727961e-03
-2.34129459e-01 1.50283292e-01 9.33256984e-01 -1.08435905e+00
8.56818736e-01 -2.78230816e-01 -1.25038564e-01 -2.21660614e-01
1.31649256e-01 -5.69076478e-01 -1.49269346e-02 -3.27855907e-03
2.55504519e-01 -1.34304500e+00 -8.16450298e-01 -7.44139671e-01
-6.99760675e-01 -7.59240314e-02 2.26184338e-01 -4.42121953e-01
8.89542699e-01 -1.44222379e+00 9.00654316e-01 8.24799299e-01
8.65115047e-01 -2.23902673e-01 -1.88229352e-01 5.81978500e-01
1.52848035e-01 -7.49701401e-03 -3.36812168e-01 -1.63258329e-01
1.14582114e-01 3.73498857e-01 -5.66808432e-02 6.88281536e-01
-9.77594405e-02 1.05397284e+00 -1.00566924e+00 -2.61865854e-01
-1.02785183e-02 6.79000139e-01 -9.82399881e-01 -4.43488099e-02
-5.46817660e-01 3.31276506e-01 -6.49778008e-01 3.87558013e-01
5.40415943e-01 -8.79748642e-01 5.92703819e-01 -1.80829257e-01
3.48714709e-01 4.15902793e-01 -1.14024985e+00 1.59960651e+00
-1.53176367e-01 1.60210967e-01 2.97340780e-01 -1.30585647e+00
7.77196586e-01 1.57729357e-01 3.60269994e-01 -2.02322021e-01
2.06228867e-01 1.45847052e-01 7.47370869e-02 -6.61123022e-02
6.29501283e-01 -4.60555941e-01 -1.97963908e-01 5.73494792e-01
6.11721814e-01 -4.86260831e-01 5.39593846e-02 8.52492392e-01
1.47724438e+00 -7.62779079e-03 1.17982393e-02 -6.86369419e-01
1.08422562e-01 -4.53212619e-01 2.31796026e-01 1.23779178e+00
5.90267144e-02 2.94325173e-01 7.28747249e-01 -5.92495978e-01
-1.16682744e+00 -1.31853497e+00 1.85636878e-01 7.17708826e-01
-1.68488398e-01 -6.86999679e-01 -9.31537569e-01 -2.47469217e-01
1.93341091e-01 4.35576677e-01 -9.45549905e-01 -2.58676916e-01
-1.96164057e-01 -1.00956047e+00 1.79856524e-01 1.58064410e-01
-1.10260677e-02 -1.15788281e+00 5.20471394e-01 1.23980038e-01
1.34610772e-01 -7.66074359e-01 -3.61113310e-01 4.78582531e-01
-8.44277859e-01 -7.81421363e-01 -2.16168150e-01 -3.37653965e-01
7.32057095e-01 -9.91125405e-02 1.35321653e+00 -6.65201023e-02
-1.94079742e-01 2.78608143e-01 4.55822617e-01 3.88587594e-01
-5.96456349e-01 2.94909418e-01 5.33128381e-01 -1.74485967e-01
1.94251969e-01 -1.08001184e+00 -3.61372441e-01 -4.42495674e-01
-7.53064215e-01 -1.57392144e-01 2.97608316e-01 1.08101404e+00
7.29682863e-01 -3.57386500e-01 6.55358005e-03 -1.43705976e+00
5.77018976e-01 -7.89088726e-01 -7.35538185e-01 -7.77237713e-02
-5.97417533e-01 9.15668428e-01 6.18829906e-01 -2.03982785e-01
-8.99779379e-01 4.15404737e-02 1.22776918e-01 -4.12369132e-01
1.75168052e-01 5.86640000e-01 2.30306521e-01 -6.03424311e-01
7.23604739e-01 3.21856737e-01 1.99169293e-01 -2.19775841e-01
5.79143047e-01 1.08483762e-01 1.59793913e-01 -1.09018922e+00
8.94129455e-01 4.93194938e-01 6.22834682e-01 -9.22595382e-01
-6.84806347e-01 1.72201525e-02 -6.70360982e-01 -2.04298720e-02
6.78205192e-01 -7.76911914e-01 -1.14774954e+00 4.17424113e-01
-9.73162055e-01 -6.07381165e-01 -3.04859340e-01 5.64198017e-01
-7.44064867e-01 3.58789265e-01 -1.20022631e+00 -8.62398684e-01
-3.45629007e-01 -1.01554096e+00 1.12473023e+00 4.36231233e-02
-4.34515104e-02 -1.26803029e+00 9.06877458e-01 1.95860878e-01
6.95615113e-01 1.37932137e-01 9.06388104e-01 -3.15873116e-01
-1.12681758e+00 -4.50282767e-02 1.39716059e-01 1.09080240e-01
-4.03674960e-01 2.35077694e-01 -6.31230175e-01 -3.53566587e-01
-3.68112057e-01 -6.66344464e-01 1.48364151e+00 9.65039194e-01
1.01971722e+00 -3.63137811e-01 -5.19820273e-01 7.08493829e-01
1.46776366e+00 -5.03356516e-01 6.86712921e-01 -6.39809847e-01
1.13389254e+00 -3.71529758e-02 -5.14633238e-01 1.79143801e-01
4.55358058e-01 8.85863006e-02 2.72116244e-01 4.30943012e-01
7.97253549e-02 -2.33534142e-01 6.52193725e-01 1.43057144e+00
-8.16029191e-01 -1.23976424e-01 -7.45472789e-01 1.69738188e-01
-1.93850791e+00 -1.40426958e+00 -2.97675550e-01 2.07487321e+00
7.37864554e-01 3.81936096e-02 -5.25762420e-03 -6.59492791e-01
5.06773651e-01 6.64327085e-01 -6.20563269e-01 -3.46885800e-01
-3.98022216e-03 9.08236027e-01 7.36244977e-01 1.02071214e+00
-7.66499758e-01 1.15669203e+00 8.17904663e+00 5.56911528e-01
-7.91048408e-01 4.91559923e-01 5.19043386e-01 -5.48136115e-01
-8.37522149e-01 3.46968383e-01 -6.10003948e-01 3.22667480e-01
1.47865546e+00 2.39319161e-01 1.24908602e+00 6.37579679e-01
-5.83408251e-02 4.35585603e-02 -9.32474673e-01 6.20627999e-01
3.96825485e-02 -1.99814129e+00 -5.51204570e-02 3.90621156e-01
1.10463464e+00 7.18304515e-01 1.01902513e-02 3.31514329e-01
1.28812110e+00 -1.28641832e+00 3.79140466e-01 1.10905004e+00
6.18384719e-01 -5.82903922e-01 3.57843369e-01 1.32307291e-01
-9.16613340e-01 6.06568277e-01 -8.14272285e-01 -2.55611688e-01
1.14422418e-01 1.06352115e+00 -2.97212631e-01 1.12847649e-01
2.71802366e-01 6.28360629e-01 -5.64649887e-03 2.29830340e-01
2.86279824e-02 9.04209077e-01 -6.46133244e-01 -5.59347987e-01
2.57140547e-01 -7.73813903e-01 7.15890646e-01 7.43090034e-01
1.67226702e-01 1.92344770e-01 3.54779273e-01 1.27833223e+00
-4.15148377e-01 -2.83396095e-01 -9.07377899e-01 -6.16754234e-01
2.30504707e-01 1.40687931e+00 -6.89152837e-01 -4.00056064e-01
-5.95443510e-02 8.27741265e-01 9.17616308e-01 7.87347913e-01
-5.33699512e-01 -5.37285134e-02 5.10531306e-01 3.12712878e-01
4.58718270e-01 -6.20191097e-01 4.23895210e-01 -1.56744337e+00
-4.88373816e-01 -7.66241312e-01 -1.45481452e-01 -6.62146389e-01
-1.51269531e+00 4.53317136e-01 -2.67009079e-01 1.05780594e-01
-3.50576282e-01 -7.35502124e-01 -5.43732762e-01 9.77333486e-01
-7.05428660e-01 -1.05570543e+00 6.69106469e-02 5.90096712e-01
-4.14260089e-01 -1.74480870e-01 9.91512060e-01 -2.98414141e-01
-5.40323913e-01 -7.97788426e-02 5.97086072e-01 -1.94802403e-01
3.78433704e-01 -1.53432643e+00 6.50331080e-01 5.12581468e-01
4.90137219e-01 9.70905185e-01 8.42706025e-01 -9.62790489e-01
-2.21863699e+00 -7.65113235e-01 3.64048392e-01 -5.90022802e-01
1.10306096e+00 -7.74706066e-01 -9.14009154e-01 9.53066587e-01
2.50476807e-01 6.73707604e-01 4.64712024e-01 6.48954868e-01
-5.82719564e-01 3.12305242e-01 -9.98547077e-01 1.81526989e-01
1.24207616e+00 -8.94201100e-01 -1.94110915e-01 8.09842706e-01
5.05569696e-01 -3.79005700e-01 -9.64549243e-01 -2.89053738e-01
4.07843709e-01 -8.24805439e-01 9.35611963e-01 -8.88880610e-01
4.11382377e-01 -3.94600704e-02 -1.88569367e-01 -1.54158664e+00
-9.77651298e-01 -9.55987155e-01 -6.51825905e-01 3.56665701e-01
2.39189148e-01 -6.32398248e-01 8.92473400e-01 4.72566396e-01
1.30809143e-01 -5.91046989e-01 -8.51783037e-01 -3.34560037e-01
1.93006337e-01 -2.64448941e-01 2.62991995e-01 6.86653912e-01
-1.24565633e-02 5.59247315e-01 -5.98215878e-01 2.93356836e-01
1.32158709e+00 1.49177507e-01 4.51141268e-01 -1.35685515e+00
-8.18911493e-01 -7.40441605e-02 -4.06016201e-01 -7.65699565e-01
8.35149109e-01 -1.37923777e+00 7.66238421e-02 -1.38258493e+00
7.89508164e-01 -2.17589960e-01 -3.57489258e-01 1.27508268e-01
1.20719627e-01 3.94520998e-01 -3.23580325e-01 4.20842171e-02
-8.32262754e-01 7.22494721e-01 1.08849931e+00 -2.76193649e-01
2.14959025e-01 -4.41701025e-01 -5.38602948e-01 5.48083603e-01
2.11584896e-01 -5.97159386e-01 -2.48773154e-02 1.49460688e-01
6.29713714e-01 2.57149011e-01 5.70428371e-01 -8.71886611e-01
1.71250135e-01 -2.33396858e-01 3.13013226e-01 -2.03674838e-01
3.48421901e-01 -1.00394264e-02 5.36270142e-01 4.05484617e-01
-1.75394163e-01 -5.62648356e-01 -3.00067097e-01 1.03438735e+00
1.78710550e-01 -1.75949112e-01 6.98260009e-01 -4.96949852e-01
-7.18101636e-02 7.48204172e-01 -5.21887064e-01 8.21434930e-02
4.33968127e-01 5.12448788e-01 -2.40386829e-01 -5.31977713e-01
-1.17174113e+00 -9.62020606e-02 6.66105330e-01 -3.68824661e-01
2.97803730e-01 -1.43821871e+00 -4.63872880e-01 2.14126453e-01
-4.19076771e-01 -2.43557945e-01 3.91795099e-01 8.22778285e-01
-4.63912219e-01 4.30799425e-01 1.10449884e-02 -3.58650774e-01
-6.33064687e-01 4.80460167e-01 4.14163530e-01 -6.18617594e-01
-4.53389585e-01 9.10350382e-01 7.88800512e-03 -5.64395845e-01
-4.32873249e-01 -4.84163612e-01 3.84672761e-01 -3.80622327e-01
2.35602900e-01 2.33786806e-01 -1.22329339e-01 -6.61169887e-01
-2.50599265e-01 2.28117049e-01 -5.95012307e-02 -1.15469858e-01
1.64650297e+00 2.22918496e-01 -8.07142735e-01 6.90711975e-01
9.37223315e-01 2.13273019e-01 -1.08697784e+00 -3.48857492e-01
-6.07732415e-01 8.35923031e-02 4.65290904e-01 -1.98282897e-01
-1.13898885e+00 7.48602390e-01 -7.55687356e-02 6.22709632e-01
8.74952301e-02 4.86409843e-01 8.30042720e-01 9.65109825e-01
7.93418825e-01 -8.86959195e-01 -6.00626767e-02 9.35269654e-01
5.17084301e-01 -1.12255573e+00 1.01245716e-01 -1.54397741e-01
-2.29895055e-01 1.07731485e+00 1.50924325e-01 -8.58797371e-01
9.67562616e-01 4.29714411e-01 -8.93355072e-01 -7.56510615e-01
-9.66931760e-01 1.60666615e-01 4.20583308e-01 4.93443638e-01
4.50452209e-01 4.71339166e-01 2.58319139e-01 4.62217927e-01
-1.09452017e-01 -1.45255566e-01 8.10455322e-01 3.52605522e-01
-6.40634298e-01 -1.07130599e+00 -1.93218753e-01 9.88576353e-01
-7.70930573e-02 -5.87860703e-01 -1.78351954e-01 3.30513388e-01
-4.66932744e-01 2.39881799e-01 1.46632791e-01 -1.81824997e-01
-4.08502728e-01 2.92865038e-01 1.23170555e+00 -9.04798985e-01
-2.05793932e-01 -2.07695007e-01 1.37107030e-01 -1.01238167e+00
-4.48362052e-01 -7.58547723e-01 -1.17404509e+00 -9.55272973e-01
-2.98270702e-01 4.11413699e-01 3.61532092e-01 1.05382860e+00
3.36432368e-01 4.55550611e-01 1.80885822e-01 -1.43733573e+00
-8.01201642e-01 -9.94198859e-01 -1.13194621e+00 7.34590366e-02
2.86583990e-01 -8.01206648e-01 -5.62654495e-01 -1.47725850e-01] | [5.729547500610352, 4.834125518798828] |
9c1bfdaa-aab1-4a4a-b120-1f05f8443fd7 | value-prediction-network | 1707.03497 | null | http://arxiv.org/abs/1707.03497v2 | http://arxiv.org/pdf/1707.03497v2.pdf | Value Prediction Network | This paper proposes a novel deep reinforcement learning (RL) architecture,
called Value Prediction Network (VPN), which integrates model-free and
model-based RL methods into a single neural network. In contrast to typical
model-based RL methods, VPN learns a dynamics model whose abstract states are
trained to make option-conditional predictions of future values (discounted sum
of rewards) rather than of future observations. Our experimental results show
that VPN has several advantages over both model-free and model-based baselines
in a stochastic environment where careful planning is required but building an
accurate observation-prediction model is difficult. Furthermore, VPN
outperforms Deep Q-Network (DQN) on several Atari games even with
short-lookahead planning, demonstrating its potential as a new way of learning
a good state representation. | ['Satinder Singh', 'Junhyuk Oh', 'Honglak Lee'] | 2017-07-11 | value-prediction-network-1 | http://papers.nips.cc/paper/7192-value-prediction-network | http://papers.nips.cc/paper/7192-value-prediction-network.pdf | neurips-2017-12 | ['value-prediction'] | ['computer-code'] | [-2.89044768e-01 3.52649003e-01 -8.30187082e-01 -2.61275887e-01
-7.13538110e-01 -3.89502406e-01 9.61939156e-01 -1.35183945e-01
-7.35605597e-01 1.29463494e+00 4.75616932e-01 -4.49537188e-01
-1.38885632e-01 -1.05753350e+00 -6.92204237e-01 -5.50857842e-01
-5.08570015e-01 8.51502419e-01 1.79870725e-01 -7.01814830e-01
3.64062190e-01 3.79221439e-01 -1.19848955e+00 -9.73778963e-02
4.43468511e-01 9.18985665e-01 2.06543460e-01 7.18678176e-01
5.53107671e-02 1.55186820e+00 -4.93340015e-01 -1.86976176e-02
3.74224782e-01 -5.08523822e-01 -9.83998716e-01 -3.48376632e-01
-4.79657382e-01 -1.01589310e+00 -6.48078620e-01 6.76207244e-01
3.03813845e-01 7.98845589e-01 3.57255995e-01 -1.23166978e+00
-4.30036128e-01 7.10789919e-01 -3.03067151e-03 2.09352165e-01
2.51846373e-01 9.05857086e-01 1.01254976e+00 -1.48033321e-01
6.89045668e-01 1.48631716e+00 2.42944956e-01 9.17429090e-01
-1.44358039e+00 -3.59863430e-01 3.94849628e-01 2.33026177e-01
-4.57815409e-01 -1.05851449e-01 5.76988757e-01 3.51648927e-02
1.55182481e+00 -2.65583634e-01 1.18679440e+00 1.53049195e+00
7.32510269e-01 1.13143623e+00 1.28327179e+00 -1.05590962e-01
6.86538041e-01 -5.84571421e-01 -3.50601882e-01 3.76992851e-01
-2.03594908e-01 1.23210108e+00 -4.47098941e-01 -1.68838248e-01
1.29740739e+00 1.56133175e-01 2.41723463e-01 -7.14696229e-01
-1.41440654e+00 1.03289950e+00 5.26827991e-01 -1.46743476e-01
-7.87421644e-01 9.91726935e-01 2.82540470e-01 6.92474365e-01
3.87392677e-02 8.60912621e-01 -6.93528056e-01 -7.42386162e-01
-6.28198326e-01 8.58862162e-01 7.15504766e-01 5.79605222e-01
6.70153022e-01 7.55566955e-01 -5.98770142e-01 1.20472550e-01
2.49376372e-01 5.29560626e-01 8.47389877e-01 -1.66857421e+00
3.60549062e-01 1.24666281e-01 7.97157764e-01 -1.87039122e-01
-6.27673388e-01 -2.76901096e-01 -2.08050057e-01 7.68692076e-01
2.62597680e-01 -6.01591289e-01 -1.06336808e+00 1.87879717e+00
-1.11248093e-02 1.94051921e-01 6.08425438e-01 7.16015935e-01
2.93562084e-01 8.94690514e-01 1.85184613e-01 -3.05579126e-01
5.04159629e-01 -1.19347584e+00 -7.48811722e-01 -3.21062565e-01
6.28735602e-01 1.64065450e-01 9.58558500e-01 5.13547659e-01
-1.46618891e+00 -4.59686190e-01 -8.95023882e-01 2.72198290e-01
-5.51298400e-03 -5.86581409e-01 9.39542472e-01 1.42741129e-01
-1.45337689e+00 1.26709807e+00 -1.26243889e+00 1.12351626e-02
2.98336595e-01 5.61227858e-01 3.84821445e-02 1.15690090e-01
-1.30058181e+00 1.31314456e+00 6.59947455e-01 -2.25003123e-01
-1.90572286e+00 -1.99526384e-01 -9.70354617e-01 2.10656464e-01
7.56850898e-01 -4.68849450e-01 2.20661163e+00 -8.59729648e-01
-2.28445220e+00 -1.00682773e-01 2.03840956e-01 -1.10682356e+00
3.70264620e-01 -2.97919124e-01 -2.75284290e-01 -1.25902653e-01
-1.79423511e-01 8.81178319e-01 6.27730370e-01 -1.03237486e+00
-5.57920218e-01 1.45992577e-01 5.47916889e-01 6.36564553e-01
4.75847393e-01 -3.31035852e-01 2.44774237e-01 -2.31562689e-01
-3.05717856e-01 -9.67500985e-01 -8.52830648e-01 -1.71332076e-01
5.17666154e-02 -4.17019457e-01 2.65275806e-01 -1.69690996e-01
8.76838326e-01 -1.49260235e+00 4.92294133e-02 -6.52543642e-03
-1.61158428e-01 2.37651333e-01 -5.17953038e-01 9.77502048e-01
1.69007927e-01 -1.96268156e-01 9.44015607e-02 -1.39986277e-01
2.38389134e-01 8.27250063e-01 -6.13219082e-01 6.14062697e-02
1.58746764e-01 1.42186999e+00 -1.50074708e+00 -8.36659446e-02
3.67657393e-01 -1.39645875e-01 -4.92341489e-01 4.54133838e-01
-1.13736951e+00 6.60023510e-01 -5.96024990e-01 3.46981764e-01
4.81739379e-02 -1.31275252e-01 5.24418414e-01 8.32480073e-01
-1.27723485e-01 7.78902769e-01 -9.63038385e-01 1.70601511e+00
-3.14549088e-01 2.06990540e-01 -4.52933580e-01 -8.68448973e-01
8.31306279e-01 3.52353573e-01 3.87102753e-01 -1.21480334e+00
1.62592426e-01 -6.11349419e-02 8.75075534e-02 -1.03133261e-01
7.32014418e-01 -4.43109334e-01 -2.51302034e-01 7.22048104e-01
1.97642148e-01 -4.55312252e-01 1.02562651e-01 7.06720948e-02
1.20257568e+00 9.10589278e-01 5.32210767e-01 2.15463769e-02
-6.23339266e-02 1.76711485e-01 8.78196180e-01 1.15110469e+00
-5.53283572e-01 -4.53887098e-02 8.66546512e-01 -7.64313757e-01
-8.18759620e-01 -1.30650139e+00 3.78698677e-01 1.17863786e+00
7.61495903e-02 -3.52142334e-01 -2.01545253e-01 -8.09580624e-01
-2.60995273e-02 1.31568885e+00 -7.46316791e-01 -2.34403268e-01
-6.78724527e-01 -3.56447428e-01 1.88249990e-01 8.42449546e-01
4.47986782e-01 -1.79483175e+00 -9.31254506e-01 6.52344286e-01
-1.60994027e-02 -3.35144192e-01 -1.69213831e-01 5.43184817e-01
-1.24704146e+00 -8.68397415e-01 -4.41661417e-01 -3.87665719e-01
-3.75551544e-02 -1.14547104e-01 1.43681860e+00 -1.05816163e-01
6.62863731e-01 6.27385557e-01 -2.45354205e-01 -4.26327497e-01
-5.66152632e-01 -3.05220872e-01 2.02735454e-01 -8.00513506e-01
1.24179594e-01 -5.50988615e-01 -7.09122658e-01 -3.71042080e-02
-5.13895571e-01 -6.96866512e-02 5.61718941e-01 1.13922131e+00
7.00652122e-01 -2.43894994e-01 6.71771646e-01 -5.19512355e-01
1.01733041e+00 -5.17105758e-01 -1.01457393e+00 6.61485344e-02
-9.42831099e-01 4.96110022e-01 7.70385087e-01 -3.80128205e-01
-1.31920612e+00 -1.39580220e-01 -1.43588141e-01 -1.78571492e-01
-8.02056119e-02 4.67026234e-01 3.96060109e-01 3.14659327e-01
5.45357585e-01 5.32042086e-01 3.81177902e-01 -1.99601382e-01
4.17013854e-01 -3.02752793e-01 2.82584250e-01 -6.00153208e-01
4.08934116e-01 6.45561293e-02 1.59847945e-01 4.71506547e-03
-6.99573040e-01 6.88372329e-02 -2.27789879e-01 -1.74033910e-01
7.37332106e-01 -1.01865983e+00 -1.27634490e+00 2.51689404e-01
-7.25857735e-01 -1.35893083e+00 -9.92626846e-01 6.41516805e-01
-1.53867865e+00 9.00881644e-03 -8.71723711e-01 -1.02815449e+00
3.11510507e-02 -1.13265729e+00 4.98555392e-01 3.91919672e-01
4.42730114e-02 -1.15439272e+00 6.68982804e-01 -2.44731188e-01
4.90314484e-01 2.23663360e-01 7.13533163e-01 -6.55133486e-01
-8.10658455e-01 1.53891936e-01 3.56989264e-01 1.58426955e-01
-3.21591288e-01 -4.98742163e-01 -5.19538343e-01 -4.51922446e-01
-6.96415752e-02 -1.11044419e+00 7.77688384e-01 6.82182133e-01
9.73447680e-01 -4.78815407e-01 -1.21675178e-01 2.25655720e-01
1.53578186e+00 5.41232526e-01 7.92670369e-01 6.71408534e-01
-3.36020440e-02 4.20191213e-02 1.03606820e+00 8.07357252e-01
7.15327501e-01 5.57737887e-01 1.13134861e+00 2.66231686e-01
3.09184462e-01 -7.03040898e-01 8.05479765e-01 1.60286725e-01
-3.10295373e-01 -1.57993644e-01 -6.79860115e-01 4.40217465e-01
-2.26630902e+00 -1.33980501e+00 4.20103163e-01 2.28423142e+00
7.88400948e-01 3.46507132e-01 5.40525556e-01 -4.87692893e-01
9.99614689e-03 4.03612375e-01 -1.24039078e+00 -8.59744310e-01
1.98173940e-01 3.95926774e-01 4.88909602e-01 4.89429533e-01
-8.43918979e-01 1.25314343e+00 7.82189560e+00 5.92560351e-01
-7.44263828e-01 -8.62862021e-02 5.02145469e-01 -4.14687097e-01
-4.00736779e-01 1.60714805e-01 -6.48211122e-01 2.79940337e-01
1.49235213e+00 -2.34143123e-01 8.25182080e-01 1.08377969e+00
3.16196203e-01 -3.63132745e-01 -1.07031167e+00 5.18069267e-01
-5.99237740e-01 -1.55467165e+00 -1.45817816e-01 2.62174070e-01
8.08769047e-01 5.38141131e-01 2.10888743e-01 1.25133669e+00
1.64853668e+00 -1.20206928e+00 7.67674387e-01 6.89444721e-01
3.53557944e-01 -1.04133797e+00 6.96232140e-01 7.68978417e-01
-8.04700732e-01 -6.72799766e-01 -5.17149508e-01 -6.64114296e-01
3.38033825e-01 -4.33399290e-01 -7.78474689e-01 2.40383565e-01
3.70417953e-01 6.99183285e-01 -1.59699598e-03 7.70819426e-01
-6.43130898e-01 6.82663441e-01 5.83117828e-02 -2.53321707e-01
9.77203488e-01 -1.94266483e-01 2.88342506e-01 5.71152806e-01
1.74200758e-01 1.38731897e-01 5.77296257e-01 8.55048001e-01
1.16920099e-01 -4.52287674e-01 -7.10310042e-01 -3.80331539e-02
3.49446148e-01 8.02642107e-01 -4.36329365e-01 -3.27315032e-01
-2.08757773e-01 6.82676792e-01 7.05376565e-01 3.88530016e-01
-8.33067179e-01 8.60318244e-02 8.84743690e-01 -2.64333814e-01
5.42109907e-01 -4.51259345e-01 2.31188595e-01 -1.05576980e+00
-5.81333041e-01 -9.02916670e-01 2.64967889e-01 -8.99920464e-01
-7.58926868e-01 4.51108694e-01 1.23496890e-01 -1.33956301e+00
-1.31316233e+00 -4.78270382e-01 -7.87085950e-01 6.50136828e-01
-1.93200326e+00 -8.05098176e-01 4.92021829e-01 6.31824672e-01
7.60280728e-01 -1.59230947e-01 1.16108787e+00 -8.73621523e-01
-2.18152016e-01 1.52631879e-01 5.07815301e-01 3.19511426e-04
2.53504038e-01 -1.65168595e+00 7.83499360e-01 5.06995320e-01
-7.49159306e-02 2.99343169e-01 7.15517223e-01 -7.00307488e-01
-1.40527511e+00 -8.43332529e-01 3.68716300e-01 -5.59889376e-01
6.18790209e-01 1.42609522e-01 -5.75168252e-01 1.00180757e+00
5.19768775e-01 -5.01707755e-02 4.03951555e-01 1.13272205e-01
9.89920422e-02 2.14152679e-01 -1.12260628e+00 6.97354555e-01
9.01803911e-01 -1.33910626e-01 -9.09398615e-01 1.91130385e-01
9.53326523e-01 -5.35180748e-01 -7.91427970e-01 2.15416849e-02
4.60626841e-01 -1.24944997e+00 9.13219333e-01 -1.12279415e+00
4.20987397e-01 1.72965750e-01 -4.64982800e-02 -1.83491921e+00
-5.06765127e-01 -1.06053650e+00 -7.47165799e-01 4.27964360e-01
2.47982323e-01 -6.27253473e-01 9.74098504e-01 7.80530393e-01
-2.15362802e-01 -9.51599121e-01 -9.15040076e-01 -1.23237038e+00
4.55526531e-01 -4.34915036e-01 9.50633168e-01 3.89580190e-01
1.55326918e-01 1.04886405e-01 -7.66952634e-01 -4.10959184e-01
4.19676095e-01 1.60914168e-01 6.93332791e-01 -9.14584100e-01
-5.95988631e-01 -3.17063153e-01 3.84908542e-02 -1.31295145e+00
3.13921720e-01 -5.09288073e-01 2.87737042e-01 -1.84716463e+00
-1.24547467e-01 -3.35341305e-01 -7.28597641e-01 6.80344641e-01
2.22285524e-01 -5.05598009e-01 5.65080345e-01 9.96982306e-02
-1.02472460e+00 1.18244171e+00 1.74750710e+00 2.06720680e-01
-4.95057821e-01 3.10185879e-01 -3.95540476e-01 6.11467659e-01
1.05119085e+00 -4.69993919e-01 -8.35532963e-01 -1.94796428e-01
3.54424834e-01 9.48800921e-01 2.36269876e-01 -9.94701505e-01
-8.57503116e-02 -1.00166750e+00 4.23896462e-01 -6.65330648e-01
6.02423966e-01 -4.84376818e-01 -2.54909873e-01 8.75681043e-01
-6.80834115e-01 4.30135489e-01 1.70814812e-01 9.80643511e-01
5.53206690e-02 -1.02258615e-01 5.96280932e-01 -6.72966063e-01
-1.18062627e+00 4.19423550e-01 -8.89994860e-01 1.07020564e-01
9.68048632e-01 -1.48587167e-01 -3.44579130e-01 -9.37412739e-01
-8.41227531e-01 6.70598090e-01 3.16458225e-01 2.38368124e-01
8.34958911e-01 -1.35134888e+00 -3.21160495e-01 6.63567036e-02
-2.05703110e-01 -1.26096725e-01 1.22915804e-01 3.62959266e-01
-2.43361264e-01 5.37297487e-01 -6.05491877e-01 4.94744675e-03
-2.69013882e-01 8.18291068e-01 4.39320534e-01 -1.12974882e+00
-6.26543105e-01 6.22835279e-01 -1.40110686e-01 -7.61332512e-01
2.46415392e-01 -4.56648558e-01 -3.63022208e-01 -3.39714468e-01
5.36590457e-01 3.25604469e-01 -5.61614335e-01 -7.56518096e-02
1.44455031e-01 -2.69094825e-01 -6.58733472e-02 -6.02601230e-01
1.59357548e+00 -1.90034043e-02 4.79721159e-01 6.54867589e-01
3.20787340e-01 -9.07476127e-01 -2.24200702e+00 -2.46404201e-01
1.13129511e-01 -2.91998804e-01 7.76900351e-02 -1.19702387e+00
-6.14363611e-01 7.83811986e-01 4.25463617e-01 2.92118844e-02
7.51071393e-01 -4.12744194e-01 8.99063110e-01 9.98679578e-01
9.10792828e-01 -1.46202171e+00 5.30592442e-01 1.03140152e+00
8.36062729e-01 -1.34859204e+00 -1.28799513e-01 9.40086722e-01
-1.28238058e+00 9.62340593e-01 9.40756261e-01 -4.89589542e-01
4.45901722e-01 -1.35004178e-01 1.74054503e-02 1.90297831e-02
-1.54357350e+00 -5.99186063e-01 -1.61528364e-01 1.00499260e+00
2.85174549e-02 1.89528689e-01 1.50204331e-01 4.22764480e-01
-1.34123757e-01 3.16033721e-01 8.59367549e-01 1.27016664e+00
-6.88926816e-01 -1.30525732e+00 2.97468212e-02 4.84878182e-01
-1.89050958e-01 7.23281577e-02 1.29857674e-01 9.12457347e-01
-5.17306149e-01 9.04307306e-01 1.62969500e-01 -3.00049663e-01
8.47897977e-02 1.23395734e-01 6.50529325e-01 -8.01053882e-01
-8.24861646e-01 -1.02413669e-01 1.97262228e-01 -1.29302812e+00
-1.68327883e-01 -5.48113704e-01 -1.56438172e+00 -3.74073088e-01
8.14042091e-02 -1.63763706e-02 3.54425192e-01 9.79391992e-01
2.77116269e-01 4.70507652e-01 7.28849888e-01 -1.09202182e+00
-1.48470414e+00 -6.59170449e-01 -7.76367068e-01 1.09395862e-01
6.32524788e-01 -8.59630108e-01 8.67491141e-02 -6.04543686e-01] | [4.0968546867370605, 1.681186318397522] |
11b1d491-0836-4fef-bf34-04fffc294098 | analysis-of-critical-parameters-of-satellite | 1905.07476 | null | https://arxiv.org/abs/1905.07476v1 | https://arxiv.org/pdf/1905.07476v1.pdf | Analysis of critical parameters of satellite stereo image for 3D reconstruction and mapping | Although nowadays advanced dense image matching (DIM) algorithms are able to produce LiDAR (Light Detection And Ranging) comparable dense point clouds from satellite stereo images, the accuracy and completeness of such point clouds heavily depend on the geometric parameters of the satellite stereo images. The intersection angle between two images are normally seen as the most important one in stereo data acquisition, as the state-of-the-art DIM algorithms work best on narrow baseline (smaller intersection angle) stereos (E.g. Semi-Global Matching regards 15-25 degrees as good intersection angle). This factor is in line with the traditional aerial photogrammetry configuration, as the intersection angle directly relates to the base-high ratio and texture distortion in the parallax direction, thus both affecting the horizontal and vertical accuracy. However, our experiments found that even with very similar (and good) intersection angles, the same DIM algorithm applied on different stereo pairs (of the same area) produced point clouds with dramatically different accuracy as compared to the ground truth LiDAR data. This raises a very practical question that is often asked by practitioners: what factors constitute a good satellite stereo pair, such that it produces accurate and optimal results for mapping purpose? In this work, we provide a comprehensive analysis on this matter by performing stereo matching over 1,000 satellite stereo pairs with different acquisition parameters including their intersection angles, off-nadir angles, sun elevation & azimuth angles, as well as time differences, thus to offer a thorough answer to this question. This work will potentially provide a valuable reference to researchers working on multi-view satellite image reconstruction, as well as industrial practitioners minimizing costs for high-quality large-scale mapping. | ['Rongjun Qin'] | 2019-05-17 | null | null | null | null | ['stereo-matching'] | ['computer-vision'] | [ 1.83961481e-01 -3.75775665e-01 2.59270612e-02 -3.15014064e-01
-3.98087919e-01 -6.32439375e-01 6.60189927e-01 1.75597459e-01
-4.87261415e-01 7.00635612e-01 -3.05062711e-01 -2.88272500e-01
-3.45600277e-01 -1.30525637e+00 -4.25843298e-01 -6.42145276e-01
3.67248803e-02 1.14966261e+00 4.68185753e-01 -6.33400440e-01
3.29078436e-01 9.84092355e-01 -2.08866215e+00 -4.64733034e-01
1.12001836e+00 7.52770960e-01 4.92126018e-01 2.12345168e-01
-1.07314974e-01 -2.79014051e-01 -1.33326888e-01 -4.55478936e-01
5.77007473e-01 3.35687958e-02 -6.70224845e-01 9.56790596e-02
8.34132969e-01 -3.44233871e-01 2.19607800e-01 1.17862380e+00
3.38231325e-01 -2.04592392e-01 5.34397781e-01 -9.62519526e-01
7.22372904e-02 -8.61020684e-02 -6.25292003e-01 7.13945329e-02
4.72725719e-01 6.77259117e-02 8.30764234e-01 -7.70062923e-01
6.15786314e-01 9.30209577e-01 7.17479825e-01 -2.41159260e-01
-1.46321046e+00 -6.46289945e-01 -3.59598935e-01 3.55933249e-01
-1.79909170e+00 -1.55755147e-01 4.47691202e-01 -4.59790081e-01
4.68019307e-01 4.92448360e-01 9.49539721e-01 3.91011924e-01
1.16217181e-01 -2.42211521e-01 1.30489326e+00 -5.00311434e-01
1.43702820e-01 3.35083812e-01 -2.49952316e-01 1.83285419e-02
4.82910156e-01 3.30903590e-01 -1.62929937e-01 -1.47201836e-01
7.98219800e-01 5.61747253e-02 -6.17788553e-01 -6.75073385e-01
-1.04612029e+00 8.92758071e-01 6.13617122e-01 5.50161123e-01
-3.22128713e-01 -4.34439570e-01 5.89051917e-02 3.08169514e-01
1.94764689e-01 6.43952712e-02 -2.66562104e-01 1.20808497e-01
-1.03741360e+00 4.36018765e-01 5.54733634e-01 8.36889803e-01
1.37265050e+00 -7.69615769e-02 5.40160179e-01 6.33491397e-01
2.44558781e-01 1.08475351e+00 7.40804896e-02 -7.39310145e-01
4.87590343e-01 6.62747264e-01 1.71882942e-01 -1.25140667e+00
-3.54677260e-01 -4.81983900e-01 -8.68978620e-01 4.24124539e-01
4.51204121e-01 3.51290464e-01 -3.23356569e-01 1.16331911e+00
5.04228652e-01 -3.51491719e-02 -2.38141622e-02 1.08300674e+00
3.61401647e-01 4.14760053e-01 -2.90267169e-01 -2.36476243e-01
1.43277228e+00 1.36354221e-02 -2.72621363e-01 -3.70743901e-01
4.30728465e-01 -1.15789652e+00 9.52723026e-01 2.31416300e-01
-7.58395076e-01 -5.04866660e-01 -9.09437656e-01 2.78906018e-01
-3.30564320e-01 -1.58990383e-01 3.56441885e-01 5.17470419e-01
-1.14690816e+00 6.93547487e-01 -3.31656992e-01 -7.18051434e-01
-1.76155433e-01 3.17207575e-01 -4.92822379e-01 -1.29419252e-01
-1.13218689e+00 1.20032048e+00 2.17758626e-01 1.95973963e-01
-1.56862304e-01 -6.91909432e-01 -6.58968627e-01 -1.00446790e-01
2.64140904e-01 -6.84330881e-01 8.45701873e-01 -7.94875383e-01
-1.03192008e+00 1.33256972e+00 -2.01942399e-01 -4.95773405e-01
5.68292618e-01 2.81187929e-02 -4.11897898e-01 8.20721462e-02
2.08703950e-01 4.67882007e-01 4.65250105e-01 -1.54816329e+00
-6.06249928e-01 -9.22097862e-01 -1.67451739e-01 4.14955139e-01
1.82889864e-01 8.58509261e-03 -8.15803632e-02 7.23010441e-03
6.01849675e-01 -8.95267844e-01 -1.26673132e-01 7.77909830e-02
-2.83699087e-03 1.30780354e-01 9.77713287e-01 -4.35630739e-01
7.57291675e-01 -2.00428152e+00 -5.97766936e-02 3.52975607e-01
-3.42267096e-01 1.98954493e-01 2.57839501e-01 5.79806566e-01
-1.26014635e-01 -7.54191503e-02 -2.63078123e-01 -1.57796405e-02
-3.49515498e-01 2.91571826e-01 -3.83311033e-01 7.85552263e-01
-2.93919146e-01 3.79604667e-01 -7.05167651e-01 -5.84514141e-01
7.23251045e-01 5.25358856e-01 -5.91946915e-02 -5.46028204e-02
1.40915692e-01 4.65072721e-01 -3.11167181e-01 5.37899077e-01
1.26041853e+00 3.69608283e-01 1.20626815e-01 -3.53780329e-01
-7.63979137e-01 1.78469047e-02 -1.52590275e+00 1.32352090e+00
-5.17176926e-01 6.71030700e-01 2.77741253e-01 -6.79221034e-01
1.37298942e+00 2.98118532e-01 3.65902245e-01 -1.02135944e+00
-3.69929746e-02 5.93364000e-01 -3.36906463e-01 -1.68733254e-01
7.75674105e-01 -6.57656372e-01 3.59609187e-01 -1.55178994e-01
-4.53857601e-01 -9.27834928e-01 7.88574219e-02 -1.87590674e-01
2.04847887e-01 -1.49790078e-01 6.63416564e-01 -5.42918444e-01
8.09884548e-01 2.03627750e-01 2.97787517e-01 2.76257485e-01
2.84928888e-01 8.15573931e-01 1.53734550e-01 -6.26120210e-01
-1.34200919e+00 -9.30638611e-01 -8.92753959e-01 1.80841479e-02
5.83885908e-01 -9.11008194e-02 -4.33433890e-01 1.37975603e-01
1.29375011e-01 5.54984689e-01 -1.17660522e-01 3.87854517e-01
-3.53645325e-01 -5.99432290e-01 9.53040570e-02 2.35886034e-02
6.34325922e-01 -5.27966082e-01 -7.42850184e-01 -4.40387204e-02
-2.50993967e-01 -1.17384648e+00 2.11388677e-01 -1.21120282e-01
-1.29909992e+00 -1.20426357e+00 -4.48885739e-01 -2.99199611e-01
4.39005703e-01 8.06101680e-01 1.29885590e+00 3.42099130e-01
-7.70302862e-02 7.15407953e-02 -2.87601769e-01 -2.58331686e-01
-2.19790012e-01 -1.07659757e-01 -2.01287922e-02 -1.73441544e-01
4.31209475e-01 -9.21460211e-01 -5.11483550e-01 8.61877799e-01
-9.39148486e-01 -4.11583483e-02 3.51956964e-01 4.65541065e-01
6.52705669e-01 1.15040459e-01 -2.86552370e-01 -4.25372541e-01
-2.28131679e-03 -2.14817956e-01 -1.28145742e+00 -2.53412928e-02
-7.13326693e-01 -2.93394625e-01 4.59279835e-01 4.53908652e-01
-6.55109763e-01 4.97243069e-02 -2.09132269e-01 -1.31911501e-01
-5.33665180e-01 3.08309913e-01 -9.15221572e-02 -4.15187389e-01
7.19257951e-01 3.95232528e-01 5.18068708e-02 -5.27666569e-01
-8.60006288e-02 7.21066952e-01 5.59454441e-01 -3.21649700e-01
1.26873255e+00 1.09731185e+00 5.69704235e-01 -1.29009593e+00
-3.50483567e-01 -8.80662441e-01 -1.13888121e+00 -2.60511667e-01
7.16009855e-01 -9.30961788e-01 -4.76669252e-01 2.65723735e-01
-1.04758966e+00 1.26981437e-01 -2.44217873e-01 5.28235912e-01
-3.92975867e-01 7.05243170e-01 1.89203143e-01 -8.79559815e-01
-1.43505424e-01 -1.29266810e+00 1.40487492e+00 1.81444854e-01
-3.55789764e-03 -9.43180561e-01 3.16135943e-01 3.86477143e-01
3.37932646e-01 2.74207115e-01 6.42871082e-01 1.65592715e-01
-8.63711238e-01 -1.35249078e-01 -2.87791669e-01 5.01506552e-02
-2.15315819e-03 2.28476018e-01 -8.42566550e-01 -3.57974708e-01
4.11110073e-02 8.16090703e-02 5.05546272e-01 3.84194374e-01
4.50279951e-01 2.38207370e-01 -3.21031123e-01 7.91878641e-01
2.15308642e+00 -9.66994762e-02 1.11812556e+00 8.25693250e-01
3.11782956e-01 7.11037219e-01 9.82045591e-01 3.87895674e-01
3.15083832e-01 1.35697031e+00 9.66855526e-01 -2.21176550e-01
2.27581769e-01 -1.77435782e-02 1.52332513e-02 3.04832906e-01
-5.56052566e-01 3.50872189e-01 -1.14055276e+00 4.58612442e-01
-1.32700562e+00 -9.93754625e-01 -8.53628278e-01 2.98201489e+00
2.81278074e-01 -1.28123015e-01 2.73600426e-02 5.15835822e-01
7.06639886e-01 1.79232955e-01 1.19446456e-01 -3.08684349e-01
-3.83384556e-01 1.33444518e-01 8.82797420e-01 7.38501668e-01
-7.30204821e-01 7.01325715e-01 5.28804827e+00 6.21475339e-01
-1.45633912e+00 -6.60080835e-03 -1.55431956e-01 3.22333246e-01
-5.67376673e-01 4.67745274e-01 -1.02331412e+00 4.30943578e-01
8.01786304e-01 -1.98256373e-01 2.21242875e-01 7.22012579e-01
5.34308374e-01 -6.35013402e-01 -5.71906388e-01 1.16051424e+00
-7.76827782e-02 -1.14519131e+00 8.06810036e-02 5.78990161e-01
6.19913936e-01 3.26910555e-01 -3.13356102e-01 -4.37775433e-01
-2.75358737e-01 -7.51893640e-01 7.20252872e-01 4.00083214e-01
8.64278674e-01 -7.64591038e-01 1.01440370e+00 5.57047725e-01
-1.48744857e+00 2.24542439e-01 -6.06103241e-01 -2.31155396e-01
4.79672790e-01 9.40789461e-01 -7.04641044e-01 1.16194832e+00
8.58081460e-01 3.43384415e-01 -6.06867015e-01 1.20780742e+00
-1.11413263e-01 -8.64473954e-02 -6.92272007e-01 3.78153503e-01
2.28306517e-01 -1.01679552e+00 7.84358323e-01 6.68246746e-01
7.38516390e-01 9.71840508e-03 -1.45736441e-01 6.86714470e-01
5.25736511e-01 3.22515428e-01 -1.01514947e+00 5.38743198e-01
6.14762425e-01 1.17728698e+00 -7.19350517e-01 -4.43037562e-02
-4.62958008e-01 5.70561707e-01 -2.41282552e-01 8.54623783e-03
-4.42278683e-01 1.58694297e-01 7.27703452e-01 8.52796435e-01
5.04326299e-02 -4.70815122e-01 -3.60787362e-01 -9.36682105e-01
1.12748489e-01 -5.47594547e-01 1.51533157e-01 -9.56227243e-01
-7.90366590e-01 5.56331694e-01 2.23840877e-01 -1.81775379e+00
-1.98077366e-01 -6.05180442e-01 -5.37253320e-01 1.05943930e+00
-1.69392157e+00 -1.04756546e+00 -8.06417286e-01 5.06382227e-01
1.97131068e-01 2.16374844e-01 8.10775876e-01 2.85685539e-01
1.81040466e-01 -2.49815166e-01 2.54500031e-01 -3.76106352e-01
5.53496420e-01 -1.04271531e+00 1.08961150e-01 6.55342817e-01
1.56365141e-01 4.03831720e-01 1.09774113e+00 -5.66933692e-01
-1.24494207e+00 -6.82632327e-01 1.09612143e+00 -3.21228623e-01
4.60357755e-01 8.79657418e-02 -9.23667252e-01 4.15819019e-01
8.94746110e-02 -1.80227548e-01 2.43588760e-01 1.37169152e-01
1.34706005e-01 -3.81434381e-01 -1.15718734e+00 4.34007734e-01
8.14719439e-01 -4.82818186e-01 -5.86949050e-01 3.05086851e-01
4.82810847e-02 -3.81257594e-01 -8.17750216e-01 6.35936320e-01
5.72068453e-01 -1.81090438e+00 1.11720550e+00 2.86125511e-01
5.29517233e-02 -6.01131976e-01 -3.07407379e-01 -1.13580525e+00
-9.46674198e-02 -1.63647463e-03 9.05622482e-01 1.04867685e+00
2.25441810e-02 -8.30719888e-01 6.54861331e-01 1.86916329e-02
8.66253898e-02 -4.00202602e-01 -1.22989440e+00 -1.07998443e+00
-1.54976293e-01 -4.11290437e-01 7.10432529e-01 8.95202816e-01
-4.89090681e-01 1.55930415e-01 -1.31630702e-02 6.06847823e-01
8.54757428e-01 5.33479810e-01 1.07000458e+00 -2.04513741e+00
2.12119147e-01 -2.13941574e-01 -7.37767875e-01 -6.43014908e-01
-6.42395690e-02 -6.45187497e-01 -3.63770247e-01 -1.33946407e+00
8.91392455e-02 -7.08603203e-01 7.33923256e-01 -2.04874262e-01
3.44943196e-01 5.15599310e-01 1.50961995e-01 5.77788293e-01
2.90383965e-01 3.95233750e-01 1.01883101e+00 1.71610743e-01
-3.48633565e-02 2.12535858e-01 -1.43020391e-01 7.92000651e-01
5.80869675e-01 -4.00067270e-01 -9.87428725e-02 -5.29994667e-01
5.27626693e-01 -3.35774124e-02 6.44033074e-01 -1.15825295e+00
1.30704671e-01 -3.07014704e-01 3.11721228e-02 -9.55611289e-01
5.04510343e-01 -1.40336549e+00 7.73564398e-01 4.93433446e-01
6.91080034e-01 1.17099084e-01 -1.51940864e-02 1.54330403e-01
-5.25124073e-01 -5.66454411e-01 1.05440235e+00 -3.08047175e-01
-8.00622225e-01 2.18980148e-01 5.47393151e-02 -4.18217033e-01
1.04522562e+00 -8.48932147e-01 -1.41134232e-01 -4.64908510e-01
-3.35807711e-01 3.15278210e-02 1.36752558e+00 2.10162960e-02
3.71373624e-01 -1.22733819e+00 -7.12540209e-01 3.62949967e-01
3.42972338e-01 2.77710021e-01 2.75563866e-01 1.03860450e+00
-9.21620131e-01 6.39729559e-01 -3.17137778e-01 -1.16054773e+00
-1.55982387e+00 2.09045768e-01 4.85233575e-01 9.11450572e-03
-5.25202513e-01 2.93552727e-01 -1.06389657e-01 -5.40579736e-01
-3.64880562e-01 -1.54190511e-01 -2.10606307e-01 2.86164343e-01
3.40184450e-01 4.66040283e-01 4.88311619e-01 -1.13182008e+00
-4.61507082e-01 1.41515803e+00 5.50464451e-01 -6.16677999e-02
1.10443044e+00 -3.22607577e-01 -2.08462358e-01 3.59262049e-01
8.18854034e-01 3.07974547e-01 -8.03429723e-01 -8.83999094e-02
-2.05215275e-01 -1.13540661e+00 1.80568472e-01 2.65735388e-03
-8.70349705e-01 1.04340041e+00 7.82572806e-01 5.18086731e-01
9.62002158e-01 -1.02557074e-02 3.55117857e-01 8.74493495e-02
9.56238270e-01 -9.15031254e-01 -7.46910036e-01 4.01318967e-01
1.03463471e+00 -1.22322786e+00 4.33844984e-01 -7.38359153e-01
-3.85110915e-01 1.30212533e+00 1.91591308e-01 4.68342863e-02
3.24947089e-01 1.07259668e-01 6.11509979e-02 -2.99785525e-01
-1.50059434e-02 -5.37670851e-01 -9.90895461e-03 7.35072792e-01
2.79562503e-01 8.26811641e-02 -5.68872929e-01 -4.50059474e-01
-5.36759079e-01 2.55452953e-02 4.08297896e-01 6.17800951e-01
-7.14925408e-01 -1.29615188e+00 -1.10090852e+00 1.27619117e-01
2.11522743e-01 1.36406794e-02 -2.11238727e-01 1.05137038e+00
2.97364086e-01 5.73076427e-01 4.88449872e-01 -1.48670465e-01
6.18378580e-01 -2.31887311e-01 4.58739340e-01 -4.65054512e-01
-2.56964445e-01 -1.90820992e-01 7.59589076e-02 -4.75092739e-01
-6.26525760e-01 -9.15650487e-01 -1.00867903e+00 -6.82241917e-01
-5.38292587e-01 3.26810300e-01 9.91245091e-01 9.27591801e-01
-2.43996698e-02 -3.56484860e-01 8.35283995e-01 -1.22419250e+00
-4.56586510e-01 -8.27163517e-01 -1.06274796e+00 3.48349214e-01
3.49010043e-02 -9.17713881e-01 -7.27040291e-01 -1.77340060e-01] | [8.414410591125488, -2.6249208450317383] |
7143477c-e879-46f8-bd05-1970edd0d0fe | evolutionary-verbalizer-search-for-prompt | 2306.10514 | null | https://arxiv.org/abs/2306.10514v1 | https://arxiv.org/pdf/2306.10514v1.pdf | Evolutionary Verbalizer Search for Prompt-based Few Shot Text Classification | Recent advances for few-shot text classification aim to wrap textual inputs with task-specific prompts to cloze questions. By processing them with a masked language model to predict the masked tokens and using a verbalizer that constructs the mapping between predicted words and target labels. This approach of using pre-trained language models is called prompt-based tuning, which could remarkably outperform conventional fine-tuning approach in the low-data scenario. As the core of prompt-based tuning, the verbalizer is usually handcrafted with human efforts or suboptimally searched by gradient descent. In this paper, we focus on automatically constructing the optimal verbalizer and propose a novel evolutionary verbalizer search (EVS) algorithm, to improve prompt-based tuning with the high-performance verbalizer. Specifically, inspired by evolutionary algorithm (EA), we utilize it to automatically evolve various verbalizers during the evolutionary procedure and select the best one after several iterations. Extensive few-shot experiments on five text classification datasets show the effectiveness of our method. | ['Hai-Lin Liu', 'Yutao Lai', 'Lei Chen', 'Tongtao Ling'] | 2023-06-18 | null | null | null | null | ['text-classification', 'few-shot-text-classification'] | ['natural-language-processing', 'natural-language-processing'] | [ 3.95424604e-01 -6.54071346e-02 -2.38711908e-01 -4.56343859e-01
-8.23187709e-01 -4.58771795e-01 5.20381927e-01 9.52377841e-02
-6.15291536e-01 5.64117789e-01 2.83790886e-01 -1.31779373e-01
-1.51670249e-02 -7.13654578e-01 -2.85008013e-01 -5.45796037e-01
6.31654441e-01 5.87803304e-01 7.84299453e-04 -4.05520022e-01
6.37007654e-01 -3.49406637e-02 -1.49047709e+00 4.92217660e-01
1.20914066e+00 8.03696334e-01 2.99494237e-01 8.04469407e-01
-6.25140965e-01 9.21018124e-01 -6.98340535e-01 -6.57497704e-01
-8.09436366e-02 -8.63779426e-01 -9.60867167e-01 2.11866394e-01
-1.65725112e-01 8.28277692e-02 2.03807920e-01 1.11898851e+00
6.54678106e-01 3.72982800e-01 6.87798917e-01 -8.79698515e-01
-8.41436923e-01 1.16485596e+00 -5.89732409e-01 2.33359262e-01
3.14031541e-01 2.66854644e-01 1.29871070e+00 -1.13713384e+00
4.59583700e-01 1.30255544e+00 6.22402489e-01 1.01667523e+00
-1.22289407e+00 -6.22026145e-01 2.77498543e-01 3.42218906e-01
-1.40706396e+00 -5.34499228e-01 9.31919456e-01 -5.61803997e-01
8.17578256e-01 3.47968727e-01 3.02997857e-01 1.11146569e+00
-2.12798163e-01 8.63301933e-01 9.38086450e-01 -8.43309879e-01
4.17134494e-01 6.85492396e-01 5.19173384e-01 9.47162747e-01
-2.87131190e-01 -1.10518590e-01 -5.94849646e-01 -3.27956706e-01
-6.19736360e-03 -1.83970511e-01 -3.15209478e-01 -2.49018893e-02
-7.70890296e-01 1.16045678e+00 2.50593752e-01 4.07521695e-01
-5.21775424e-01 -5.41595742e-02 3.79831642e-01 3.13023001e-01
7.09904075e-01 9.01497543e-01 -3.41676652e-01 -3.47406417e-02
-1.29349124e+00 9.30455625e-02 6.74017131e-01 6.53647125e-01
8.50086570e-01 2.60984097e-02 -9.32628036e-01 1.12934899e+00
1.17532067e-01 -3.87319364e-03 1.20611763e+00 -3.30093652e-01
3.99528772e-01 8.49549830e-01 8.94320011e-02 -6.78106368e-01
-3.18542480e-01 -4.94750172e-01 -4.88629460e-01 -7.05237389e-02
-4.70266528e-02 -3.50254714e-01 -8.56064677e-01 1.73377061e+00
2.69087046e-01 5.06024882e-02 1.17769830e-01 8.97136986e-01
7.89673746e-01 7.48978674e-01 3.83044422e-01 -4.23195064e-01
1.48285329e+00 -1.27962732e+00 -8.40325236e-01 -4.20930296e-01
8.60748410e-01 -4.32981938e-01 1.76089919e+00 2.72944301e-01
-8.73500049e-01 -5.01895785e-01 -1.14057279e+00 1.60967499e-01
-3.19925010e-01 2.74464369e-01 1.31108344e-01 8.32639813e-01
-6.42044067e-01 4.53167379e-01 -2.62733936e-01 -2.96603799e-01
2.53273249e-01 5.12546077e-02 3.26313823e-01 2.34600276e-01
-1.49230576e+00 8.61209691e-01 5.68922698e-01 -2.13530496e-01
-8.96097839e-01 -6.86454415e-01 -5.46816468e-01 2.84796149e-01
6.74950421e-01 -6.78888619e-01 1.41593528e+00 -1.28899789e+00
-1.89392531e+00 8.90124679e-01 -2.65146166e-01 -4.78020191e-01
5.47827959e-01 -1.42633900e-01 -1.77479997e-01 -1.26382574e-01
1.60665531e-02 3.70799333e-01 1.16740131e+00 -1.26272178e+00
-4.88466442e-01 -7.52307326e-02 -1.88037470e-01 2.21656263e-01
-8.98824096e-01 3.81948978e-01 -1.17854737e-01 -6.31043375e-01
-2.67137587e-01 -6.19278967e-01 -2.58870244e-01 -4.22025502e-01
-3.69825751e-01 -6.14573777e-01 4.72732991e-01 -3.92910630e-01
1.78536499e+00 -1.95064425e+00 2.00520962e-01 -5.40761799e-02
2.74943441e-01 6.49568737e-01 -2.24222809e-01 2.68996090e-01
2.65289694e-02 2.09048077e-01 -1.77933887e-01 -3.47572327e-01
1.79751083e-01 -7.06847981e-02 -5.48414290e-01 1.10017374e-01
1.56927004e-01 1.05032539e+00 -9.63108242e-01 -6.95295095e-01
-4.58560213e-02 -5.08824959e-02 -6.67992890e-01 6.09969258e-01
-7.41003811e-01 7.51606096e-03 -5.91320217e-01 3.03567767e-01
2.96218097e-01 -5.00051200e-01 -7.42528439e-02 9.92988795e-02
-1.45938620e-01 1.93129063e-01 -9.52233374e-01 1.47164547e+00
-6.64273977e-01 3.20377618e-01 -3.48307602e-02 -8.71019065e-01
1.30397177e+00 2.95993507e-01 -6.98574781e-02 -5.10360479e-01
5.83011389e-01 6.86250851e-02 -2.46148363e-01 -8.17294955e-01
6.22864425e-01 -4.49057639e-01 -1.30124062e-01 8.35562408e-01
6.73862621e-02 -8.65997151e-02 1.38118148e-01 2.22972646e-01
9.31097984e-01 -8.20042714e-02 5.67425072e-01 -1.39700890e-01
8.75484288e-01 1.53010398e-01 3.36430848e-01 8.88225198e-01
-1.20286241e-01 3.34896058e-01 2.81957924e-01 -3.25925797e-01
-8.88460338e-01 -4.55127656e-01 2.57022291e-01 1.88731217e+00
-8.58412609e-02 -6.26787126e-01 -1.09768307e+00 -7.45098531e-01
-3.35020036e-01 1.24186671e+00 -7.88144290e-01 -4.99503195e-01
-3.69382620e-01 -8.25361907e-01 6.48468733e-01 2.16078341e-01
3.26313049e-01 -1.25122809e+00 -7.48830080e-01 4.76737797e-01
-1.35995805e-01 -6.33051336e-01 -5.95383584e-01 5.94769478e-01
-3.46473545e-01 -6.27228916e-01 -5.91756880e-01 -7.30232775e-01
5.25972903e-01 7.86432717e-03 9.51969326e-01 3.20954263e-01
-1.62066206e-01 -6.87734559e-02 -5.69416821e-01 -3.61666113e-01
-5.94953418e-01 3.25748712e-01 -8.81451666e-02 3.78151655e-01
6.71824217e-01 -3.02212387e-01 -1.90426901e-01 1.34466782e-01
-6.88029945e-01 9.36868563e-02 5.01059234e-01 1.19875073e+00
2.12916821e-01 -1.72880113e-01 5.87011516e-01 -1.19435906e+00
1.44029737e+00 -4.55699593e-01 -3.70749116e-01 8.16092014e-01
-9.49144542e-01 5.94998002e-01 1.03272438e+00 -8.48878264e-01
-1.21954775e+00 -2.12159082e-02 1.62377387e-01 -6.30746841e-01
1.37628004e-01 4.42766130e-01 6.81025013e-02 1.92593977e-01
1.21397471e+00 3.27888906e-01 -5.34920096e-01 -4.33705330e-01
6.02087796e-01 1.08776009e+00 3.55992883e-01 -7.72632420e-01
6.79454386e-01 9.56943780e-02 -7.42033124e-01 -3.92446488e-01
-1.31777883e+00 -3.61138344e-01 -4.64836985e-01 -3.04554850e-01
9.07987714e-01 -4.87243861e-01 -5.99555671e-01 5.34390146e-03
-1.30936790e+00 -2.83874959e-01 -1.67939827e-01 1.95493579e-01
-3.70833158e-01 -1.55664608e-02 -4.49912399e-01 -9.87360954e-01
-7.16319919e-01 -9.62731361e-01 8.25617433e-01 3.80217791e-01
-6.19007409e-01 -8.00346434e-01 3.18451405e-01 3.40924382e-01
4.77099508e-01 -2.48294398e-01 1.09932232e+00 -1.12742484e+00
-4.28523421e-02 -6.81397766e-02 -3.12199295e-02 3.85058187e-02
-4.47000861e-02 -5.63124456e-02 -1.01749325e+00 -7.15550929e-02
2.23014608e-01 -6.87579513e-01 8.79322529e-01 -2.79005058e-02
1.24018812e+00 -4.58547562e-01 -2.74238706e-01 6.97964907e-01
1.30122519e+00 8.90553519e-02 1.01934690e-02 4.05846119e-01
5.24190068e-01 6.69389427e-01 6.73800707e-01 6.79160535e-01
7.06282482e-02 5.37571609e-01 -5.53868711e-03 2.72092134e-01
1.52018655e-03 -5.40464044e-01 2.49567837e-01 9.26779687e-01
3.37104470e-01 -3.61655831e-01 -9.44829047e-01 2.80814022e-01
-1.78542387e+00 -8.48751903e-01 3.56615603e-01 1.63457775e+00
1.22608793e+00 1.68968201e-01 -4.29333188e-02 -1.83802247e-02
1.12294793e+00 1.72163993e-01 -7.35686541e-01 -6.66680396e-01
-1.49145704e-02 3.19329828e-01 1.86032742e-01 7.20297813e-01
-7.97170818e-01 1.41794050e+00 5.33234692e+00 1.09051526e+00
-1.29498613e+00 4.47152048e-01 4.92630422e-01 -3.11383516e-01
-4.46565568e-01 4.24344465e-02 -9.94948030e-01 4.22923833e-01
1.01552141e+00 -7.47345269e-01 6.48634434e-01 8.90895367e-01
2.84366965e-01 3.23557377e-01 -1.08251107e+00 8.69058132e-01
4.56877589e-01 -1.21247721e+00 1.03266314e-01 -5.90311348e-01
8.17682505e-01 -5.22735357e-01 1.67488642e-02 6.21413231e-01
5.53609848e-01 -1.03041601e+00 9.24819350e-01 3.95944148e-01
7.40501165e-01 -6.72489345e-01 3.59975308e-01 7.72057831e-01
-8.86826813e-01 -3.13496172e-01 -4.21652496e-01 7.89947137e-02
3.37274894e-02 1.41634673e-01 -1.16574347e+00 9.26382467e-02
4.25796449e-01 1.58800006e-01 -7.37671316e-01 6.18154228e-01
-5.03044367e-01 8.22037160e-01 2.19980374e-01 -8.94399643e-01
4.90153700e-01 1.15564205e-01 5.43922961e-01 1.46762872e+00
4.46359031e-02 3.22102010e-01 1.25157401e-01 1.17479467e+00
-1.74741089e-01 4.76820201e-01 -4.01523821e-02 -1.53048798e-01
7.71467984e-01 1.29373765e+00 -6.43738627e-01 -6.02544188e-01
-1.63816344e-02 1.05993247e+00 7.88473904e-01 2.27094933e-01
-8.79284084e-01 -6.57892168e-01 1.48394302e-01 -9.02023315e-02
3.53837401e-01 4.02248859e-01 -4.46462095e-01 -9.65373695e-01
-2.14736119e-01 -1.07046247e+00 4.48126227e-01 -8.28128934e-01
-1.35235274e+00 9.43346262e-01 -2.61632025e-01 -9.82059360e-01
-3.61500561e-01 -3.23371172e-01 -7.97745347e-01 8.78582776e-01
-1.21516383e+00 -7.19383419e-01 -3.94757897e-01 5.98760664e-01
9.95626926e-01 -2.70975888e-01 7.84842968e-01 -8.02916065e-02
-7.65665650e-01 8.43934953e-01 -1.25588953e-01 -4.07647192e-02
8.85997295e-01 -1.10408735e+00 2.84807663e-02 7.93921947e-01
1.57473445e-01 6.67886138e-01 1.02827644e+00 -5.54728091e-01
-9.26241696e-01 -1.10389614e+00 9.94215488e-01 -2.70048887e-01
8.21870387e-01 -5.08382797e-01 -1.07498872e+00 3.77414674e-01
3.25422823e-01 -3.21349949e-01 8.01550090e-01 2.29806527e-02
-4.04060513e-01 3.53172682e-02 -9.53686476e-01 7.13385880e-01
8.95919263e-01 -5.50774872e-01 -1.12436426e+00 4.37998891e-01
1.01460850e+00 -1.69636235e-01 -1.27951736e-02 -4.39877100e-02
1.18405245e-01 -6.67090118e-01 4.47958201e-01 -9.85978961e-01
3.92291844e-01 -9.71981138e-02 1.16319142e-01 -1.62950945e+00
-4.40219492e-01 -8.92661929e-01 -2.97901686e-03 1.44055676e+00
4.64801908e-01 -3.86556834e-01 8.16362619e-01 5.90415537e-01
-5.16791940e-02 -6.79444551e-01 -6.16256654e-01 -6.11263037e-01
6.32396340e-02 -1.61600649e-01 7.05573976e-01 1.04136920e+00
1.88460499e-01 8.78163576e-01 -5.87090373e-01 -1.67159155e-01
3.13235462e-01 2.20826268e-01 5.23669660e-01 -1.10902858e+00
-4.16381657e-01 -7.06232369e-01 1.96397245e-01 -1.02592897e+00
6.00098908e-01 -1.10014105e+00 5.21807671e-01 -1.02361619e+00
3.70621651e-01 -1.00601114e-01 -3.47834885e-01 5.41020155e-01
-7.88621724e-01 -1.42330796e-01 5.46029471e-02 -2.40394957e-02
-8.11513782e-01 7.52203524e-01 8.26169372e-01 -2.27549478e-01
-4.46415931e-01 -1.92986000e-02 -9.99055743e-01 5.74451923e-01
7.70899177e-01 -8.87247622e-01 -3.72510433e-01 -2.89571464e-01
3.12761635e-01 -3.56029198e-02 8.12656898e-03 -6.37257099e-01
6.56787097e-01 -2.73833483e-01 -5.07381139e-03 -2.89041549e-01
-5.11605584e-04 -4.83980566e-01 -3.27159226e-01 5.45204937e-01
-1.18047810e+00 5.92056923e-02 -2.25575104e-01 3.85628402e-01
-9.80277061e-02 -9.41046834e-01 1.00494885e+00 -2.98549861e-01
-7.28707194e-01 8.14786106e-02 -4.14854765e-01 3.87920916e-01
9.90589142e-01 -7.54224658e-02 -1.95609376e-01 -2.00498149e-01
-4.69662398e-01 4.23217624e-01 1.70942217e-01 4.54240888e-01
4.78243649e-01 -1.18162072e+00 -9.20152009e-01 6.64851516e-02
3.04064840e-01 -4.82411414e-01 4.85503227e-02 3.84717762e-01
-6.43843934e-02 4.05869156e-01 1.74071595e-01 -3.32961410e-01
-1.38135362e+00 9.55626905e-01 3.24085921e-01 -4.05344516e-01
-2.82565862e-01 1.33744740e+00 -1.49793416e-01 -3.51083308e-01
3.12290162e-01 4.41262424e-02 -4.61270899e-01 3.39463979e-01
7.11052001e-01 2.00465202e-01 8.17838684e-02 -2.66367465e-01
-2.82150805e-01 4.69710797e-01 -1.81750566e-01 -2.65420526e-01
1.22117758e+00 -1.19017668e-01 -9.87432599e-02 5.55906713e-01
1.16663456e+00 6.52842000e-02 -1.03940606e+00 -6.90896511e-01
4.47545767e-01 -2.63703912e-01 8.64029452e-02 -8.12463760e-01
-6.39472485e-01 8.33989859e-01 2.89721340e-01 2.92468250e-01
1.00691020e+00 2.04127934e-03 5.88255346e-01 5.66871464e-01
-1.63353026e-01 -1.28379893e+00 5.27907073e-01 6.54887557e-01
6.45125806e-01 -1.01911390e+00 -3.09313625e-01 -3.18612084e-02
-9.47420716e-01 1.09274757e+00 9.47226048e-01 5.20553403e-02
2.36250624e-01 2.44327307e-01 2.19455481e-01 -1.75301298e-01
-1.27799535e+00 -2.08277836e-01 2.06616357e-01 1.18474685e-01
4.66748089e-01 -6.54132366e-02 -5.93053758e-01 1.00464606e+00
-4.90016848e-01 -8.05999041e-02 1.84154764e-01 7.64654040e-01
-1.00625992e+00 -8.67547035e-01 -2.92954952e-01 3.22065115e-01
-2.00138792e-01 -3.80591422e-01 -6.57542467e-01 8.31563100e-02
7.62160718e-02 1.12385750e+00 -9.17013064e-02 -6.00683987e-01
3.06835860e-01 5.80309629e-01 1.13983229e-01 -8.87170255e-01
-1.07389271e+00 -3.23939025e-01 -7.07867444e-02 -2.04388469e-01
-4.72079068e-02 -4.13816363e-01 -1.17450082e+00 -1.09025404e-01
-5.93069792e-01 6.45817578e-01 5.05732775e-01 1.16166723e+00
8.35821778e-02 5.74841678e-01 9.23469126e-01 -4.63372648e-01
-1.17895496e+00 -1.17332566e+00 -1.08638987e-01 4.48505610e-01
2.35032946e-01 -5.63828170e-01 -5.80877185e-01 -4.95313816e-02] | [10.703187942504883, 7.905996799468994] |
1354928c-69b3-48f6-be4f-b30412d8638f | exploiting-uncertainty-in-regression-forests | null | null | http://openaccess.thecvf.com/content_cvpr_2015/html/Valentin_Exploiting_Uncertainty_in_2015_CVPR_paper.html | http://openaccess.thecvf.com/content_cvpr_2015/papers/Valentin_Exploiting_Uncertainty_in_2015_CVPR_paper.pdf | Exploiting Uncertainty in Regression Forests for Accurate Camera Relocalization | Recent advances in camera relocalization use predictions from a regression forest to guide the camera pose optimization procedure. In these methods, each tree associates one pixel with a point in the scene's 3D world coordinate frame. In previous work, these predictions were point estimates and the subsequent camera pose optimization implicitly assumed an isotropic distribution of these estimates. In this paper, we train a regression forest to predict mixtures of anisotropic 3D Gaussians and show how the predicted uncertainties can be taken into account for continuous pose optimization. Experiments show that our proposed method is able to relocalize up to 40% more frames than the state-of-the-art. | ['Jamie Shotton', 'Julien Valentin', 'Philip H. S. Torr', 'Matthias Niessner', 'Shahram Izadi', 'Andrew Fitzgibbon'] | 2015-06-01 | null | null | null | cvpr-2015-6 | ['camera-relocalization'] | ['computer-vision'] | [ 7.17780665e-02 1.02315657e-01 -3.46179485e-01 -3.54394346e-01
-6.96314633e-01 -5.81708729e-01 5.65032363e-01 -1.90320641e-01
-6.01232827e-01 6.39088631e-01 -3.79845058e-03 -1.62739202e-01
2.64181376e-01 -5.80377400e-01 -9.84041035e-01 -6.58806324e-01
4.65963066e-01 9.37376976e-01 5.79200923e-01 3.66880715e-01
3.35705966e-01 8.78851116e-01 -1.12066233e+00 -3.43506962e-01
5.45588434e-01 8.93394947e-01 3.33687901e-01 9.25476313e-01
7.61695206e-02 4.93382365e-01 -3.22346419e-01 -4.55015659e-01
3.14772397e-01 1.70807183e-01 -5.10269761e-01 4.17194724e-01
7.85065830e-01 -3.79289746e-01 -7.31977373e-02 1.02061141e+00
5.01996949e-02 2.47714996e-01 8.61828625e-01 -9.72242892e-01
-1.85044348e-01 3.82213801e-01 -8.22815061e-01 -1.92163914e-01
4.54629123e-01 -5.79597466e-02 6.67139232e-01 -9.14702535e-01
7.28806853e-01 1.16140497e+00 8.40566933e-01 1.88579544e-01
-1.45858753e+00 -4.96734142e-01 5.12723029e-01 8.21456835e-02
-1.65620470e+00 -2.18852013e-01 7.08086669e-01 -6.91105187e-01
1.05372214e+00 1.06924484e-02 9.17320967e-01 9.75767553e-01
3.81305218e-01 4.86207098e-01 9.34587598e-01 -4.91208106e-01
2.73099303e-01 1.44773468e-01 -1.03398442e-01 7.17739046e-01
2.99222320e-01 1.41667917e-01 -4.79337603e-01 8.79690275e-02
1.06720364e+00 -1.10736117e-01 -6.73804656e-02 -9.75253463e-01
-1.40781677e+00 7.06336796e-01 5.08355737e-01 -2.59164840e-01
-3.56421053e-01 4.35039759e-01 -2.07568765e-01 -5.91179729e-01
7.03622997e-01 3.37161541e-01 -3.10278624e-01 -1.22742113e-02
-9.73366797e-01 -5.09050733e-04 5.80906272e-01 1.32764101e+00
9.60352361e-01 -1.13332771e-01 2.69561470e-01 6.01094484e-01
6.68701649e-01 9.75656927e-01 -3.27195711e-02 -1.52026129e+00
3.96145999e-01 1.32975236e-01 3.11830103e-01 -1.08601809e+00
-3.79457414e-01 -2.53448635e-01 -5.87845981e-01 4.18194503e-01
4.78088319e-01 -7.95296729e-02 -9.94243503e-01 1.17703927e+00
4.55692917e-01 4.72352713e-01 -1.61094874e-01 8.81326377e-01
2.18792975e-01 5.81765175e-01 -1.62012339e-01 -9.72910374e-02
9.62726951e-01 -1.03125203e+00 -3.75829518e-01 -5.01807511e-01
9.64442119e-02 -1.15056813e+00 4.23773199e-01 5.77447593e-01
-9.29371357e-01 -6.77994192e-01 -8.34925711e-01 1.18647531e-01
5.97344227e-02 4.36017185e-01 4.02158320e-01 6.72117114e-01
-1.12325108e+00 3.64023358e-01 -1.24104798e+00 -6.42077923e-01
2.12980568e-01 4.16560173e-01 -4.47762311e-01 7.65795931e-02
-4.51148182e-01 1.27850640e+00 2.51735479e-01 1.39991283e-01
-7.57905364e-01 -3.75134647e-01 -8.22797000e-01 -3.08467746e-01
4.93218750e-01 -1.01120198e+00 1.24817622e+00 -4.74559337e-01
-1.80575264e+00 8.67301881e-01 -4.70222801e-01 -4.58470881e-01
6.38317764e-01 -6.32130742e-01 1.23024277e-01 4.59985286e-02
1.25683174e-01 9.91812408e-01 1.06075418e+00 -1.44338143e+00
-8.52641106e-01 -3.57589662e-01 9.45365131e-02 4.44908023e-01
3.53176594e-01 -1.41361117e-01 -1.02833855e+00 -5.00842214e-01
6.03456557e-01 -1.54321253e+00 -5.78757465e-01 1.74052760e-01
-6.26587033e-01 2.24992633e-01 5.93774974e-01 -3.92528594e-01
7.22347260e-01 -1.82429707e+00 6.42585635e-01 2.97734171e-01
1.02143802e-01 -4.35813576e-01 2.11859956e-01 -1.02524430e-01
1.08922124e-01 -1.45023754e-02 8.95510316e-02 -8.70021164e-01
-1.80868492e-01 2.03968391e-01 -1.72857583e-01 8.53961706e-01
-1.58694655e-01 2.42732942e-01 -7.30354905e-01 -5.78802586e-01
8.36511731e-01 7.88788676e-01 -6.83710396e-01 2.76198119e-01
-1.23485520e-01 6.28340960e-01 -3.14343691e-01 4.48196769e-01
9.07728732e-01 1.45570740e-01 -3.73290218e-02 -5.01196206e-01
-2.51771748e-01 -1.04814246e-01 -1.28204834e+00 1.81827867e+00
-3.56268436e-01 6.68946385e-01 -3.22936336e-03 -3.78842145e-01
9.74758625e-01 -1.17329225e-01 5.63779950e-01 2.90271699e-01
8.60978067e-02 -1.23757437e-01 -4.84783351e-01 1.37195677e-01
8.39343071e-01 4.78278734e-02 1.19329885e-01 9.72528011e-02
2.02939227e-01 -9.69406545e-01 -1.15084574e-01 -5.02132662e-02
5.75957000e-01 7.59383440e-01 3.33279490e-01 -3.67963687e-02
5.62391818e-01 1.86922178e-01 5.91409624e-01 6.26934230e-01
-3.22809964e-02 1.00283527e+00 1.23790599e-01 -4.51185197e-01
-1.19927049e+00 -1.21587574e+00 -9.58158597e-02 6.56567097e-01
4.67123836e-01 -4.22203153e-01 -9.55349207e-01 -6.98031723e-01
-1.22478776e-01 9.23259020e-01 -3.57664317e-01 2.42745146e-01
-5.07699311e-01 -4.45175469e-01 -4.03269790e-02 6.43187106e-01
3.07377368e-01 -4.09987181e-01 -6.93791270e-01 -3.34828980e-02
-9.35340822e-02 -1.36652362e+00 -5.43741763e-01 2.85361916e-01
-1.01116705e+00 -1.15274990e+00 -6.98900521e-01 -3.34919721e-01
9.20885861e-01 3.83095145e-01 1.05317318e+00 -5.00113308e-01
2.43501645e-02 6.39293194e-01 -2.33662397e-01 -3.28624308e-01
-3.48711312e-01 1.37764812e-01 2.82478303e-01 -2.51809597e-01
3.95100981e-01 -3.73448104e-01 -3.67668122e-01 6.18429482e-01
-4.04235929e-01 3.78767252e-01 3.81091774e-01 5.67076743e-01
1.01908410e+00 5.21990806e-02 -7.41510153e-01 -9.51097548e-01
-4.21018340e-02 -1.09269045e-01 -1.21231484e+00 1.63695097e-01
-5.57489514e-01 3.19882870e-01 1.16270736e-01 -4.41492319e-01
-1.20155931e+00 9.56539452e-01 1.48422673e-01 -7.31113076e-01
-4.37012166e-01 1.13968074e-01 -1.25012621e-01 -2.59139150e-01
4.30765063e-01 -2.18672812e-01 -3.40928793e-01 -3.73252183e-01
7.06550717e-01 2.54447311e-01 7.04810739e-01 -7.08175838e-01
1.07366872e+00 5.56092024e-01 2.92130947e-01 -6.04921639e-01
-7.63198376e-01 -5.67881405e-01 -1.45133829e+00 -5.63997924e-01
1.24057674e+00 -1.13434303e+00 -5.16218841e-01 3.78269911e-01
-1.45818436e+00 -1.77545428e-01 -1.14248969e-01 8.98116648e-01
-8.43825519e-01 3.53109419e-01 -1.57552958e-01 -7.89886653e-01
2.02714354e-01 -1.55994463e+00 1.47151875e+00 2.88604289e-01
-3.02320957e-01 -8.88552547e-01 2.78777838e-01 3.77911657e-01
2.68363226e-02 -6.19299039e-02 2.60867625e-01 5.15094735e-02
-9.75183547e-01 -2.52991885e-01 -5.11855148e-02 1.80772498e-01
-5.33355735e-02 5.95669150e-01 -1.00446320e+00 3.42934728e-02
-2.85606720e-02 3.33761126e-01 6.95456326e-01 8.71678591e-01
8.73144329e-01 2.32226744e-01 -4.70151871e-01 1.03768468e+00
1.43401384e+00 4.06732112e-02 4.37694132e-01 5.30468941e-01
8.75694871e-01 2.94211894e-01 7.58816898e-01 3.88002723e-01
4.54632401e-01 1.02945697e+00 7.89266646e-01 1.93059653e-01
-2.00075284e-03 -4.55431670e-01 2.64519572e-01 4.15428251e-01
-6.47191942e-01 -3.26598994e-02 -9.49025035e-01 1.82737872e-01
-1.92021215e+00 -7.15702891e-01 -2.19851896e-01 2.32991076e+00
3.32323968e-01 2.74789721e-01 -8.97287279e-02 -3.68671924e-01
9.18609142e-01 1.56658232e-01 -5.50526619e-01 -3.46036591e-02
-4.63441052e-02 -2.29188457e-01 1.41479039e+00 9.39238012e-01
-1.51180518e+00 1.24801481e+00 7.27453184e+00 3.60698313e-01
-8.35803032e-01 -7.75752217e-02 4.23173010e-01 1.04919806e-01
-6.36786148e-02 3.52725714e-01 -1.26421821e+00 1.11291841e-01
6.33165538e-01 9.06265005e-02 4.07098860e-01 1.21823692e+00
1.92510784e-01 -6.96107447e-01 -1.09389400e+00 9.81393754e-01
2.74430305e-01 -1.11957157e+00 -2.71924913e-01 2.34979481e-01
8.98211896e-01 1.82680741e-01 2.63660960e-02 -1.04057416e-01
7.89939940e-01 -7.66789317e-01 1.10818338e+00 9.53685641e-01
6.21688545e-01 -6.73587680e-01 6.89230502e-01 4.19191867e-01
-1.15169442e+00 1.67774498e-01 -6.51926816e-01 1.49866164e-01
5.46272039e-01 6.55860662e-01 -1.18392575e+00 3.67805600e-01
7.61762798e-01 8.49426627e-01 -7.97919810e-01 1.16851544e+00
-5.28286159e-01 3.54128867e-01 -7.00426877e-01 2.47089043e-01
7.39043280e-02 -5.96589386e-01 6.81055307e-01 1.09381068e+00
6.68898940e-01 -2.20185995e-01 3.52587789e-01 6.95915043e-01
2.22537026e-01 -2.13525459e-01 -5.37744343e-01 6.30632222e-01
4.00149405e-01 9.84140217e-01 -9.79296803e-01 -1.23730078e-01
-3.42890561e-01 1.01486981e+00 2.27386788e-01 3.32028091e-01
-1.04645288e+00 3.68085355e-01 6.28496110e-01 1.03937022e-01
4.66206104e-01 -6.49728119e-01 -2.28562728e-01 -1.36718094e+00
-2.22905651e-01 -2.78795421e-01 -2.50952896e-02 -1.36224329e+00
-1.08965504e+00 6.63830400e-01 4.22902137e-01 -1.29637611e+00
-4.51314956e-01 -9.01517332e-01 -1.84271380e-01 9.11556542e-01
-1.13261509e+00 -1.33646452e+00 -3.10837746e-01 3.16397071e-01
6.21252835e-01 -7.52119422e-02 8.18408489e-01 -3.21000576e-01
-4.62462306e-01 -5.94432987e-02 1.47716522e-01 -9.96530987e-04
8.93903553e-01 -1.46400142e+00 3.98075432e-01 9.43402946e-01
3.73622447e-01 4.40932512e-01 1.03869212e+00 -6.80280924e-01
-1.21542525e+00 -9.87311840e-01 4.08807129e-01 -8.34259272e-01
6.51821136e-01 -2.02908874e-01 -2.64492422e-01 1.05575383e+00
1.56701840e-02 2.60463804e-01 3.56945276e-01 1.61983818e-01
-1.94754660e-01 -9.91285965e-02 -8.77745330e-01 3.92905891e-01
7.87732124e-01 -3.64380062e-01 -3.49107146e-01 2.90257663e-01
4.91830140e-01 -9.75197792e-01 -8.05298328e-01 3.25566739e-01
6.19652152e-01 -1.14405572e+00 1.30619121e+00 -6.17365390e-02
1.84825156e-02 -7.21940339e-01 -3.60046625e-01 -1.34318471e+00
-3.00623357e-01 -2.39809752e-01 -1.33026198e-01 9.76722002e-01
2.75168061e-01 -1.42172530e-01 1.13651764e+00 7.52287507e-01
1.16714435e-02 -2.09171161e-01 -1.05318105e+00 -4.53852355e-01
-1.39280215e-01 -7.41891086e-01 2.35837266e-01 3.35452408e-01
-5.74327290e-01 3.54201555e-01 -4.56390619e-01 6.74382269e-01
8.88987839e-01 -9.25331265e-02 1.03509378e+00 -1.36646771e+00
-3.76631886e-01 -4.02491480e-01 -6.60799265e-01 -1.50857520e+00
4.63756531e-01 -2.95398235e-01 2.31749997e-01 -1.38200951e+00
1.42526880e-01 -3.51023883e-01 2.70620227e-01 -2.56525762e-02
-1.16660863e-01 1.74560472e-01 3.91522497e-01 1.56746492e-01
-5.88300526e-01 2.85237044e-01 8.19357455e-01 -2.11312220e-01
-7.61852264e-02 4.76814419e-01 -9.27110389e-02 1.18409216e+00
4.37249839e-01 -4.74120378e-01 -2.70666003e-01 -6.63824081e-01
2.05949605e-01 -1.15497142e-01 2.83651173e-01 -1.20342362e+00
4.65783626e-01 -3.55127454e-01 5.97443998e-01 -1.08206153e+00
8.18124175e-01 -1.57946861e+00 6.31289184e-01 5.84296621e-02
-1.07693793e-02 9.85496715e-02 5.91397993e-02 8.45655143e-01
-3.11792400e-02 -3.07653129e-01 7.13940620e-01 -7.82827139e-02
-6.88525140e-01 7.21328855e-02 -4.34992403e-01 -5.71066082e-01
1.18432915e+00 -3.67132664e-01 9.03310254e-02 -4.80206758e-01
-1.01147461e+00 -1.62851572e-01 9.81215477e-01 1.36715114e-01
5.55024385e-01 -1.14236259e+00 -4.43364680e-01 7.68936872e-02
8.44129734e-03 5.78853726e-01 -8.14851299e-02 8.42573106e-01
-1.01733017e+00 4.87634301e-01 5.01383282e-02 -1.22997999e+00
-1.37663877e+00 6.12186372e-01 4.12091106e-01 -1.45729899e-01
-3.79140109e-01 8.58518183e-01 2.66573094e-02 -4.83676672e-01
2.54406393e-01 -6.16345704e-01 5.17492229e-03 -1.97886199e-01
1.02368422e-01 2.95777112e-01 -2.62961656e-01 -1.11695743e+00
-3.70341957e-01 1.34344959e+00 3.27212572e-01 -3.18221837e-01
1.37884939e+00 -5.54897785e-01 -3.58118303e-02 5.61469078e-01
7.67720282e-01 2.79607654e-01 -1.81649649e+00 -1.46384075e-01
-1.20842338e-01 -7.32999504e-01 3.49277079e-01 -3.54378223e-01
-1.11945093e+00 7.11840868e-01 5.10242581e-01 -3.34422201e-01
8.76221716e-01 1.36129990e-01 2.48332620e-02 4.10037488e-01
7.55486965e-01 -9.50479388e-01 -4.04709160e-01 5.08940816e-01
5.95503390e-01 -1.23162758e+00 4.54219818e-01 -7.53989816e-01
-7.68973947e-01 1.37183309e+00 5.28585672e-01 -2.10549071e-01
8.09789240e-01 4.67208624e-01 -4.59718294e-02 3.20646942e-01
-3.58778715e-01 -2.81451028e-02 3.71830016e-01 9.03200686e-01
2.61566758e-01 6.81774840e-02 3.92500281e-01 1.31891996e-01
-2.70348579e-01 -1.92907527e-01 5.67645669e-01 4.79499280e-01
-4.72179294e-01 -1.06242478e+00 -1.06672728e+00 -1.05978407e-01
3.45324613e-02 3.41714621e-02 -2.32305035e-01 7.27337003e-01
9.88244414e-02 7.69172609e-01 2.39734977e-01 -3.52284640e-01
3.18360627e-01 -5.87196946e-02 6.28126919e-01 -6.52121902e-01
1.11156873e-01 6.24689996e-01 -1.21473469e-01 -7.67859697e-01
-7.33508468e-01 -9.11883831e-01 -8.72326791e-01 -2.41637275e-01
-6.14987552e-01 -1.31744593e-01 9.99687433e-01 6.94906175e-01
-7.23914504e-02 2.39960581e-01 3.66090328e-01 -1.55300951e+00
-1.25622556e-01 -8.10737133e-01 -4.66885179e-01 -1.90952420e-01
1.75887898e-01 -8.75082672e-01 -4.14561808e-01 5.89860857e-01] | [7.773956298828125, -2.3706836700439453] |
bc4eda9f-3fe9-46bb-8f05-537a468be0e7 | machine-learning-based-detection-of-clickbait | 1710.01977 | null | http://arxiv.org/abs/1710.01977v1 | http://arxiv.org/pdf/1710.01977v1.pdf | Machine Learning Based Detection of Clickbait Posts in Social Media | Clickbait (headlines) make use of misleading titles that hide critical
information from or exaggerate the content on the landing target pages to
entice clicks. As clickbaits often use eye-catching wording to attract viewers,
target contents are often of low quality. Clickbaits are especially widespread
on social media such as Twitter, adversely impacting user experience by causing
immense dissatisfaction. Hence, it has become increasingly important to put
forward a widely applicable approach to identify and detect clickbaits. In this
paper, we make use of a dataset from the clickbait challenge 2017
(clickbait-challenge.com) comprising of over 21,000 headlines/titles, each of
which is annotated by at least five judgments from crowdsourcing on how
clickbait it is. We attempt to build an effective computational clickbait
detection model on this dataset. We first considered a total of 331 features,
filtered out many features to avoid overfitting and improve the running time of
learning, and eventually selected the 60 most important features for our final
model. Using these features, Random Forest Regression achieved the following
results: MSE=0.035 MSE, Accuracy=0.82, and F1-sore=0.61 on the clickbait class. | ['Thai Le', 'Xinyue Cao', 'Jason', 'Zhang'] | 2017-10-05 | null | null | null | null | ['clickbait-detection'] | ['natural-language-processing'] | [-2.68796057e-01 -2.17625290e-01 -3.39099020e-01 -3.91988248e-01
-1.06717563e+00 -8.05763304e-01 5.92486620e-01 4.17169034e-01
-7.00392723e-01 9.17249978e-01 1.09722661e-02 -3.79777163e-01
1.05694316e-01 -5.53550243e-01 -6.95058763e-01 -1.41588911e-01
5.02622649e-02 1.02420449e-01 6.80938840e-01 -2.39844751e-02
6.89338803e-01 5.47293620e-03 -1.54403520e+00 2.24675924e-01
1.04831445e+00 1.34605038e+00 1.27045680e-02 5.87484419e-01
-9.60339755e-02 8.72240901e-01 -8.33085120e-01 -6.40183151e-01
1.93795204e-01 -1.30246758e-01 -4.28953528e-01 -1.23762853e-01
7.06753492e-01 -3.80061120e-01 -2.97998250e-01 9.88952994e-01
2.49239549e-01 -3.56843919e-02 1.71811849e-01 -1.24699473e+00
-5.29713392e-01 2.73906142e-01 -5.66884816e-01 4.92630929e-01
3.01518828e-01 1.21301509e-01 1.58743358e+00 -9.43592966e-01
5.82915127e-01 8.82627070e-01 4.17835295e-01 5.53908236e-02
-1.31853259e+00 -7.51834452e-01 5.55109270e-02 1.89938083e-01
-1.42369986e+00 -1.01737790e-01 3.03618908e-01 -7.72546649e-01
2.09329277e-01 9.45097923e-01 6.24693394e-01 1.05718124e+00
3.12833264e-02 1.17259097e+00 1.45214665e+00 -7.89817572e-02
2.04054087e-01 5.12422144e-01 5.37836969e-01 2.34846622e-01
4.64423001e-01 -1.10235780e-01 -7.08205581e-01 -4.48376179e-01
1.56492487e-01 -2.83843484e-02 -1.14739098e-01 1.03767984e-01
-1.00412011e+00 1.19006538e+00 7.44230330e-01 1.35154009e-01
-3.10522079e-01 5.33327311e-02 1.25717223e-01 1.98197335e-01
7.22264469e-01 7.24707067e-01 -1.52001128e-01 -3.67572606e-01
-9.60269988e-01 8.23551059e-01 7.28451610e-01 5.44553816e-01
7.94721425e-01 -5.52831411e-01 -3.01777929e-01 9.50288653e-01
1.25013709e-01 7.31968105e-01 3.43465418e-01 -5.27056932e-01
5.39645135e-01 6.81012571e-01 6.16626561e-01 -1.31565726e+00
-6.29364327e-02 -6.08222306e-01 -2.85300225e-01 -7.31481537e-02
5.20730197e-01 -2.94472091e-02 -8.42322350e-01 1.11233151e+00
4.66528945e-02 -6.14651740e-01 -9.28770840e-01 1.16421509e+00
3.93245071e-01 4.56451684e-01 2.29387254e-01 1.22156866e-01
1.50973320e+00 -6.91601932e-01 -8.40717673e-01 -3.93206149e-01
6.47299469e-01 -1.16457975e+00 1.44704449e+00 3.82375360e-01
-8.29730213e-01 -2.10562468e-01 -1.00851631e+00 4.54292297e-02
-6.52438879e-01 1.47375003e-01 4.83194023e-01 4.82417643e-01
-3.04079950e-01 4.55634952e-01 -3.27035636e-01 -7.66690299e-02
6.15529895e-01 -9.46232677e-02 5.35518192e-02 5.74393272e-02
-1.25961876e+00 8.43732595e-01 -3.44344787e-03 3.36335041e-02
-5.30314445e-01 -7.65224755e-01 -3.46950628e-02 -1.89249530e-01
8.15976858e-01 -7.91927427e-02 1.43617034e+00 -9.04232383e-01
-7.15698540e-01 7.36007750e-01 -2.17251629e-02 -4.00819659e-01
8.44307661e-01 -5.92832863e-01 -5.26359677e-01 -1.19815014e-01
5.65645516e-01 2.09778890e-01 9.32907879e-01 -8.47488344e-01
-1.12562144e+00 -2.79912591e-01 -7.52273723e-02 -2.62952387e-01
-2.97330052e-01 1.82471097e-01 -4.00629282e-01 -4.71306384e-01
7.93895721e-02 -1.10356939e+00 1.98472124e-02 -1.18958764e-01
-8.74203384e-01 -3.73512536e-01 5.28444052e-01 -9.46211934e-01
1.84800398e+00 -2.06487060e+00 -5.57443440e-01 3.17333341e-01
7.77106225e-01 4.47160840e-01 2.64008671e-01 5.17494678e-01
3.06424439e-01 6.58507705e-01 4.60528821e-01 2.77225077e-01
2.60466605e-01 -4.60547507e-01 -4.06865895e-01 2.52448797e-01
-1.07572619e-02 8.67931545e-01 -6.88874662e-01 -2.84076154e-01
-2.00914159e-01 -5.27750887e-02 -4.05343503e-01 2.04161182e-01
-3.37464929e-01 -1.72293000e-02 -9.14550424e-01 6.49423361e-01
6.94129348e-01 -4.76473838e-01 -4.21985447e-01 -3.62567380e-02
-5.23119867e-01 6.41815782e-01 -7.96158552e-01 5.30941606e-01
-3.13353747e-01 1.06133592e+00 -2.33061016e-01 -1.77163314e-02
8.29273224e-01 -2.24253833e-01 1.49000168e-01 -9.79960263e-01
2.06493102e-02 4.82118696e-01 -4.59995791e-02 -7.07359672e-01
5.73722482e-01 2.45961562e-01 -1.65796861e-01 4.68240261e-01
-7.92145669e-01 2.98102677e-01 3.63095284e-01 4.46859241e-01
1.18848300e+00 -4.76053417e-01 1.31158188e-01 2.88192439e-03
2.76930958e-01 3.43372673e-01 7.92574435e-02 8.80041718e-01
-3.22241992e-01 4.33807939e-01 7.38474667e-01 -5.46685338e-01
-1.09238458e+00 -7.70535767e-01 -1.34848226e-02 1.45377898e+00
-1.36906013e-01 -5.28537631e-01 -5.09973407e-01 -7.06535697e-01
3.03630769e-01 9.87792253e-01 -5.22165596e-01 5.82174547e-02
-2.76659399e-01 -5.73384345e-01 4.67599213e-01 1.44711956e-01
6.78816020e-01 -4.19231236e-01 -1.73936322e-01 1.33552521e-01
-4.23004389e-01 -9.30096626e-01 -7.66854703e-01 -1.60674080e-01
-2.92374074e-01 -1.14740551e+00 -9.40035284e-01 -3.27649593e-01
4.76749241e-01 4.98427123e-01 7.66980886e-01 1.47722498e-01
-3.49354774e-01 -1.40458822e-01 -3.95442277e-01 -5.28340876e-01
-6.78150728e-02 3.72101158e-01 -3.18913281e-01 2.06587672e-01
7.15658486e-01 9.30443034e-02 -6.48300886e-01 7.78778195e-01
-8.05882275e-01 -3.22920948e-01 4.52342808e-01 6.26772821e-01
4.99397852e-02 -2.34001145e-01 5.33354998e-01 -7.75459170e-01
1.02354014e+00 -5.15485108e-01 -8.44770372e-01 -6.65638922e-03
-5.39211869e-01 -3.68316412e-01 2.04339802e-01 -5.87774634e-01
-6.01337433e-01 -3.32602412e-01 6.92164451e-02 1.84728146e-01
2.83120573e-01 6.37007713e-01 3.54237556e-01 -8.41196328e-02
8.92555058e-01 7.32629374e-02 5.28886914e-03 -6.31612659e-01
-2.98200697e-02 1.02782798e+00 -1.86401054e-01 1.73160031e-01
9.93549585e-01 6.75379857e-02 -5.04606187e-01 -9.97935236e-01
-1.38801897e+00 -7.15264261e-01 -4.06230316e-02 -3.85486215e-01
6.81746423e-01 -7.74228156e-01 -1.10035062e+00 2.66461104e-01
-8.45384121e-01 1.94420442e-01 3.72276396e-01 5.00206292e-01
2.09286019e-01 2.15750933e-01 -3.57079655e-01 -8.94252777e-01
2.12485828e-02 -8.69825363e-01 5.92391849e-01 9.90959033e-02
-5.97747445e-01 -5.55236280e-01 2.81367842e-02 7.70301580e-01
8.44607353e-01 -3.76512222e-02 6.60454154e-01 -1.21066058e+00
-8.40203643e-01 -1.06821561e+00 -5.86951792e-01 3.41243923e-01
-1.22189522e-01 4.82133441e-02 -9.39182639e-01 7.21231252e-02
-2.92059302e-01 -5.35834372e-01 7.16914177e-01 8.47734064e-02
1.15958297e+00 -4.48349327e-01 -3.62689316e-01 -2.84399271e-01
1.07274127e+00 -1.33045584e-01 4.71473038e-01 7.74844408e-01
4.11147505e-01 6.07678235e-01 1.05788469e+00 5.21418154e-01
1.26007780e-01 8.42921078e-01 4.60843235e-01 1.41651064e-01
2.93789476e-01 -6.08225822e-01 1.35428339e-01 4.04992938e-01
1.10627018e-01 -1.34636387e-01 -7.46486902e-01 3.66283655e-01
-1.68709290e+00 -6.15848899e-01 -5.37069678e-01 2.57899642e+00
7.55540431e-01 4.69555706e-01 4.81077313e-01 -5.33708697e-03
9.37098503e-01 1.91865548e-01 -3.13134402e-01 5.74840754e-02
2.16975123e-01 -2.81725615e-01 9.75926936e-01 4.24141496e-01
-1.07304227e+00 8.01533103e-01 5.57423067e+00 1.00884187e+00
-9.65815783e-01 -4.43112068e-02 6.71928763e-01 -1.57914624e-01
-2.10510895e-01 -1.05786696e-01 -1.14237404e+00 9.34176922e-01
7.60196447e-01 -2.76650757e-01 2.73163170e-01 1.01606727e+00
6.63657784e-01 -4.86669779e-01 -6.82057321e-01 7.55241096e-01
-1.51054010e-01 -1.29296052e+00 -3.68804097e-01 2.91389465e-01
4.40288603e-01 4.66374569e-02 1.44294679e-01 4.82640773e-01
8.13112110e-02 -9.17072773e-01 7.19805837e-01 3.03190529e-01
2.94391632e-01 -2.15178683e-01 6.96579278e-01 4.56864715e-01
-5.66975594e-01 8.88505660e-04 -2.32271492e-01 -1.65910468e-01
1.40181288e-01 9.79101717e-01 -1.08220589e+00 -2.86155760e-01
4.94348705e-01 2.79769421e-01 -1.15021443e+00 1.54815102e+00
-1.74949199e-01 8.46186101e-01 -3.50446075e-01 -1.07317460e+00
4.88773704e-01 2.38066137e-01 5.95730782e-01 8.25718462e-01
1.62181146e-02 -4.67656553e-01 9.30441245e-02 8.04291129e-01
-1.58014789e-01 2.24984989e-01 -2.57690012e-01 -6.05777860e-01
5.52643657e-01 1.37642372e+00 -5.15802324e-01 -3.70979816e-01
-3.81854087e-01 6.23878360e-01 1.81443289e-01 3.07945669e-01
-1.15324926e+00 -5.84767640e-01 3.71929735e-01 7.81785071e-01
2.65123993e-01 -1.29954368e-01 -1.51401147e-01 -1.03302276e+00
3.47802430e-01 -8.89767766e-01 3.67200114e-02 -7.08264112e-01
-1.40806103e+00 4.58576798e-01 -3.12142491e-01 -1.14226568e+00
2.35593736e-01 -4.18617934e-01 -3.08965236e-01 8.56664717e-01
-1.42010403e+00 -8.17354679e-01 -5.58123827e-01 8.03631470e-02
4.94827896e-01 1.53431207e-01 1.08374245e-01 7.19123065e-01
-5.00739932e-01 4.35306579e-01 -1.72198180e-03 7.67915621e-02
9.39046800e-01 -1.03889716e+00 3.77811581e-01 2.89392918e-01
-8.82313922e-02 8.61314833e-01 9.56546247e-01 -8.52235138e-01
-1.14036381e+00 -9.14678514e-01 1.39308357e+00 -5.01087189e-01
1.45262384e+00 -4.58154798e-01 -7.99760997e-01 6.57355666e-01
-2.81741768e-01 -3.11722398e-01 6.47418380e-01 2.63036817e-01
-3.19679141e-01 -1.43496081e-01 -6.68433070e-01 7.11791575e-01
3.94931495e-01 -4.54564810e-01 -4.64263111e-01 7.38910794e-01
3.74729007e-01 3.56209539e-02 -4.48197633e-01 -3.25641066e-01
6.40204668e-01 -8.10442686e-01 6.53593838e-01 -6.82452440e-01
3.51937741e-01 -5.03065549e-02 3.97276785e-03 -1.11220789e+00
-3.28886330e-01 -5.78463554e-01 4.12625223e-01 1.28362525e+00
7.56690562e-01 -7.34349668e-01 8.02908242e-01 9.00727034e-01
2.19264060e-01 -4.77706403e-01 -7.16408610e-01 -7.17604637e-01
-3.83705556e-01 -3.34556282e-01 2.08891124e-01 4.28877681e-01
-1.98692113e-01 5.43464661e-01 -5.83876193e-01 -2.52096653e-01
5.59674263e-01 -1.09947652e-01 1.05402601e+00 -1.15574872e+00
5.81250079e-02 -3.53931636e-01 -1.53639823e-01 -1.38555622e+00
-5.19881666e-01 -6.26975834e-01 1.02840308e-02 -8.37183833e-01
1.89682379e-01 -5.02515852e-01 4.72758971e-02 2.33375832e-01
-3.25841546e-01 4.96664524e-01 2.25742772e-01 5.95173359e-01
-9.69513774e-01 3.00923228e-01 1.20361936e+00 -9.99867991e-02
-1.04504891e-01 7.04124570e-01 -8.15807998e-01 7.20440328e-01
7.38299429e-01 -5.49696982e-01 1.13458745e-01 2.42283382e-02
8.97569597e-01 -4.37155932e-01 5.41326761e-01 -5.93964100e-01
1.24486342e-01 -1.98977098e-01 1.79552317e-01 -7.01136708e-01
2.61284858e-01 -6.72874928e-01 -1.56406373e-01 4.62004334e-01
-8.51770878e-01 -2.27642298e-01 -3.04166734e-01 7.60069430e-01
-2.96634495e-01 -5.72032154e-01 4.63755906e-01 -2.56605111e-02
-4.06375647e-01 -2.30939183e-02 -7.10205317e-01 2.31239989e-01
9.70881283e-01 4.00934219e-02 -5.87581813e-01 -5.51919281e-01
-4.87472177e-01 2.38712132e-01 1.16804935e-01 5.02314806e-01
2.58153051e-01 -1.20530140e+00 -6.21159613e-01 4.33238260e-02
3.91387224e-01 -7.28964925e-01 -1.63202621e-02 9.67741668e-01
-6.21652782e-01 7.03510106e-01 2.19156295e-01 -5.22488356e-01
-1.31038463e+00 1.61174223e-01 -1.50494292e-01 -3.92878130e-02
-1.64007768e-01 7.94926286e-01 -6.52888268e-02 9.02360901e-02
3.12481672e-01 -1.96796268e-01 9.81457755e-02 4.82252061e-01
8.36108267e-01 8.62971604e-01 5.16558141e-02 -3.15971881e-01
-2.97532767e-01 8.77383873e-02 -5.09089828e-01 -1.33167624e-01
1.02820706e+00 -2.61395246e-01 -2.65592653e-02 5.39205492e-01
1.27781475e+00 3.81141245e-01 -9.04313564e-01 -1.11930847e-01
5.40592313e-01 -1.11185598e+00 1.57902434e-01 -9.66917336e-01
-5.64696550e-01 5.68522155e-01 2.02098906e-01 1.10335732e+00
3.19137692e-01 2.51403078e-02 9.55568969e-01 3.83442402e-01
1.09698802e-01 -1.25100517e+00 3.40357691e-01 2.79919475e-01
9.23137784e-01 -1.53111756e+00 2.67046094e-01 -4.63811070e-01
-7.11199939e-01 8.54653120e-01 4.34234560e-01 -1.05226152e-01
5.26638269e-01 -4.29630995e-01 -4.79519181e-02 -2.59981781e-01
-7.50289679e-01 -7.16608539e-02 4.35649097e-01 6.79927021e-02
4.27057385e-01 8.40395764e-02 -9.36956644e-01 6.15355432e-01
-8.53895545e-02 -8.90101213e-03 4.10991669e-01 8.09086025e-01
-1.06791580e+00 -6.02478325e-01 -3.90674710e-01 9.19519603e-01
-8.42763484e-01 -1.08635738e-01 -5.82428575e-01 9.80884194e-01
-4.61687505e-01 9.90837812e-01 -1.65531471e-01 -3.32382262e-01
3.46073747e-01 -9.61936191e-02 -3.07177752e-01 -3.19443882e-01
-5.83347797e-01 2.45577991e-01 5.92527509e-01 -6.07372880e-01
7.52964467e-02 -4.75330621e-01 -5.15287578e-01 -4.40039843e-01
-8.10971856e-01 3.53032678e-01 1.06415141e+00 7.16720641e-01
3.40092093e-01 9.20917466e-02 8.61484289e-01 -3.40605259e-01
-1.00000167e+00 -1.14994466e+00 -7.15715587e-01 4.81217474e-01
2.72524625e-01 -6.67612374e-01 -7.65255094e-01 -3.32931161e-01] | [7.754286289215088, 9.787952423095703] |
efcb6643-4f90-4b7c-bdcd-9513b28ebd06 | sparse-needlets-for-lighting-estimation-with | 2106.13090 | null | https://arxiv.org/abs/2106.13090v1 | https://arxiv.org/pdf/2106.13090v1.pdf | Sparse Needlets for Lighting Estimation with Spherical Transport Loss | Accurate lighting estimation is challenging yet critical to many computer vision and computer graphics tasks such as high-dynamic-range (HDR) relighting. Existing approaches model lighting in either frequency domain or spatial domain which is insufficient to represent the complex lighting conditions in scenes and tends to produce inaccurate estimation. This paper presents NeedleLight, a new lighting estimation model that represents illumination with needlets and allows lighting estimation in both frequency domain and spatial domain jointly. An optimal thresholding function is designed to achieve sparse needlets which trims redundant lighting parameters and demonstrates superior localization properties for illumination representation. In addition, a novel spherical transport loss is designed based on optimal transport theory which guides to regress lighting representation parameters with consideration of the spatial information. Furthermore, we propose a new metric that is concise yet effective by directly evaluating the estimated illumination maps rather than rendered images. Extensive experiments show that NeedleLight achieves superior lighting estimation consistently across multiple evaluation metrics as compared with state-of-the-art methods. | ['Ling Shao', 'Xuansong Xie', 'Feiying Ma', 'Shijian Lu', 'WenBo Hu', 'Changgong Zhang', 'Fangneng Zhan'] | 2021-06-24 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Zhan_Sparse_Needlets_for_Lighting_Estimation_With_Spherical_Transport_Loss_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Zhan_Sparse_Needlets_for_Lighting_Estimation_With_Spherical_Transport_Loss_ICCV_2021_paper.pdf | iccv-2021-1 | ['lighting-estimation'] | ['computer-vision'] | [ 2.69884437e-01 -6.94266081e-01 1.99294731e-01 -5.58834612e-01
-6.76571012e-01 -4.01140392e-01 3.83628845e-01 -5.08580267e-01
-3.27113010e-02 6.51781738e-01 4.61736433e-02 -6.38642460e-02
2.28043333e-01 -5.53387702e-01 -5.09591818e-01 -8.78244400e-01
4.06893194e-01 -1.04641177e-01 7.75963664e-02 -5.32659963e-02
2.48844147e-01 6.28239334e-01 -1.64197707e+00 -1.29054442e-01
1.05376148e+00 1.16910660e+00 3.52527916e-01 6.69170558e-01
-2.30957419e-01 4.84839499e-01 -5.01098990e-01 -2.48637810e-01
5.21102488e-01 -4.58839774e-01 -1.93805486e-01 1.48160249e-01
8.25008631e-01 -4.81573820e-01 -1.56596407e-01 1.00807750e+00
6.11760736e-01 3.50116491e-01 7.49731719e-01 -1.03684163e+00
-8.78388047e-01 -4.19177979e-01 -9.73994136e-01 -2.13772058e-02
5.85110128e-01 1.23028673e-01 6.70672238e-01 -1.08007276e+00
4.02908802e-01 1.24778676e+00 6.80366457e-01 1.94995150e-01
-1.33352280e+00 -4.62440372e-01 1.30892947e-01 4.64054421e-02
-1.62487543e+00 -2.69934237e-01 1.00901198e+00 -1.61012784e-01
6.39968812e-01 5.10979652e-01 5.70930779e-01 8.23400497e-01
2.91141599e-01 2.69225389e-01 1.65988481e+00 -5.19351780e-01
9.02012736e-02 1.85973793e-01 -1.37978360e-01 9.82865155e-01
-2.41493862e-02 1.27883464e-01 -6.16249084e-01 -8.74461234e-02
9.93505001e-01 -3.08264941e-01 -5.74252248e-01 -1.75369918e-01
-1.21017027e+00 4.67051059e-01 5.00416696e-01 -1.63891375e-01
-2.33146161e-01 3.75056416e-01 8.72172266e-02 -1.30915552e-01
7.16368854e-01 1.95346832e-01 -1.55876219e-01 1.62580580e-01
-8.53077352e-01 -9.42109227e-02 4.71730739e-01 9.71985936e-01
9.10350144e-01 2.83996701e-01 -3.44210386e-01 1.11349165e+00
4.99112904e-01 1.04149175e+00 -1.95597515e-01 -1.23707712e+00
1.95767954e-01 2.26879716e-01 2.87641525e-01 -1.07152092e+00
-2.84418166e-01 -1.37841716e-01 -9.08864737e-01 3.28353792e-01
1.94556415e-01 1.91342741e-01 -9.82548654e-01 1.53814924e+00
6.90350592e-01 4.24450010e-01 -2.11745635e-01 1.23786330e+00
8.33214402e-01 9.10864711e-01 -3.12404990e-01 -4.41865951e-01
1.25056291e+00 -9.01255727e-01 -1.13897121e+00 1.16809681e-01
3.91393937e-02 -1.22563052e+00 1.59501994e+00 7.11425602e-01
-1.10704136e+00 -4.30453688e-01 -9.94321883e-01 -6.02706552e-01
-2.60379463e-01 1.22607172e-01 5.98309994e-01 7.57076502e-01
-1.13284993e+00 2.36221775e-01 -4.12189573e-01 -3.01393956e-01
1.70528919e-01 -2.88478471e-02 2.05307052e-01 -4.59695876e-01
-8.48922491e-01 9.01141107e-01 -3.57304484e-01 1.72736928e-01
-6.80224240e-01 -7.02991128e-01 -8.28633785e-01 -3.31453651e-01
-4.31843400e-02 -7.66293228e-01 7.41865456e-01 -5.06391168e-01
-1.83740079e+00 5.71491241e-01 -6.65079474e-01 1.18200734e-01
6.71023607e-01 -1.15148257e-02 -3.18039000e-01 2.98813134e-01
-9.06256735e-02 5.14582753e-01 1.27556014e+00 -1.71907282e+00
-3.40335816e-01 -2.05890134e-01 -2.31677562e-01 5.59427321e-01
-3.08206707e-01 -1.56855006e-02 -7.37915337e-01 -5.87114275e-01
2.62388468e-01 -6.75659418e-01 6.76831976e-03 5.11636734e-01
-4.84876752e-01 1.51278913e-01 1.03882158e+00 -6.70990705e-01
1.14941525e+00 -1.98650920e+00 -8.90785381e-02 3.50929260e-01
-4.26462740e-02 -1.75181702e-01 -1.72579754e-02 7.48987123e-02
1.90016732e-01 -2.78333336e-01 -1.86785355e-01 -5.68719685e-01
1.72624856e-01 2.83103526e-01 -4.77202207e-01 9.17424381e-01
-2.46409267e-01 4.92929578e-01 -6.77647829e-01 -5.82374871e-01
8.92879248e-01 1.28835762e+00 -3.27117443e-01 1.76758900e-01
-1.53440312e-02 6.94404721e-01 -1.10582963e-01 8.77861142e-01
1.14829457e+00 -1.84419993e-02 -3.17067593e-01 -8.57194126e-01
-3.47267956e-01 -2.97728568e-01 -1.28721440e+00 1.62795794e+00
-8.82861316e-01 7.50168800e-01 8.44847262e-02 -2.07502067e-01
9.85018730e-01 -5.07813990e-02 3.46309215e-01 -1.05521774e+00
2.18450516e-01 1.31835505e-01 -9.22226429e-01 -4.54599589e-01
6.48655295e-01 -2.53422484e-02 3.36288065e-01 1.90188184e-01
-5.58715999e-01 -6.28699005e-01 -7.81174153e-02 -1.82037666e-01
6.04555726e-01 4.89971727e-01 2.96806451e-02 -2.10712999e-01
5.00827491e-01 -3.85802925e-01 5.40733635e-01 4.01288271e-01
-8.63560587e-02 1.01175451e+00 -5.38254008e-02 -3.65408659e-01
-1.05962384e+00 -1.05956757e+00 -5.06690085e-01 9.57123756e-01
6.48218453e-01 -6.32975101e-02 -7.52800167e-01 -1.80540934e-01
-1.34183392e-01 8.82898569e-01 -5.15983939e-01 2.20549181e-01
-3.52229625e-01 -8.75355542e-01 1.15487829e-01 1.88341126e-01
8.29292238e-01 -3.75951767e-01 -4.59106207e-01 -2.94967443e-01
-4.66693550e-01 -1.19054186e+00 -7.70716012e-01 -1.14639245e-01
-3.55724573e-01 -9.22184050e-01 -8.07749033e-01 -5.44214487e-01
9.70078588e-01 8.27418387e-01 1.09253263e+00 -8.36355537e-02
-7.93023407e-01 5.73301017e-01 -2.16070071e-01 -2.80566305e-01
1.15546964e-01 -5.46238482e-01 -1.31166562e-01 3.11757714e-01
-2.86226608e-02 -2.76001394e-01 -1.09162652e+00 7.72753000e-01
-8.09316278e-01 1.84411794e-01 2.49467567e-01 6.58461690e-01
9.37982023e-01 1.86053827e-01 -8.56612474e-02 -4.39064145e-01
4.31070060e-01 -6.96498081e-02 -9.11884725e-01 3.81580025e-01
-6.09465122e-01 -2.45448276e-01 6.66167080e-01 -3.53555501e-01
-1.59564674e+00 -1.03814021e-01 1.95549086e-01 -6.28553927e-01
7.55792111e-02 -3.54245961e-01 -4.86333519e-02 -7.51945555e-01
6.03425682e-01 2.93047279e-01 -2.87950605e-01 -3.32131624e-01
3.90085757e-01 4.38457817e-01 4.11342949e-01 -7.20032871e-01
9.54605460e-01 1.00020015e+00 4.49873358e-01 -1.12206256e+00
-9.50375021e-01 -5.26141286e-01 -2.52082020e-01 -5.63206255e-01
8.11946809e-01 -8.01756203e-01 -9.22278225e-01 4.56360936e-01
-1.11355603e+00 -4.06764925e-01 -1.06674917e-01 3.90047491e-01
-6.13506675e-01 4.22193706e-01 -4.81576592e-01 -9.90686536e-01
-2.41075113e-01 -1.26527834e+00 1.46932447e+00 1.63912743e-01
1.31452188e-01 -1.04355800e+00 8.89182463e-03 4.17802095e-01
5.45016646e-01 3.23922276e-01 6.82235718e-01 7.72504926e-01
-9.02900994e-01 1.58578277e-01 -6.97354555e-01 4.94501650e-01
1.20656922e-01 4.06452864e-01 -1.26932240e+00 -2.94115603e-01
-9.12008528e-03 -1.62576046e-02 6.76279902e-01 6.63604140e-01
1.49694014e+00 -8.28632489e-02 -2.75225993e-02 1.28481531e+00
2.05262661e+00 -2.68523060e-02 8.63802373e-01 2.01725781e-01
9.44921255e-01 5.36863387e-01 7.77353048e-01 5.50017059e-01
6.40200436e-01 7.88063645e-01 4.79241610e-01 -6.99359715e-01
-5.26611090e-01 -3.19208354e-02 1.42598212e-01 6.51263237e-01
-7.28364363e-02 -2.38021314e-01 -4.12303060e-01 1.86113730e-01
-1.47614241e+00 -5.42739272e-01 -2.72280723e-01 2.41120839e+00
7.41961539e-01 -4.94946331e-01 -3.36750507e-01 -3.29052843e-02
6.78856790e-01 2.23652318e-01 -6.68051600e-01 -6.33803368e-01
-3.71695042e-01 -2.22536009e-02 8.66021633e-01 7.03529477e-01
-6.62615716e-01 6.42316282e-01 7.00803328e+00 9.09241796e-01
-1.00354576e+00 2.05150425e-01 6.47791505e-01 -1.99454073e-02
-6.14560485e-01 -8.69055986e-02 -7.88380563e-01 6.28685832e-01
3.67889315e-01 2.09307179e-01 8.25404823e-01 5.12655377e-01
6.25187993e-01 -4.44321752e-01 -6.73116386e-01 1.35753870e+00
4.23514277e-01 -7.38257527e-01 3.85643207e-02 7.21796975e-02
9.67056632e-01 -2.49306932e-01 4.86171663e-01 -2.97305524e-01
1.95359096e-01 -8.28044891e-01 5.74327350e-01 7.91562438e-01
1.16236937e+00 -7.08118498e-01 2.96712041e-01 -1.63552836e-01
-1.38114095e+00 1.61103576e-01 -5.15100837e-01 3.46764565e-01
3.82939219e-01 6.93537354e-01 -5.90336561e-01 3.93969744e-01
7.61246741e-01 6.27862573e-01 -3.08986485e-01 1.06230497e+00
-3.54238898e-01 2.15356648e-01 -4.57369030e-01 2.36541092e-01
4.54551214e-03 -6.65584028e-01 4.89297092e-01 1.14848769e+00
5.62225044e-01 3.44076753e-02 2.60679543e-01 8.28778088e-01
9.65234190e-02 2.61686683e-01 -5.31591296e-01 6.72978580e-01
4.51590359e-01 1.44297886e+00 -6.90308571e-01 -1.06925601e-02
-5.02600908e-01 1.17391956e+00 -3.99767458e-02 9.28290963e-01
-1.18194222e+00 -4.07978594e-01 6.06208324e-01 2.07428753e-01
-2.92194039e-02 -3.04108679e-01 -4.94345337e-01 -1.06224227e+00
-6.23952318e-03 -2.34013975e-01 -1.08890511e-01 -1.48032510e+00
-1.27934456e+00 4.60140854e-01 3.95544060e-02 -1.34396207e+00
4.07900155e-01 -5.76684833e-01 -5.34419596e-01 9.54623938e-01
-2.29898667e+00 -1.46467984e+00 -8.56505573e-01 8.80973458e-01
4.95685875e-01 3.93803865e-01 6.42902911e-01 3.58096898e-01
-7.48652279e-01 4.80607241e-01 4.10305619e-01 -5.15526712e-01
1.01846790e+00 -1.31786430e+00 -7.15247989e-02 9.60176885e-01
-3.54947150e-01 5.46950579e-01 9.93619978e-01 -3.04865211e-01
-1.61559761e+00 -1.16369355e+00 3.66704732e-01 -1.68818787e-01
2.67463565e-01 -2.91687191e-01 -7.04361200e-01 2.72631139e-01
2.57993132e-01 2.21751139e-01 5.92834592e-01 -2.78744221e-01
-5.35327494e-01 -4.39977467e-01 -1.49831116e+00 7.08242059e-01
8.46488833e-01 -4.17148113e-01 8.89984742e-02 6.01348341e-01
6.40988171e-01 -5.80569625e-01 -9.01883721e-01 2.51206994e-01
5.27983248e-01 -1.12786949e+00 1.16858435e+00 5.41638136e-01
-2.32876861e-04 -7.46422470e-01 -5.76957762e-01 -1.25469983e+00
-1.36976168e-01 -9.30943668e-01 -1.61613330e-01 1.20477176e+00
-3.91011201e-02 -8.47642839e-01 4.03427422e-01 6.49289310e-01
-1.10499471e-01 -6.88496411e-01 -6.38783574e-01 -6.49313271e-01
-5.97223401e-01 -3.91560435e-01 6.20240510e-01 8.64844322e-01
-7.42320120e-01 1.94083005e-02 -5.77554226e-01 2.91353494e-01
1.40224183e+00 1.63396180e-01 7.50144303e-01 -8.42796206e-01
1.39452651e-01 -1.11101918e-01 -9.82894599e-02 -1.03401756e+00
6.31907880e-02 -3.64586949e-01 1.90877676e-01 -1.76852667e+00
1.30148619e-01 -5.95021665e-01 -1.27442449e-01 1.70779809e-01
4.69875969e-02 7.99714386e-01 -8.88012946e-02 1.27443671e-01
-5.28082371e-01 7.29843676e-01 1.61522543e+00 9.79091823e-02
-5.88740930e-02 -4.35680211e-01 -4.24662679e-01 7.54322886e-01
5.81813991e-01 1.32190093e-01 -6.97997808e-01 -7.10873544e-01
2.92609066e-01 -3.59089643e-01 4.98213768e-01 -8.62345040e-01
3.13094370e-02 -2.94747651e-01 7.55066276e-01 -6.69651806e-01
7.60460854e-01 -9.88809645e-01 1.75857529e-01 -2.03805566e-01
-4.57947738e-02 -1.37817293e-01 -7.41260573e-02 6.11189842e-01
4.27136049e-02 2.62265831e-01 1.29257166e+00 1.02721043e-01
-6.06058419e-01 4.18751448e-01 1.60797596e-01 -1.54640555e-01
1.02221787e+00 -6.13877475e-01 -4.98673677e-01 -3.61386687e-01
-1.23343490e-01 7.27237388e-02 8.46107364e-01 -9.39976275e-02
9.14130509e-01 -1.53412426e+00 -4.60461527e-01 3.10200214e-01
7.93322772e-02 -5.22446223e-02 4.64641631e-01 8.68713379e-01
-9.48254406e-01 1.46175072e-01 1.78574964e-01 -7.20439017e-01
-1.35105443e+00 3.12264740e-01 1.52396142e-01 1.49349302e-01
-6.89624012e-01 8.41867208e-01 4.70487803e-01 -2.58792937e-01
3.03392410e-01 -3.93510461e-01 4.63702939e-02 -1.84371904e-01
5.97378314e-01 6.27848744e-01 -2.03252900e-02 -7.97671914e-01
-9.40856189e-02 1.28120530e+00 2.82656223e-01 -7.02569867e-03
1.05842936e+00 -9.26624835e-01 -1.77760109e-01 3.86011302e-01
1.45502961e+00 -1.66540027e-01 -1.54701102e+00 -2.77198046e-01
-8.89175177e-01 -1.21479237e+00 5.82576513e-01 -8.28436136e-01
-1.00082207e+00 8.55678916e-01 8.37205827e-01 2.68730968e-02
1.56468678e+00 -5.85894227e-01 9.98248339e-01 7.23507404e-02
5.69605291e-01 -1.43267477e+00 2.18193978e-01 2.18003020e-01
7.83444166e-01 -1.29433954e+00 3.14471751e-01 -6.62144482e-01
-5.12389362e-01 9.07688677e-01 3.93505514e-01 1.22331657e-01
5.17120183e-01 3.50433886e-01 1.14544205e-01 -1.02897778e-01
-3.53499562e-01 -1.84548378e-01 4.64843124e-01 6.60650134e-01
2.74926543e-01 -1.22488417e-01 2.07526945e-02 -4.47735786e-01
1.05596125e-01 -2.49221846e-01 3.09246331e-01 5.75043857e-01
-5.04541874e-01 -6.36642516e-01 -8.26246142e-01 2.28648737e-01
-3.57039243e-01 -1.35267556e-01 1.50384277e-01 4.19043213e-01
3.37223100e-05 9.37667668e-01 -8.34084526e-02 2.82670222e-02
2.98698395e-01 -4.35543656e-01 6.88275099e-01 -1.82239398e-01
-3.97720300e-02 5.40471554e-01 -1.92025006e-01 -8.57221484e-01
-6.16244793e-01 -2.28039905e-01 -9.72609937e-01 -4.74962234e-01
-4.81069624e-01 -3.39523941e-01 1.27160168e+00 3.57448339e-01
1.21470593e-01 6.07789457e-01 1.12904382e+00 -1.22725475e+00
1.82494447e-02 -4.56763923e-01 -9.30483997e-01 3.14536572e-01
4.58302855e-01 -9.51820076e-01 -7.08328605e-01 1.16014220e-01] | [9.95829963684082, -2.886890411376953] |
497f0b14-bee9-415d-9553-2c9cd72a4e78 | universal-domain-adaptation-in-ordinal | 2106.11576 | null | https://arxiv.org/abs/2106.11576v2 | https://arxiv.org/pdf/2106.11576v2.pdf | Universal Domain Adaptation in Ordinal Regression | We address the problem of universal domain adaptation (UDA) in ordinal regression (OR), which attempts to solve classification problems in which labels are not independent, but follow a natural order. We show that the UDA techniques developed for classification and based on the clustering assumption, under-perform in OR settings. We propose a method that complements the OR classifier with an auxiliary task of order learning, which plays the double role of discriminating between common and private instances, and expanding class labels to the private target images via ranking. Combined with adversarial domain discrimination, our model is able to address the closed set, partial and open set configurations. We evaluate our method on three face age estimation datasets, and show that it outperforms the baseline methods. | ['Boris Chidlovskii', 'Christian Wolf', 'Assem Sadek'] | 2021-06-22 | null | null | null | null | ['age-estimation', 'universal-domain-adaptation', 'age-estimation'] | ['computer-vision', 'computer-vision', 'miscellaneous'] | [ 5.68046749e-01 2.01043665e-01 -5.39465964e-01 -5.42665601e-01
-8.06956172e-01 -7.36221313e-01 8.10693920e-01 5.74257821e-02
-3.42431813e-01 8.97077739e-01 -8.27334076e-02 -1.95138454e-01
-3.90148491e-01 -6.09455526e-01 -6.55289710e-01 -9.06989813e-01
-9.83003229e-02 1.08426452e+00 -7.23373890e-02 7.75558203e-02
-3.24202403e-02 4.76604283e-01 -1.72862291e+00 2.20330819e-01
8.87221992e-01 1.16192305e+00 -5.80134869e-01 5.49180388e-01
2.10593611e-01 7.32180774e-01 -5.12322426e-01 -8.09258699e-01
3.63748372e-01 -3.58411968e-01 -1.02494395e+00 6.17533736e-02
9.39646363e-01 -3.98701251e-01 -1.02019638e-01 9.67847347e-01
6.19176805e-01 3.65085080e-02 1.48708189e+00 -1.68274105e+00
-8.74120533e-01 3.67621392e-01 -6.39367223e-01 -1.08178876e-01
3.54036391e-01 -5.48102379e-01 1.17065537e+00 -5.41694522e-01
5.20127296e-01 1.49778283e+00 7.10486710e-01 1.25861299e+00
-1.81145990e+00 -9.53559637e-01 -2.32489612e-02 1.97522476e-01
-1.22900498e+00 -5.37168801e-01 7.94411778e-01 -6.16150022e-01
1.77747965e-01 2.21297935e-01 -2.56985515e-01 1.65527630e+00
-3.89209628e-01 5.22728920e-01 1.71916664e+00 -6.16260886e-01
3.03377897e-01 2.53393114e-01 4.29439634e-01 4.71542776e-01
1.57479066e-02 1.91021785e-01 -3.51296842e-01 -4.53515023e-01
4.34721172e-01 -2.83922702e-01 1.71846405e-01 -5.74080348e-01
-5.00145435e-01 9.80442941e-01 -9.95626859e-03 -9.10893753e-02
-1.46740034e-01 -1.84380352e-01 3.13358635e-01 6.06898129e-01
6.52019083e-01 4.43341345e-01 -7.51472652e-01 3.28683525e-01
-6.43963277e-01 5.26771210e-02 7.89277256e-01 7.13471711e-01
5.37599742e-01 -4.88484114e-01 -1.44206598e-01 1.24660313e+00
-7.86114782e-02 4.18768108e-01 4.86656487e-01 -1.19494784e+00
6.34353161e-02 3.16602498e-01 -3.35392207e-02 -3.81265074e-01
-3.54796588e-01 -9.02923644e-02 -8.77051055e-01 4.49512392e-01
9.39752579e-01 7.42352158e-02 -7.78680027e-01 2.22560596e+00
4.01488602e-01 2.91346103e-01 1.32388428e-01 3.40512127e-01
5.99656403e-01 1.19017303e-01 2.98390031e-01 -4.19668555e-01
1.48382664e+00 -4.83661622e-01 -4.47974950e-01 1.10368103e-01
5.63163579e-01 -3.56246650e-01 1.00694895e+00 6.13071322e-01
-8.06473255e-01 -6.46030903e-01 -9.39895928e-01 -6.12530746e-02
-4.73457485e-01 -2.88783177e-03 5.38355529e-01 1.11068892e+00
-9.76851702e-01 7.97259152e-01 -2.02096164e-01 -3.31714094e-01
7.74363518e-01 7.32600212e-01 -7.91899323e-01 -2.02058211e-01
-1.11261332e+00 5.68612337e-01 1.62657604e-01 -7.07957745e-01
-7.08442986e-01 -7.36599624e-01 -7.91981101e-01 -2.30429634e-01
2.24679396e-01 -5.67656875e-01 9.90424037e-01 -1.18053424e+00
-1.52336919e+00 1.53105581e+00 -5.36499247e-02 -5.31925917e-01
5.23834646e-01 -1.25228137e-01 -3.72622430e-01 5.76611795e-02
1.04633346e-01 7.47870445e-01 1.14573705e+00 -1.45350063e+00
-5.17047644e-01 -9.60627615e-01 5.60606737e-03 6.34338260e-02
-6.50678158e-01 1.27793074e-01 -6.10604249e-02 -6.47360563e-01
-1.08316846e-01 -1.05916893e+00 1.29648015e-01 -1.73783496e-01
-2.04384342e-01 -7.50672698e-01 6.64802015e-01 -7.10277081e-01
1.09250844e+00 -2.31297469e+00 3.09223413e-01 3.92423600e-01
1.19015306e-01 -1.19682774e-01 -1.02082714e-01 -5.88617437e-02
-6.00614071e-01 2.14489132e-01 -3.63811761e-01 -6.17917478e-01
2.07344785e-01 5.51559746e-01 -4.11509842e-01 4.17912900e-01
2.28250489e-01 3.51832747e-01 -6.25598609e-01 -7.73947537e-01
-2.77421474e-01 7.05574988e-04 -6.82684481e-01 2.49026820e-01
-1.21258534e-01 7.28989959e-01 9.26123783e-02 6.21695995e-01
8.39714885e-01 -1.49159864e-01 5.36988199e-01 3.50126326e-01
5.92788279e-01 4.00982276e-02 -1.20964301e+00 1.14277518e+00
-4.18191999e-01 1.66485503e-01 6.66859448e-02 -1.31237996e+00
8.22326541e-01 1.54718667e-01 5.42571425e-01 -5.90509832e-01
7.78100640e-02 1.27568513e-01 3.21028419e-02 -3.18954557e-01
-1.33870587e-01 -2.81752199e-01 -4.06123757e-01 3.13796222e-01
3.26973706e-01 1.97330430e-01 -1.04351737e-01 1.37845829e-01
1.02540112e+00 8.77068415e-02 3.63436282e-01 -1.95097849e-01
8.49224508e-01 -6.05383635e-01 7.18190372e-01 8.22814524e-01
-3.80982369e-01 6.07976437e-01 1.03770339e+00 -1.43186167e-01
-8.39814842e-01 -1.41707349e+00 -7.60382593e-01 1.68364239e+00
-2.90012322e-02 9.67459660e-03 -6.50858760e-01 -1.44265497e+00
3.22429478e-01 5.22025824e-01 -1.13901794e+00 -1.86746165e-01
-4.57021058e-01 -7.39917815e-01 5.58168352e-01 5.53031266e-01
1.58728883e-01 -7.46409297e-01 3.71091962e-01 -2.50809461e-01
-5.25313690e-02 -1.24014187e+00 -2.06717774e-01 4.05991316e-01
-7.52933621e-01 -1.24577892e+00 -7.28989959e-01 -9.15130913e-01
7.49956369e-01 -5.82556784e-01 1.20868945e+00 -2.28198409e-01
-4.16291170e-02 4.83210891e-01 -3.26464683e-01 -2.50989974e-01
-6.90872490e-01 1.39212191e-01 4.83495414e-01 3.65983754e-01
5.42201817e-01 -8.53434324e-01 -2.27872550e-01 4.18508172e-01
-7.54213095e-01 -4.08013880e-01 4.08705324e-01 1.15609694e+00
3.11891675e-01 1.05530895e-01 8.19118440e-01 -1.56982470e+00
3.09223980e-01 -5.94845116e-01 -4.05423462e-01 4.09092903e-01
-8.28884363e-01 2.79449105e-01 7.00130880e-01 -6.72270000e-01
-8.10649753e-01 3.25206637e-01 -1.69516191e-01 -4.91391242e-01
-5.28372765e-01 -2.39610374e-01 -6.14175797e-01 -1.48830473e-01
9.12824273e-01 -1.39666051e-01 1.70153514e-01 -6.33408725e-01
2.41584033e-01 8.78694475e-01 6.19807124e-01 -1.01679158e+00
8.72852504e-01 5.49078465e-01 3.60373825e-01 -5.72618127e-01
-8.06923687e-01 -4.83734220e-01 -1.01313114e+00 3.20489928e-02
7.91851044e-01 -7.62182355e-01 -6.71334028e-01 4.04760540e-01
-8.76394153e-01 -1.84261471e-01 -2.48129413e-01 2.97005147e-01
-6.80104375e-01 5.36719620e-01 -5.21821797e-01 -8.58517051e-01
1.38546273e-01 -8.84111226e-01 1.01691175e+00 -4.28356826e-02
-2.98757523e-01 -9.50814068e-01 -1.63160905e-01 6.19905293e-01
-3.23843867e-01 3.01814765e-01 1.28419662e+00 -1.34885335e+00
1.81699991e-01 -1.81110069e-01 -1.17476471e-01 6.64172590e-01
1.34798726e-02 -2.86497176e-01 -1.24452770e+00 -2.31313571e-01
-2.27361903e-01 -7.15614676e-01 9.42618430e-01 1.88260406e-01
1.67575812e+00 -3.71634066e-01 -1.41587988e-01 5.42000532e-01
1.09571493e+00 3.24333385e-02 5.77076852e-01 6.70231953e-02
5.17435372e-01 9.85555351e-01 5.56313813e-01 3.52561682e-01
2.02344060e-01 7.10554421e-01 3.83020759e-01 1.05012521e-01
-6.54744804e-02 -2.02717587e-01 2.80311882e-01 2.61723757e-01
-1.75877452e-01 1.42426670e-01 -6.93932831e-01 3.49454492e-01
-1.56106019e+00 -9.50653255e-01 1.42531171e-01 2.33860803e+00
1.15775967e+00 3.58022898e-02 5.86635351e-01 4.24100518e-01
7.22147703e-01 -1.79632902e-01 -5.05797446e-01 -5.15536308e-01
-1.56688645e-01 6.34688795e-01 6.11469030e-01 5.07130206e-01
-1.53083599e+00 5.55373013e-01 6.83816528e+00 1.09207988e+00
-5.16415715e-01 3.48758042e-01 9.65540767e-01 2.72154212e-01
1.09427407e-01 -2.80757815e-01 -7.17771828e-01 5.50447106e-01
8.33795071e-01 2.72051752e-01 5.38498759e-01 9.85808790e-01
-6.78431451e-01 2.05277815e-01 -1.64359486e+00 9.02038753e-01
1.69516161e-01 -4.96875316e-01 -1.09208383e-01 3.25345993e-01
9.69497144e-01 -6.07751310e-01 6.37920439e-01 5.14246762e-01
5.92331052e-01 -1.12339163e+00 2.75397003e-01 8.76400098e-02
1.23803723e+00 -7.20217407e-01 5.65784216e-01 4.22264487e-01
-5.03975451e-01 -5.65373182e-01 -2.29003340e-01 -1.63182616e-02
-7.23517954e-01 2.00991794e-01 -6.39539242e-01 5.05122423e-01
7.87971318e-01 5.97091079e-01 -6.44091129e-01 4.83685941e-01
-3.62147510e-01 5.76020181e-01 -1.81426495e-01 4.65708375e-01
-3.04284006e-01 -3.91723439e-02 2.19211355e-01 8.05772603e-01
-2.65494343e-02 9.86645222e-02 4.40286770e-02 2.39936233e-01
-5.23662806e-01 1.30569413e-01 -7.20859170e-01 3.56942564e-01
4.96549308e-01 1.00420296e+00 -3.96586090e-01 -1.68339461e-01
-3.20641190e-01 1.31428599e+00 4.55102503e-01 2.42779166e-01
-7.51237571e-01 -3.85720730e-02 6.05436504e-01 1.86937645e-01
1.39192954e-01 2.43128270e-01 -3.99062812e-01 -9.15520608e-01
-1.10044092e-01 -1.00790882e+00 1.11197603e+00 -2.14287266e-01
-1.89528167e+00 8.84155929e-02 1.49444297e-01 -1.15785122e+00
-3.09393913e-01 -9.78083134e-01 -2.56494403e-01 5.97981393e-01
-1.46409500e+00 -1.34378207e+00 1.19939826e-01 8.20643365e-01
1.33580074e-01 -4.26464885e-01 9.49263990e-01 4.93947178e-01
-5.23256898e-01 1.25644732e+00 3.82641286e-01 2.89519668e-01
1.22153616e+00 -1.70161045e+00 -3.13224077e-01 3.28041703e-01
1.43808406e-02 4.35456187e-01 4.87561464e-01 -2.40030274e-01
-5.99717617e-01 -9.29752827e-01 9.46363807e-01 -9.59257424e-01
5.04219234e-01 -8.22034478e-01 -6.93786263e-01 7.72971928e-01
-1.29799575e-01 2.33089119e-01 1.08477151e+00 6.76080287e-01
-9.94037807e-01 -3.49874139e-01 -1.55436242e+00 2.00959325e-01
1.23713720e+00 -6.38460696e-01 -5.47302604e-01 3.93129647e-01
5.31769931e-01 -1.99634619e-02 -1.07111883e+00 6.19923651e-01
6.93110228e-01 -7.99493313e-01 1.28492486e+00 -1.03222609e+00
7.14838147e-01 1.50085449e-01 -2.35684186e-01 -9.99894738e-01
-3.80801767e-01 -4.67470318e-01 -2.17015103e-01 1.53252351e+00
2.38021851e-01 -8.46192837e-01 8.51194561e-01 3.86364222e-01
5.26112437e-01 -3.25114250e-01 -1.36108792e+00 -1.03118968e+00
5.52598655e-01 -4.93719168e-02 5.79061031e-01 1.18285513e+00
-3.96150738e-01 5.90969622e-01 -5.20960748e-01 3.66913766e-01
9.33918715e-01 -3.40265669e-02 5.80175221e-01 -2.01800179e+00
-3.93611073e-01 -4.21447039e-01 -4.83542055e-01 -5.80306411e-01
9.33518291e-01 -9.06603515e-01 -2.26634383e-01 -6.36010647e-01
1.62489384e-01 -8.35515261e-01 -6.51302457e-01 4.34808761e-01
-1.80723533e-01 4.27585572e-01 -1.67109489e-01 2.99402654e-01
-6.32075369e-01 2.17175677e-01 8.02821040e-01 -2.49384269e-01
9.26523060e-02 2.37676322e-01 -8.78645003e-01 7.44719565e-01
6.38285697e-01 -6.19189501e-01 -2.70174563e-01 1.50175080e-01
-6.79124966e-02 -2.85196695e-02 4.15015042e-01 -8.40310216e-01
-6.65318072e-02 2.52443328e-02 7.21982419e-01 -1.47499800e-01
3.02014232e-01 -1.12605238e+00 -3.53362620e-01 3.73411506e-01
-6.49519026e-01 -2.70357847e-01 -2.24121049e-01 5.63961327e-01
1.53613150e-01 -2.24284992e-01 1.18339944e+00 2.78961629e-01
-6.42130494e-01 4.52920288e-01 2.86951363e-01 1.82896405e-01
1.15444124e+00 -3.75945009e-02 -1.62546515e-01 -2.35807732e-01
-1.17584610e+00 1.67788908e-01 4.28598046e-01 3.69224191e-01
9.31604505e-02 -1.48272586e+00 -9.60803986e-01 3.50892484e-01
4.34689730e-01 -3.44781160e-01 1.15561724e-01 5.91659665e-01
3.80879454e-02 2.50967219e-02 -2.85272419e-01 -4.01644260e-01
-1.63836682e+00 9.04559970e-01 1.54169753e-01 -4.77855414e-01
1.03596151e-01 8.70836616e-01 5.29159725e-01 -8.16725433e-01
4.85283673e-01 4.93291855e-01 -5.97935319e-01 3.93163860e-01
2.99244642e-01 4.66443270e-01 -1.18141174e-01 -5.84074199e-01
-3.13853025e-01 5.25946677e-01 -1.00413524e-01 -1.08410418e-01
1.13621986e+00 -2.34671608e-02 -1.87660798e-01 3.53115261e-01
1.33590221e+00 1.11565381e-01 -9.21741843e-01 -3.27052057e-01
1.10498570e-01 -3.27758491e-01 -4.80541557e-01 -9.02481914e-01
-5.81466794e-01 8.42987597e-01 8.21604192e-01 3.30052793e-01
1.31833708e+00 3.53870720e-01 2.02141672e-01 1.13459744e-01
9.90366787e-02 -1.19128001e+00 2.31597334e-01 3.57822061e-01
7.15106308e-01 -1.55102181e+00 -1.28196388e-01 -5.42253852e-01
-4.29500908e-01 8.04673791e-01 8.25822592e-01 -6.79513291e-02
6.76121950e-01 3.41404565e-02 -2.10735396e-01 3.40981603e-01
-5.84557533e-01 -2.85707802e-01 3.75326037e-01 1.15087402e+00
6.47469684e-02 1.83568791e-01 -4.46886301e-01 9.32059586e-01
-8.80225748e-02 -3.17302018e-01 -8.16676840e-02 4.38049406e-01
7.25890547e-02 -1.62017369e+00 -6.94780827e-01 5.29762328e-01
-8.68228793e-01 3.70264612e-02 -6.80763006e-01 7.96796024e-01
5.90896726e-01 7.37585545e-01 2.33897835e-01 -5.79079211e-01
1.88152850e-01 5.18740177e-01 7.44325578e-01 -5.08327127e-01
-2.95145929e-01 -3.49753827e-01 1.86378673e-01 -2.27256924e-01
-3.70302290e-01 -9.98656332e-01 -7.00846195e-01 -1.38501287e-01
-8.87863040e-02 -1.47343446e-02 3.77480477e-01 7.43207753e-01
-5.70615791e-02 1.92543179e-01 1.15269458e+00 -3.62870276e-01
-8.77003074e-01 -8.83959770e-01 -8.65073383e-01 1.06469667e+00
3.62116009e-01 -8.97663116e-01 -5.56155622e-01 4.69269305e-02] | [10.339712142944336, 3.243985414505005] |
923e64bf-c776-4bfe-a6e0-550b45700105 | deep-understanding-based-multi-document | 2204.03494 | null | https://arxiv.org/abs/2204.03494v1 | https://arxiv.org/pdf/2204.03494v1.pdf | Deep Understanding based Multi-Document Machine Reading Comprehension | Most existing multi-document machine reading comprehension models mainly focus on understanding the interactions between the input question and documents, but ignore following two kinds of understandings. First, to understand the semantic meaning of words in the input question and documents from the perspective of each other. Second, to understand the supporting cues for a correct answer from the perspective of intra-document and inter-documents. Ignoring these two kinds of important understandings would make the models oversee some important information that may be helpful for inding correct answers. To overcome this deiciency, we propose a deep understanding based model for multi-document machine reading comprehension. It has three cascaded deep understanding modules which are designed to understand the accurate semantic meaning of words, the interactions between the input question and documents, and the supporting cues for the correct answer. We evaluate our model on two large scale benchmark datasets, namely TriviaQA Web and DuReader. Extensive experiments show that our model achieves state-of-the-art results on both datasets. | ['Bingchao Wang', 'Chunchao Liu', 'Jiaqi Wang', 'Huimin Wu', 'Shilei Liu', 'Yu Guo', 'Zhibo Wang', 'Bochao Li', 'Yongkang Liu', 'Feiliang Ren'] | 2022-02-25 | null | null | null | null | ['triviaqa'] | ['miscellaneous'] | [ 2.00856254e-01 3.19792569e-01 -1.77316274e-02 -6.89551830e-01
-8.76273036e-01 -7.27148652e-01 6.38453424e-01 7.56596923e-01
-1.81455597e-01 1.60224408e-01 7.18552589e-01 -5.48010528e-01
-2.27738813e-01 -9.18189466e-01 -8.58654976e-01 -2.08727509e-01
4.65514004e-01 7.40425289e-01 4.56035584e-01 -5.02911448e-01
5.97613335e-01 -4.76866186e-01 -1.45795143e+00 1.02724624e+00
1.16299784e+00 7.34453917e-01 4.05774623e-01 8.26256573e-01
-7.64021099e-01 9.33177471e-01 -6.44639313e-01 -4.67361033e-01
-2.93340564e-01 -6.44771159e-01 -1.35450602e+00 -4.45984602e-02
6.51787162e-01 -7.39133358e-01 -3.62622254e-02 1.08515632e+00
-1.34678438e-01 2.30785180e-03 8.22841465e-01 -7.32555091e-01
-1.03263974e+00 6.99940324e-01 -5.30010343e-01 2.77994245e-01
5.83563983e-01 -1.90352306e-01 1.43849039e+00 -6.55996144e-01
1.64640486e-01 1.79984379e+00 6.04335107e-02 4.71480072e-01
-5.80366135e-01 -2.28782177e-01 7.97310114e-01 6.17837548e-01
-4.77429986e-01 -2.81074107e-01 4.90650088e-01 -3.16544682e-01
6.87484920e-01 1.51691675e-01 1.75014064e-01 9.33547914e-01
4.60317641e-01 1.19256938e+00 9.78116512e-01 -5.47568262e-01
-2.41907127e-02 3.38468924e-02 1.17943156e+00 6.35523975e-01
7.06029162e-02 -3.32298696e-01 -4.60432768e-01 7.94853717e-02
2.13414311e-01 1.80912036e-02 -4.52124804e-01 2.91139167e-02
-1.03233266e+00 1.01855934e+00 2.04807490e-01 2.76230901e-01
-1.57702595e-01 -4.80500944e-02 2.24434495e-01 4.24125582e-01
3.35906118e-01 2.77907193e-01 -5.73670089e-01 -3.56597751e-02
-3.18580300e-01 1.96790934e-01 9.02476072e-01 6.89949751e-01
5.32144070e-01 -7.68218040e-01 -4.25728410e-01 6.48590028e-01
8.10850143e-01 5.93477309e-01 3.05539131e-01 -6.67906642e-01
9.40068126e-01 8.34816456e-01 6.98654726e-02 -1.05187225e+00
-2.67349511e-01 -4.51797336e-01 -5.19559562e-01 -3.27342808e-01
5.69301724e-01 4.01797006e-03 -1.00123668e+00 1.75176036e+00
2.24186182e-01 -3.42015684e-01 2.50093758e-01 8.65204036e-01
1.18042779e+00 6.78181827e-01 4.91834283e-02 1.83074296e-01
1.76240575e+00 -1.30256271e+00 -9.31298256e-01 -7.12177455e-01
6.82977557e-01 -8.50846589e-01 1.44690692e+00 1.32868469e-01
-1.04545689e+00 -5.69886029e-01 -9.27631557e-01 -7.08161294e-01
-3.83625567e-01 6.42135516e-02 2.70226896e-01 6.48212358e-02
-7.01472223e-01 1.55749291e-01 -4.68596429e-01 -2.84039140e-01
2.56988138e-01 -8.69745314e-02 2.02230826e-01 -6.57353878e-01
-1.34021783e+00 8.50111663e-01 1.19916096e-01 7.93675780e-02
-9.06172574e-01 -5.64551830e-01 -6.49857402e-01 5.54427862e-01
6.26470208e-01 -9.09509063e-01 1.58982527e+00 -1.00977767e+00
-9.27915752e-01 6.73485160e-01 -7.71824121e-01 3.96339707e-02
1.90064579e-01 -8.55777681e-01 -2.39898697e-01 2.99070150e-01
3.82060468e-01 4.83119279e-01 6.72902465e-01 -1.36257350e+00
-6.22440219e-01 -8.90475094e-01 5.19299209e-01 4.03123498e-01
-1.44715950e-01 -1.74412623e-01 -5.34601212e-01 -3.47041577e-01
3.70062679e-01 -2.68746316e-01 4.28387314e-01 -1.52144387e-01
-5.09964466e-01 -4.86180246e-01 8.20443273e-01 -1.00623381e+00
1.02181554e+00 -1.86975002e+00 2.42704719e-01 -1.68176472e-01
4.70741153e-01 -6.30314946e-02 -6.47828937e-01 7.13383734e-01
1.09285891e-01 1.85773432e-01 3.72670181e-02 -2.69355357e-01
-1.46842906e-02 3.19980711e-01 -7.96162546e-01 -2.53893852e-01
-1.74360871e-02 9.98061895e-01 -1.05557466e+00 -5.62938638e-02
-2.94418037e-01 1.24415927e-01 -2.57821560e-01 5.78980565e-01
-9.09685552e-01 1.66969344e-01 -8.08911860e-01 2.27445766e-01
5.18347502e-01 -5.90536058e-01 -7.98695982e-02 4.06264663e-02
4.78813857e-01 8.19169581e-01 -7.99740970e-01 1.52182543e+00
-4.34619993e-01 3.95763308e-01 7.82233626e-02 -6.69605792e-01
5.58987975e-01 1.24006279e-01 -4.71133977e-01 -1.00951731e+00
-2.43760701e-02 -1.63536668e-02 9.72531289e-02 -8.09440732e-01
3.11231971e-01 -1.04749009e-01 1.50719866e-01 1.10485280e+00
-9.91320685e-02 2.70496160e-02 1.68470442e-01 9.73610401e-01
7.08808661e-01 -2.89313942e-01 -1.19134121e-01 -3.63720834e-01
6.40736341e-01 -1.52458087e-01 -5.65672405e-02 1.01144063e+00
2.24990278e-01 5.44251382e-01 9.85291362e-01 -3.19209844e-01
-4.28679228e-01 -1.00612497e+00 4.95748818e-01 1.44627285e+00
5.72597027e-01 -6.03547990e-01 -7.81711698e-01 -8.62940371e-01
-1.91513509e-01 1.20849538e+00 -9.22414839e-01 -2.48216793e-01
-2.73037821e-01 -2.94813782e-01 1.22165471e-01 7.06442416e-01
7.71138132e-01 -6.29528463e-01 -4.27951068e-01 1.49099650e-02
-7.76642442e-01 -1.06102240e+00 -4.90051419e-01 -2.16183439e-01
-8.38020384e-01 -1.51525283e+00 -2.21231952e-01 -7.00915158e-01
7.11135924e-01 8.48862052e-01 1.53282702e+00 7.76575685e-01
3.23540926e-01 6.45221710e-01 -6.02703512e-01 -9.00824189e-01
-4.26947147e-01 7.23188161e-04 -5.87729692e-01 -1.41913310e-01
7.63490736e-01 -1.67041063e-01 -5.16133189e-01 4.15116102e-01
-9.78595853e-01 4.31945324e-01 4.62948322e-01 6.59344733e-01
2.68895537e-01 7.27173686e-02 5.25079608e-01 -1.04809046e+00
1.15303028e+00 -6.89107656e-01 -1.08094975e-01 9.77358043e-01
-3.14973801e-01 3.20550859e-01 5.42707622e-01 -1.33552228e-03
-1.27854550e+00 -9.40270364e-01 -2.41803169e-01 3.19778204e-01
-2.18610466e-01 5.51736891e-01 -4.21494961e-01 8.12602222e-01
1.94564521e-01 3.72581333e-01 -2.43610963e-01 -6.82801485e-01
4.80232954e-01 5.78469813e-01 4.10095811e-01 -8.39992583e-01
4.21145588e-01 4.26363647e-01 -4.98931170e-01 -5.20155489e-01
-1.85480034e+00 -6.31534278e-01 -5.46042383e-01 1.16058782e-01
9.53331530e-01 -7.39230573e-01 -5.87133825e-01 6.02724433e-01
-1.68234360e+00 7.53276870e-02 5.59264980e-02 3.42581496e-02
-2.66695738e-01 3.93549085e-01 -5.26439369e-01 -5.44261634e-01
-3.66118997e-01 -1.08162260e+00 1.24243057e+00 3.99822295e-01
-2.59362876e-01 -1.29685247e+00 -1.57769114e-01 9.28989947e-01
3.26863796e-01 -3.87234658e-01 1.80309784e+00 -9.86828566e-01
-7.19783664e-01 3.74321640e-03 -5.58802962e-01 1.71982467e-01
1.38629496e-01 -3.39322031e-01 -9.63528812e-01 -5.40217152e-04
3.79654080e-01 -6.26949131e-01 1.11037695e+00 1.36293828e-01
1.15064526e+00 -3.17686409e-01 -1.69034034e-01 1.31846480e-02
9.66607332e-01 -6.15009628e-02 4.12170261e-01 4.51671071e-02
4.68509018e-01 1.02887833e+00 5.99050283e-01 -7.92077631e-02
1.10588229e+00 1.62238762e-01 6.37113512e-01 6.87236786e-02
-8.74938965e-02 -6.34134650e-01 2.14279771e-01 9.05385256e-01
4.87100840e-01 -6.23670101e-01 -9.10522461e-01 7.43011475e-01
-1.98953772e+00 -6.39241099e-01 -4.60534900e-01 1.69654918e+00
6.34051263e-01 1.60954550e-01 -4.75832522e-01 -2.06920609e-01
4.09195662e-01 3.58458757e-01 -7.22209454e-01 -5.08072555e-01
-4.26943675e-02 -1.71859175e-01 -5.35421073e-01 8.07264805e-01
-6.10349894e-01 8.10031235e-01 5.61533260e+00 5.21074951e-01
-4.34205383e-01 3.00952066e-02 8.58198524e-01 2.23764405e-01
-9.28068578e-01 1.47516459e-01 -9.67150927e-01 4.18366008e-02
6.19463086e-01 -4.36036102e-02 2.14673832e-01 3.19714278e-01
1.13897786e-01 -7.47729123e-01 -1.38493848e+00 3.44765276e-01
3.91179591e-01 -1.05230176e+00 6.34306669e-01 -3.29912096e-01
3.34022045e-01 -2.36603081e-01 -1.14021309e-01 2.65030950e-01
3.20432007e-01 -1.17247975e+00 5.81957638e-01 5.17942309e-01
1.29991263e-01 -6.32771313e-01 9.52100754e-01 7.62189984e-01
-7.94470727e-01 6.82677403e-02 -3.53384316e-01 -3.42470020e-01
2.15547457e-01 4.22145307e-01 -4.01569396e-01 5.32660902e-01
5.19324243e-01 6.82307482e-01 -8.08124483e-01 3.74983639e-01
-9.53437686e-01 6.41490996e-01 5.69010712e-02 -1.54294074e-01
4.49140400e-01 -1.60171222e-02 2.69883662e-01 7.02966094e-01
4.32390757e-02 5.54329932e-01 -7.99321532e-02 9.90730941e-01
-2.59264410e-01 5.15843295e-02 -1.49376273e-01 -2.38339767e-01
1.77175924e-01 8.27589393e-01 -4.91912842e-01 -4.29174155e-01
-5.62492192e-01 9.35406983e-01 4.87325519e-01 4.88828987e-01
-4.59403664e-01 -4.01998848e-01 6.51481450e-01 5.00093922e-02
6.60188794e-02 -2.23109201e-01 -5.22963941e-01 -1.36019707e+00
3.73475552e-01 -1.22945976e+00 7.23333955e-01 -1.03539729e+00
-1.24698460e+00 2.54111409e-01 -1.01334408e-01 -3.94975185e-01
-1.11756381e-02 -6.47320330e-01 -1.09304070e+00 1.00951934e+00
-1.87235713e+00 -1.10563231e+00 -3.60418916e-01 3.64214092e-01
1.13516581e+00 2.92329878e-01 6.95918918e-01 -3.91386330e-01
-2.07274646e-01 1.44182086e-01 6.65830076e-02 2.40240544e-01
6.36018932e-01 -1.29764986e+00 5.62276781e-01 8.80622804e-01
1.79361537e-01 9.36681747e-01 5.69075286e-01 -6.36585832e-01
-1.33496475e+00 -4.45846230e-01 1.33298349e+00 -8.69036615e-01
6.05898201e-01 -4.01229173e-01 -1.38971281e+00 6.80198550e-01
6.23091459e-01 -8.93182516e-01 9.17510688e-01 5.31624615e-01
-7.67738402e-01 3.12350899e-01 -5.77497900e-01 4.76988554e-01
7.72195756e-01 -6.10322237e-01 -1.41473091e+00 5.60359716e-01
1.08636451e+00 -2.80489773e-01 -2.67605335e-01 1.09612659e-01
4.31954384e-01 -1.13541734e+00 7.61797369e-01 -1.09738731e+00
1.17145860e+00 -2.35569719e-02 -1.69660106e-01 -1.47709215e+00
4.48074527e-02 1.05804816e-01 -2.22371653e-01 1.10184443e+00
4.93735701e-01 -3.92032832e-01 1.41895354e-01 7.77553320e-01
-1.17039569e-01 -8.38504791e-01 -4.54819471e-01 -1.53477162e-01
4.64425147e-01 -4.38242257e-01 7.85853922e-01 6.37968600e-01
8.10324401e-02 9.51315582e-01 1.35150090e-01 2.99298257e-01
4.32932049e-01 6.44954443e-01 5.19498229e-01 -1.16152394e+00
-4.48013932e-01 -2.06043154e-01 4.34528410e-01 -1.88466704e+00
1.94778100e-01 -5.30722976e-01 -2.77557038e-02 -2.05125880e+00
6.97119772e-01 1.41157597e-01 -4.97477129e-02 2.79175371e-01
-7.64935553e-01 -6.72475100e-01 2.55233020e-01 1.18300214e-01
-9.65623200e-01 5.58950543e-01 1.65881348e+00 -2.54787058e-01
3.67657095e-01 -5.51106408e-02 -1.24550724e+00 8.91098261e-01
5.16334713e-01 -2.32806802e-01 -8.83369386e-01 -1.50421083e+00
4.46692318e-01 1.16381474e-01 3.97037119e-01 -2.99509466e-01
3.50213349e-01 -2.33099371e-01 2.88984299e-01 -7.23421395e-01
6.50487617e-02 -5.50165772e-01 -8.95987809e-01 2.78934360e-01
-8.32631469e-01 3.22844326e-01 1.48784146e-01 8.66522253e-01
-4.32177454e-01 -4.57512289e-01 3.88055146e-01 -3.67220014e-01
-5.36468446e-01 -8.89679492e-02 -2.26046965e-01 6.02181256e-01
4.96810943e-01 7.34639391e-02 -9.77248490e-01 -1.00521743e+00
-4.62524801e-01 9.94559884e-01 1.40516236e-01 9.36663449e-01
7.26536512e-01 -7.36016929e-01 -7.64705718e-01 1.36971250e-01
2.58052111e-01 1.66870981e-01 3.80902708e-01 3.43757004e-01
-7.17564449e-02 8.49551797e-01 2.54452080e-01 -4.03241307e-01
-1.39002502e+00 2.85884529e-01 3.89742970e-01 -4.62444544e-01
-2.45323673e-01 9.62722719e-01 9.06576693e-01 -5.10078490e-01
3.51487488e-01 -5.43739259e-01 -6.33752286e-01 8.21456537e-02
1.19738936e+00 1.58095568e-01 -4.22173925e-02 -2.53652364e-01
-3.82814482e-02 7.26104438e-01 -5.72361290e-01 8.09515566e-02
1.17292285e+00 -7.03391373e-01 -3.47091675e-01 4.80410069e-01
9.59441662e-01 -2.93130666e-01 -8.34868193e-01 -4.19282317e-01
1.85898423e-01 -4.40308362e-01 -1.58689055e-03 -1.31837678e+00
-7.16979921e-01 1.63223493e+00 -5.64420335e-02 4.13662158e-02
9.89546776e-01 3.49437684e-01 1.07838297e+00 6.38594627e-01
-8.65543932e-02 -8.64312530e-01 5.04347503e-01 9.00788307e-01
1.09428787e+00 -1.50815046e+00 -2.05918595e-01 -7.00860262e-01
-5.15289068e-01 1.42446673e+00 1.09110808e+00 5.10137141e-01
3.00015479e-01 -2.64304578e-01 3.75766724e-01 -6.13806367e-01
-1.17057586e+00 4.74623218e-02 4.81157959e-01 2.44647056e-01
6.58456802e-01 -1.80335715e-01 -1.25465244e-01 8.26028466e-01
6.17022999e-02 -3.53534937e-01 5.50720036e-01 8.03602040e-01
-7.92034447e-01 -9.60385203e-01 -2.52263725e-01 3.92738253e-01
-2.93820024e-01 -2.97516286e-01 -8.35495949e-01 4.28767949e-01
-3.77168864e-01 1.38019824e+00 3.52292918e-02 4.20605130e-02
2.75643140e-01 2.62284458e-01 2.70584881e-01 -7.97200143e-01
-5.34914911e-01 -2.87815928e-01 1.82265162e-01 -4.48915720e-01
-1.59824356e-01 -1.99174508e-01 -1.46904194e+00 -2.04178378e-01
-3.30346704e-01 3.37397844e-01 5.58276355e-01 1.75513351e+00
5.00815094e-01 5.46219826e-01 1.55154750e-01 2.97937751e-01
-7.50711977e-01 -8.64675581e-01 -2.03514889e-01 5.05946994e-01
4.25499201e-01 -2.18256667e-01 -2.37124875e-01 -1.34930998e-01] | [11.226245880126953, 8.050590515136719] |
da7828ef-9b1f-4ac4-a0af-714b9627f400 | benchmark-data-and-evaluation-framework-for | 2205.11966 | null | https://arxiv.org/abs/2205.11966v2 | https://arxiv.org/pdf/2205.11966v2.pdf | Benchmark Data and Evaluation Framework for Intent Discovery Around COVID-19 Vaccine Hesitancy | The COVID-19 pandemic has made a huge global impact and cost millions of lives. As COVID-19 vaccines were rolled out, they were quickly met with widespread hesitancy. To address the concerns of hesitant people, we launched VIRA, a public dialogue system aimed at addressing questions and concerns surrounding the COVID-19 vaccines. Here, we release VIRADialogs, a dataset of over 8k dialogues conducted by actual users with VIRA, providing a unique real-world conversational dataset. In light of rapid changes in users' intents, due to updates in guidelines or in response to new information, we highlight the important task of intent discovery in this use-case. We introduce a novel automatic evaluation framework for intent discovery, leveraging the existing intent classifier of VIRA. We use this framework to report baseline intent discovery results over VIRADialogs, that highlight the difficulty of this task. | ['Noam Slonim', 'Yoav Katz', 'Pooja Sangha', 'João Sedoc', 'Naor Bar-Zeev', 'Rose Weeks', 'Dan Lahav', 'Roni Friedman', 'Assaf Toledo', 'Shai Gretz'] | 2022-05-24 | null | null | null | null | ['intent-discovery'] | ['natural-language-processing'] | [-3.00567448e-02 5.73816262e-02 -2.86568075e-01 -2.27303147e-01
-6.83954656e-01 -1.09840679e+00 9.45329070e-01 4.46121365e-01
-4.33867872e-01 8.17975521e-01 8.79071712e-01 -6.33214593e-01
2.69390762e-01 -5.38560629e-01 -4.51904312e-02 -7.79008195e-02
-2.83586293e-01 8.43857765e-01 -8.11939463e-02 -7.34245300e-01
2.13054731e-01 1.89164251e-01 -9.05016184e-01 3.73060763e-01
9.32876408e-01 3.03558379e-01 -1.65537357e-01 7.19628274e-01
-5.46338521e-02 7.68798649e-01 -9.19640481e-01 -5.15782297e-01
-1.55582234e-01 -2.45413676e-01 -1.11266387e+00 -5.90771735e-01
-1.22604780e-01 -9.33683634e-01 -1.00303814e-01 5.46512961e-01
5.80514789e-01 -3.58274169e-02 5.89517236e-01 -1.35162783e+00
-2.48332292e-01 3.91976207e-01 -2.79014558e-01 3.47928971e-01
1.04468739e+00 3.65021348e-01 9.79089677e-01 -3.14938962e-01
9.65328336e-01 1.15726328e+00 1.03288174e+00 7.65703142e-01
-1.07271838e+00 -5.30141950e-01 -1.54451877e-01 -3.56425881e-01
-1.06088805e+00 -5.30926824e-01 2.71822721e-01 -8.35620522e-01
1.31356263e+00 6.69743657e-01 5.96126139e-01 1.68422413e+00
2.23438859e-01 6.19417071e-01 8.09096634e-01 2.26717964e-01
2.59359181e-01 1.74836293e-01 4.42789495e-01 4.90859836e-01
1.18067347e-01 -1.81220740e-01 -2.82922328e-01 -1.21765327e+00
1.62595902e-02 -1.16165346e-02 -3.93159658e-01 3.25050235e-01
-9.56180513e-01 1.28529668e+00 -1.99933812e-01 2.49937162e-01
-5.94552100e-01 -3.27450752e-01 7.76615560e-01 2.16544777e-01
6.17653608e-01 6.25615716e-01 -6.36785626e-01 -7.98398852e-01
-3.23147535e-01 5.46697676e-01 1.43704009e+00 4.46811646e-01
4.09962475e-01 -5.32003462e-01 -2.00162023e-01 7.48056054e-01
2.19977930e-01 4.74380165e-01 1.21681362e-01 -5.08708000e-01
2.86485404e-01 5.07038414e-01 4.70705867e-01 -1.25248957e+00
-9.14854646e-01 5.26621751e-02 -5.19640446e-01 -5.03058195e-01
2.06174150e-01 -7.52613008e-01 -4.71877217e-01 1.87229037e+00
4.64218825e-01 -1.72577992e-01 9.15422738e-02 5.44009686e-01
8.77115965e-01 7.57992566e-01 1.17323361e-01 -3.81937355e-01
1.67011082e+00 -1.41580224e-01 -8.45707774e-01 1.61037699e-03
8.30996275e-01 -8.39525342e-01 6.88842595e-01 9.66182724e-02
-4.95388567e-01 3.22334766e-01 -6.60235465e-01 3.33707064e-01
-5.53038180e-01 -6.84470177e-01 7.16183066e-01 1.03051281e+00
-1.04207206e+00 1.40289769e-01 -5.50035954e-01 -1.03681850e+00
1.42832994e-02 1.19005747e-01 -1.27307519e-01 2.55091846e-01
-1.40002859e+00 9.38814342e-01 1.47783548e-01 -2.16140181e-01
-7.03635931e-01 -6.98688447e-01 -6.91505373e-01 -5.06951392e-01
5.16189873e-01 -5.05645573e-01 1.40104198e+00 -1.23158887e-01
-1.12682235e+00 7.07965434e-01 -3.00990865e-02 -2.41187245e-01
4.27124828e-01 -2.61354238e-01 -6.98123038e-01 -5.36204092e-02
2.76719511e-01 3.34726781e-01 2.98694700e-01 -6.21899426e-01
-3.17615360e-01 -1.60659224e-01 2.22706005e-01 -5.31402081e-02
-1.20963506e-01 5.39213657e-01 -1.18332252e-01 -5.34841061e-01
-9.09563422e-01 -9.48657095e-01 -8.33493918e-02 -8.10911000e-01
-5.04354298e-01 -5.64526737e-01 1.00987792e+00 -8.57452929e-01
1.15179729e+00 -1.78476107e+00 -3.75090659e-01 4.73443195e-02
4.76236433e-01 5.21151960e-01 -3.26161027e-01 1.14924836e+00
3.47368747e-01 5.61459005e-01 1.05174720e-01 7.32443035e-02
-7.41430447e-02 1.22973897e-01 -4.69761282e-01 3.01283091e-01
4.35423863e-04 8.09690535e-01 -1.32685232e+00 -1.57688186e-01
-6.44424036e-02 4.65658814e-01 -7.22642303e-01 5.91823876e-01
-6.81314707e-01 4.83311653e-01 -7.72639155e-01 6.36747897e-01
5.89145720e-01 -5.12216926e-01 4.79618609e-01 -7.54756629e-02
-3.13759178e-01 6.42215192e-01 -3.87236513e-02 1.23833728e+00
-2.09830239e-01 5.11134744e-01 1.73814386e-01 -3.46546054e-01
4.85295147e-01 5.66894293e-01 6.82193458e-01 -3.15781772e-01
3.58026326e-01 -1.48401752e-01 -2.58318245e-01 -8.12084377e-01
6.54882908e-01 3.04178987e-02 -4.14899200e-01 9.86655414e-01
-3.37387800e-01 1.01192497e-01 -3.17280293e-02 5.30365705e-01
1.51882613e+00 -5.43083668e-01 5.68794250e-01 -2.72106081e-01
5.63844666e-02 3.86001021e-01 3.31898093e-01 7.65666664e-01
-5.42457342e-01 1.98879570e-01 5.55595696e-01 -6.00036860e-01
-5.02170920e-01 -7.12597668e-01 1.56211168e-01 1.23068964e+00
-2.25226566e-01 -8.33696842e-01 -6.54880345e-01 -9.26342845e-01
-5.41473832e-03 6.90483809e-01 -6.95612192e-01 2.42519900e-01
-4.15406853e-01 -1.03586876e+00 8.09156239e-01 -8.19259360e-02
3.71634632e-01 -8.28281462e-01 -6.18408382e-01 5.14494836e-01
-7.94458091e-01 -1.21126306e+00 -6.98201895e-01 -3.88071805e-01
-5.01308925e-02 -1.25898051e+00 -5.52706242e-01 -1.94860041e-01
1.89467326e-01 2.71518588e-01 1.09761429e+00 1.29731432e-01
-4.83631045e-01 6.68208897e-01 -4.57311839e-01 -3.48319739e-01
-8.99818182e-01 1.70529664e-01 3.29167604e-01 -3.64446968e-01
5.93698263e-01 -1.96297437e-01 -6.70222163e-01 3.68758380e-01
-7.39898503e-01 -7.82145858e-02 -8.94004256e-02 4.17975724e-01
-7.31797040e-01 -4.11320210e-01 8.92167509e-01 -1.15691900e+00
1.16919255e+00 -1.10384333e+00 -4.72757250e-01 3.37700434e-02
-4.25534248e-01 -3.74307543e-01 3.00612122e-01 -1.40378252e-01
-1.03782248e+00 -6.75177813e-01 -6.19142950e-01 5.88208675e-01
-4.64514866e-02 7.48869956e-01 5.01652956e-01 1.94208473e-01
7.29917824e-01 -1.78167343e-01 3.48057747e-01 -5.64960182e-01
3.92400354e-01 1.30077577e+00 2.13349387e-01 -3.23098809e-01
4.56482410e-01 2.97168612e-01 -8.94275129e-01 -1.28429449e+00
-5.90320945e-01 -7.21903741e-01 1.73954889e-01 -2.67075151e-01
9.21955526e-01 -9.14185107e-01 -1.37561762e+00 7.53742635e-01
-1.58268034e+00 -3.07603985e-01 4.57575977e-01 1.00227222e-01
-4.60313410e-02 4.72637028e-01 -9.76107657e-01 -8.02430093e-01
-6.93562984e-01 -9.12088037e-01 5.81706107e-01 1.75091147e-01
-1.19096112e+00 -1.20398688e+00 8.06805551e-01 6.14622474e-01
9.74902213e-01 3.56557816e-01 9.32116866e-01 -1.39918280e+00
-2.23063100e-02 -1.56049505e-01 -3.07300091e-01 -1.95421234e-01
7.80819237e-01 -2.49502808e-01 -8.08756113e-01 -5.72133005e-01
-1.13505527e-01 -7.65613735e-01 1.93525538e-01 -2.37820476e-01
3.30962420e-01 -9.59671974e-01 -5.02429605e-01 -1.33205369e-01
9.79834318e-01 4.13526714e-01 1.82745248e-01 1.69190448e-02
2.09342048e-01 6.85050964e-01 2.84248531e-01 9.99781370e-01
7.20642090e-01 7.53167868e-01 1.45465389e-01 1.39510483e-01
3.46076041e-01 -2.71840662e-01 3.86973977e-01 8.70309353e-01
-3.16964686e-02 -6.05361640e-01 -1.21328068e+00 5.28048515e-01
-1.57014894e+00 -9.06149507e-01 2.41115302e-01 1.90399206e+00
1.15385175e+00 -1.37459829e-01 4.91246670e-01 -8.16862106e-01
4.33767349e-01 5.18206120e-01 -3.48974884e-01 -7.34082997e-01
2.77443767e-01 -2.77518511e-01 1.70871153e-01 7.41154552e-01
-9.88702059e-01 6.69099987e-01 7.99864197e+00 3.65567058e-01
-1.09715581e+00 1.52848542e-01 4.55431879e-01 1.27512485e-01
-5.44802725e-01 -2.39093363e-01 -5.92771053e-01 5.98563612e-01
1.03921831e+00 -4.10533875e-01 4.57144201e-01 4.59543556e-01
2.62481600e-01 -1.09404467e-01 -1.00410450e+00 5.51699817e-01
2.25080356e-01 -1.62711763e+00 -1.74114764e-01 1.39370590e-01
4.36301529e-01 5.68004131e-01 -1.75037667e-01 4.09791112e-01
8.60701144e-01 -8.80829334e-01 -2.21496299e-01 1.04978077e-01
7.34349191e-01 -5.03298581e-01 7.78178155e-01 5.13147950e-01
-7.14019299e-01 3.83937240e-01 2.06089869e-01 -1.18859053e-01
5.02414584e-01 3.79017830e-01 -1.68495119e+00 6.15588576e-02
3.89749825e-01 4.15990084e-01 -2.21809626e-01 5.31275749e-01
3.77705812e-01 8.27036142e-01 -3.57615173e-01 -4.66696888e-01
2.55953670e-01 -7.52111003e-02 8.12367737e-01 1.55059826e+00
-2.59958804e-01 4.20058817e-01 2.03882620e-01 5.55966735e-01
-4.54002991e-02 2.73571610e-02 -9.97917593e-01 -7.56364048e-01
3.78595531e-01 1.22859740e+00 -3.42829525e-01 -2.78603256e-01
-4.14847940e-01 8.85273874e-01 1.85808882e-01 3.71147156e-01
-7.53252745e-01 -2.88904846e-01 1.11026549e+00 -1.15425838e-03
-1.03809498e-01 -1.61186587e-02 4.32406485e-01 -9.08792734e-01
-5.38061857e-01 -1.34008658e+00 5.75505376e-01 -4.18458074e-01
-1.43429422e+00 5.51728487e-01 2.13755742e-01 -4.85654503e-01
-7.98485994e-01 -1.94652110e-01 -7.53193498e-01 6.56601191e-01
-8.10828447e-01 -8.27358842e-01 6.05082512e-02 1.89641967e-01
4.37551171e-01 -6.07299954e-02 1.04121339e+00 1.06908619e-01
-4.24760044e-01 4.27580416e-01 -7.37806782e-02 1.53559476e-01
8.37939024e-01 -6.94328368e-01 8.79170477e-01 2.55213916e-01
-7.07248628e-01 1.08083618e+00 1.20830429e+00 -1.06081212e+00
-1.63879204e+00 -8.71972799e-01 1.18896222e+00 -8.58421981e-01
9.70830500e-01 -7.88049459e-01 -8.39646518e-01 9.15943742e-01
6.90247059e-01 -1.01126456e+00 1.10758030e+00 4.58747089e-01
-4.40528721e-01 5.64523160e-01 -1.33220935e+00 8.39764953e-01
8.44906449e-01 -7.26544976e-01 -7.38046110e-01 7.84229338e-01
1.04298615e+00 -2.36660928e-01 -7.03438997e-01 4.52805936e-01
6.13574624e-01 -8.46377969e-01 6.59934938e-01 -8.87815237e-01
-4.84513119e-02 2.64311045e-01 7.95090795e-02 -1.29313648e+00
-3.35610360e-02 -1.37981200e+00 8.52453858e-02 1.05991387e+00
3.81735712e-01 -1.02666843e+00 5.88138759e-01 8.14785898e-01
4.13123637e-01 -4.92094815e-01 -6.90579236e-01 -4.21369642e-01
-2.13821933e-01 -3.36014003e-01 4.99629140e-01 1.34126806e+00
7.56214976e-01 6.48488522e-01 -8.21738482e-01 -1.90918490e-01
4.29723144e-01 -1.02028146e-01 9.10735428e-01 -1.03735387e+00
-3.41544211e-01 -1.67472377e-01 -3.44146788e-02 -8.83594334e-01
-2.86647409e-01 -5.32882154e-01 1.23164570e-02 -1.29997206e+00
5.02589107e-01 -4.85999584e-01 3.02050471e-01 6.98253274e-01
-1.90791324e-01 8.90723616e-02 5.85225113e-02 1.30601257e-01
-5.93223214e-01 1.80910051e-01 4.50527400e-01 -2.22389266e-01
-3.87881517e-01 -2.05403790e-02 -9.51500118e-01 6.80435956e-01
9.58337724e-01 -1.50293306e-01 -4.27415788e-01 3.06792203e-02
7.37753153e-01 1.02424763e-01 -2.13662665e-02 -3.47167403e-01
5.14885560e-02 -3.84015977e-01 -2.32884422e-01 -8.94236505e-01
3.52319807e-01 -2.05758542e-01 3.49086374e-01 7.50362754e-01
-1.06208712e-01 1.12750426e-01 3.53156030e-01 4.96226877e-01
3.58610898e-01 1.42562479e-01 3.38541299e-01 1.61171928e-01
-4.34799343e-01 1.30450293e-01 -1.18826628e+00 6.73241138e-01
9.22320426e-01 3.52984607e-01 -1.06345570e+00 -7.51863956e-01
-2.43170023e-01 5.60955346e-01 5.65699041e-01 8.41885030e-01
4.62237388e-01 -8.69384587e-01 -8.50673497e-01 9.54297110e-02
3.49217296e-01 -6.71265066e-01 3.18628252e-01 8.12848210e-01
-4.83587921e-01 5.93507767e-01 -1.22403093e-01 -2.98225671e-01
-1.41845620e+00 5.50940156e-01 1.18407741e-01 -3.48904431e-01
-6.27380371e-01 5.57566583e-01 2.51056194e-01 -7.04773903e-01
1.81561202e-01 -4.61516939e-02 -2.72877395e-01 2.36228809e-01
9.97297883e-01 3.94166440e-01 1.44105433e-02 -4.46458727e-01
-7.85754621e-01 -1.91108510e-01 -5.40547073e-01 -2.41393059e-01
9.41062868e-01 -1.04892475e-03 -2.61005133e-01 1.96566150e-01
1.38378203e+00 3.90484661e-01 -3.98615062e-01 -6.38497323e-02
-1.06667206e-01 -2.88975686e-01 -5.97155094e-01 -1.17651057e+00
-2.63051033e-01 1.31723076e-01 2.51800627e-01 6.02283835e-01
5.05413890e-01 4.58921231e-02 1.25945616e+00 4.76611972e-01
4.94985282e-01 -7.92652786e-01 4.70173396e-02 8.53602111e-01
9.14917886e-01 -1.34254599e+00 -1.41078845e-01 -1.84513271e-01
-5.61570525e-01 5.12955964e-01 2.23381400e-01 7.35810637e-01
6.50990844e-01 2.22390592e-01 5.09544551e-01 -5.95206916e-01
-1.07988155e+00 2.84946293e-01 -2.53151327e-01 8.22427690e-01
4.99505788e-01 2.63594866e-01 -5.68921447e-01 4.50556785e-01
-1.09772384e-02 -5.95959164e-02 6.86948299e-01 1.09420049e+00
-4.38518494e-01 -9.74326372e-01 -2.74637818e-01 5.42421579e-01
-4.64115411e-01 -1.13418981e-01 -8.13440740e-01 5.53788126e-01
-3.70384127e-01 1.39442444e+00 -1.45807490e-01 -6.24148130e-01
-5.17286640e-03 -1.20662954e-02 -1.55259207e-01 -5.29206872e-01
-7.08556831e-01 -3.20547491e-01 9.45683062e-01 -6.01035297e-01
-2.63390511e-01 -6.33792341e-01 -1.16138518e+00 -8.28992188e-01
1.13918968e-01 4.00242209e-01 5.86921275e-01 1.04784119e+00
5.88139832e-01 -2.10956216e-01 6.92554653e-01 -3.04374278e-01
-6.15064144e-01 -8.10611606e-01 -7.96655491e-02 3.58875722e-01
6.00659430e-01 -2.19662756e-01 -2.84790158e-01 -3.57774526e-01] | [8.452095031738281, 9.435580253601074] |
51f758c8-0b8e-4b73-bd42-c4c376235674 | autonomous-uav-exploration-of-dynamic | 2010.07429 | null | https://arxiv.org/abs/2010.07429v3 | https://arxiv.org/pdf/2010.07429v3.pdf | Autonomous UAV Exploration of Dynamic Environments via Incremental Sampling and Probabilistic Roadmap | Autonomous exploration requires robots to generate informative trajectories iteratively. Although sampling-based methods are highly efficient in unmanned aerial vehicle exploration, many of these methods do not effectively utilize the sampled information from the previous planning iterations, leading to redundant computation and longer exploration time. Also, few have explicitly shown their exploration ability in dynamic environments even though they can run real-time. To overcome these limitations, we propose a novel dynamic exploration planner (DEP) for exploring unknown environments using incremental sampling and Probabilistic Roadmap (PRM). In our sampling strategy, nodes are added incrementally and distributed evenly in the explored region, yielding the best viewpoints. To further shortening exploration time and ensuring safety, our planner optimizes paths locally and refine them based on the Euclidean Signed Distance Function (ESDF) map. Meanwhile, as the multi-query planner, PRM allows the proposed planner to quickly search alternative paths to avoid dynamic obstacles for safe exploration. Simulation experiments show that our method safely explores dynamic environments and outperforms the benchmark planners in terms of exploration time, path length, and computational time. | ['Kenji Shimada', 'Di Deng', 'Zhefan Xu'] | 2020-10-14 | null | null | null | null | ['safe-exploration'] | ['robots'] | [ 9.47163329e-02 2.34961510e-01 -1.61711082e-01 -1.84012383e-01
-5.33654332e-01 -9.24751818e-01 2.55608350e-01 2.55872644e-02
-4.31066036e-01 1.04573536e+00 1.02480590e-01 -3.96564454e-01
-5.75626433e-01 -1.30313241e+00 -4.73788649e-01 -6.00472033e-01
-5.16401947e-01 7.73564935e-01 6.41272366e-01 -2.89060742e-01
2.71215171e-01 6.63958311e-01 -1.34381413e+00 -5.74568033e-01
1.37479389e+00 6.44309461e-01 5.67634106e-01 2.30598286e-01
2.13628814e-01 1.86019227e-01 -2.97926247e-01 4.58361506e-01
6.03815317e-01 -8.56017023e-02 -9.64421868e-01 -1.61562920e-01
-4.19476092e-01 -7.35332072e-01 -4.13447678e-01 1.08828104e+00
3.50071698e-01 6.35510981e-01 2.30330288e-01 -1.50942886e+00
-3.61780897e-02 6.33354902e-01 -6.27935946e-01 -3.30660075e-01
5.82834005e-01 3.94719362e-01 5.53007185e-01 -6.50093734e-01
8.74163747e-01 1.26363564e+00 2.48209029e-01 1.92130417e-01
-7.32547045e-01 -6.95937037e-01 5.75121999e-01 1.82361856e-01
-1.52068937e+00 1.04686670e-01 5.41108191e-01 6.50916025e-02
6.20621800e-01 1.25743702e-01 9.43998575e-01 5.76008558e-01
3.82549852e-01 6.50984168e-01 6.46320462e-01 1.02424145e-01
6.12591863e-01 -2.84632444e-01 -5.87676346e-01 7.90400207e-01
4.40606087e-01 4.37936395e-01 -3.05373639e-01 -2.02503681e-01
6.54767096e-01 -1.12211762e-03 -5.45574546e-01 -9.65623021e-01
-1.70771217e+00 7.04269469e-01 6.53062940e-01 -3.45438033e-01
-5.67287982e-01 4.43738475e-02 3.52337927e-01 6.06208183e-02
-3.20822567e-01 7.20079839e-01 -3.30328166e-01 -3.83139461e-01
-5.73054135e-01 6.16400599e-01 6.16553664e-01 1.56695783e+00
9.43696260e-01 2.17232723e-02 1.09749578e-01 2.03670934e-01
2.04155639e-01 6.19592607e-01 1.34049371e-01 -1.34190226e+00
5.56450248e-01 8.70935559e-01 6.28246546e-01 -1.30638731e+00
-5.99275589e-01 -1.58322275e-01 -5.46901345e-01 5.99297702e-01
3.46659720e-02 -3.69386733e-01 -6.41856551e-01 1.47280324e+00
7.51664519e-01 -3.18990529e-01 3.63882959e-01 1.08170414e+00
2.15529844e-01 6.81378782e-01 -1.85698077e-01 -1.72162354e-01
8.10671210e-01 -1.13753164e+00 -4.84374851e-01 -2.77104467e-01
5.44053137e-01 -2.08580181e-01 9.81182635e-01 5.93728065e-01
-8.22690785e-01 -2.04363897e-01 -1.23549139e+00 2.20793977e-01
-1.34032533e-01 9.89911854e-02 4.61883575e-01 2.09273666e-01
-9.93207812e-01 5.60206294e-01 -1.00399184e+00 -3.02483022e-01
3.57944131e-01 2.72436023e-01 -1.29820466e-01 -3.06470484e-01
-9.57515597e-01 7.47882307e-01 7.59565949e-01 1.59022212e-01
-1.64711726e+00 -1.83631971e-01 -8.63919437e-01 -1.33658960e-01
8.75361621e-01 -3.48236144e-01 1.05821788e+00 5.72696794e-03
-1.69795108e+00 -1.41942546e-01 -3.25175762e-01 -3.57097387e-01
6.27629340e-01 -3.41232628e-01 7.79061615e-02 1.77420601e-01
3.35536331e-01 1.06840658e+00 6.30299598e-02 -1.22154200e+00
-9.22609866e-01 -2.87685066e-01 4.20939207e-01 7.63093114e-01
-6.32566214e-02 -7.81557262e-01 -3.45074117e-01 -7.00871199e-02
7.54227877e-01 -1.24283147e+00 -1.02272356e+00 1.92571461e-01
-5.30452907e-01 -3.37832305e-03 1.00450659e+00 -1.79670811e-01
1.04548192e+00 -1.84646165e+00 4.60047722e-01 3.97837490e-01
-8.40293095e-02 -2.09201828e-01 -1.35113388e-01 6.22160196e-01
7.26109564e-01 -9.59638432e-02 -5.15675008e-01 1.48789048e-01
-1.16675302e-01 4.63726282e-01 -5.56002438e-01 4.24633592e-01
-4.47856277e-01 6.97008848e-01 -1.45317197e+00 -4.36962634e-01
2.00543448e-01 2.07182243e-01 -5.31151175e-01 -5.79024106e-02
-5.15476644e-01 6.65009081e-01 -9.79304254e-01 9.40236747e-01
8.39625180e-01 3.00342739e-01 1.84141666e-01 3.48904461e-01
-7.00765848e-01 1.71474054e-01 -1.29721224e+00 2.15879464e+00
-3.17195803e-01 3.53273213e-01 5.15754558e-02 -3.78837615e-01
1.19570577e+00 -8.55400264e-02 3.71824831e-01 -3.94705683e-01
-5.39710075e-02 4.69929218e-01 -4.04615223e-01 -1.40791878e-01
1.05901575e+00 5.82435310e-01 -1.71682566e-01 2.05771402e-01
-7.26390064e-01 -4.34517026e-01 1.16090395e-01 -1.07696624e-02
1.18271649e+00 6.28768981e-01 5.01061738e-01 -2.51618296e-01
4.48004186e-01 8.09896111e-01 9.92409110e-01 4.47948962e-01
-3.65296364e-01 4.39838246e-02 2.61871189e-01 -5.75542092e-01
-6.45934701e-01 -1.14275384e+00 1.29432172e-01 4.95093852e-01
1.28272498e+00 -4.00435835e-01 -5.27182281e-01 -4.69129950e-01
-1.65505052e-01 7.83501387e-01 -1.45670414e-01 -2.72750705e-02
-7.26802588e-01 -2.63991684e-01 2.26551384e-01 2.38519326e-01
6.89776838e-01 -1.04860044e+00 -1.71586859e+00 4.44517046e-01
-3.97609383e-01 -7.14781046e-01 -3.03553492e-01 -1.45035461e-01
-9.98635292e-01 -1.33037126e+00 -4.34706002e-01 -5.21304488e-01
8.65815520e-01 7.26499498e-01 4.15321380e-01 -1.43380553e-01
-1.57079324e-01 1.39033362e-01 -5.86987853e-01 -1.90239549e-01
-2.84271836e-02 2.05686912e-01 1.54885694e-01 -7.70720959e-01
3.37835290e-02 -5.75486362e-01 -8.47302079e-01 7.42025912e-01
-5.27244091e-01 1.22045554e-01 6.86774313e-01 6.29841983e-01
9.93131280e-01 6.28269434e-01 2.46292293e-01 -1.16282277e-01
5.05769849e-01 -6.04675412e-01 -1.11079252e+00 1.30697012e-01
-7.74448514e-01 -2.76255496e-02 6.59229279e-01 -3.35814178e-01
-8.36798608e-01 2.87672132e-01 2.87410051e-01 -2.11453691e-01
9.82600544e-03 5.12112021e-01 -3.26954722e-01 -3.25245529e-01
4.27083075e-01 3.47681642e-01 -1.86634049e-01 -4.49539982e-02
3.67923200e-01 1.39916450e-01 4.34940487e-01 -4.13213760e-01
9.87853587e-01 6.70094848e-01 3.86669725e-01 -4.03883398e-01
-1.15014181e-01 -1.53504759e-01 -5.74386954e-01 -2.68849492e-01
5.39294124e-01 -7.20740795e-01 -9.58637297e-01 -6.74387813e-02
-1.00105178e+00 -3.22088003e-01 -3.11731726e-01 5.34995317e-01
-7.20024407e-01 3.99349093e-01 1.26637578e-01 -1.14699292e+00
-1.85550332e-01 -1.68293440e+00 8.50966573e-01 4.91598487e-01
-1.31861925e-01 -4.92181182e-01 4.92851250e-02 -5.36952853e-01
4.18470562e-01 9.72145140e-01 4.30261880e-01 7.43998587e-02
-1.42370307e+00 9.37544852e-02 -3.87299210e-02 -8.69411767e-01
1.34966061e-01 -2.98826516e-01 -1.73894733e-01 -5.58957338e-01
-4.11468774e-01 -1.60696968e-01 5.77352941e-01 2.66796261e-01
9.67556357e-01 -4.96743560e-01 -1.01682603e+00 6.97058022e-01
1.38924992e+00 7.84991860e-01 4.05985266e-01 8.79650176e-01
1.86144710e-01 7.02409029e-01 1.63333750e+00 8.99474084e-01
6.36805058e-01 5.43705344e-01 1.23266160e+00 4.27932888e-01
7.16764271e-01 -7.74836779e-01 3.95346254e-01 1.55621618e-01
8.89909491e-02 -3.43181908e-01 -9.48438525e-01 8.30073535e-01
-2.00362396e+00 -5.94725788e-01 1.73484832e-01 2.30086064e+00
5.31446338e-01 9.16428044e-02 -1.05055764e-01 -9.30862501e-02
2.55193383e-01 1.67786658e-01 -9.91068602e-01 -7.04488307e-02
2.11515397e-01 -5.80501914e-01 9.04697537e-01 8.46929431e-01
-8.38962317e-01 1.06851876e+00 5.90348959e+00 4.42606211e-01
-6.69942081e-01 -2.80144185e-01 -7.77318478e-02 -2.05027252e-01
-6.13684952e-01 4.48700219e-01 -9.01621342e-01 2.21969545e-01
4.55515176e-01 -3.83418918e-01 7.36121655e-01 1.45318997e+00
1.76440895e-01 -6.36178911e-01 -4.88868117e-01 5.61509252e-01
-3.43053937e-01 -1.15450764e+00 -1.10372351e-02 8.12933221e-02
7.18777061e-01 -4.97437790e-02 -4.80137840e-02 9.66956988e-02
7.88446248e-01 -8.01317751e-01 7.24388182e-01 3.87071520e-01
5.98348618e-01 -1.12886667e+00 5.53121746e-01 6.46246374e-01
-1.59949493e+00 -3.28043550e-01 -7.48751819e-01 6.57167062e-02
7.61894405e-01 3.18259239e-01 -1.11347628e+00 7.74035871e-01
7.21129715e-01 4.70917135e-01 -1.78979069e-01 1.32469857e+00
-4.78957564e-01 -1.41054764e-01 -4.92798686e-01 -3.34282041e-01
6.66240335e-01 -5.40592015e-01 1.25208616e+00 4.09069479e-01
7.33584344e-01 2.49614924e-01 7.86705673e-01 9.26516294e-01
6.30242825e-01 -2.04094738e-01 -7.86195517e-01 7.37405941e-02
1.08992755e+00 9.95919108e-01 -8.59453619e-01 -8.61139446e-02
2.68537790e-01 7.26883054e-01 3.57698686e-02 3.59876513e-01
-8.55558336e-01 -8.56486261e-01 6.71028614e-01 -9.33968648e-02
-9.51960534e-02 -6.88433290e-01 -1.80718064e-01 -4.75650758e-01
1.62760634e-02 -5.48050642e-01 8.11019614e-02 -5.32559991e-01
-2.32279450e-01 1.00120103e+00 3.25285673e-01 -1.60432637e+00
-3.95405650e-01 -1.84260324e-01 -3.89467448e-01 6.60441339e-01
-1.67450881e+00 -9.28529501e-01 -8.66823256e-01 2.22755790e-01
7.99629331e-01 -1.97067201e-01 6.63352311e-01 -4.25634205e-01
-1.42886132e-01 6.14934079e-02 -8.94540325e-02 -5.29296219e-01
3.39213371e-01 -9.78664219e-01 6.34593487e-01 9.78355050e-01
-4.29668576e-01 6.05587184e-01 8.18356216e-01 -1.07701635e+00
-1.49348772e+00 -1.09634507e+00 1.32241786e-01 7.06748292e-02
2.04891771e-01 5.79574257e-02 -6.76635504e-01 5.00247955e-01
-7.24201500e-02 -5.30144095e-01 1.82001367e-02 -3.37373823e-01
2.52835125e-01 1.12130888e-01 -1.15017498e+00 1.04151475e+00
1.39304507e+00 2.96842694e-01 -6.57558963e-02 2.88993418e-01
1.18943202e+00 -8.53143096e-01 -4.47723776e-01 7.56387472e-01
6.18423343e-01 -1.01268041e+00 7.06992447e-01 1.04999349e-01
-1.16695156e-02 -9.03016567e-01 -1.76641494e-01 -1.45230973e+00
-7.64086992e-02 -9.38297093e-01 1.20431416e-01 6.44866526e-01
4.38889056e-01 -8.04958761e-01 8.56552184e-01 5.00425100e-01
-5.32936573e-01 -1.19793928e+00 -9.47508752e-01 -9.25596476e-01
-4.18973684e-01 -1.64385155e-01 1.18554270e+00 3.50225747e-01
2.00146124e-01 -5.53229034e-01 -2.44021609e-01 6.73399985e-01
7.03688800e-01 3.42134178e-01 1.05083466e+00 -9.23927009e-01
1.95392132e-01 -1.35000348e-01 -9.56533849e-02 -1.28672218e+00
-5.41839935e-02 -4.08539593e-01 6.74634933e-01 -1.96140301e+00
-3.16724300e-01 -9.98873651e-01 3.29946637e-01 4.22370791e-01
1.59449533e-01 -4.93233502e-01 -1.66227698e-01 2.71026015e-01
-6.63281202e-01 9.61822987e-01 1.49175489e+00 5.67293465e-02
-7.58072674e-01 -1.18707851e-01 -3.21369082e-01 7.32763886e-01
8.71654153e-01 -3.10311586e-01 -9.80079532e-01 -5.14662147e-01
2.07510009e-01 4.68047678e-01 7.38395900e-02 -1.31619215e+00
4.95950371e-01 -7.81355143e-01 2.60640159e-02 -1.25016844e+00
4.82061863e-01 -9.09859061e-01 3.09757352e-01 7.30853677e-01
-1.40400946e-01 4.15962636e-01 1.84742302e-01 9.97781336e-01
-2.62172818e-01 -2.54301339e-01 4.88991588e-01 -1.93326786e-01
-1.02451551e+00 6.35375619e-01 -3.85089815e-01 -1.98391229e-01
1.54374731e+00 -4.38601941e-01 -1.56843856e-01 -2.58583337e-01
-2.54846007e-01 1.05248213e+00 1.04162467e+00 2.39453331e-01
1.08978057e+00 -1.14771378e+00 -1.56734988e-01 1.79839015e-01
3.23237926e-02 8.07597935e-01 3.01956028e-01 6.98127925e-01
-9.67202604e-01 5.53575337e-01 -3.74668270e-01 -4.44079459e-01
-8.68901253e-01 6.19723558e-01 -2.28641331e-02 -1.28653273e-01
-8.27045262e-01 7.55916774e-01 6.13634326e-02 -6.61305428e-01
2.71865308e-01 -4.86964762e-01 -1.54903293e-01 -3.90771866e-01
5.18852413e-01 8.24741364e-01 -6.87714517e-01 -1.80971742e-01
-5.29930472e-01 6.92143261e-01 2.83511490e-01 -5.46891630e-01
9.66385126e-01 -4.88999248e-01 -5.48386015e-02 5.70031777e-02
5.11056066e-01 1.40139729e-01 -1.64863288e+00 1.63405314e-01
-1.81865003e-02 -8.45825136e-01 -1.85627207e-01 -5.32330394e-01
-6.10200644e-01 4.73372638e-01 2.17811316e-01 -2.38802671e-01
1.20875061e+00 -5.40818155e-01 1.04777575e+00 7.62642443e-01
1.33637118e+00 -9.92350280e-01 -8.47986639e-02 5.44376910e-01
9.68826532e-01 -9.52191532e-01 2.52722740e-01 -5.97458720e-01
-8.91985118e-01 1.09381199e+00 8.77934277e-01 -1.04664907e-01
2.25376353e-01 1.74008980e-01 -2.72747993e-01 -3.92728904e-03
-6.34020269e-01 -4.57155547e-04 -3.03481013e-01 8.70504439e-01
-5.36460876e-01 5.89658543e-02 -2.34311894e-01 1.62126258e-01
-5.30733347e-01 -1.71964437e-01 8.38825345e-01 1.09062552e+00
-1.02432799e+00 -8.56926858e-01 -3.26716095e-01 -2.50655394e-02
5.55540383e-01 3.91316980e-01 5.08848159e-03 8.95733178e-01
-2.74723768e-02 9.69936609e-01 -1.77157924e-01 -4.81196612e-01
3.23983580e-01 -4.91113752e-01 2.05802601e-02 -2.71788806e-01
2.41432860e-01 -3.53361011e-01 7.46391043e-02 -1.03915119e+00
-1.74491014e-02 -6.27156079e-01 -1.68079114e+00 -2.88553033e-02
-2.78862506e-01 4.72809643e-01 6.13499820e-01 6.06340230e-01
4.80814785e-01 4.23477679e-01 6.70742214e-01 -8.60959232e-01
-5.40452063e-01 -5.35170257e-01 -2.63046473e-01 -6.08352780e-01
2.03355521e-01 -1.13693368e+00 -2.47113351e-02 -6.40943825e-01] | [4.874289512634277, 1.3525300025939941] |
0ba56209-02ec-48fa-ad87-4c983f82dd42 | using-n-gram-and-word-network-features-for | null | null | https://aclanthology.org/W13-1732 | https://aclanthology.org/W13-1732.pdf | Using N-gram and Word Network Features for Native Language Identification | null | ['Shibamouli Lahiri', 'Rada Mihalcea'] | 2013-06-01 | null | null | null | ws-2013-6 | ['native-language-identification'] | ['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] | [-7.384732246398926, 3.7719781398773193] |
4d894448-6bf5-46a6-80df-96f30c658ad1 | coupling-adversarial-learning-with-selective | 2207.08145 | null | https://arxiv.org/abs/2207.08145v1 | https://arxiv.org/pdf/2207.08145v1.pdf | Coupling Adversarial Learning with Selective Voting Strategy for Distribution Alignment in Partial Domain Adaptation | In contrast to a standard closed-set domain adaptation task, partial domain adaptation setup caters to a realistic scenario by relaxing the identical label set assumption. The fact of source label set subsuming the target label set, however, introduces few additional obstacles as training on private source category samples thwart relevant knowledge transfer and mislead the classification process. To mitigate these issues, we devise a mechanism for strategic selection of highly-confident target samples essential for the estimation of class-importance weights. Furthermore, we capture class-discriminative and domain-invariant features by coupling the process of achieving compact and distinct class distributions with an adversarial objective. Experimental findings over numerous cross-domain classification tasks demonstrate the potential of the proposed technique to deliver superior and comparable accuracy over existing methods. | ['Arunabha Sen', 'Hemanth Venkateswara', 'Sandipan Choudhuri'] | 2022-07-17 | null | null | null | null | ['partial-domain-adaptation'] | ['methodology'] | [ 5.75809062e-01 3.21525455e-01 -4.61653769e-01 -5.88441849e-01
-1.12346196e+00 -8.91720235e-01 7.03010440e-01 7.31283501e-02
-4.38608408e-01 1.21851647e+00 -4.96292040e-02 1.78234708e-02
-2.76804358e-01 -6.33592069e-01 -5.53915441e-01 -1.02683485e+00
1.55850977e-01 5.05446553e-01 9.58128124e-02 5.29965125e-02
2.03250349e-01 4.64820176e-01 -1.30988610e+00 1.75575316e-01
8.88749361e-01 1.21132541e+00 -1.27839923e-01 1.07182646e-02
1.85294114e-02 3.80953848e-01 -9.36432838e-01 -7.97996938e-01
7.02814460e-01 -2.83417225e-01 -6.12263739e-01 1.65241122e-01
4.84290600e-01 -2.58786768e-01 3.89818065e-02 1.18459892e+00
3.97924900e-01 1.77432477e-01 1.21206987e+00 -1.49858642e+00
-7.24653900e-01 3.53766769e-01 -5.06112456e-01 1.13405570e-01
1.99337795e-01 -1.18463330e-01 9.01073933e-01 -6.70457006e-01
6.62688553e-01 8.43112290e-01 5.99482536e-01 7.83838987e-01
-1.35087454e+00 -1.00568330e+00 2.00214580e-01 -1.57530546e-01
-1.58426869e+00 -5.75274169e-01 1.04851043e+00 -3.78280044e-01
1.22256391e-01 9.39010084e-02 9.93526541e-03 1.51902246e+00
-4.64020483e-03 6.32681668e-01 1.43228936e+00 -3.80651712e-01
5.40069044e-01 8.28339279e-01 9.75409895e-02 2.02764392e-01
4.33741927e-01 2.04149663e-01 -2.88417488e-01 -6.17658198e-01
5.76091886e-01 5.58766397e-03 -3.03470284e-01 -1.10277653e+00
-9.34044838e-01 9.91253078e-01 1.99688897e-01 1.18255772e-01
-2.98036277e-01 -2.62034357e-01 5.23416162e-01 5.41860998e-01
5.95817983e-01 2.96163857e-01 -6.48048460e-01 3.26591790e-01
-8.12094152e-01 1.76623553e-01 7.35121906e-01 1.12691176e+00
6.93457603e-01 -8.50731507e-02 -3.31534714e-01 7.89397776e-01
6.70761317e-02 4.63255078e-01 3.96641701e-01 -9.44789886e-01
5.30470014e-01 2.65778035e-01 2.96927482e-01 -6.51642799e-01
1.07428089e-01 -7.93005347e-01 -7.71573603e-01 2.94174224e-01
7.85313845e-01 -1.17517851e-01 -7.81452000e-01 2.00757337e+00
5.38677096e-01 1.29871786e-01 2.51652300e-01 7.50500143e-01
2.73202062e-01 1.95198163e-01 3.81523788e-01 -2.19548881e-01
1.26665545e+00 -4.63862211e-01 -5.86405039e-01 -1.70831252e-02
3.18890780e-01 -3.89785498e-01 9.55911875e-01 2.65732169e-01
-6.64718628e-01 -3.77370059e-01 -1.11563087e+00 3.14136058e-01
-4.59850103e-01 -1.64508790e-01 5.91853321e-01 1.07117152e+00
-5.07697880e-01 3.32011431e-01 -2.70360202e-01 -1.33032173e-01
9.61320758e-01 4.97195274e-01 -6.64076269e-01 -1.89447597e-01
-1.18251812e+00 7.58240938e-01 5.97640753e-01 -5.35306394e-01
-7.58019149e-01 -9.65512395e-01 -7.91930079e-01 6.01560809e-02
3.40042174e-01 -6.25449002e-01 1.04586768e+00 -1.21155953e+00
-1.60282004e+00 1.16731358e+00 1.05787620e-01 -6.48861706e-01
7.38963187e-01 1.47329971e-01 -3.97291869e-01 2.08384678e-01
2.03303635e-01 6.56265080e-01 1.23814738e+00 -1.42085731e+00
-7.32993543e-01 -2.73956716e-01 -6.89647272e-02 3.36087555e-01
-7.11804807e-01 -8.64527076e-02 8.24550092e-02 -8.13094616e-01
-2.24958315e-01 -6.77579284e-01 -1.27637252e-01 1.73844010e-01
-1.41347080e-01 -5.46530522e-02 8.46731603e-01 -2.95975357e-01
8.46477985e-01 -2.35565972e+00 -9.24987867e-02 3.99798810e-01
1.25121802e-01 2.62722671e-01 5.53153977e-02 1.55829996e-01
-1.18646279e-01 -1.03182755e-01 -4.81665105e-01 -2.39252090e-01
1.53360143e-01 3.26323301e-01 -5.76597512e-01 6.67717695e-01
2.96114713e-01 5.46676278e-01 -7.95662045e-01 -5.81618667e-01
7.01924190e-02 2.96577156e-01 -5.10027945e-01 1.15493827e-01
-1.69241056e-01 6.26615107e-01 -6.98104858e-01 5.57833195e-01
9.27265644e-01 -1.91976041e-01 1.66805342e-01 1.86512291e-01
6.71902657e-01 1.28388599e-01 -1.25975430e+00 1.52287042e+00
-2.13063747e-01 3.35980132e-02 1.48752630e-01 -1.37980568e+00
1.05309463e+00 2.18265876e-01 5.00869930e-01 -4.61687535e-01
1.90942049e-01 1.07879005e-01 -2.80554056e-01 1.19464852e-01
2.76527613e-01 -6.93640530e-01 -5.38473845e-01 4.59077984e-01
3.47726971e-01 5.19239344e-02 -4.32537466e-01 3.38912234e-02
7.32989311e-01 1.78555101e-01 6.39908612e-01 -3.77052844e-01
6.14412248e-01 -1.05481148e-01 7.60756671e-01 6.85741007e-01
-6.23020470e-01 5.22607148e-01 2.78600663e-01 -5.35281934e-02
-8.15873086e-01 -1.34430230e+00 -4.61811602e-01 1.17333090e+00
1.79429874e-01 2.94874281e-01 -6.84986115e-01 -1.46053016e+00
2.87747234e-01 9.01292741e-01 -7.94212341e-01 -4.85472411e-01
-2.60593355e-01 -6.24908745e-01 6.75176084e-01 4.20691967e-01
3.40805173e-01 -6.99790835e-01 -2.31262133e-01 4.12844494e-02
-3.68641131e-02 -1.05390680e+00 -4.87115115e-01 4.81335282e-01
-7.69765794e-01 -9.41042364e-01 -8.27432513e-01 -9.19051707e-01
7.45105863e-01 1.66655853e-01 1.08475757e+00 -4.95580435e-01
3.69370393e-02 3.44203353e-01 -2.97091573e-01 -5.73044956e-01
-5.88771164e-01 7.98155740e-02 2.13083755e-02 3.05150867e-01
7.21352994e-01 -4.95816171e-01 -4.68862087e-01 3.12672913e-01
-7.76982605e-01 -4.44066763e-01 3.90592635e-01 1.01767051e+00
5.08854032e-01 2.16468915e-01 1.14863944e+00 -1.13810384e+00
5.09872675e-01 -8.30867946e-01 -4.93734419e-01 2.27490932e-01
-7.72916496e-01 -1.31707527e-02 8.23680580e-01 -7.16334939e-01
-1.38055694e+00 1.04678713e-01 2.89680660e-01 -4.52545166e-01
-5.26124656e-01 -1.43128082e-01 -4.72780824e-01 -2.22026780e-01
8.34309578e-01 3.27006400e-01 2.41467893e-01 -3.69406044e-01
3.40346485e-01 7.62145996e-01 4.90546107e-01 -7.83252537e-01
1.16569936e+00 6.36614680e-01 -1.54717416e-01 -2.19041437e-01
-9.43646371e-01 -4.95513231e-01 -7.56632864e-01 3.05248201e-01
3.67659807e-01 -1.14748681e+00 -2.17518091e-01 1.29786521e-01
-7.03508794e-01 1.44202514e-02 -7.16635108e-01 3.34007412e-01
-7.03871191e-01 3.47254336e-01 -2.07958758e-01 -6.82816029e-01
-1.40306562e-01 -8.16074491e-01 9.63867366e-01 6.95362091e-02
-2.72332996e-01 -1.23263490e+00 -8.59834999e-02 4.22201842e-01
2.62991160e-01 4.17332232e-01 7.52401292e-01 -1.35335898e+00
-3.24891210e-01 -4.36183482e-01 -6.98471665e-02 6.24323487e-01
2.12302238e-01 -6.23028219e-01 -1.22921753e+00 -4.60805953e-01
7.87454173e-02 -6.53925896e-01 6.17201626e-01 1.67111695e-01
1.10490012e+00 -2.18947098e-01 -4.61977839e-01 5.13041139e-01
1.40622020e+00 6.23182096e-02 3.38241816e-01 3.06438297e-01
3.01989734e-01 7.11225092e-01 9.17582512e-01 6.12459123e-01
2.11977690e-01 6.13665581e-01 1.23367973e-01 -3.35111432e-02
-7.53815025e-02 -3.05156738e-01 2.00118750e-01 -1.49276853e-02
4.55762118e-01 -2.11021423e-01 -5.01308441e-01 6.21639431e-01
-1.40533149e+00 -1.02852452e+00 4.98402327e-01 2.39780974e+00
1.16427350e+00 1.13194562e-01 2.28199452e-01 2.44508982e-01
7.74071157e-01 -3.25090922e-02 -6.50629401e-01 -1.58612475e-01
-1.57311544e-01 2.92190701e-01 7.26125300e-01 2.62088448e-01
-1.26086795e+00 8.30072284e-01 6.64788818e+00 1.23736334e+00
-8.90512586e-01 3.03312451e-01 5.31552434e-01 1.71464607e-02
-3.41983825e-01 -1.58137485e-01 -9.28780198e-01 5.19665539e-01
7.61255145e-01 -3.17810982e-01 1.76037624e-01 1.02105451e+00
-3.24577063e-01 1.55108288e-01 -1.19337928e+00 6.54592574e-01
6.65631890e-02 -9.44851398e-01 -1.18439116e-01 3.07968259e-01
7.66448319e-01 -3.30153018e-01 3.80938739e-01 4.06680405e-01
6.30794823e-01 -7.66389191e-01 6.80819929e-01 3.03918272e-02
1.06606328e+00 -9.29687202e-01 6.20102048e-01 4.29475546e-01
-6.41990066e-01 -2.72400111e-01 -4.72568512e-01 9.52740908e-02
-2.05231890e-01 5.88746965e-01 -1.22001767e+00 6.14762366e-01
4.32705611e-01 4.59894925e-01 -3.61344308e-01 7.34297037e-01
7.70377964e-02 6.71927869e-01 -2.20462069e-01 3.07569444e-01
6.95016906e-02 -8.89982209e-02 5.60547173e-01 1.14998412e+00
1.34314418e-01 -1.97753131e-01 1.90630198e-01 7.41373897e-01
-2.01528996e-01 2.19080243e-02 -1.00305283e+00 2.09004894e-01
7.17430413e-01 1.00340676e+00 -5.75408340e-01 -2.95990318e-01
-3.59151274e-01 1.10141170e+00 3.57168764e-01 4.60958034e-01
-9.05175209e-01 -2.52702117e-01 8.83632958e-01 4.79669832e-02
5.61039388e-01 1.82165682e-01 -5.69815099e-01 -1.11465573e+00
5.19982427e-02 -8.58587980e-01 7.28656530e-01 -6.01705387e-02
-1.74894309e+00 2.31113359e-01 1.85776591e-01 -1.49651945e+00
-2.59972841e-01 -4.79722977e-01 -3.54414314e-01 9.72182155e-01
-1.83436370e+00 -1.22041130e+00 8.71875808e-02 9.80469823e-01
2.01921001e-01 -5.00568271e-01 1.16638410e+00 2.03184053e-01
-2.21539006e-01 1.26340663e+00 4.77799237e-01 5.65350726e-02
1.15261209e+00 -1.24794388e+00 -1.11546434e-01 5.75387955e-01
-1.54281065e-01 6.01450264e-01 6.09999061e-01 -3.77378672e-01
-7.75013506e-01 -1.25091457e+00 5.70581317e-01 -7.03555107e-01
5.94330966e-01 -6.92859650e-01 -9.86534536e-01 6.24739587e-01
-1.09327734e-01 2.42386103e-01 1.17522991e+00 3.59468628e-03
-7.99520612e-01 -2.31843919e-01 -1.89481759e+00 2.71884143e-01
9.54141974e-01 -5.39559424e-01 -7.83317745e-01 2.87745118e-01
6.37581766e-01 -1.26847327e-01 -1.09584439e+00 3.04158151e-01
3.63474518e-01 -7.36748338e-01 1.22782111e+00 -7.38399923e-01
1.21323638e-01 -1.15163475e-01 -3.78783554e-01 -1.30547178e+00
-3.63744050e-01 -4.99149948e-01 -8.85162428e-02 1.56792986e+00
1.61316082e-01 -9.16504860e-01 9.79179978e-01 6.31515026e-01
2.46250153e-01 -2.92205304e-01 -1.14750803e+00 -1.06367731e+00
5.57228565e-01 -6.99234158e-02 6.99384451e-01 1.21781743e+00
-2.07513765e-01 6.44948408e-02 -2.25534663e-01 2.83310622e-01
1.07524991e+00 2.28416398e-01 6.06119692e-01 -1.48426318e+00
-2.82434016e-01 -2.52319485e-01 -3.61305565e-01 -7.19617963e-01
6.68576956e-01 -9.51213360e-01 -4.84187007e-02 -6.74693644e-01
2.15815812e-01 -1.04658318e+00 -7.95146704e-01 4.02152777e-01
-1.95946991e-01 3.19595486e-01 2.81283334e-02 1.94364026e-01
-5.49325824e-01 5.17061114e-01 1.18254066e+00 -8.37195665e-03
1.03417128e-01 2.22333670e-01 -1.24471080e+00 4.06413555e-01
7.34330475e-01 -7.90858150e-01 -8.53092492e-01 1.60213158e-01
-4.83247012e-01 -5.64516746e-02 3.84131104e-01 -7.33738244e-01
-7.36435801e-02 -4.02869999e-01 4.50051665e-01 -5.32040596e-02
3.61466914e-01 -1.25156283e+00 -4.86738309e-02 2.67213821e-01
-4.74119633e-01 -7.92308927e-01 8.91003758e-02 1.06920290e+00
-1.74107045e-01 -3.07515889e-01 1.09354210e+00 7.07942992e-02
-6.03442550e-01 2.37027436e-01 -5.88794164e-02 2.94634849e-01
1.43561649e+00 -6.31008565e-01 -4.99508865e-02 -1.02793276e-01
-4.41491544e-01 -1.49693154e-02 7.34366000e-01 3.19589347e-01
2.17231825e-01 -1.36262906e+00 -6.49110675e-01 4.49933529e-01
4.51554328e-01 -4.10169512e-02 2.80256141e-02 3.32477778e-01
2.87819684e-01 2.74409026e-01 -3.34452242e-01 -3.55857730e-01
-9.87517357e-01 7.88060784e-01 2.04646692e-01 -3.94404918e-01
-5.52880704e-01 8.45305920e-01 6.41199529e-01 -6.52428031e-01
3.54185253e-01 2.51152784e-01 -1.24487191e-01 5.12466393e-02
5.01111448e-01 8.43755007e-02 -1.58474538e-02 -5.26432216e-01
-3.55633795e-01 2.92557687e-01 -2.98387557e-01 1.89375076e-02
1.16133094e+00 -3.28271240e-01 4.83673036e-01 1.54500932e-01
1.22548127e+00 5.81884524e-03 -1.57599676e+00 -5.98674119e-01
-4.95535173e-02 -5.41096032e-01 -2.16217190e-01 -1.13432395e+00
-7.04928398e-01 6.16151571e-01 7.01349914e-01 -5.67951202e-02
1.21414733e+00 -1.54311210e-01 5.78710973e-01 1.68638885e-01
6.45853162e-01 -1.10592651e+00 -7.34078288e-02 1.41641244e-01
5.48565984e-01 -1.53101540e+00 -1.39207214e-01 -5.96262157e-01
-9.28477347e-01 6.61553085e-01 5.52231967e-01 -1.45397499e-01
5.99057853e-01 1.76227376e-01 5.64288236e-02 1.57888532e-01
-4.73710924e-01 1.37957737e-01 3.32246155e-01 1.27859628e+00
6.24998398e-02 2.66607434e-01 -1.00900054e-01 7.91708231e-01
7.59110674e-02 1.01502791e-01 2.46719748e-01 8.87289166e-01
-2.59678513e-01 -1.28764403e+00 -6.78787351e-01 3.06846648e-01
-7.43616939e-01 8.68699178e-02 -2.83813268e-01 8.52849305e-01
1.28438562e-01 7.41134465e-01 5.65834418e-02 -8.64048824e-02
1.11329474e-01 3.62387151e-01 4.15189147e-01 -6.66254699e-01
-4.67467099e-01 -1.91032752e-01 -3.19702625e-02 -1.22165643e-01
-2.85664231e-01 -7.14856982e-01 -1.00024164e+00 -3.11774965e-02
-3.66833627e-01 2.70194441e-01 3.89051408e-01 8.57797503e-01
4.33415920e-01 1.39015123e-01 8.18362057e-01 -5.25330663e-01
-1.38707089e+00 -7.56255805e-01 -8.18202317e-01 8.84977758e-01
3.06254357e-01 -1.05101657e+00 -3.87750298e-01 1.28322139e-01] | [10.35435962677002, 3.204477548599243] |
7a488fdc-d3bf-4ca8-ba0f-92e08755399d | avlen-audio-visual-language-embodied | 2210.07940 | null | https://arxiv.org/abs/2210.07940v1 | https://arxiv.org/pdf/2210.07940v1.pdf | AVLEN: Audio-Visual-Language Embodied Navigation in 3D Environments | Recent years have seen embodied visual navigation advance in two distinct directions: (i) in equipping the AI agent to follow natural language instructions, and (ii) in making the navigable world multimodal, e.g., audio-visual navigation. However, the real world is not only multimodal, but also often complex, and thus in spite of these advances, agents still need to understand the uncertainty in their actions and seek instructions to navigate. To this end, we present AVLEN~ -- an interactive agent for Audio-Visual-Language Embodied Navigation. Similar to audio-visual navigation tasks, the goal of our embodied agent is to localize an audio event via navigating the 3D visual world; however, the agent may also seek help from a human (oracle), where the assistance is provided in free-form natural language. To realize these abilities, AVLEN uses a multimodal hierarchical reinforcement learning backbone that learns: (a) high-level policies to choose either audio-cues for navigation or to query the oracle, and (b) lower-level policies to select navigation actions based on its audio-visual and language inputs. The policies are trained via rewarding for the success on the navigation task while minimizing the number of queries to the oracle. To empirically evaluate AVLEN, we present experiments on the SoundSpaces framework for semantic audio-visual navigation tasks. Our results show that equipping the agent to ask for help leads to a clear improvement in performance, especially in challenging cases, e.g., when the sound is unheard during training or in the presence of distractor sounds. | ['Anoop Cherian', 'Amit K. Roy-Chowdhury', 'Sudipta Paul'] | 2022-10-14 | null | null | null | null | ['hierarchical-reinforcement-learning'] | ['methodology'] | [ 4.38599326e-02 2.65070885e-01 3.05775583e-01 -1.04161121e-01
-1.01055145e+00 -9.48823988e-01 5.76412737e-01 7.04848170e-02
-7.33545661e-01 3.73562396e-01 2.08209768e-01 -4.97945100e-01
4.70367037e-02 -7.32897937e-01 -6.95406377e-01 -5.03278971e-01
-2.25448087e-01 6.62510037e-01 1.79012209e-01 -3.42762649e-01
2.65845865e-01 3.43512386e-01 -1.69438255e+00 5.05554006e-02
5.52403390e-01 8.33568931e-01 5.64012170e-01 9.70502019e-01
-2.09517285e-01 9.01787877e-01 -3.43475938e-01 2.22971186e-01
2.74143671e-03 -4.74289089e-01 -1.06370592e+00 -1.34025037e-01
-1.64301068e-01 -4.88895625e-01 -2.86682636e-01 9.64583099e-01
4.15831953e-01 4.96962428e-01 5.62475622e-01 -1.54136264e+00
-4.28539485e-01 4.01289761e-01 -1.86049324e-02 -1.71671778e-01
9.73661542e-01 6.91182017e-01 1.14883256e+00 -7.56018698e-01
6.44671500e-01 1.56254935e+00 1.11498058e-01 5.96768439e-01
-1.14589107e+00 -2.76218712e-01 5.32894135e-01 7.03403801e-02
-1.10200870e+00 -6.13276124e-01 4.44463044e-01 -5.03960848e-01
8.87811720e-01 9.92158651e-02 5.60577393e-01 1.23849833e+00
-1.02726430e-01 8.08724105e-01 7.84809411e-01 -5.49618185e-01
5.81251562e-01 2.07023621e-01 -3.55189949e-01 9.20221388e-01
-5.10813832e-01 5.39656103e-01 -8.09532762e-01 -2.02738956e-01
7.14147925e-01 -3.57143849e-01 -2.89698958e-01 -6.85556173e-01
-1.21637189e+00 7.42281795e-01 5.35486281e-01 1.12044513e-01
-4.00312036e-01 3.08343321e-01 2.13694111e-01 5.87050259e-01
-4.30729240e-01 7.92329967e-01 -2.72118121e-01 -4.00841922e-01
-1.39499933e-01 2.16598585e-01 7.34402299e-01 8.15365553e-01
7.50088096e-01 1.68233871e-01 2.38863528e-02 6.77253306e-01
6.27035499e-01 5.42370021e-01 2.90015131e-01 -1.40151942e+00
4.71924484e-01 3.35411608e-01 5.13866365e-01 -7.09228694e-01
-4.43386942e-01 7.35113993e-02 -1.76356852e-01 8.30294490e-01
6.49716020e-01 -2.13619202e-01 -8.63599300e-01 2.11465716e+00
3.05032492e-01 -2.83321112e-01 2.66930848e-01 1.15071189e+00
5.25001109e-01 7.24584103e-01 2.45988935e-01 1.98869273e-01
1.38961053e+00 -7.77086675e-01 -4.53561366e-01 -6.25152886e-01
5.80299139e-01 -3.33803922e-01 1.71620226e+00 4.30056959e-01
-8.33609879e-01 -3.67386937e-01 -9.09175754e-01 8.51019323e-02
-3.81223768e-01 -3.19562644e-01 3.33764285e-01 1.19770922e-01
-1.11805236e+00 1.25550732e-01 -9.13617253e-01 -4.84417856e-01
-9.94357541e-02 2.41681606e-01 -6.26092792e-01 -7.94916674e-02
-1.28703392e+00 7.46447563e-01 2.12604165e-01 -5.62961698e-02
-1.51982415e+00 -6.04385026e-02 -1.15122628e+00 1.64927766e-01
7.56733775e-01 -7.18180001e-01 1.70261359e+00 -1.02061236e+00
-1.63869858e+00 5.07963002e-01 -4.81812768e-02 -1.17782332e-01
5.35182297e-01 -1.68295011e-01 -5.04029403e-03 3.63546699e-01
3.13228101e-01 1.06380987e+00 6.66356564e-01 -1.56631982e+00
-8.12365890e-01 -3.66439551e-01 5.29695272e-01 6.98177516e-01
1.88493356e-02 -1.80415690e-01 -5.74396193e-01 -1.96496904e-01
6.43940419e-02 -1.07396758e+00 -2.89093494e-01 3.08243066e-01
-4.05555606e-01 -1.04493856e-01 6.02533996e-01 -1.89860001e-01
5.81462622e-01 -2.27565551e+00 3.32122475e-01 3.30053896e-01
-6.04523011e-02 -3.31825972e-01 -4.04531896e-01 6.61082566e-01
1.92099094e-01 8.79000351e-02 -3.44485566e-02 -4.48701650e-01
3.19983423e-01 2.70841390e-01 -3.46386075e-01 -3.33975591e-02
-1.61913976e-01 6.63217247e-01 -1.05728614e+00 -2.98726916e-01
1.52274027e-01 5.07373631e-01 -6.22247398e-01 5.65198481e-01
-4.93952632e-01 7.63710380e-01 -5.82706571e-01 5.78166246e-01
-1.96100995e-01 -3.68871130e-02 4.47997004e-02 3.78933847e-01
-2.22258613e-01 3.14220130e-01 -1.32611191e+00 1.99195647e+00
-7.08224654e-01 5.56056559e-01 5.84164023e-01 -4.89453822e-01
4.16351050e-01 3.46437603e-01 1.06982984e-01 -1.05558014e+00
-1.46440446e-01 1.14520840e-01 -1.67855859e-01 -6.47888839e-01
3.71753305e-01 1.34691894e-02 -2.68913686e-01 4.83765483e-01
-5.17372601e-02 -2.64574766e-01 -1.76665150e-02 4.76515979e-01
1.13974071e+00 2.99092799e-01 -1.27964973e-01 2.74687290e-01
2.95510262e-01 3.44422869e-02 1.29597619e-01 1.17951381e+00
-1.62677422e-01 1.98659390e-01 5.67467630e-01 -3.44563313e-02
-5.02341449e-01 -1.31353474e+00 4.73399550e-01 1.67331600e+00
3.83121610e-01 -2.90391237e-01 -6.93786442e-01 -4.56597060e-01
-1.41305521e-01 9.50009227e-01 -4.06204492e-01 -1.60473913e-01
-3.72169465e-01 1.66251153e-01 3.31070691e-01 5.24340272e-01
2.78020859e-01 -1.59266436e+00 -1.16282749e+00 9.79985669e-02
-3.32666218e-01 -1.02297211e+00 -4.86048341e-01 5.31374812e-01
-3.78033310e-01 -9.51139927e-01 -3.59156877e-01 -8.65573525e-01
4.89834934e-01 1.22298606e-01 9.43767369e-01 7.33976886e-02
-6.96668103e-02 1.23223019e+00 -3.49768549e-01 -2.17312485e-01
-3.72280568e-01 -8.63720477e-02 -9.64155719e-02 -2.97207236e-01
-9.86015573e-02 -4.66261923e-01 -4.95573640e-01 4.45996106e-01
-7.04371989e-01 -5.02604842e-02 3.53384614e-01 7.86483228e-01
3.96202862e-01 -1.36963844e-01 4.72282052e-01 -2.90268898e-01
8.01158369e-01 -4.26583439e-01 -8.19583535e-01 5.46883158e-02
-2.46822983e-01 2.87295491e-01 5.22458613e-01 -4.09854084e-01
-7.68228292e-01 3.54446262e-01 -2.36556157e-01 -7.34333545e-02
-5.61816573e-01 6.70344472e-01 -2.64987975e-01 1.48128234e-02
6.73791528e-01 1.70539126e-01 1.19365342e-02 -1.34985328e-01
5.68307102e-01 5.00515401e-01 6.92208469e-01 -9.16807413e-01
6.65044308e-01 2.18388394e-01 -2.99772441e-01 -8.11179280e-01
-1.56409547e-01 -1.14318565e-01 -8.65494683e-02 -4.03141767e-01
1.02563596e+00 -7.75110841e-01 -1.36452639e+00 1.28105849e-01
-9.43870544e-01 -8.30097675e-01 -2.55159196e-02 4.13408995e-01
-9.41843331e-01 -2.83092447e-02 -4.57582831e-01 -1.06732142e+00
2.43747890e-01 -1.71983063e+00 1.03858137e+00 2.78092563e-01
-3.16168904e-01 -6.66161001e-01 -1.61896527e-01 -2.28029732e-02
4.19592947e-01 -8.06850046e-02 1.22006845e+00 -5.97516119e-01
-9.26108599e-01 -4.71437424e-02 1.98003910e-02 -3.60159665e-01
-2.11316068e-02 -3.17172378e-01 -9.29567397e-01 -1.27092749e-01
-3.48317087e-01 -7.83454180e-01 4.22798455e-01 9.71002281e-02
7.67465532e-01 -1.28666803e-01 -2.49097109e-01 3.74603659e-01
1.05150795e+00 8.77323270e-01 3.19905639e-01 6.23282075e-01
2.20200941e-01 7.53578424e-01 7.36281276e-01 3.72207850e-01
6.32018089e-01 8.74747694e-01 9.49910164e-01 1.74451947e-01
2.34662563e-01 -7.30444074e-01 4.56510842e-01 1.03840083e-01
2.99082249e-01 -4.23275560e-01 -9.62243021e-01 3.25026065e-01
-1.87609220e+00 -7.83577621e-01 6.79976583e-01 2.21177554e+00
7.02147961e-01 2.50156045e-01 2.54558653e-01 -1.66619420e-02
2.22820058e-01 1.47395385e-02 -8.09796453e-01 -2.59581834e-01
2.03773946e-01 -2.50825465e-01 -1.30722538e-01 1.04742801e+00
-9.34502602e-01 1.10552943e+00 5.79697371e+00 2.56442696e-01
-1.08489692e+00 -1.67360410e-01 2.04779848e-01 6.76023355e-03
-4.21484709e-01 6.12385087e-02 -4.11125958e-01 1.83197439e-01
5.81459939e-01 2.06444144e-01 1.06020474e+00 7.45941520e-01
1.54752120e-01 -5.23257494e-01 -1.37111664e+00 9.98651862e-01
-2.83536106e-01 -8.18954885e-01 -8.64257962e-02 -1.92136154e-01
-2.31713861e-01 -7.93268383e-02 1.94750220e-01 5.14527321e-01
6.89172864e-01 -1.18120551e+00 1.05239356e+00 4.97845352e-01
6.42965615e-01 -7.59773552e-01 1.31003395e-01 6.71311975e-01
-1.15239000e+00 -3.12712640e-01 2.06135824e-01 3.34386490e-02
3.37178171e-01 -4.34086174e-01 -7.52098262e-01 9.47925746e-02
7.27422535e-01 -9.01419520e-02 -9.09081772e-02 8.86033952e-01
-5.49954057e-01 1.94982663e-01 -3.25261831e-01 -2.38822252e-01
4.17489290e-01 -1.19962767e-01 6.91937327e-01 8.50427449e-01
3.41726899e-01 2.29053199e-01 4.99385566e-01 7.84298658e-01
2.42180571e-01 -8.82614776e-02 -8.82821441e-01 -2.56168693e-01
5.98372638e-01 8.76275003e-01 -6.04562104e-01 9.30103138e-02
-2.08870351e-01 1.05962145e+00 3.14319700e-01 8.57152939e-01
-6.17595971e-01 -6.42257154e-01 6.60602093e-01 -3.47992219e-02
4.43991125e-02 -5.30095994e-01 1.34508684e-01 -6.56355560e-01
-2.09141567e-01 -1.10838318e+00 3.14102530e-01 -1.25630498e+00
-7.48683035e-01 8.41989934e-01 -1.39449984e-01 -8.27631831e-01
-6.58204317e-01 -5.24006605e-01 -3.85875255e-01 8.09934080e-01
-1.17924261e+00 -7.22069502e-01 -3.05644751e-01 6.63609684e-01
6.14977896e-01 -2.22304553e-01 9.99261677e-01 4.10737544e-02
-1.13583513e-01 3.08669329e-01 -5.84558785e-01 5.70615679e-02
6.51971817e-01 -1.27547967e+00 2.01899722e-01 2.94556946e-01
3.09428662e-01 5.51031590e-01 9.17597175e-01 -4.13055569e-01
-1.58581614e+00 -3.99503082e-01 4.71417099e-01 -3.45559120e-01
5.56547880e-01 -4.61628377e-01 -7.77907789e-01 7.25083113e-01
2.76946843e-01 -3.16151708e-01 4.39179301e-01 2.17582714e-02
-3.89165193e-01 1.25204682e-01 -1.00338769e+00 1.04640961e+00
9.93142068e-01 -9.67876136e-01 -5.71575463e-01 1.91719636e-01
8.75328004e-01 -3.71311516e-01 -2.27775455e-01 1.93546098e-02
6.57781243e-01 -9.38184261e-01 9.88432407e-01 -6.87445998e-01
1.24820948e-01 -5.94055474e-01 -4.53795582e-01 -1.56200135e+00
4.55455966e-02 -6.32240832e-01 2.78228015e-01 8.80810142e-01
5.23540437e-01 -6.33164525e-01 4.91263986e-01 7.69053698e-01
-7.97472224e-02 -3.80399644e-01 -9.71671402e-01 -3.31209540e-01
-2.51121163e-01 -6.41126931e-01 2.94511855e-01 5.11907697e-01
1.27848446e-01 5.13798237e-01 -2.81847030e-01 4.90591764e-01
2.17948377e-01 -8.25808123e-02 7.25792170e-01 -9.70041811e-01
-4.18284267e-01 -4.94913757e-01 -3.76408398e-02 -1.40829754e+00
3.37258369e-01 -6.33040428e-01 6.97187304e-01 -1.58133972e+00
-3.05077881e-01 -5.59876621e-01 -7.65787512e-02 7.41085231e-01
1.25244543e-01 -3.64392757e-01 4.09576237e-01 -8.17961693e-02
-7.28772998e-01 6.60089254e-01 1.09743476e+00 -1.84727296e-01
-5.29425323e-01 4.15244810e-02 -7.07964420e-01 7.77948976e-01
5.61289370e-01 -3.07011217e-01 -7.90105939e-01 -5.89739680e-01
4.35299307e-01 8.41828227e-01 5.88096738e-01 -8.57931912e-01
5.50504446e-01 -1.99690506e-01 1.30379185e-01 -3.18249196e-01
7.74556577e-01 -9.57075357e-01 -1.36660725e-01 3.13688517e-01
-7.27294803e-01 3.33127409e-01 2.62980253e-01 7.04609215e-01
-1.70776069e-01 -1.77750260e-01 4.50563312e-01 -3.16702574e-01
-8.79384041e-01 -9.39205959e-02 -1.06242239e+00 9.71786231e-02
7.35441804e-01 -8.84008110e-02 -3.59992057e-01 -1.08244324e+00
-1.14193571e+00 6.76139653e-01 5.10737002e-01 4.32810426e-01
7.84111321e-01 -1.19508731e+00 1.39016192e-02 2.49296963e-01
2.55454659e-01 -7.79861063e-02 -2.84236693e-03 5.04081249e-01
-2.37455070e-01 3.00565720e-01 -8.56878459e-02 -6.51625037e-01
-1.06979811e+00 5.34523726e-01 5.90772450e-01 3.57871741e-01
-3.13414365e-01 9.18659687e-01 5.60135961e-01 -7.28303850e-01
1.03801906e+00 -1.85669437e-01 -1.76897049e-01 -6.06302954e-02
4.37099427e-01 -9.11519453e-02 -2.63311803e-01 -4.50796455e-01
-4.21120048e-01 4.94483173e-01 3.07097822e-01 -9.41270351e-01
9.32108462e-01 -2.78237730e-01 2.98320830e-01 6.49991453e-01
7.64034986e-01 1.46480110e-02 -1.44300461e+00 -1.12962835e-01
-2.49489974e-02 -3.08760315e-01 -1.53755859e-01 -1.11138093e+00
-6.67869210e-01 9.85637009e-01 5.80121994e-01 2.88690388e-01
8.16843987e-01 7.07852319e-02 3.22009176e-01 8.49168897e-01
7.16737986e-01 -8.90107214e-01 5.62200189e-01 7.59872973e-01
1.05784965e+00 -1.16744769e+00 -6.75077915e-01 8.66748095e-02
-1.03737164e+00 1.06281960e+00 8.20372045e-01 3.84641349e-01
3.48602027e-01 2.10545942e-01 4.76031035e-01 -1.92105412e-01
-9.72185791e-01 -3.10003638e-01 -5.61068356e-02 8.13605845e-01
6.39299750e-02 -1.38353482e-01 6.80873156e-01 3.07234436e-01
-8.59569982e-02 -3.76169503e-01 2.51446396e-01 1.10665119e+00
-6.98601067e-01 -8.03481042e-01 -4.38240975e-01 -1.13029383e-01
7.77780190e-02 9.28315595e-02 -6.24264896e-01 8.19682360e-01
-2.13516444e-01 1.29332387e+00 -2.65603866e-02 -2.78879106e-01
5.75532556e-01 1.88659698e-01 2.37797305e-01 -5.11925101e-01
-3.88356507e-01 1.41867459e-01 1.12988137e-01 -8.18520725e-01
5.39043173e-02 -5.59268415e-01 -1.74669588e+00 2.49806702e-01
-7.90008437e-03 3.63513172e-01 6.89860523e-01 9.17118013e-01
1.73222065e-01 4.75878805e-01 4.64569241e-01 -1.08303547e+00
-5.34210563e-01 -4.13543791e-01 -2.20816225e-01 -2.75333114e-05
6.36692107e-01 -8.04669678e-01 -4.52429771e-01 -3.46528471e-01] | [4.409423828125, 0.7278719544410706] |
6333ed10-6912-477f-93b1-13f7d46af40a | joint-open-knowledge-base-canonicalization-1 | 2212.01207 | null | https://arxiv.org/abs/2212.01207v1 | https://arxiv.org/pdf/2212.01207v1.pdf | Joint Open Knowledge Base Canonicalization and Linking | Open Information Extraction (OIE) methods extract a large number of OIE triples (noun phrase, relation phrase, noun phrase) from text, which compose large Open Knowledge Bases (OKBs). However, noun phrases (NPs) and relation phrases (RPs) in OKBs are not canonicalized and often appear in different paraphrased textual variants, which leads to redundant and ambiguous facts. To address this problem, there are two related tasks: OKB canonicalization (i.e., convert NPs and RPs to canonicalized form) and OKB linking (i.e., link NPs and RPs with their corresponding entities and relations in a curated Knowledge Base (e.g., DBPedia). These two tasks are tightly coupled, and one task can benefit significantly from the other. However, they have been studied in isolation so far. In this paper, we explore the task of joint OKB canonicalization and linking for the first time, and propose a novel framework JOCL based on factor graph model to make them reinforce each other. JOCL is flexible enough to combine different signals from both tasks, and able to extend to fit any new signals. A thorough experimental study over two large scale OIE triple data sets shows that our framework outperforms all the baseline methods for the task of OKB canonicalization (OKB linking) in terms of average F1 (accuracy). | ['Xiaojie Yuan', 'Zhenglu Yang', 'Jianyong Wang', 'Yuanfei Wang', 'Wei Shen', 'Yinan Liu'] | 2022-12-02 | joint-open-knowledge-base-canonicalization | https://dl.acm.org/doi/10.1145/3448016.3452776 | https://dl.acm.org/doi/10.1145/3448016.3452776 | proceedings-of-the-2021-international | ['open-information-extraction'] | ['natural-language-processing'] | [-2.56604642e-01 4.04886872e-01 -4.27669138e-01 -1.53726861e-01
-5.80914259e-01 -7.66224265e-01 4.55207467e-01 3.11693817e-01
-2.88756996e-01 1.13880110e+00 3.00603807e-01 -1.98908716e-01
-3.11863244e-01 -9.66796875e-01 -1.00047231e+00 -2.96049416e-01
1.64601132e-01 7.01192319e-01 4.81945783e-01 -5.29719830e-01
-3.25880945e-01 4.52577621e-02 -1.18183517e+00 3.35875064e-01
1.15572071e+00 6.16782010e-01 8.81986022e-02 -1.18129365e-01
-4.81485367e-01 4.33752656e-01 -4.42965865e-01 -1.05399263e+00
9.59162489e-02 2.40464620e-02 -1.01996768e+00 -4.54685003e-01
3.31657499e-01 2.94668406e-01 -4.12782013e-01 1.00633073e+00
2.73524493e-01 1.27119005e-01 6.69482350e-01 -1.26132941e+00
-9.61961091e-01 1.15945387e+00 -5.30718923e-01 1.10983565e-01
5.88952303e-01 -5.11532843e-01 1.57960856e+00 -1.17035854e+00
9.39098537e-01 1.29962432e+00 7.19840765e-01 3.96057159e-01
-1.25489879e+00 -7.61089504e-01 1.49404392e-01 4.85841900e-01
-1.47304058e+00 -7.99221098e-02 3.55576217e-01 -2.03926146e-01
1.16297185e+00 3.25732231e-01 5.93617678e-01 1.14406395e+00
-5.58029599e-02 7.16145217e-01 9.00910139e-01 -5.74278295e-01
-2.11135998e-01 2.02148676e-01 7.62230515e-01 4.12212998e-01
8.67108405e-01 -2.35091791e-01 -5.98254859e-01 -2.15512584e-03
2.21899033e-01 -2.62314349e-01 -4.70279157e-01 -1.74597293e-01
-1.34247386e+00 5.54892361e-01 6.37789905e-01 2.67809749e-01
-2.85115331e-01 -1.65785238e-01 3.57505560e-01 1.56595662e-01
3.09532955e-02 5.38955867e-01 -6.91313922e-01 1.82554960e-01
-2.25033000e-01 5.22353590e-01 1.19170105e+00 1.58817530e+00
9.58826303e-01 -7.02079356e-01 -2.47169822e-01 1.20666683e+00
4.54057038e-01 5.56435764e-01 3.09671134e-01 -3.61370236e-01
9.78330791e-01 7.60542691e-01 2.65313350e-02 -1.24997568e+00
-4.94930685e-01 -1.70113876e-01 -6.37594998e-01 -9.84600186e-01
1.79706365e-01 -1.75303683e-01 -7.15824783e-01 1.78843343e+00
4.44904566e-01 4.28502411e-02 3.46268684e-01 6.99211895e-01
1.59986341e+00 8.85562241e-01 1.29174501e-01 -3.33784640e-01
1.79841042e+00 -1.00111485e+00 -1.04573834e+00 -4.45342153e-01
3.13522756e-01 -6.52085483e-01 8.54489863e-01 1.74542397e-01
-6.28423572e-01 -3.08088869e-01 -1.14128113e+00 -4.74253625e-01
-7.67901599e-01 -1.50293589e-01 5.36931932e-01 3.07286382e-01
-3.76421899e-01 2.72325724e-01 -3.93782586e-01 -4.18643773e-01
1.47726044e-01 1.51484132e-01 -6.40426338e-01 -2.80511856e-01
-1.98169553e+00 1.02097595e+00 1.10882521e+00 2.23709673e-01
-9.93325710e-02 -4.94141161e-01 -8.73274505e-01 1.18815526e-01
1.00316691e+00 -9.59371507e-01 7.61168301e-01 -1.13044187e-01
-8.27555001e-01 6.44229949e-01 -1.05542198e-01 -3.91509354e-01
-9.97253656e-02 -4.47476327e-01 -7.66171753e-01 -2.19123945e-01
4.30024385e-01 5.66163838e-01 5.13437986e-01 -1.41696453e+00
-7.85550416e-01 -3.98130536e-01 2.40929589e-01 2.29795799e-01
-5.20336747e-01 4.79396507e-02 -9.51203346e-01 -7.96479523e-01
3.45941037e-01 -9.56457436e-01 1.55155212e-01 -9.25839484e-01
-7.19334722e-01 -6.96180582e-01 4.56734598e-01 -9.10912395e-01
1.77223098e+00 -1.96133137e+00 5.15879214e-01 2.70369768e-01
4.57907051e-01 3.04998189e-01 -5.97493630e-03 6.97563529e-01
-1.97958261e-01 3.69980216e-01 -1.37980627e-02 1.24823093e-01
2.30697364e-01 7.34232128e-01 -5.59261262e-01 -2.48016685e-01
-5.83795421e-02 1.31899190e+00 -1.08787084e+00 -7.76379287e-01
-5.19322574e-01 -1.02045067e-01 -2.50588417e-01 -1.41090631e-01
-3.52186918e-01 4.28752378e-02 -3.74538273e-01 7.20928550e-01
5.34645438e-01 -7.21800104e-02 4.67329264e-01 -7.22105980e-01
2.02813685e-01 4.45721626e-01 -1.19798303e+00 1.33511055e+00
-3.01059663e-01 4.26956505e-01 -4.92415667e-01 -8.55537295e-01
9.16912794e-01 3.57467145e-01 2.81634182e-01 -3.28204274e-01
4.41599078e-03 4.89252895e-01 -1.63922176e-01 -6.23140275e-01
6.99538827e-01 -2.26527989e-01 -4.73406583e-01 7.82865062e-02
5.94083905e-01 3.25511061e-02 7.55073547e-01 4.23990846e-01
1.10567939e+00 4.75740135e-02 7.60469198e-01 -3.44342291e-02
3.90900940e-01 1.23295411e-01 1.04235792e+00 6.24015689e-01
2.49482930e-01 9.90370065e-02 6.53701484e-01 -2.92147905e-01
-8.70810270e-01 -1.13084173e+00 -3.80724192e-01 7.61090994e-01
4.41716433e-01 -1.02618837e+00 -4.01513189e-01 -9.74882543e-01
3.19227099e-01 6.06242716e-01 -4.50770438e-01 -3.36550325e-02
-6.59950614e-01 -7.86908150e-01 6.91276670e-01 4.59414303e-01
2.24398747e-01 -9.64353859e-01 2.84960568e-01 3.37984651e-01
-8.06876481e-01 -1.59608722e+00 -2.04475239e-01 3.18320304e-01
-4.47635025e-01 -1.12890017e+00 -2.02646077e-01 -8.67370903e-01
4.84299183e-01 1.78184584e-01 1.43639529e+00 -2.15797499e-01
1.65212408e-01 6.80981800e-02 -9.17919338e-01 -6.25552416e-01
-2.62301475e-01 2.23646820e-01 2.80030876e-01 -2.40138099e-01
7.55600333e-01 -5.71074247e-01 3.99809778e-02 5.13038039e-01
-9.19252634e-01 1.95363574e-02 6.24299169e-01 1.07658720e+00
6.65385008e-01 1.69209510e-01 8.00666273e-01 -1.20500648e+00
8.35293233e-01 -6.28124416e-01 -4.23705280e-01 7.83708155e-01
-6.36852562e-01 6.77909851e-02 4.61575687e-01 -1.86153829e-01
-1.02818036e+00 -3.43013704e-01 1.24710731e-01 -8.16922709e-02
1.57034755e-01 1.24055159e+00 -5.70419252e-01 2.51176447e-01
6.23001397e-01 -1.11633115e-01 -6.10692918e-01 -4.98031735e-01
6.90919995e-01 8.99899185e-01 6.98035657e-01 -8.15120757e-01
9.03877735e-01 8.46887901e-02 -7.74337277e-02 -6.63106978e-01
-1.19669604e+00 -5.66113234e-01 -6.68222189e-01 1.59438223e-01
5.83227694e-01 -1.05995286e+00 -3.63394052e-01 1.08917132e-01
-1.44203901e+00 3.97619396e-01 -2.73940027e-01 3.62030447e-01
-6.94768652e-02 4.96422589e-01 -4.08805132e-01 -7.74905235e-02
-1.81006461e-01 -6.13626361e-01 8.03325951e-01 3.16867471e-01
-2.48474851e-01 -6.68879986e-01 1.95817664e-01 5.74213862e-01
-4.60043281e-01 -1.07848737e-02 1.18682814e+00 -9.73761141e-01
-4.67264503e-01 -2.44134203e-01 -4.12839383e-01 3.39969307e-01
1.48171872e-01 -2.29542032e-02 -5.26275098e-01 3.88064273e-02
-3.52788985e-01 -2.04856396e-01 9.56752360e-01 -1.72612309e-01
3.89881641e-01 -4.51850832e-01 -6.75644994e-01 3.90027970e-01
1.23214066e+00 -2.41453061e-03 7.15901256e-01 3.16200763e-01
9.18966651e-01 6.94050670e-01 8.25802982e-01 2.60836184e-02
6.77951097e-01 8.10811102e-01 -2.63849236e-02 3.26088250e-01
-2.73379534e-01 -5.30318618e-01 2.89134592e-01 1.19203949e+00
-1.88827351e-01 -3.27031612e-01 -1.08442390e+00 6.67015433e-01
-2.11260223e+00 -7.70662367e-01 -4.99135435e-01 1.74003375e+00
1.29562104e+00 3.78611982e-02 -1.57570913e-01 -1.30590275e-01
8.15924942e-01 -9.55063477e-02 -1.41275480e-01 -1.09251328e-01
-4.09087718e-01 1.55845702e-01 3.39990944e-01 2.26534754e-01
-1.03339243e+00 1.16661298e+00 5.15277910e+00 1.21252334e+00
-3.22119206e-01 1.89839050e-01 -2.49115750e-01 2.93283105e-01
-5.78191996e-01 4.77196693e-01 -1.33934259e+00 5.03795743e-01
4.83906120e-01 -3.20418894e-01 4.91044492e-01 5.78725457e-01
-4.73440230e-01 -4.55771871e-02 -1.21700454e+00 1.01572526e+00
-5.03945444e-03 -1.16414821e+00 3.88451636e-01 -1.17757469e-01
7.84585476e-01 -8.08232278e-02 -6.53735518e-01 7.92206764e-01
3.64923596e-01 -7.49487758e-01 6.04237258e-01 4.49852705e-01
5.02999246e-01 -4.69500363e-01 8.77524793e-01 3.51577908e-01
-1.34664893e+00 -1.44507095e-01 -5.55892110e-01 2.95739204e-01
2.58680403e-01 7.91232049e-01 -5.92086673e-01 1.26161718e+00
7.68261373e-01 8.14484179e-01 -6.05203152e-01 6.78843796e-01
-8.99801552e-01 3.79018009e-01 -4.36695516e-01 -2.25084409e-01
2.01575812e-02 -1.93494678e-01 7.82570124e-01 1.09174061e+00
8.67522284e-02 2.63152331e-01 8.39496553e-02 8.78774226e-01
-3.86608452e-01 3.73505235e-01 -4.80030447e-01 -1.48892388e-01
8.51559043e-01 1.25269139e+00 -4.62935477e-01 -2.87247777e-01
-7.71004558e-01 5.88372469e-01 7.44250715e-01 3.49110335e-01
-1.00550270e+00 -5.73377728e-01 3.05775940e-01 -2.59636581e-01
4.29655135e-01 -2.93899439e-02 5.93216950e-03 -1.59626544e+00
4.33348954e-01 -7.71072268e-01 7.28226900e-01 -7.01859236e-01
-1.79007697e+00 5.24090111e-01 2.86846519e-01 -1.12413514e+00
4.66315337e-02 -5.08630335e-01 -1.33434394e-02 7.53364325e-01
-1.40230584e+00 -1.24818766e+00 -3.01687121e-01 5.38709104e-01
1.76354378e-01 5.11674909e-04 7.58933485e-01 6.39567912e-01
-7.66382456e-01 5.54203093e-01 1.53312758e-01 3.88118505e-01
9.78176594e-01 -1.23976898e+00 3.62761945e-01 9.47431803e-01
6.02152705e-01 1.06456554e+00 4.93103027e-01 -1.15943587e+00
-1.23587513e+00 -9.26987946e-01 1.30999243e+00 -6.48152709e-01
8.71673465e-01 -3.95833462e-01 -9.66866076e-01 9.43335891e-01
5.34619018e-03 1.23763241e-01 5.96088290e-01 7.52772033e-01
-6.62731290e-01 -2.06488535e-01 -4.86767083e-01 6.52754486e-01
1.21710002e+00 -4.23513025e-01 -1.37462294e+00 2.75589705e-01
1.05907810e+00 -6.11395955e-01 -9.86474335e-01 7.83248901e-01
4.73187506e-01 -5.55901587e-01 1.07007825e+00 -9.68173742e-01
4.00836587e-01 -4.68998224e-01 -2.31635377e-01 -1.37644660e+00
-2.26954326e-01 -4.27683264e-01 -4.79457676e-01 1.61380458e+00
8.63835931e-01 -5.29545009e-01 1.38928309e-01 4.11364108e-01
-1.46362111e-01 -7.99683690e-01 -1.00527632e+00 -1.14832532e+00
-1.42556429e-01 -3.97956967e-01 6.95577979e-01 1.00935948e+00
6.18285179e-01 1.08121049e+00 -3.32629204e-01 2.27967501e-01
2.01511592e-01 4.56486613e-01 7.45636225e-01 -1.27950943e+00
-1.83270812e-01 -1.16257980e-01 -2.56477147e-01 -1.05246031e+00
2.15606898e-01 -1.18631494e+00 -6.67230114e-02 -1.67659879e+00
4.00217801e-01 -7.54813433e-01 -3.13695073e-02 7.68383563e-01
-6.08664155e-01 3.58623713e-02 1.95487440e-01 4.05199617e-01
-6.47233009e-01 5.26904583e-01 1.02713275e+00 -1.97203115e-01
-2.36972019e-01 -2.88242936e-01 -9.54180956e-01 6.70664847e-01
3.96974981e-01 -6.52351022e-01 -2.09072560e-01 -3.70262176e-01
8.61105919e-01 -1.17552998e-02 3.27358007e-01 -5.56199491e-01
3.16700548e-01 -6.94294870e-02 -1.72349751e-01 -5.03551245e-01
3.75998393e-02 -8.83415520e-01 3.02869707e-01 -9.62162092e-02
1.01142690e-01 -1.63265720e-01 1.60300389e-01 8.92241895e-01
-4.58454520e-01 -6.27678454e-01 1.31330907e-01 4.01945300e-02
-8.91907215e-01 1.55747160e-01 1.73054278e-01 5.11212111e-01
9.58918810e-01 2.20410917e-02 -6.53195858e-01 7.91321322e-02
-8.66023004e-01 4.59754437e-01 -7.75901973e-02 6.87710524e-01
4.57488358e-01 -1.57368004e+00 -6.38058305e-01 -1.46368876e-01
5.12492001e-01 3.52015495e-01 5.89605346e-02 8.92490804e-01
-2.32794553e-01 6.35013223e-01 1.15484394e-01 -1.73886165e-01
-1.05064392e+00 7.97959864e-01 -4.80006374e-02 -7.32807934e-01
-5.37152588e-01 7.49805808e-01 -9.37364921e-02 -6.05241358e-01
2.18162209e-01 -3.20383340e-01 -4.69780147e-01 4.97945458e-01
1.42835751e-01 2.22071648e-01 2.06774354e-01 -6.25494957e-01
-4.66280937e-01 4.72940505e-01 -1.62789628e-01 2.78662950e-01
1.31759286e+00 -2.46159256e-01 -7.23601997e-01 3.72197241e-01
9.07729089e-01 3.69769156e-01 -3.16960931e-01 -6.22660100e-01
4.30956542e-01 -3.05929750e-01 -3.10854793e-01 -7.54493177e-01
-6.69844687e-01 2.42567733e-01 -2.83500463e-01 3.77797365e-01
7.15198040e-01 4.51754540e-01 1.17200017e+00 7.62769699e-01
5.36224961e-01 -1.13920736e+00 -3.20622623e-01 8.58292758e-01
1.04403067e+00 -1.04183900e+00 1.79959759e-01 -1.10664940e+00
-6.17440164e-01 1.04609001e+00 7.65713215e-01 2.89893836e-01
6.14806950e-01 3.69262137e-02 -2.49878570e-01 -4.55140680e-01
-5.91253698e-01 -3.94880980e-01 5.94121754e-01 5.09797335e-01
6.73755035e-02 1.32928684e-01 -8.67047727e-01 1.46233463e+00
-2.36858413e-01 -2.25885764e-01 3.10644656e-01 6.31163895e-01
-1.85224742e-01 -1.39008152e+00 -3.73632967e-01 4.29704607e-01
-1.58634245e-01 -4.02084023e-01 -5.13843000e-01 9.49387491e-01
4.53874469e-01 8.08523238e-01 -3.89358759e-01 -5.02890110e-01
5.58756471e-01 1.46026835e-01 5.21397114e-01 -8.22814286e-01
-3.32024664e-01 -1.59972548e-01 5.34068525e-01 -2.53908336e-01
-5.41369915e-01 -5.17288506e-01 -1.25402880e+00 -1.52476788e-01
-8.61726820e-01 2.96209693e-01 2.39238322e-01 1.12557817e+00
1.95411533e-01 5.62239707e-01 2.82351613e-01 -9.68316942e-03
-3.56839985e-01 -9.18256223e-01 -7.58919597e-01 5.20533383e-01
-3.04607391e-01 -1.10054636e+00 6.62364513e-02 -5.64780906e-02] | [9.071941375732422, 8.330933570861816] |
dcefbc3f-3827-4bac-9bb9-c3fe62dee960 | accurate-facial-image-parsing-at-real-time | null | null | https://ieeexplore.ieee.org/document/8682072 | https://ieeexplore.ieee.org/document/8682072 | Accurate facial image parsing at real-time speed | In this paper, we propose a design scheme for deep learning networks in the face parsing task with promising accuracy and real-time inference speed. By analyzing the differences between the general image parsing task and face parsing task, we first revisit the structure of traditional FCN and make improvements to adapt to the unique properties of the face parsing task. Especially, the concept of Normalized Receptive Field is proposed to give more insights on designing the network. Then, a novel loss function called Statistical Contextual Loss is introduced, which integrates richer contextual information and regularizes features during training. For further model acceleration, we propose a semi-supervised distillation scheme that effectively transfers the learned knowledge to a lighter network. Extensive experiments on LFW and Helen dataset demonstrate the significant superiority of the new design scheme on both efficacy and efficiency. | ['Hefei Ling', 'Yao Sun', 'Si Liu', 'Zhen Wei'] | 2019-04-09 | null | null | null | ieee-transactions-on-image-processing-2019-4 | ['face-parsing'] | ['computer-vision'] | [ 2.81020105e-01 3.69722009e-01 -3.77940059e-01 -1.06876051e+00
-3.77665490e-01 -1.63122952e-01 3.64437699e-01 -5.47196090e-01
-3.46216679e-01 6.19685829e-01 2.57613808e-01 -2.69371390e-01
-1.94741741e-01 -7.57601023e-01 -7.91449368e-01 -7.17812598e-01
-5.22478446e-02 -1.81856025e-02 -1.43015936e-01 1.70708627e-01
-1.34053854e-02 6.40763342e-01 -1.27901769e+00 1.94190517e-01
8.08144331e-01 1.38148308e+00 5.21806717e-01 2.27779388e-01
-3.52502227e-01 1.04578590e+00 -3.02752554e-01 -5.74354947e-01
2.56927073e-01 -4.03213590e-01 -1.02563965e+00 -3.43427099e-02
5.28043389e-01 -5.20614088e-01 -2.86314607e-01 1.00393319e+00
6.12015426e-01 1.18561246e-01 2.30865508e-01 -8.98895264e-01
-7.56716192e-01 7.33377218e-01 -4.07909989e-01 1.84890628e-01
1.29787400e-02 -1.13092817e-01 1.01390111e+00 -7.90875316e-01
3.56571466e-01 1.61962426e+00 6.81846678e-01 1.10780823e+00
-8.83225203e-01 -8.90700281e-01 6.16212785e-01 1.72883898e-01
-9.62244749e-01 -5.25413215e-01 9.75542188e-01 9.75622311e-02
6.11882865e-01 -3.49902883e-02 3.06398094e-01 1.05488312e+00
-2.04980135e-01 1.12662613e+00 8.86786222e-01 -4.60641921e-01
1.84898257e-01 -1.08085260e-01 1.38753816e-01 1.08058107e+00
4.99901548e-02 -2.49819867e-02 -3.68047655e-01 1.32231474e-01
8.23669553e-01 5.75590320e-02 -1.10438123e-01 -3.37694883e-01
-4.87451702e-01 1.00694060e+00 7.17578590e-01 2.44914919e-01
-1.31317064e-01 2.90138721e-01 3.55579197e-01 -2.08427683e-02
5.72837293e-01 -2.35493742e-02 -5.35469234e-01 3.42401862e-01
-7.81749904e-01 5.04132472e-02 4.95082200e-01 9.02680099e-01
7.15329409e-01 2.73798797e-02 -3.74112546e-01 1.08564091e+00
3.83862197e-01 3.09393793e-01 2.89090544e-01 -1.35089183e+00
4.24974203e-01 5.18984795e-01 -4.54762518e-01 -8.83718729e-01
-4.51544285e-01 -4.49500054e-01 -9.95364785e-01 -2.46637836e-01
1.43734962e-01 -2.03112021e-01 -9.44233596e-01 2.17127490e+00
4.74541277e-01 3.80611539e-01 -6.01153597e-02 7.72245646e-01
1.03435588e+00 4.47346658e-01 4.81120169e-01 -1.26656696e-01
1.49240255e+00 -1.12001204e+00 -8.08171690e-01 -2.36834362e-01
4.05613959e-01 -3.00204694e-01 1.10534346e+00 4.50439006e-03
-1.14183271e+00 -8.11052084e-01 -8.20388615e-01 -2.66849607e-01
-2.26745650e-01 2.91509271e-01 1.14450157e+00 6.29505277e-01
-1.17137086e+00 6.67845190e-01 -7.38330126e-01 -2.63172716e-01
1.01741588e+00 5.95047474e-01 -2.20546275e-01 -1.78454936e-01
-1.14070845e+00 3.60241026e-01 3.85927171e-01 4.45520341e-01
-6.94809735e-01 -7.69249439e-01 -9.87727761e-01 2.58216828e-01
5.71374536e-01 -8.21860909e-01 1.33841074e+00 -1.10914242e+00
-1.89123034e+00 7.68263698e-01 -3.18363458e-01 -2.87513644e-01
2.48664975e-01 -2.22152755e-01 -2.47292861e-01 3.47577900e-01
-9.18091610e-02 1.00823534e+00 8.69787335e-01 -1.03820002e+00
-6.19547188e-01 -3.62188876e-01 4.07781184e-01 -2.05366462e-02
-5.29096842e-01 6.54740483e-02 -6.06241763e-01 -7.33540356e-01
-7.21422955e-02 -5.03517926e-01 -1.66636020e-01 9.55172330e-02
-3.65555137e-01 -6.02737367e-01 7.81945229e-01 -5.12921333e-01
1.15718102e+00 -2.05555296e+00 -6.05197288e-02 1.16369687e-01
3.46442342e-01 5.60166240e-01 -4.55810368e-01 2.88666915e-02
-2.25624159e-01 7.60087371e-02 -4.11416322e-01 -5.46148479e-01
-2.74322536e-02 5.28328419e-01 -2.26668894e-01 1.20516263e-01
6.92990005e-01 1.14611936e+00 -6.73476219e-01 -5.43559253e-01
-1.70622796e-01 5.95268011e-01 -8.62367570e-01 3.72234106e-01
-2.02395141e-01 5.00655413e-01 -8.26278329e-01 7.80901909e-01
9.85766709e-01 -3.84712368e-01 3.50852907e-01 -3.65912169e-01
2.34819755e-01 2.64622092e-01 -8.06168079e-01 1.86264479e+00
-6.24333143e-01 1.40594453e-01 4.76906061e-01 -1.44189084e+00
8.40010762e-01 3.07447687e-02 2.15044975e-01 -9.06387866e-01
3.36770743e-01 -2.31801331e-01 -3.16215366e-01 -4.54231262e-01
9.40127205e-03 -1.58501759e-01 2.30897397e-01 2.79637188e-01
3.69898319e-01 4.60675120e-01 5.55047132e-02 5.00217080e-02
7.80175924e-01 2.60016382e-01 -4.76882001e-03 -5.79781473e-01
8.77108872e-01 -6.33793056e-01 9.78088856e-01 5.50282359e-01
-2.83466339e-01 2.66520232e-01 6.71592951e-01 -6.75142527e-01
-5.81689596e-01 -8.15168262e-01 -2.32988968e-01 1.47544456e+00
2.02255324e-02 -4.66890484e-01 -1.16890335e+00 -1.17947745e+00
-9.32260901e-02 2.85710931e-01 -8.40019763e-01 -2.18180478e-01
-9.29359078e-01 -9.49461222e-01 4.80258286e-01 1.02206278e+00
1.03673363e+00 -1.08202708e+00 -2.95906305e-01 -1.17282286e-01
-1.04590274e-01 -1.29324400e+00 -3.30315799e-01 -4.78266664e-02
-1.02326024e+00 -9.62163985e-01 -3.72606456e-01 -1.26557040e+00
7.84792244e-01 1.31656930e-01 1.12546623e+00 3.88476253e-01
-2.87437022e-01 2.23545954e-01 -2.43199140e-01 -1.60301074e-01
-3.81780900e-02 4.11977023e-02 -2.06056505e-01 6.49461057e-03
3.37572962e-01 -5.19436419e-01 -7.43740618e-01 1.07634589e-01
-8.46101344e-01 5.47025539e-02 6.86528802e-01 9.34051335e-01
4.68187243e-01 -2.68552341e-02 6.60094619e-01 -1.12284708e+00
5.25055170e-01 -2.50900209e-01 -5.45588315e-01 4.14043248e-01
-5.15612483e-01 2.70341188e-01 7.26617873e-01 -2.22526178e-01
-1.57909358e+00 1.53230891e-01 -4.11840290e-01 -1.41276523e-01
-1.04100658e-02 2.59762973e-01 -5.49892843e-01 -2.30373800e-01
-1.08910792e-01 7.85024837e-03 -5.23089617e-02 -7.35165894e-01
5.37813067e-01 4.19706911e-01 4.92311239e-01 -1.00096416e+00
6.69422984e-01 5.47184944e-01 8.64379555e-02 -5.58316946e-01
-1.10011995e+00 9.40835103e-02 -6.62764549e-01 9.99021307e-02
8.99668634e-01 -1.00144923e+00 -1.07841432e+00 4.15568382e-01
-1.30722487e+00 -4.70151573e-01 -9.63949263e-02 1.70485035e-01
-4.65311855e-01 4.70151305e-01 -8.60771000e-01 -6.29086673e-01
-4.33279723e-01 -1.01871848e+00 1.10841894e+00 4.12242562e-01
3.36234033e-01 -1.16843271e+00 -1.96447909e-01 4.61099803e-01
5.24462700e-01 -2.83633769e-02 1.14694536e+00 -4.57018256e-01
-5.11342466e-01 2.06636384e-01 -8.18781793e-01 6.19344413e-01
1.99784096e-02 -2.90316433e-01 -1.23988962e+00 -2.75239021e-01
2.26235792e-01 -4.72996652e-01 1.15287673e+00 4.33208615e-01
1.88848615e+00 -2.99652576e-01 -3.11210006e-01 1.09486246e+00
1.31756473e+00 7.99567178e-02 5.20829737e-01 -4.15756181e-02
7.42449045e-01 8.48884284e-01 3.05349678e-01 1.96510926e-01
4.05826211e-01 3.25309694e-01 4.10739809e-01 -3.29967648e-01
-1.58254802e-01 -3.51010263e-01 2.04768896e-01 7.28634119e-01
-1.08348198e-01 -4.88053188e-02 -4.23513353e-01 1.30340338e-01
-1.68980181e+00 -5.86144984e-01 5.70544660e-01 1.64323461e+00
7.49714613e-01 -1.95481703e-02 -4.60863002e-02 -3.45133126e-01
8.55660617e-01 3.06804776e-01 -5.37823081e-01 -3.25196862e-01
-4.85687889e-02 5.69874823e-01 2.51104861e-01 5.03044844e-01
-1.23948252e+00 1.26685691e+00 6.86439323e+00 1.06416380e+00
-1.10798752e+00 2.14573815e-01 9.43380117e-01 7.49301687e-02
-6.19641207e-02 -1.83012798e-01 -1.05675435e+00 3.61216903e-01
7.98623085e-01 3.99111301e-01 5.92149317e-01 9.10293400e-01
-1.46515341e-03 2.45999664e-01 -1.11904204e+00 9.09972310e-01
-3.36385183e-02 -1.33380210e+00 1.89715356e-01 -1.79726347e-01
2.77958751e-01 -1.42157555e-01 8.78835842e-02 4.53020692e-01
3.06489468e-01 -1.06047547e+00 4.27217722e-01 2.58051336e-01
6.59738839e-01 -8.81753385e-01 5.89351416e-01 2.05938712e-01
-1.37814271e+00 -4.30643350e-01 -6.32404506e-01 -2.28330895e-01
6.39281643e-04 5.47410905e-01 -3.86119902e-01 4.82038081e-01
8.50457907e-01 7.27492094e-01 -5.77587843e-01 4.22316492e-01
-4.15079176e-01 6.19028211e-01 -1.03682473e-01 1.77528068e-01
3.09778184e-01 -1.94915563e-01 -3.75501513e-02 1.23125100e+00
3.90440561e-02 2.50016421e-01 8.38184133e-02 9.24647450e-01
-4.51052129e-01 5.66746891e-02 -4.01572585e-01 1.21193327e-01
4.69239831e-01 1.53575337e+00 -6.82800412e-01 -2.59129554e-01
-5.46159029e-01 7.70723939e-01 8.19443583e-01 3.46484542e-01
-9.23113108e-01 -2.51091808e-01 6.57328606e-01 -1.33522883e-01
6.26951754e-01 3.45395319e-02 -7.83744752e-02 -1.03454447e+00
7.91084245e-02 -4.62009281e-01 3.15587938e-01 -3.60046625e-01
-1.23722267e+00 6.97100759e-01 4.26355638e-02 -5.76404691e-01
1.38602471e-02 -1.00616944e+00 -8.76981616e-01 6.18116140e-01
-2.07621121e+00 -1.41338480e+00 -2.19005495e-01 6.09608412e-01
4.69512850e-01 -1.86193168e-01 6.50607765e-01 5.15895784e-01
-1.05109966e+00 9.38803673e-01 -1.59021512e-01 4.53247666e-01
4.14971143e-01 -1.07998204e+00 4.02565092e-01 7.30593324e-01
-2.01721117e-01 8.37973595e-01 1.06856778e-01 -3.04965258e-01
-1.33933342e+00 -1.35526133e+00 7.47219980e-01 2.51240563e-02
4.29115802e-01 -6.88175619e-01 -8.48378420e-01 6.71412885e-01
2.32715607e-01 2.37756819e-01 6.17919385e-01 2.27229938e-01
-4.61440206e-01 -4.89318013e-01 -1.29273164e+00 3.87939155e-01
1.45762360e+00 -5.13848722e-01 -4.99747634e-01 2.57525474e-01
9.41121280e-01 -7.48606473e-02 -7.96174765e-01 7.97266901e-01
5.28270185e-01 -9.72681165e-01 1.13499689e+00 -7.83356190e-01
3.91325235e-01 1.77156046e-01 -1.43919826e-01 -8.40286911e-01
-6.23601437e-01 -6.11156642e-01 -1.67048708e-01 1.52631211e+00
9.35098082e-02 -6.35870099e-01 1.02575350e+00 5.40017128e-01
-7.18154991e-03 -1.10101116e+00 -8.24104488e-01 -6.21052682e-01
2.68177152e-01 -2.30627477e-01 8.65225554e-01 8.56795967e-01
-3.18062156e-01 3.57597202e-01 -3.75262469e-01 4.14076336e-02
6.18247330e-01 1.81125864e-01 3.85055542e-01 -1.26982296e+00
-3.04313987e-01 -4.08305317e-01 -2.80235754e-03 -1.43421340e+00
6.75496519e-01 -8.95722151e-01 7.48581858e-03 -1.07446694e+00
4.75659907e-01 -3.06715816e-01 -5.45822680e-01 6.25325501e-01
-3.55716437e-01 1.57350138e-01 1.94165617e-01 -1.25320971e-01
-7.07989514e-01 6.30610049e-01 1.47089291e+00 6.31392971e-02
1.67636365e-01 -6.11346699e-02 -1.00092125e+00 8.69693041e-01
7.06557512e-01 -2.90711761e-01 -6.34834945e-01 -8.37774873e-01
-1.50630981e-01 -7.63343573e-02 3.44403505e-01 -7.14520156e-01
3.63442972e-02 -3.08435187e-02 4.60688621e-01 -3.79008830e-01
1.97970465e-01 -7.19347715e-01 -5.18396676e-01 3.85415763e-01
-4.56479281e-01 -1.70755759e-01 8.03477615e-02 5.73800981e-01
-1.89746305e-01 -1.07787251e-01 8.91337156e-01 -1.47693977e-01
-7.90454388e-01 6.00055158e-01 1.86573237e-01 9.86683294e-02
7.04289317e-01 -8.12714547e-03 -2.00342000e-01 -1.15054756e-01
-6.26972079e-01 4.10315871e-01 -7.47968489e-03 3.19839388e-01
5.34217834e-01 -1.39494050e+00 -4.48891789e-01 5.28807640e-01
-3.14885616e-01 2.03144029e-01 3.91665429e-01 5.95623374e-01
-3.68340194e-01 5.49678564e-01 -2.89891362e-01 -3.57742578e-01
-1.02222872e+00 5.04901230e-01 2.19757825e-01 -3.46935034e-01
-5.26146293e-01 1.13198411e+00 7.83432543e-01 -6.27765596e-01
5.43806612e-01 -3.18311870e-01 -3.15291494e-01 -1.42014503e-01
8.02747786e-01 1.20211348e-01 -2.25638635e-02 -4.37884361e-01
-3.72232080e-01 7.60568500e-01 -2.22574264e-01 5.27461171e-01
1.49237037e+00 -2.25241169e-01 -3.00483495e-01 -3.00442964e-01
1.35014307e+00 -3.29069167e-01 -1.60076082e+00 -4.61635262e-01
5.92567325e-02 -2.06225261e-01 1.08890444e-01 -6.73636556e-01
-1.72584045e+00 1.12963820e+00 6.10035777e-01 -1.01984151e-01
1.56287515e+00 1.02140285e-01 9.66695249e-01 5.40575147e-01
1.04742773e-01 -1.04829550e+00 5.12649305e-02 5.43342888e-01
4.96606767e-01 -1.17779517e+00 -1.30572662e-01 -7.08998263e-01
-3.66642356e-01 1.16514099e+00 8.24746668e-01 -1.65005922e-01
7.49680758e-01 3.06952745e-01 -1.52257845e-01 -1.86466664e-01
-5.81904888e-01 -1.18323408e-01 8.83256868e-02 5.61083317e-01
4.43254620e-01 -1.21453159e-01 -3.01843822e-01 1.00466478e+00
1.01420075e-01 1.57663345e-01 -2.31114343e-01 9.14534569e-01
-3.95643234e-01 -1.32112825e+00 7.50873759e-02 1.38691351e-01
-6.18805766e-01 -9.25880745e-02 -2.52592146e-01 7.60562062e-01
3.70385826e-01 6.65057182e-01 2.20078439e-01 -2.84386158e-01
1.17993332e-01 7.81211853e-02 6.21409893e-01 -4.51516002e-01
-2.83130556e-01 6.11076355e-02 -1.38393223e-01 -9.43875194e-01
-7.42893279e-01 -2.35251829e-01 -1.28435910e+00 -2.50015765e-01
-3.25640827e-01 1.44331887e-01 6.03012204e-01 1.15568733e+00
5.35404325e-01 5.26627243e-01 7.57087052e-01 -8.94900799e-01
-5.83159387e-01 -7.81928539e-01 -4.93932784e-01 2.78201252e-01
2.19506815e-01 -7.46114314e-01 -1.50586933e-01 -5.24354726e-02] | [13.453548431396484, 0.7058273553848267] |
4c32a54f-c373-47d2-884c-cb953a90b463 | dialogue-act-classification-in-domain | null | null | https://aclanthology.org/C16-1189 | https://aclanthology.org/C16-1189.pdf | Dialogue Act Classification in Domain-Independent Conversations Using a Deep Recurrent Neural Network | In this study, we applied a deep LSTM structure to classify dialogue acts (DAs) in open-domain conversations. We found that the word embeddings parameters, dropout regularization, decay rate and number of layers are the parameters that have the largest effect on the final system accuracy. Using the findings of these experiments, we trained a deep LSTM network that outperforms the state-of-the-art on the Switchboard corpus by 3.11{\%}, and MRDA by 2.2{\%}. | ['Nishitha la', 'Rodney Nielsen', 'Hamed Khanpour', 'Guntak'] | 2016-12-01 | dialogue-act-classification-in-domain-1 | https://aclanthology.org/C16-1189 | https://aclanthology.org/C16-1189.pdf | coling-2016-12 | ['dialogue-act-classification', 'dialogue-interpretation'] | ['natural-language-processing', 'natural-language-processing'] | [-2.43741080e-01 4.23996329e-01 -7.39070699e-02 -4.93069917e-01
-3.54726851e-01 -4.20106560e-01 5.71254432e-01 -1.24506213e-01
-8.05189252e-01 1.03338909e+00 5.23451626e-01 -6.71416461e-01
3.01826090e-01 -5.86478233e-01 -8.83495286e-02 -4.02681023e-01
-1.29736960e-01 5.66576719e-01 5.01693822e-02 -5.50237656e-01
2.59694040e-01 1.75616235e-01 -7.99392402e-01 4.20160383e-01
8.47410798e-01 1.06745493e+00 -2.01883867e-01 8.22831929e-01
-4.20969397e-01 1.24123371e+00 -1.08798516e+00 -4.10042852e-01
-1.83268487e-01 -3.97121310e-01 -1.29075336e+00 -6.35117013e-03
2.30445221e-01 -8.82938802e-01 -9.28119600e-01 3.88050556e-01
7.30408728e-01 7.91859627e-01 6.84148729e-01 -5.52371025e-01
-8.01466882e-01 7.65803993e-01 2.31602769e-02 5.77224195e-01
3.00397217e-01 5.27518950e-02 1.04241419e+00 -7.47134328e-01
5.11590242e-01 1.44399953e+00 5.72754502e-01 8.93374145e-01
-1.10039580e+00 -4.83389854e-01 4.01633866e-02 -2.68036187e-01
-7.10245311e-01 -6.16787493e-01 5.05441070e-01 -2.09500819e-01
1.68445742e+00 -4.25008163e-02 1.50715366e-01 1.56572998e+00
2.05028385e-01 7.71654487e-01 8.70862782e-01 -6.36970997e-01
9.14298147e-02 3.17929924e-01 7.24362195e-01 8.54376793e-01
-1.54114410e-01 -3.92218113e-01 -2.50609696e-01 -3.45623374e-01
6.65523767e-01 -3.53324920e-01 -1.15898959e-01 5.23436904e-01
-5.56217134e-01 1.29222405e+00 1.33530080e-01 5.93067408e-01
-1.33345753e-01 1.64804254e-02 6.77761018e-01 6.92958415e-01
9.82802033e-01 8.58960032e-01 -7.42207527e-01 -8.54104996e-01
-3.02068949e-01 -3.41294333e-02 1.18357873e+00 3.45123500e-01
2.52939880e-01 4.33791339e-01 -2.88843006e-01 1.54622960e+00
5.56889512e-02 3.87895517e-02 6.96758091e-01 -9.98761773e-01
6.67369664e-01 5.71473837e-01 1.72382027e-01 -5.54496646e-01
-6.02768719e-01 -1.78070918e-01 -4.22815979e-01 -2.59482265e-01
7.03920364e-01 -1.17743123e+00 -8.69991720e-01 1.61122227e+00
-2.52120644e-01 -4.14553374e-01 2.70727992e-01 6.15909457e-01
1.11110044e+00 7.49187231e-01 -2.08188058e-03 5.75273111e-02
8.90314519e-01 -9.95223999e-01 -1.18894684e+00 -3.66212875e-01
1.09377789e+00 -7.00093925e-01 1.19298983e+00 3.96381140e-01
-1.11560273e+00 -4.51412231e-01 -9.71742034e-01 -2.08223179e-01
-2.69553423e-01 1.84403718e-01 5.81197739e-01 6.03440881e-01
-1.04199886e+00 5.66313088e-01 -7.37018049e-01 -5.48406899e-01
3.40414867e-02 3.29800755e-01 -5.42052761e-02 2.64668971e-01
-1.49146450e+00 1.50411046e+00 2.28114232e-01 1.06836982e-01
-7.21542358e-01 -2.98094898e-01 -6.83161914e-01 4.32018265e-02
-1.01792730e-01 2.32264530e-02 1.66936970e+00 -5.56265652e-01
-2.05625176e+00 6.89687133e-01 -9.00648609e-02 -6.37391269e-01
2.08338588e-01 -2.97102481e-01 -3.23936671e-01 -4.49222773e-02
-5.07915199e-01 4.80315775e-01 1.71660662e-01 -9.12188530e-01
-4.37753558e-01 -2.92049527e-01 4.16067094e-01 2.87116915e-01
-7.29433119e-01 1.71706676e-01 -2.73090247e-02 -1.27969921e-01
-2.83160239e-01 -8.87170970e-01 -1.55302107e-01 -6.23805523e-01
-1.94231182e-01 -9.08978462e-01 7.61770070e-01 -1.07123196e+00
1.29324627e+00 -2.15387821e+00 3.06614041e-01 -2.15984508e-01
2.17275426e-01 6.87223494e-01 -4.36690561e-02 6.08156443e-01
4.33905631e-01 1.97172418e-01 3.52204531e-01 -5.37339747e-01
6.13994077e-02 4.49096620e-01 -1.62688881e-01 1.67708933e-01
5.28624803e-02 5.45373857e-01 -6.10869944e-01 -1.47282586e-01
2.94161469e-01 4.47869569e-01 -2.83809394e-01 6.40475094e-01
-9.22678635e-02 1.85855836e-01 -2.76497096e-01 3.47706117e-02
3.76117676e-01 -2.21877359e-02 3.69871408e-01 2.51475692e-01
-1.45372331e-01 9.67089474e-01 -3.79584789e-01 1.56738639e+00
-8.29061329e-01 1.17697930e+00 1.29745170e-01 -5.47167718e-01
1.43023562e+00 6.79349720e-01 1.14394993e-01 -5.36283135e-01
4.47798967e-01 2.19367906e-01 5.37046373e-01 -5.58208704e-01
7.20628381e-01 -4.27255668e-02 -2.34551243e-02 6.06717527e-01
4.05156642e-01 1.61854401e-01 2.24556234e-02 1.03209510e-01
1.26318300e+00 -3.99176985e-01 -2.13746950e-01 -2.62762249e-01
3.68488163e-01 -2.25064293e-01 5.04144132e-01 6.14536881e-01
-4.55861896e-01 3.07072073e-01 9.52806830e-01 -7.30240703e-01
-9.73645926e-01 -7.97329724e-01 -1.30518034e-01 1.58735919e+00
-5.83325803e-01 -1.39954880e-01 -9.63672400e-01 -6.40346944e-01
-2.56430924e-01 1.15795517e+00 -5.54576457e-01 -4.45617214e-02
-7.35564351e-01 -6.11852348e-01 1.04735434e+00 4.99514788e-01
6.26468539e-01 -1.31885481e+00 -3.77015024e-01 3.92805576e-01
-3.03878933e-01 -1.32850432e+00 -3.40102464e-01 4.16604072e-01
-9.14851725e-01 -7.12467372e-01 -6.15527153e-01 -9.28456724e-01
4.22296792e-01 -5.78792617e-02 1.11026299e+00 1.86344758e-02
1.55077562e-01 -1.27456322e-01 -4.98473197e-01 -2.43634418e-01
-4.88867134e-01 5.90585172e-01 2.54033506e-02 -4.44052190e-01
6.26919568e-01 -3.82729739e-01 -1.14981741e-01 1.18159071e-01
-5.26245534e-01 -4.41059142e-01 1.40897363e-01 1.08025730e+00
-5.52549183e-01 -2.34506682e-01 6.03112340e-01 -1.12306154e+00
1.48860610e+00 -2.30376288e-01 -7.51911774e-02 4.34818454e-02
-4.21573758e-01 1.30748570e-01 6.13603234e-01 -4.62066859e-01
-1.31661832e+00 -4.88476038e-01 -4.64027107e-01 -3.88820097e-02
-2.64298558e-01 4.21082765e-01 3.71755093e-01 2.10247934e-01
7.26720154e-01 -1.51011989e-01 9.19538829e-03 -7.11280882e-01
2.03907549e-01 1.22227836e+00 -2.44926997e-02 -3.01709384e-01
-1.73240170e-01 -1.23759210e-01 -8.07282090e-01 -1.16388869e+00
-9.27009106e-01 -2.76480038e-02 -5.96767187e-01 -4.66291159e-02
1.16767514e+00 -7.27436244e-01 -6.80383980e-01 5.25548160e-01
-1.51042366e+00 -7.81985402e-01 1.21229187e-01 4.34061646e-01
-1.28056556e-01 1.38274252e-01 -1.42589104e+00 -1.01150489e+00
-5.24123490e-01 -1.04405868e+00 2.26420864e-01 1.33394301e-01
-7.15631008e-01 -1.36385965e+00 2.44209528e-01 6.69242144e-01
7.44468212e-01 -8.22732747e-02 9.02912378e-01 -1.10494947e+00
3.68667424e-01 4.12849057e-03 -7.74271265e-02 7.14578807e-01
3.08712155e-01 -2.42851805e-02 -1.23521471e+00 -2.19326124e-01
1.47492826e-01 -6.99694455e-01 8.03732634e-01 4.61748987e-01
8.04579675e-01 -3.32552522e-01 2.61701923e-02 1.45655617e-01
6.18696570e-01 5.24637043e-01 4.65712637e-01 1.81651101e-01
5.03001153e-01 6.75423384e-01 9.24644321e-02 5.34361482e-01
3.73445302e-01 3.93163830e-01 5.38687669e-02 -1.76060528e-01
1.02568261e-01 -5.89629896e-02 5.76173186e-01 1.22695148e+00
7.16716349e-02 -4.75435287e-01 -1.01839316e+00 4.33958113e-01
-1.68123710e+00 -6.55585587e-01 4.45127785e-02 1.70664132e+00
8.26234221e-01 5.31246245e-01 2.50568718e-01 -8.17792341e-02
6.45808518e-01 6.56805754e-01 -1.64329275e-01 -1.54650533e+00
1.37590200e-01 2.96895325e-01 2.93431997e-01 8.17719936e-01
-8.99987400e-01 1.17068315e+00 7.17163038e+00 3.61120611e-01
-9.54272687e-01 2.20234290e-01 8.02554548e-01 3.36155742e-02
-9.37709659e-02 -3.62642020e-01 -8.54750633e-01 4.83983666e-01
1.61032414e+00 1.81286290e-01 3.90268922e-01 6.01938426e-01
4.00704384e-01 -6.63160235e-02 -9.36634898e-01 4.01547134e-01
8.74625966e-02 -9.67887282e-01 -2.45968610e-01 1.81412980e-01
6.88764691e-01 3.05236280e-01 -6.65486157e-02 6.70057356e-01
9.62935448e-01 -1.19273651e+00 -1.08634673e-01 7.52464905e-02
3.89914542e-01 -7.12423146e-01 1.28499556e+00 3.12184244e-01
-3.58955145e-01 -8.31881911e-02 -4.71019715e-01 -4.96744186e-01
1.88313380e-01 3.91151637e-01 -1.34354353e+00 -7.19492435e-02
4.46144402e-01 3.46518815e-01 -1.03450559e-01 2.14514300e-01
-3.02769303e-01 9.55248177e-01 -1.20110594e-01 -4.22820210e-01
6.97896004e-01 -2.16142863e-01 5.55543266e-02 1.49661469e+00
-2.26951033e-01 3.58954221e-01 8.49860385e-02 4.73693818e-01
-4.29628313e-01 -3.64328593e-01 -6.78029835e-01 -2.82658994e-01
6.04070365e-01 9.25183177e-01 -2.03852758e-01 -2.12981418e-01
-3.61857861e-01 8.50009859e-01 7.06313491e-01 4.82546151e-01
-6.52204752e-01 -6.31457031e-01 9.37675476e-01 -4.09680724e-01
4.06306714e-01 -5.28180838e-01 -2.59973735e-01 -8.37710917e-01
-2.78665483e-01 -9.46618795e-01 2.76628345e-01 -1.99697047e-01
-1.37675190e+00 1.16176724e+00 -3.63640964e-01 -5.76482475e-01
-5.63172340e-01 -8.11610878e-01 -8.45381677e-01 8.72646093e-01
-1.12445784e+00 -6.36778295e-01 7.03069046e-02 1.34827375e-01
8.78001988e-01 -4.45259571e-01 1.20034504e+00 2.50135452e-01
-1.14711082e+00 8.04046214e-01 2.90495843e-01 6.63354874e-01
5.04945099e-01 -1.20239890e+00 6.27412856e-01 2.94710994e-01
-2.52740294e-01 6.09451354e-01 6.55933976e-01 -2.24803925e-01
-8.27720284e-01 -5.28750598e-01 1.11941469e+00 -3.97048205e-01
8.15634727e-01 -5.81675172e-01 -8.43848169e-01 9.56181347e-01
8.88595223e-01 -3.76529753e-01 8.21989596e-01 6.43066347e-01
-9.78037417e-02 1.73457772e-01 -1.21903419e+00 5.08822799e-01
7.26785958e-01 -7.65065014e-01 -8.25625122e-01 1.26235485e-01
8.82054269e-01 -6.24840438e-01 -1.37654328e+00 -5.25705926e-02
4.17357177e-01 -8.08549523e-01 5.63091755e-01 -1.07444894e+00
4.41997707e-01 7.79379725e-01 -5.67918643e-02 -1.60524011e+00
-3.62816215e-01 -6.63496196e-01 -2.32015371e-01 1.22688341e+00
9.35239196e-01 -8.35136473e-01 6.69624209e-01 9.89311635e-01
-3.76076728e-01 -7.67264485e-01 -9.95396614e-01 -3.99333298e-01
5.93253672e-01 8.64975248e-03 1.54831678e-01 8.90363276e-01
4.04678464e-01 6.88198209e-01 -4.84029800e-01 -6.05115473e-01
-1.21950701e-01 -4.91675913e-01 5.50647020e-01 -1.09757197e+00
-8.27637985e-02 -3.05428475e-01 -8.53301063e-02 -1.29362547e+00
5.75593889e-01 -3.02597374e-01 6.81580752e-02 -1.66801500e+00
-2.62703478e-01 -2.79294252e-01 -4.18932468e-01 6.53322041e-01
8.66794493e-03 -1.99560747e-01 1.88522100e-01 -4.22590524e-02
-4.59896684e-01 8.05567801e-01 9.46181417e-01 -1.01233423e-01
-4.32653904e-01 -4.77571748e-02 -5.44420123e-01 7.05447614e-01
1.30463910e+00 -2.38242373e-01 -1.64608791e-01 -1.01328504e+00
-5.30355930e-01 1.12651721e-01 -5.51703274e-01 -6.62378669e-01
1.90190561e-02 4.80085015e-02 1.48745865e-01 -2.56759614e-01
7.73805618e-01 -1.08667269e-01 -8.77212048e-01 3.88808131e-01
-1.07337356e+00 2.47430801e-01 3.85690808e-01 -4.63380888e-02
-1.33165151e-01 -4.32416677e-01 5.93496561e-01 -2.18218297e-01
-2.55620092e-01 -3.13611150e-01 -1.02107155e+00 4.34121817e-01
4.08837646e-01 1.68759719e-01 -5.86390793e-01 -7.28308499e-01
-5.17614901e-01 2.90049285e-01 -1.42288819e-01 5.89520633e-01
3.86474997e-01 -9.85921323e-01 -7.36043334e-01 8.54645297e-02
-5.25814354e-01 -3.36329430e-01 -1.61052376e-01 4.18031186e-01
-5.02748311e-01 8.14894319e-01 -3.11395437e-01 -6.95159435e-02
-1.20168352e+00 -4.74909246e-01 5.54296613e-01 -4.54070330e-01
-3.15483093e-01 1.08821368e+00 -3.62643898e-01 -8.91580820e-01
7.32391059e-01 -4.27302927e-01 -4.55038488e-01 7.47525692e-02
3.24266642e-01 6.59887791e-01 5.65789938e-02 -2.81803250e-01
-1.06478684e-01 -2.18184531e-01 -5.41195571e-01 -3.50663930e-01
1.42512465e+00 1.03020817e-01 4.86641936e-03 8.85231793e-01
1.22048032e+00 -3.37973058e-01 -1.17256331e+00 -2.67190814e-01
-1.16771450e-02 -2.72564352e-01 3.26041013e-01 -8.96631598e-01
-8.06662202e-01 1.00966060e+00 4.66443419e-01 6.59831822e-01
4.38744277e-01 -2.65900999e-01 1.14349651e+00 6.77014053e-01
3.41053540e-03 -1.49785733e+00 2.45510757e-01 1.21480691e+00
6.60348713e-01 -1.19589186e+00 -1.65860370e-01 8.27103183e-02
-9.66460764e-01 1.16368902e+00 1.13069987e+00 -2.74913788e-01
4.45347071e-01 2.72592574e-01 3.92550141e-01 -9.50169340e-02
-1.13584924e+00 2.40711331e-01 -6.32312521e-02 3.14839751e-01
1.16202080e+00 -8.31663981e-03 -4.64900136e-01 3.08598042e-01
-2.02422813e-01 -1.17277272e-01 6.97271585e-01 7.95848191e-01
-6.27415478e-01 -1.20163310e+00 -1.95757989e-02 5.18085897e-01
-5.52805722e-01 -2.35514641e-02 -1.03595173e+00 7.67242670e-01
-3.39104027e-01 1.47079349e+00 3.99976313e-01 -7.47014940e-01
4.25787091e-01 3.44300300e-01 2.03020573e-01 -7.42320836e-01
-1.16249001e+00 -2.68469274e-01 1.06078219e+00 -1.51086569e-01
-6.05840422e-02 -2.71963239e-01 -1.35819590e+00 -7.51464427e-01
-4.92594272e-01 3.65199864e-01 4.61041957e-01 1.06872141e+00
2.88842022e-01 6.28135443e-01 7.45416820e-01 -4.31137085e-01
-9.33298826e-01 -1.97108877e+00 -3.73177081e-01 1.93052918e-01
2.76590139e-01 -3.96182150e-01 -6.10248685e-01 -6.04260743e-01] | [12.848832130432129, 7.848941326141357] |
7500af4c-68e6-499b-b527-b5d3160d2397 | es-net-an-efficient-stereo-matching-network | 2103.03922 | null | https://arxiv.org/abs/2103.03922v1 | https://arxiv.org/pdf/2103.03922v1.pdf | ES-Net: An Efficient Stereo Matching Network | Dense stereo matching with deep neural networks is of great interest to the research community. Existing stereo matching networks typically use slow and computationally expensive 3D convolutions to improve the performance, which is not friendly to real-world applications such as autonomous driving. In this paper, we propose the Efficient Stereo Network (ESNet), which achieves high performance and efficient inference at the same time. ESNet relies only on 2D convolution and computes multi-scale cost volume efficiently using a warping-based method to improve the performance in regions with fine-details. In addition, we address the matching ambiguity issue in the occluded region by proposing ESNet-M, a variant of ESNet that additionally estimates an occlusion mask without supervision. We further improve the network performance by proposing a new training scheme that includes dataset scheduling and unsupervised pre-training. Compared with other low-cost dense stereo depth estimation methods, our proposed approach achieves state-of-the-art performance on the Scene Flow [1], DrivingStereo [2], and KITTI-2015 dataset [3]. Our code will be made available. | ['Panqu Wang', 'Theodore B. Norris', 'Zhengyu Huang'] | 2021-03-05 | null | null | null | null | ['stereo-depth-estimation', 'unsupervised-pre-training'] | ['computer-vision', 'methodology'] | [ 5.47173843e-02 -1.57761395e-01 3.39978002e-02 -6.00445271e-01
-1.85221851e-01 -1.87747538e-01 4.98972088e-01 -3.26180965e-01
-6.84159994e-01 6.33166373e-01 1.86221376e-01 -2.87146121e-01
1.84088200e-01 -1.16493154e+00 -7.68185973e-01 -3.91857445e-01
1.85978517e-01 3.56390506e-01 7.80975997e-01 -3.25468719e-01
3.44290644e-01 4.96159017e-01 -1.83001411e+00 8.70241150e-02
1.07970226e+00 1.22580957e+00 4.41996396e-01 3.25449288e-01
-1.44328654e-01 8.70294750e-01 -1.43365543e-02 -3.22755873e-01
6.58166230e-01 1.48199340e-02 -6.89146996e-01 -1.10639088e-01
1.05067813e+00 -8.64863157e-01 -9.74690080e-01 1.00921273e+00
5.67342341e-01 3.04288536e-01 1.62423447e-01 -1.16199028e+00
2.45510012e-01 3.41264866e-02 -5.58736742e-01 2.82608628e-01
4.30143215e-02 3.63317728e-01 6.34742975e-01 -8.07062447e-01
5.57903349e-01 1.37255633e+00 7.86213815e-01 3.53923529e-01
-9.28789496e-01 -1.06355762e+00 2.92480774e-02 5.06706417e-01
-1.22481334e+00 -5.32716453e-01 8.99352431e-01 -4.14361089e-01
9.76563096e-01 -2.13938385e-01 7.45837808e-01 7.10866690e-01
-2.52800751e-02 7.44481564e-01 1.00996757e+00 7.44547769e-02
1.39356256e-01 -4.49476689e-01 -1.07576087e-01 8.42084885e-01
2.37440318e-01 6.53425574e-01 -5.68902969e-01 1.90949768e-01
1.13245499e+00 1.66154832e-01 -2.98401594e-01 -5.16480267e-01
-1.16468239e+00 8.63036692e-01 8.20083320e-01 -1.14144035e-01
-2.60557532e-01 3.26916337e-01 5.59344471e-01 4.55701612e-02
5.45498073e-01 -3.60861793e-02 -2.96128273e-01 -2.38092616e-01
-1.04739118e+00 4.28530008e-01 5.33489168e-01 9.66508806e-01
1.22416258e+00 1.60765499e-02 -3.77620794e-02 8.23634684e-01
1.35274768e-01 3.64782423e-01 2.20619500e-01 -1.38593507e+00
8.20289731e-01 5.08852124e-01 -6.71471134e-02 -1.07747018e+00
-4.81496066e-01 -3.84704560e-01 -1.06397033e+00 4.01331395e-01
3.49132210e-01 1.65636521e-02 -8.24433804e-01 1.53082752e+00
4.27174747e-01 7.32911587e-01 -2.64319647e-02 1.20608056e+00
1.10555720e+00 4.93509620e-01 -2.31659621e-01 3.16332459e-01
9.38098848e-01 -1.31260419e+00 -5.05466580e-01 -5.64174592e-01
4.59211141e-01 -7.37925589e-01 7.55618513e-01 5.98292947e-02
-9.73308921e-01 -7.75283873e-01 -1.10333288e+00 -5.82748353e-01
-1.32273301e-01 -2.25228921e-01 1.10360205e+00 2.88324237e-01
-1.00293195e+00 7.84731865e-01 -8.92045975e-01 -3.49584222e-01
6.47518694e-01 4.28951353e-01 -4.15320635e-01 -3.59315872e-01
-1.18656337e+00 6.72573566e-01 4.52446967e-01 3.18787009e-01
-7.17183292e-01 -7.43551493e-01 -1.36338949e+00 -3.37067693e-02
3.63410115e-01 -9.94634509e-01 1.19488823e+00 -5.77835023e-01
-1.58783805e+00 9.00134265e-01 -4.69611734e-01 -5.02671480e-01
7.85401881e-01 -3.41676027e-01 1.28892455e-02 1.54513806e-01
2.44497821e-01 1.11300302e+00 4.91285264e-01 -8.41628611e-01
-9.23925102e-01 -4.03562248e-01 3.41466874e-01 1.91805422e-01
-3.53815709e-03 -3.72495562e-01 -6.71337187e-01 -6.10723197e-01
3.97066236e-01 -7.34271646e-01 -3.34218383e-01 3.89847994e-01
-2.10679621e-01 -2.36221608e-02 9.51276302e-01 -5.24003088e-01
8.65416586e-01 -2.10201311e+00 -7.95137584e-02 -1.87700931e-02
4.50703740e-01 3.14254671e-01 -1.72777604e-02 -4.95247468e-02
2.21586734e-01 -4.90823090e-01 -4.74686444e-01 -6.41831219e-01
-1.18771188e-01 2.83951014e-01 -1.72275886e-01 6.54037297e-01
1.27245367e-01 7.44113386e-01 -9.23605084e-01 -7.05896258e-01
8.05052102e-01 5.02422631e-01 -9.06375051e-01 3.30170065e-01
5.22779413e-02 5.91528237e-01 -2.57312000e-01 5.77051640e-01
1.22698212e+00 -1.59261510e-01 -3.17138821e-01 -4.67187673e-01
-4.78272438e-01 3.72341245e-01 -1.28538108e+00 2.15406084e+00
-6.01769745e-01 9.33871508e-01 1.49151608e-01 -9.38899755e-01
9.30484295e-01 -1.56562448e-01 4.12133753e-01 -9.77856874e-01
2.04283044e-01 3.71791363e-01 -1.48261994e-01 -4.26779330e-01
7.40765512e-01 -2.29739118e-02 2.06087515e-01 -6.36364520e-02
-8.98105875e-02 -2.76347339e-01 2.41085008e-01 -1.47032337e-02
1.03722548e+00 2.85564721e-01 7.14951828e-02 -3.28721374e-01
7.23302960e-01 -1.62454247e-02 9.78984773e-01 5.69998860e-01
-3.89556080e-01 7.75857449e-01 4.81224023e-02 -7.50297010e-01
-1.03490877e+00 -8.73398185e-01 -3.06751370e-01 3.95685077e-01
7.82463372e-01 -1.87630147e-01 -5.31905353e-01 -3.66508454e-01
2.51563579e-01 2.16135874e-01 -4.35835272e-01 5.44494539e-02
-8.80803764e-01 -2.98028350e-01 3.75557244e-01 7.07185626e-01
1.36134875e+00 -9.76637900e-01 -7.46017039e-01 2.86904901e-01
-5.66483259e-01 -1.52995026e+00 -5.62484562e-01 -2.97685545e-02
-1.16553402e+00 -1.11202943e+00 -6.52636826e-01 -7.22639143e-01
4.72554058e-01 5.89008093e-01 1.09952247e+00 1.15614727e-01
-1.20553233e-01 -2.20976636e-01 4.11401354e-02 -9.81099382e-02
1.27739817e-01 2.55181998e-01 -2.70057112e-01 -8.15966129e-02
4.47701782e-01 -7.36605048e-01 -9.19951916e-01 4.79684651e-01
-8.06827962e-01 4.97099936e-01 3.23075086e-01 7.94799924e-01
4.73583430e-01 -1.40722051e-01 -5.44628575e-02 -6.55486107e-01
-6.37749135e-02 -1.53487578e-01 -1.08003950e+00 -4.01161969e-01
-2.94514090e-01 1.58022076e-01 5.64782798e-01 -1.39937298e-02
-1.15303981e+00 2.56467044e-01 -4.37285960e-01 -7.08846569e-01
-1.33019313e-01 1.78038217e-02 -3.44852023e-02 -5.23782611e-01
3.75326484e-01 9.19657499e-02 -8.00085589e-02 -4.19710010e-01
-4.43329662e-02 4.73281622e-01 8.81534398e-01 -3.62712890e-01
9.06196475e-01 1.01014864e+00 2.94094980e-01 -5.14808118e-01
-9.66863036e-01 -6.79323852e-01 -6.72030568e-01 -2.27071583e-01
8.04670870e-01 -1.32375932e+00 -8.69791925e-01 8.81012917e-01
-1.36650240e+00 -4.33317065e-01 1.00584313e-01 7.99803615e-01
-6.26259208e-01 4.81219560e-01 -6.89779878e-01 -3.40853810e-01
-2.23749295e-01 -1.36633825e+00 1.30615294e+00 3.00069034e-01
1.34936273e-01 -9.32345152e-01 1.44363297e-02 5.57779074e-01
6.08143032e-01 3.50827426e-01 2.32528001e-01 1.52906001e-01
-1.13063681e+00 2.72358209e-01 -7.78703332e-01 1.88872144e-01
-1.25925541e-01 -3.11554223e-01 -1.16138268e+00 -2.61348516e-01
-2.46790528e-01 -3.00208479e-01 1.24225152e+00 6.20398223e-01
1.45644510e+00 2.94542581e-01 -2.43811205e-01 1.51290822e+00
1.54122174e+00 -1.72755085e-02 8.68874788e-01 7.11044312e-01
9.67462897e-01 6.78219557e-01 6.70702696e-01 4.59333330e-01
7.76383877e-01 7.41829455e-01 6.40169978e-01 -4.30184603e-01
-2.28815988e-01 -2.94111341e-01 1.61030833e-02 5.62112451e-01
-1.79486468e-01 1.91022173e-01 -7.49066830e-01 5.05276501e-01
-2.06427455e+00 -1.00621724e+00 -2.97632992e-01 2.05119038e+00
5.56966543e-01 3.04823816e-01 -1.62015751e-01 9.29445997e-02
7.32658625e-01 4.94052052e-01 -7.08513379e-01 -6.60995916e-02
-1.78144693e-01 3.96575660e-01 9.39487159e-01 7.84995437e-01
-1.27610600e+00 1.17817974e+00 5.50476551e+00 8.06388736e-01
-1.26255131e+00 1.11295111e-01 5.09027243e-01 1.50908362e-02
-1.88604578e-01 1.58120796e-01 -9.01812732e-01 5.50501227e-01
2.59536415e-01 5.80780804e-02 3.17225397e-01 8.28692555e-01
4.25358742e-01 -3.35808605e-01 -9.98201728e-01 1.44506967e+00
-1.24509260e-01 -1.61835003e+00 -3.07193607e-01 1.46351591e-01
7.92519867e-01 4.55814153e-01 -3.97734404e-01 1.55205861e-01
1.89184949e-01 -7.11432457e-01 6.98142231e-01 2.85153806e-01
8.04442942e-01 -7.39384770e-01 9.33520019e-01 1.96828917e-01
-1.52332032e+00 3.93436514e-02 -4.55540568e-01 -3.90150696e-01
4.37870324e-01 8.26697111e-01 -2.06614763e-01 4.20931250e-01
8.81104112e-01 1.21749485e+00 -3.94515067e-01 1.34184372e+00
-9.33093503e-02 1.51182890e-01 -4.46908295e-01 3.68055254e-01
3.95015866e-01 -2.49966606e-01 3.57291877e-01 1.07219005e+00
3.06054205e-01 2.10117102e-02 2.05050528e-01 8.06580722e-01
-9.00236145e-02 -2.58987725e-01 -7.84504652e-01 6.67623162e-01
4.61030275e-01 1.08332002e+00 -4.84125108e-01 -5.13422728e-01
-5.66874981e-01 1.02402461e+00 4.62723315e-01 1.15252160e-01
-7.40054548e-01 -6.18715346e-01 9.62874889e-01 1.94071293e-01
3.74051303e-01 -3.85956347e-01 -3.89291048e-01 -1.34321713e+00
1.98821232e-01 -3.58611912e-01 1.77726984e-01 -6.79594398e-01
-1.05505013e+00 4.39004630e-01 -1.30917847e-01 -1.47194743e+00
-1.73392773e-01 -4.87026900e-01 -5.56735694e-01 9.22043025e-01
-2.26399279e+00 -7.64249802e-01 -1.12615454e+00 8.33784640e-01
5.67030072e-01 6.57565072e-02 2.23894313e-01 7.73783624e-01
-4.65000093e-01 3.56030673e-01 -2.62199581e-01 1.49086744e-01
7.48635054e-01 -8.85102630e-01 7.99402416e-01 9.26791370e-01
-5.37442207e-01 2.11480066e-01 4.41532731e-01 -4.22186583e-01
-1.13498127e+00 -1.34825563e+00 1.03016853e+00 8.50907564e-02
3.52990180e-01 -3.00679386e-01 -8.06945205e-01 4.70043123e-01
6.97339028e-02 3.93073648e-01 -1.94416083e-02 -2.08016962e-01
-1.83382124e-01 -4.94321585e-01 -1.22316492e+00 5.18944085e-01
1.69653404e+00 -4.93583173e-01 -3.66812438e-01 1.24671839e-01
7.43170142e-01 -1.00884640e+00 -5.35216451e-01 7.49054492e-01
4.96675521e-01 -1.65074611e+00 9.93184030e-01 1.96051914e-02
6.63719773e-01 -4.53250080e-01 -1.27791166e-01 -9.91101682e-01
-2.65743405e-01 -5.17991006e-01 -4.03561480e-02 7.29213715e-01
-1.70721024e-01 -8.69052589e-01 1.15817142e+00 4.04666364e-01
-4.99665946e-01 -4.68231738e-01 -1.02950859e+00 -7.33473361e-01
-3.82042378e-01 -5.40992916e-01 7.04278409e-01 7.82147646e-01
-5.22457600e-01 4.96528372e-02 -3.74106228e-01 1.44388676e-01
1.02184832e+00 2.61375248e-01 1.15398514e+00 -1.19230473e+00
-1.50474891e-01 -4.28821653e-01 -8.04407716e-01 -1.59319270e+00
3.95642996e-01 -6.53098941e-01 1.40928254e-01 -1.38755965e+00
-1.39946193e-02 -4.68013585e-01 1.27542526e-01 2.55085438e-01
-9.71052349e-02 3.30929846e-01 6.98740780e-02 1.26595616e-01
-4.10613060e-01 7.61043012e-01 1.52372015e+00 -1.51648372e-01
-5.99248074e-02 -1.93448111e-01 -1.80005625e-01 7.15900242e-01
6.94457591e-01 -2.23347038e-01 -3.83527875e-01 -7.71715224e-01
-1.75731882e-01 6.07580356e-02 6.71625972e-01 -1.39906490e+00
3.62263829e-01 -1.04959391e-01 1.90039381e-01 -9.94088948e-01
4.75819230e-01 -8.55885863e-01 5.33808162e-03 5.72422683e-01
6.32924885e-02 1.44240811e-01 2.92954445e-01 2.84088403e-01
-6.80488884e-01 2.36562174e-02 9.93475199e-01 -1.76978230e-01
-1.21675587e+00 9.28351521e-01 1.46146849e-01 1.63270056e-01
8.66361856e-01 -4.34020668e-01 -3.95847857e-01 -3.40098739e-01
-4.15772684e-02 4.55510587e-01 6.45618975e-01 4.01850820e-01
6.92040741e-01 -1.43388283e+00 -6.15126550e-01 4.04505759e-01
1.47289142e-01 7.06985474e-01 3.96259665e-01 6.81825757e-01
-9.98820007e-01 5.55657029e-01 -4.50423777e-01 -9.00484204e-01
-9.24004734e-01 1.23595640e-01 4.64021266e-01 -1.18002012e-01
-8.86853814e-01 7.24920809e-01 5.26704192e-01 -5.19815564e-01
3.28777373e-01 -3.45699102e-01 5.43617597e-03 -4.16243076e-01
4.30362523e-01 4.34739321e-01 7.26834983e-02 -5.93804300e-01
-2.93992460e-01 8.70399296e-01 1.28627881e-01 1.18781753e-01
1.21736872e+00 -1.37681976e-01 -8.51669684e-02 4.93096597e-02
1.46723568e+00 -4.26602304e-01 -1.76817763e+00 -4.54768211e-01
-4.62882221e-01 -9.86977696e-01 5.34130752e-01 4.06656461e-03
-1.56303596e+00 1.13525581e+00 6.76482201e-01 -4.99743283e-01
1.11114049e+00 -3.06725860e-01 1.34801459e+00 1.87171310e-01
5.16136706e-01 -1.05243504e+00 -2.21811265e-01 7.70126164e-01
5.96206605e-01 -1.63279378e+00 -1.16048038e-01 -6.48734689e-01
-9.70056355e-02 1.20874214e+00 1.07740104e+00 -2.22667456e-01
6.98471963e-01 2.68903196e-01 -8.51302221e-03 -1.42408758e-01
-4.34326231e-01 -4.53813404e-01 7.49842152e-02 4.30640250e-01
1.49151877e-01 -2.06341624e-01 -1.57298580e-01 -2.00683385e-01
-2.27614701e-01 3.41296226e-01 2.19347015e-01 8.61586630e-01
-2.97179878e-01 -9.29957867e-01 -1.61830783e-01 2.87942618e-01
3.17520425e-02 -2.41847172e-01 1.43697634e-01 7.21075356e-01
3.54020864e-01 7.12871909e-01 5.44650912e-01 -4.34437096e-01
4.86108541e-01 -5.36002755e-01 4.58822191e-01 -2.98620343e-01
-2.67568797e-01 -2.45881185e-01 7.78209791e-02 -1.16318274e+00
-7.71522760e-01 -6.02745950e-01 -1.19773424e+00 -8.76606703e-01
-1.48755717e-04 -3.75855744e-01 5.85350335e-01 9.62372482e-01
4.30708826e-01 3.16061169e-01 6.91286027e-01 -1.22534502e+00
5.73838456e-03 -7.73355186e-01 -3.85666400e-01 4.09504265e-01
5.37927628e-01 -9.99891877e-01 -3.82578731e-01 -1.85518980e-01] | [8.79751205444336, -2.268230438232422] |
78e3007d-eb19-4967-96fc-92053821d56b | densely-connected-graph-convolutional-1 | 1908.05957 | null | https://arxiv.org/abs/1908.05957v2 | https://arxiv.org/pdf/1908.05957v2.pdf | Densely Connected Graph Convolutional Networks for Graph-to-Sequence Learning | We focus on graph-to-sequence learning, which can be framed as transducing graph structures to sequences for text generation. To capture structural information associated with graphs, we investigate the problem of encoding graphs using graph convolutional networks (GCNs). Unlike various existing approaches where shallow architectures were used for capturing local structural information only, we introduce a dense connection strategy, proposing a novel Densely Connected Graph Convolutional Networks (DCGCNs). Such a deep architecture is able to integrate both local and non-local features to learn a better structural representation of a graph. Our model outperforms the state-of-the-art neural models significantly on AMRto-text generation and syntax-based neural machine translation. | ['Zhijiang Guo', 'Zhiyang Teng', 'Yan Zhang', 'Wei Lu'] | 2019-08-16 | densely-connected-graph-convolutional | https://aclanthology.org/Q19-1019 | https://aclanthology.org/Q19-1019.pdf | tacl-2019-3 | ['graph-to-sequence'] | ['natural-language-processing'] | [ 6.92028105e-01 7.21922636e-01 -3.00482064e-01 -2.82077163e-01
-5.16012251e-01 -5.74703217e-01 9.31475878e-01 1.97433278e-01
1.01733811e-01 8.22581768e-01 4.95691031e-01 -7.85689592e-01
3.80780697e-01 -1.44563520e+00 -1.15336680e+00 -3.79605323e-01
4.07803208e-02 6.12124741e-01 -1.33735046e-01 -6.60423219e-01
7.58083398e-03 2.73849696e-01 -8.13525975e-01 6.27358258e-01
9.86688673e-01 6.42766595e-01 1.66536301e-01 8.90044391e-01
-6.41659379e-01 9.54985321e-01 -5.46478570e-01 -6.51518404e-01
4.59077209e-02 -8.79366577e-01 -1.12309825e+00 3.13927047e-02
5.35777688e-01 -1.11294769e-01 -5.27774394e-01 1.05781102e+00
4.09146070e-01 -6.97279274e-02 6.82091475e-01 -1.10002899e+00
-1.40124226e+00 1.11114204e+00 -2.81447023e-02 -8.56898278e-02
6.98794842e-01 2.70847112e-01 1.38194585e+00 -5.02898633e-01
9.16986763e-01 1.52458858e+00 5.19656479e-01 7.45982289e-01
-1.36552787e+00 -1.15870953e-01 2.56362587e-01 3.78497727e-02
-8.57771575e-01 -1.30532935e-01 9.54844236e-01 -3.46391082e-01
1.31205249e+00 8.27643946e-02 8.79473031e-01 1.52687192e+00
3.74156594e-01 7.77922332e-01 6.82295561e-01 -5.70907891e-01
-9.61090345e-03 -5.47992885e-01 -4.21565063e-02 1.23267770e+00
3.74588698e-01 4.76919487e-02 -2.76796311e-01 -2.67157182e-02
9.01512325e-01 -2.30680287e-01 -2.92063743e-01 -2.61663884e-01
-1.27500474e+00 1.23110342e+00 9.83593166e-01 3.82445604e-01
-2.38503143e-01 7.74342656e-01 3.41861904e-01 5.12028277e-01
4.93892372e-01 6.05314016e-01 -1.90634444e-01 1.09424256e-01
-4.71743613e-01 3.48873794e-01 8.47961605e-01 1.19767237e+00
9.72028375e-01 6.02040529e-01 -6.14153564e-01 3.43786895e-01
3.36693078e-01 3.28858644e-01 4.17860657e-01 -2.28181750e-01
9.03370440e-01 9.94830370e-01 -5.13455808e-01 -8.31882477e-01
-2.92375058e-01 -2.93045104e-01 -1.03063154e+00 -3.81615579e-01
1.96816310e-01 -3.95095766e-01 -1.25122809e+00 1.63631630e+00
-7.08853453e-02 1.80418566e-01 1.93572968e-01 7.42691219e-01
1.19599795e+00 8.15988421e-01 -6.71529844e-02 2.04847604e-01
1.00097609e+00 -1.09221840e+00 -6.56403184e-01 -4.48974818e-01
1.05270338e+00 -4.23039854e-01 9.52153862e-01 -2.69216090e-01
-1.10988832e+00 -4.77315724e-01 -1.00017333e+00 -2.46834174e-01
-6.91457868e-01 -1.44755363e-01 9.09117401e-01 4.29040402e-01
-1.67131305e+00 6.42082453e-01 -4.87966210e-01 -3.72403204e-01
3.96480441e-01 2.77301908e-01 -5.10999322e-01 -2.03141913e-01
-1.52599978e+00 6.48996055e-01 7.74148881e-01 2.48498872e-01
-9.37138796e-01 -4.07531291e-01 -1.33825517e+00 1.84161946e-01
2.68762082e-01 -1.28501070e+00 9.63111520e-01 -1.14093792e+00
-1.62803555e+00 6.90848351e-01 -6.18390441e-02 -7.57226348e-01
9.30872932e-02 2.72240371e-01 -7.17199147e-02 1.74010292e-01
-3.06678772e-01 6.97762311e-01 9.20388818e-01 -8.72931719e-01
9.37685445e-02 -8.33867714e-02 3.87366056e-01 1.23159155e-01
-1.38401330e-01 -6.89170957e-02 -3.76024731e-02 -9.05541480e-01
-1.89415514e-01 -9.14019406e-01 -4.33245391e-01 -5.58825135e-01
-7.41746426e-01 -2.60902911e-01 6.25433326e-01 -7.33000457e-01
1.01226246e+00 -1.34299755e+00 6.36953175e-01 5.08596264e-02
5.08256793e-01 5.76273799e-01 -7.16649711e-01 9.44919407e-01
-1.93413123e-01 3.95701110e-01 -2.65351027e-01 -2.05287874e-01
1.51679873e-01 2.49200597e-01 -2.75409818e-01 -6.05823286e-02
8.26978028e-01 1.93522894e+00 -1.15728116e+00 -2.29824260e-01
-1.34501057e-02 3.77795666e-01 -5.93508244e-01 4.82581466e-01
-8.66652071e-01 3.80676538e-01 -5.51876545e-01 3.90875757e-01
5.24305761e-01 -5.75227618e-01 5.08955836e-01 7.91922137e-02
5.56842804e-01 6.36384070e-01 -4.17603225e-01 1.99501562e+00
-3.96647960e-01 4.94098306e-01 -2.57237673e-01 -1.40289497e+00
1.21180880e+00 4.01362866e-01 -2.33715326e-02 -6.92861080e-01
7.88363218e-02 7.63492212e-02 3.90704758e-02 -6.45223558e-02
5.70224464e-01 -9.01879966e-02 -1.34286821e-01 5.10012686e-01
5.45298457e-01 -2.23128900e-01 1.87371939e-01 7.19225407e-01
1.42581189e+00 2.20725939e-01 1.72361791e-01 -1.75331190e-01
4.26587194e-01 -1.22950025e-01 -1.53892227e-02 6.11605883e-01
2.43183434e-01 6.00530744e-01 9.14580345e-01 -5.04028976e-01
-9.95477855e-01 -8.88135493e-01 7.87651539e-01 1.05112231e+00
-3.82989883e-01 -7.02801764e-01 -9.78782415e-01 -1.04738629e+00
-2.34229729e-01 5.15576541e-01 -7.15773940e-01 -5.25322795e-01
-8.45527828e-01 -5.58167815e-01 7.71920919e-01 5.14419198e-01
2.32747167e-01 -1.24645686e+00 2.13102564e-01 3.30381870e-01
-2.47468024e-01 -1.31024051e+00 -7.35356569e-01 6.93300292e-02
-1.01171589e+00 -8.35017800e-01 -6.57912374e-01 -1.07884574e+00
7.53141463e-01 -8.24727789e-02 1.58505845e+00 4.42197442e-01
5.35818376e-02 1.13776878e-01 -6.36352599e-01 -1.35154068e-01
-1.00559306e+00 7.09641993e-01 -6.66704655e-01 8.16238485e-03
1.49985298e-03 -6.08394623e-01 -3.30682278e-01 -3.82070005e-01
-9.91743922e-01 6.00709915e-01 6.49734199e-01 8.90492678e-01
2.19888374e-01 -6.71163201e-01 6.83872879e-01 -1.29766464e+00
1.11499739e+00 -4.01710451e-01 -5.24012983e-01 4.06874359e-01
-5.10472953e-01 5.48577070e-01 9.39170361e-01 -1.99945629e-01
-7.54312396e-01 1.79424987e-03 -1.91058233e-01 -3.37516278e-01
-5.76942600e-02 8.71029913e-01 -1.20813273e-01 -4.81347293e-02
6.50572181e-01 3.40780228e-01 -6.60585910e-02 -1.11853316e-01
9.65214252e-01 2.95437306e-01 1.96252838e-01 -6.25974059e-01
8.59973729e-01 -9.42149013e-02 3.52158457e-01 -4.05579120e-01
-6.02582455e-01 -1.40161347e-02 -7.97539353e-01 1.06612027e-01
1.07474649e+00 -8.35034013e-01 -2.56515980e-01 3.15400988e-01
-1.51268148e+00 -7.62021065e-01 -2.07072467e-01 2.31481660e-02
-6.73610330e-01 3.48548681e-01 -9.17931616e-01 -3.82372767e-01
-6.61007166e-01 -1.04001498e+00 1.38343859e+00 -1.92346349e-01
-1.89512921e-03 -1.67228305e+00 2.72632718e-01 1.74468860e-01
5.59602082e-01 6.90945923e-01 1.26921046e+00 -7.36709714e-01
-8.46844733e-01 -1.42821759e-01 -4.08737570e-01 1.53258070e-01
1.48960054e-01 -3.31142455e-01 -6.60799742e-01 -3.86959434e-01
-7.71423280e-01 -4.38634276e-01 9.73624289e-01 1.03621453e-01
8.81174266e-01 -7.57447541e-01 -2.64926285e-01 7.19442964e-01
1.29376626e+00 -3.01820427e-01 9.35093582e-01 -7.53153935e-02
1.49057841e+00 2.74051964e-01 -1.11760251e-01 -1.04945086e-01
5.44628263e-01 5.28949559e-01 6.06125116e-01 -4.56924975e-01
-5.88742316e-01 -8.83715689e-01 6.19819105e-01 1.08392572e+00
-1.39761433e-01 -7.38770723e-01 -8.52599978e-01 4.94014233e-01
-2.03294683e+00 -6.20212913e-01 -2.58321464e-01 1.46834302e+00
7.57599652e-01 -7.04663247e-02 -3.80620435e-02 -3.10451984e-01
8.22021484e-01 5.42626739e-01 -1.31141633e-01 -8.51239324e-01
-2.27417156e-01 6.53102636e-01 4.14613783e-01 5.59285045e-01
-7.62175739e-01 1.46641493e+00 6.95466900e+00 7.45700121e-01
-1.08336949e+00 -2.38646850e-01 4.73662585e-01 4.79391843e-01
-8.78671408e-01 1.25546545e-01 -4.00156975e-01 1.95411369e-01
1.37852407e+00 -3.60486433e-02 7.76312232e-01 6.09466136e-01
-1.23299390e-01 7.30611324e-01 -1.26562631e+00 6.12502337e-01
2.09987849e-01 -1.91246152e+00 9.49741364e-01 2.34105568e-02
9.71204877e-01 2.62472946e-02 -2.52094507e-01 4.91286427e-01
9.80174780e-01 -1.57730663e+00 5.06868899e-01 5.32480001e-01
8.42911184e-01 -4.12425011e-01 6.53085589e-01 1.46570966e-01
-1.38062823e+00 2.82066047e-01 -4.52695668e-01 -2.60600954e-01
1.00369222e-01 4.31843907e-01 -1.24347031e+00 1.18312621e+00
-6.57712221e-02 9.36289549e-01 -7.84767926e-01 2.88632542e-01
-5.94842911e-01 5.74187577e-01 2.96195358e-01 -2.84163564e-01
6.98118567e-01 -4.27757144e-01 3.99831116e-01 1.29452229e+00
3.74295682e-01 -2.79224426e-01 3.16503137e-01 1.24854779e+00
-7.01676488e-01 2.74362117e-02 -1.15939188e+00 -7.20434427e-01
-1.02530830e-01 1.27335215e+00 -5.57177246e-01 -5.36559820e-01
-3.58030051e-01 1.13274419e+00 9.37259912e-01 5.55995405e-01
-6.55412674e-01 -4.60749060e-01 2.17856407e-01 1.12465575e-01
4.53050733e-01 -3.96367043e-01 5.21558977e-04 -1.41937387e+00
-3.10424924e-01 -1.09939158e+00 3.94627273e-01 -9.36281621e-01
-1.15326560e+00 8.21116805e-01 -1.45860016e-01 -6.87237442e-01
-6.36977255e-01 -8.03385019e-01 -9.76009905e-01 9.97070551e-01
-1.45927060e+00 -1.81542051e+00 -1.01905048e-01 7.27215886e-01
1.80110693e-01 -2.92668104e-01 9.98011887e-01 -1.11963779e-01
-2.03170136e-01 6.08183563e-01 -3.58739883e-01 5.68670809e-01
2.55810201e-01 -1.55200970e+00 1.29862499e+00 9.63923097e-01
5.32790720e-01 6.51401997e-01 3.09711337e-01 -8.10107529e-01
-1.87498188e+00 -1.61209583e+00 1.03727674e+00 -4.18431431e-01
8.00189257e-01 -8.02872539e-01 -8.58853400e-01 1.10404956e+00
5.51639795e-01 1.61099985e-01 3.82169276e-01 -7.00829725e-04
-5.41855097e-01 4.31606591e-01 -7.99888849e-01 6.56409204e-01
1.57594204e+00 -7.43834078e-01 -4.46387887e-01 5.19060016e-01
1.43760610e+00 -3.95426333e-01 -8.33749712e-01 2.31674910e-01
-9.95674357e-02 -4.76225466e-01 7.73997784e-01 -1.17864752e+00
6.61795676e-01 5.84408827e-02 1.52049186e-02 -1.78659141e+00
-4.58808780e-01 -1.16487265e+00 -3.89042079e-01 8.65441680e-01
7.06716180e-01 -6.73735797e-01 8.44781518e-01 -1.32700190e-01
-4.84117627e-01 -6.24891639e-01 -5.56262136e-01 -5.58478713e-01
4.01544422e-01 -4.96464111e-02 9.64059055e-01 1.04143047e+00
1.10823624e-01 1.06118560e+00 -3.19957763e-01 -2.04044431e-01
2.64636368e-01 2.06030756e-01 8.33223939e-01 -1.11568463e+00
-2.82458872e-01 -4.81767029e-01 -4.35746282e-01 -9.92713988e-01
6.32302225e-01 -1.67485297e+00 -1.65217504e-01 -2.24202561e+00
-4.00104970e-02 5.21231778e-02 -1.12004183e-01 4.44077700e-01
-3.74349594e-01 1.46851763e-01 1.98513806e-01 -3.55478287e-01
-5.41694462e-01 7.70419657e-01 1.67782438e+00 -4.98495877e-01
6.40193149e-02 -3.70208532e-01 -8.24388921e-01 1.38495013e-01
8.17195892e-01 -1.77821159e-01 -5.45029521e-01 -7.19080567e-01
5.19916415e-01 2.90051401e-01 2.66799122e-01 -4.42475051e-01
-1.11609302e-01 -1.80819824e-01 7.15246946e-02 -1.52970403e-01
3.68246324e-02 -2.40905225e-01 6.48441166e-02 4.64839906e-01
-5.78674793e-01 1.30776420e-01 8.10645148e-02 5.96419334e-01
-2.27576673e-01 -4.18069065e-02 3.47227335e-01 -5.87660015e-01
-2.88190097e-01 6.19369388e-01 -3.53208274e-01 4.42756619e-03
6.20746613e-01 7.34593645e-02 -7.30931759e-01 -6.44257784e-01
-6.06024742e-01 5.38297137e-03 3.74358445e-01 5.12194574e-01
7.51047313e-01 -1.58193076e+00 -8.73820901e-01 2.06957623e-01
2.06896588e-01 -6.49569184e-02 -2.81229526e-01 4.18464512e-01
-7.60765851e-01 7.71067679e-01 -1.02053009e-01 -3.12422186e-01
-8.94527435e-01 6.71233714e-01 3.34660202e-01 -8.94801736e-01
-5.26010454e-01 7.15637803e-01 2.47542322e-01 -8.61349642e-01
-1.67565182e-01 -7.50816584e-01 -1.70796320e-01 -3.31043124e-01
1.10366717e-01 -7.05827773e-02 1.95764259e-01 -5.29284894e-01
2.59202020e-03 2.80029923e-01 -2.31147371e-02 1.60587043e-01
1.17220604e+00 2.16990009e-01 -4.08055246e-01 1.24283694e-01
1.26000178e+00 -3.88797730e-01 -8.59299004e-01 -3.09913248e-01
1.02947548e-01 -6.24216488e-03 -3.71102877e-02 -5.74063480e-01
-1.01141107e+00 1.12733305e+00 -1.53622562e-02 3.00613463e-01
7.33890951e-01 -1.86509527e-02 1.05015194e+00 6.47814155e-01
2.32756585e-01 -6.12834036e-01 5.52522779e-01 9.82774615e-01
1.10657382e+00 -1.09400463e+00 -2.30451778e-01 -4.78900731e-01
-5.78889191e-01 1.29805076e+00 3.98032725e-01 -3.59511197e-01
2.74193585e-01 6.66159019e-02 -2.80902773e-01 -3.30341637e-01
-9.36895430e-01 -5.09143412e-01 6.70883298e-01 9.41181123e-01
6.39131427e-01 3.19339126e-01 -8.97652134e-02 2.55008042e-01
-3.02524686e-01 -5.72913773e-02 5.62665641e-01 7.50079811e-01
-1.82461783e-01 -1.47590244e+00 1.09892771e-01 4.08943564e-01
-3.07414621e-01 -5.80190718e-01 -9.75693941e-01 4.84551072e-01
-3.11411351e-01 8.17444026e-01 7.54832895e-03 -5.86471140e-01
9.28410441e-02 1.57982916e-01 7.03386307e-01 -1.06381428e+00
-8.08767915e-01 -1.62797377e-01 4.48779970e-01 -5.06591558e-01
-2.83834398e-01 2.63794046e-02 -1.15814781e+00 -4.88105923e-01
-2.08478481e-01 3.64191420e-02 3.56122822e-01 7.36582339e-01
4.38254684e-01 9.90967035e-01 3.86861056e-01 -7.72739589e-01
-1.55258879e-01 -1.13641834e+00 -2.33340129e-01 5.50329864e-01
2.94094563e-01 -1.89922489e-02 4.84052710e-02 9.13059935e-02] | [10.25270938873291, 8.322868347167969] |
99701d07-0ca5-47fa-b1cf-0e06d7636c9d | local-to-global-learning-for-iterative | null | null | https://aclanthology.org/2022.naacl-industry.13 | https://aclanthology.org/2022.naacl-industry.13.pdf | Local-to-global learning for iterative training of production SLU models on new features | In production SLU systems, new training data becomes available with time so that ML models need to be updated on a regular basis. Specifically, releasing new features adds new classes of data while the old data remains constant. However, retraining the full model each time from scratch is computationally expensive. To address this problem, we propose to consider production releases from the curriculum learning perspective and to adapt the local-to-global learning (LGL) schedule (Cheng et. al, 2019) for a statistical model that starts with fewer output classes and adds more classes with each iteration. We report experiments for the tasks of intent classification and slot filling in the context of a production voice-assistant. First, we apply the original LGL schedule on our data and then adapt LGL to the production setting where the full data is not available at initial training iterations. We demonstrate that our method improves model error rates by 7.3% and saves up to 25% training time for individual iterations. | ['Daniil Sorokin', 'Yulia Grishina'] | null | null | null | null | naacl-acl-2022-7 | ['intent-classification', 'slot-filling'] | ['natural-language-processing', 'natural-language-processing'] | [ 5.86380363e-01 4.00245279e-01 -5.75869739e-01 -4.63602155e-01
-1.06044722e+00 -6.45625889e-01 1.11057244e-01 2.27218434e-01
-3.99966419e-01 9.49778676e-01 1.33159935e-01 -5.84828556e-01
2.61150718e-01 -5.18822491e-01 -5.67979097e-01 -2.22621724e-01
1.69683918e-01 6.77673280e-01 3.66745859e-01 -4.75853533e-02
-1.23271942e-01 3.04993659e-01 -1.46260738e+00 5.03871441e-01
7.80329108e-01 6.95466995e-01 7.08905160e-01 9.10164654e-01
-2.79330224e-01 7.99003780e-01 -8.73195469e-01 1.63110659e-01
1.62159458e-01 -4.87764657e-01 -1.03462601e+00 4.51392889e-01
4.81177151e-01 -3.35905254e-01 -1.68265611e-01 5.08464754e-01
2.72220641e-01 5.83698809e-01 1.98794618e-01 -1.07390583e+00
-3.47689807e-01 9.24233854e-01 -1.90653041e-01 6.84476122e-02
9.94725153e-02 -1.20562010e-01 1.03770244e+00 -7.53707111e-01
7.18648970e-01 1.03437793e+00 4.59786057e-01 8.35543692e-01
-1.31836438e+00 -5.40351212e-01 4.97200400e-01 -2.55276710e-01
-9.67563570e-01 -7.03602970e-01 6.12975001e-01 -2.43998140e-01
1.12099457e+00 3.35068852e-01 4.11927491e-01 7.24307656e-01
-4.50538620e-02 1.23045349e+00 8.36160421e-01 -8.27366590e-01
3.17658395e-01 2.58840024e-01 3.15547168e-01 8.74620974e-01
-1.57084212e-01 -2.45185107e-01 -8.31692815e-01 -2.10880324e-01
4.95402634e-01 -2.58913428e-01 -2.95995455e-02 -3.33106339e-01
-8.17397594e-01 7.70832956e-01 -4.42900099e-02 2.42104262e-01
1.09079210e-02 5.67685217e-02 2.83652037e-01 7.14774191e-01
4.20724630e-01 7.35672474e-01 -1.05926180e+00 -4.37387079e-01
-8.74173939e-01 2.45293006e-01 8.91985953e-01 1.20981276e+00
8.59791398e-01 8.37380663e-02 -1.66034088e-01 1.22341418e+00
6.34641796e-02 2.92980254e-01 6.76904142e-01 -1.06923747e+00
5.11811376e-01 3.30772042e-01 -1.07679307e-01 -6.35165721e-02
-1.96974456e-01 -5.84993124e-01 -2.57667065e-01 -1.92764759e-01
3.70592356e-01 -5.28307199e-01 -1.01888812e+00 1.92371356e+00
3.84511113e-01 1.25091538e-01 1.27528042e-01 1.43217489e-01
2.54236698e-01 8.08998168e-01 -4.70121205e-02 -7.03638971e-01
8.61446738e-01 -1.27714992e+00 -9.80227113e-01 -6.00289941e-01
1.07399905e+00 -8.23522329e-01 1.33133578e+00 6.01345956e-01
-1.05818975e+00 -8.73395562e-01 -1.04538512e+00 -9.68711451e-02
-1.17209166e-01 2.37820536e-01 7.47550964e-01 5.96040905e-01
-8.98983300e-01 6.51285529e-01 -9.73077238e-01 -1.29537746e-01
9.94072035e-02 4.98701006e-01 -9.93166342e-02 -1.15885213e-01
-9.66044486e-01 5.33182144e-01 5.27233481e-01 -2.83520892e-02
-6.19128704e-01 -8.03600788e-01 -9.36328590e-01 3.17686833e-02
7.49031961e-01 -2.28439152e-01 2.10843515e+00 -6.81869805e-01
-1.84740388e+00 1.31746069e-01 -4.26594317e-01 -2.31962323e-01
1.36366040e-01 -3.63057852e-01 -1.06121503e-01 -4.61070210e-01
-8.93694460e-02 5.86734951e-01 7.02622533e-01 -1.11570430e+00
-9.51065838e-01 -6.04787841e-02 4.38716300e-02 3.33214790e-01
-3.55189055e-01 -1.58328399e-01 -4.61787224e-01 -5.67420483e-01
3.34981620e-01 -1.19477844e+00 -1.81760862e-02 -6.39678359e-01
-6.20066561e-02 -4.11260605e-01 7.76003242e-01 -6.59193397e-01
1.72221315e+00 -2.01841831e+00 2.32481658e-01 4.38573472e-02
1.31731583e-02 2.53703207e-01 -2.81054944e-01 2.26316735e-01
1.17911488e-01 -2.46227738e-02 -2.99562305e-01 -9.00205433e-01
-1.57262683e-01 5.42775929e-01 -2.48301491e-01 -2.60688484e-01
5.96206486e-02 5.96294641e-01 -9.92400825e-01 -3.95272106e-01
-6.64519444e-02 -2.52810985e-01 -1.15483999e+00 3.01236510e-01
-6.18222177e-01 2.90282160e-01 -1.84043229e-01 5.67453682e-01
2.00057149e-01 -6.08119629e-02 7.77949452e-01 3.80429685e-01
-6.74705580e-02 7.61605024e-01 -1.17313397e+00 1.99908602e+00
-8.54898989e-01 7.36491323e-01 -1.12120556e-02 -6.34413362e-01
7.30299592e-01 2.76118308e-01 3.72079670e-01 -2.79928595e-01
-1.94062904e-01 1.63882598e-01 2.57665873e-01 3.01911216e-02
8.00275028e-01 -2.35459059e-01 -4.01147157e-01 6.47580922e-01
2.98219234e-01 -2.34104723e-01 1.69250295e-01 6.47859201e-02
1.07581913e+00 1.63181797e-01 2.85204083e-01 5.25651276e-02
1.76698878e-01 -1.85344238e-02 7.11558461e-01 8.65198374e-01
-2.74617728e-02 2.74872839e-01 3.92212987e-01 -1.61426082e-01
-7.98128963e-01 -9.32528079e-01 4.33950423e-04 1.71700406e+00
-5.29296160e-01 -7.45219409e-01 -8.00645649e-01 -8.95780623e-01
1.53408915e-01 1.03991437e+00 -2.10669562e-01 -2.87112802e-01
-9.18734133e-01 -3.32233518e-01 1.97567523e-01 5.01033008e-01
2.19718307e-01 -1.31928539e+00 -3.27906966e-01 6.08741462e-01
-6.40481487e-02 -9.31626439e-01 -6.84956670e-01 7.18115926e-01
-1.07929921e+00 -6.42109513e-01 -1.84599997e-03 -8.43969107e-01
7.70248711e-01 -8.01084638e-02 7.99245894e-01 1.79351538e-01
-3.98937687e-02 3.38910371e-02 -3.53262395e-01 -2.89130658e-01
-8.65967155e-01 7.98059642e-01 2.16413289e-01 -2.86638081e-01
8.23218152e-02 -2.95661002e-01 3.00680518e-01 -8.49633813e-02
-7.03387201e-01 2.54710495e-01 5.24145186e-01 1.10955417e+00
5.57372689e-01 4.48071510e-02 8.15571606e-01 -1.42419696e+00
7.91403174e-01 -2.11066425e-01 -4.34993893e-01 3.36603761e-01
-7.50732124e-01 4.01970685e-01 6.62618339e-01 -6.34888291e-01
-1.16750062e+00 2.45603010e-01 -5.36807999e-02 -4.41833466e-01
-3.75454649e-02 6.89104438e-01 -1.90253869e-01 3.86224210e-01
4.61132705e-01 6.93810955e-02 -1.48460269e-01 -9.51332092e-01
2.69281834e-01 9.48483884e-01 2.55027801e-01 -7.80473173e-01
6.77624226e-01 -4.93684471e-01 -4.92988199e-01 -7.74291694e-01
-1.06325054e+00 -5.14033675e-01 -9.21359539e-01 4.36258167e-02
1.70668483e-01 -6.26270652e-01 -3.37635279e-01 4.29995269e-01
-9.91390765e-01 -1.12042224e+00 -5.52139640e-01 3.23399365e-01
-5.80819845e-01 1.94681153e-01 -6.32741451e-01 -8.12145770e-01
1.28229382e-02 -1.11707556e+00 7.84157991e-01 -9.67687815e-02
-4.11893010e-01 -8.85969877e-01 4.97435108e-02 3.44866246e-01
2.04158798e-01 -5.15918672e-01 1.18826795e+00 -7.18210220e-01
-3.91871810e-01 -7.50341192e-02 5.61183751e-01 5.11617720e-01
5.49109161e-01 -3.80598336e-01 -9.69618678e-01 -4.93187547e-01
-3.82349156e-02 -5.78769386e-01 6.89740896e-01 7.33619854e-02
1.30931640e+00 -3.38454902e-01 -2.37315878e-01 2.58721501e-01
8.32504332e-01 5.54984570e-01 -1.86713114e-02 1.59467645e-02
5.93130887e-01 3.68171304e-01 8.41884375e-01 3.67543727e-01
1.12804554e-01 8.22878718e-01 -3.21049869e-01 2.10701138e-01
-3.30557406e-01 -7.18448579e-01 5.17798483e-01 1.27480388e+00
3.94347757e-01 -2.77698249e-01 -7.79285133e-01 4.44041938e-01
-1.87328243e+00 -5.55976152e-01 5.60067952e-01 2.52264261e+00
1.36819565e+00 4.58435625e-01 -1.21296927e-01 4.93927635e-02
3.73904526e-01 1.25693262e-01 -5.42284012e-01 -6.78189039e-01
3.74900371e-01 5.19846678e-01 3.26341212e-01 1.11189175e+00
-9.53904569e-01 1.52731013e+00 6.72070312e+00 7.36865222e-01
-8.77693117e-01 2.71105587e-01 4.95110333e-01 -3.31348509e-01
-2.66321272e-01 2.44527549e-01 -1.17680180e+00 2.35558152e-01
1.24721348e+00 4.38677222e-02 7.55093634e-01 1.02928710e+00
-1.40861094e-01 -2.26435572e-01 -1.25154293e+00 6.58834100e-01
3.19587737e-02 -1.05955029e+00 -4.98172082e-02 -7.77764898e-03
8.63222957e-01 -1.62667841e-01 -1.36267856e-01 9.95598018e-01
4.30678546e-01 -6.63827300e-01 5.63306868e-01 2.48980135e-01
7.26871967e-01 -7.68937826e-01 4.12125856e-01 7.70544589e-01
-1.05829823e+00 -1.72793880e-01 -3.41056824e-01 -2.09294498e-01
1.04430169e-01 1.27711862e-01 -1.44401670e+00 1.76869810e-01
5.17456159e-02 1.66395888e-01 -5.17962098e-01 4.88935620e-01
-2.25103989e-01 9.64069903e-01 -3.90960038e-01 -5.72410338e-02
8.74699429e-02 1.48162752e-01 2.01731384e-01 9.50561166e-01
2.00400740e-01 -1.46613002e-03 7.88968980e-01 4.98457044e-01
-1.76628873e-01 4.94605750e-02 -3.56971353e-01 -7.03937709e-02
6.83312535e-01 9.53247547e-01 -3.67714852e-01 -4.75153655e-01
-3.26249659e-01 1.10073972e+00 6.06247485e-01 2.50859678e-01
-3.71268630e-01 -3.84459466e-01 5.54751575e-01 2.80176736e-02
1.87646210e-01 -4.26741481e-01 -4.71324623e-02 -1.01573336e+00
-1.25872910e-01 -8.78742456e-01 2.96117455e-01 -5.30411780e-01
-6.90645278e-01 5.14056802e-01 7.78414905e-02 -8.29317868e-01
-8.62412095e-01 -3.72688979e-01 -8.99496302e-02 7.24511921e-01
-1.32200122e+00 -7.72779524e-01 2.38104865e-01 7.11102858e-02
1.27636170e+00 -3.06967586e-01 9.24909532e-01 -1.96989048e-02
-5.12633860e-01 8.74353766e-01 1.63335204e-01 1.03630603e-03
6.68557227e-01 -1.38974333e+00 5.34579873e-01 6.73901796e-01
4.29419875e-01 8.25986564e-01 5.24874806e-01 -8.16945553e-01
-1.18408918e+00 -8.62732708e-01 1.38725829e+00 -5.49866438e-01
5.43623865e-01 -7.17643380e-01 -8.44755173e-01 1.08552337e+00
-7.46783242e-02 -2.25502968e-01 8.30095351e-01 6.82349980e-01
-1.18455626e-02 -8.48634318e-02 -7.81330645e-01 5.64072907e-01
1.05236256e+00 -6.00135505e-01 -6.27821147e-01 3.83750230e-01
1.06641924e+00 -8.04605961e-01 -8.15986156e-01 1.64991796e-01
3.62864375e-01 -3.27862084e-01 3.19023281e-01 -8.68980587e-01
-1.26489088e-01 -5.53167053e-02 -2.09924698e-01 -1.45850408e+00
-2.11758077e-01 -9.57908869e-01 -4.62880641e-01 1.02428043e+00
8.67901206e-01 -4.26004976e-01 7.14507580e-01 6.42288506e-01
-6.06085539e-01 -8.08318019e-01 -8.11052382e-01 -6.70336127e-01
-3.35396901e-02 -6.89615726e-01 4.82759476e-01 7.95040429e-01
3.78911972e-01 6.93035901e-01 -4.19042587e-01 -1.89248323e-02
1.33648247e-01 1.99201182e-01 9.77236152e-01 -1.21052325e+00
-6.72990859e-01 7.40255415e-02 2.80643910e-01 -1.49940729e+00
3.79698098e-01 -1.00228786e+00 3.88306260e-01 -1.13835192e+00
8.87403861e-02 -9.39291179e-01 -4.32329446e-01 9.06743407e-01
-3.28289270e-01 -3.82966459e-01 5.49870431e-01 1.10735551e-01
-4.79316682e-01 3.01318347e-01 1.12813473e+00 8.55792984e-02
-7.77463078e-01 4.21908885e-01 -6.08475626e-01 5.02131641e-01
9.32118893e-01 -5.36938429e-01 -8.25926185e-01 -3.08940619e-01
-1.47200733e-01 2.84066051e-01 -4.16366577e-01 -1.00412989e+00
2.59032220e-01 -3.47576797e-01 1.15560733e-01 -5.85466444e-01
4.89422768e-01 -6.05056226e-01 -1.56452239e-01 3.89912188e-01
-7.29979873e-01 -1.41096324e-01 3.96394461e-01 3.70446086e-01
-1.39483633e-02 -6.40028059e-01 6.13934457e-01 -3.64253297e-02
-5.85256815e-01 6.00686632e-02 -5.70525944e-01 -1.26506062e-02
6.36398792e-01 -6.72168583e-02 1.05176210e-01 -2.76429683e-01
-1.13199520e+00 1.36723951e-01 1.69440046e-01 5.72415531e-01
2.55936652e-01 -1.10111964e+00 -1.56673625e-01 6.30992830e-01
3.75460498e-02 2.97978431e-01 3.66442092e-02 3.10488105e-01
-7.82302022e-02 7.50786722e-01 2.18481943e-01 -3.87238860e-01
-1.34217799e+00 3.21087718e-01 2.11028054e-01 -6.06256127e-01
-3.59863281e-01 9.12766099e-01 -6.10731319e-02 -9.10780549e-01
5.27776897e-01 -8.08878481e-01 8.67882147e-02 4.73662391e-02
2.08384812e-01 1.03617191e-01 2.12589949e-01 -2.18203571e-02
1.93240702e-01 1.74004417e-02 -5.44779658e-01 -4.39615756e-01
1.00851178e+00 -3.51850763e-02 3.52897048e-01 1.02621269e+00
8.41313004e-01 3.09619933e-01 -1.33762264e+00 -6.07152045e-01
2.74191886e-01 -3.31391931e-01 1.12730607e-01 -9.99080479e-01
-7.20581532e-01 6.06201172e-01 2.20929757e-01 -1.65535100e-02
7.83876657e-01 -9.91793536e-03 8.00030231e-01 8.99381280e-01
3.98501277e-01 -1.52569914e+00 2.29235262e-01 9.27669704e-01
5.72264910e-01 -8.60630631e-01 -1.57903135e-01 -3.27636629e-01
-6.38402700e-01 9.70925510e-01 8.81479204e-01 4.44909751e-01
4.35592502e-01 2.90015042e-01 -7.92548880e-02 4.11162138e-01
-1.23213148e+00 4.06731404e-02 5.03411405e-02 4.27250773e-01
5.57484031e-01 2.72470385e-01 -8.87987167e-02 6.06026590e-01
-3.98391366e-01 2.22860053e-02 4.63042974e-01 1.32145488e+00
-8.07759523e-01 -1.70693433e+00 -4.67406511e-02 6.17040634e-01
-8.38315580e-03 -2.56599039e-01 -2.16414094e-01 7.53030837e-01
1.83628127e-01 8.66052926e-01 2.89276004e-01 -3.21387380e-01
2.95569956e-01 8.50252211e-01 5.76754332e-01 -1.35379183e+00
-4.90792155e-01 3.81606817e-01 3.53254616e-01 -1.48062989e-01
8.34706947e-02 -9.67751026e-01 -1.48178315e+00 7.07052499e-02
-4.77096260e-01 4.01639104e-01 5.70657372e-01 1.00788891e+00
2.10602880e-01 7.92336047e-01 7.13107884e-01 -4.12900776e-01
-8.88590455e-01 -1.34825957e+00 -4.68763798e-01 -1.66800901e-01
3.86091083e-01 -5.95787525e-01 -2.71916956e-01 2.49483287e-01] | [10.848596572875977, 8.42949390411377] |
103aa8d1-5ed3-4796-8314-93a8a4440991 | simulation-based-counterfactual-causal | 2306.03354 | null | https://arxiv.org/abs/2306.03354v1 | https://arxiv.org/pdf/2306.03354v1.pdf | Simulation-Based Counterfactual Causal Discovery on Real World Driver Behaviour | Being able to reason about how one's behaviour can affect the behaviour of others is a core skill required of intelligent driving agents. Despite this, the state of the art struggles to meet the need of agents to discover causal links between themselves and others. Observational approaches struggle because of the non-stationarity of causal links in dynamic environments, and the sparsity of causal interactions while requiring the approaches to work in an online fashion. Meanwhile interventional approaches are impractical as a vehicle cannot experiment with its actions on a public road. To counter the issue of non-stationarity we reformulate the problem in terms of extracted events, while the previously mentioned restriction upon interventions can be overcome with the use of counterfactual simulation. We present three variants of the proposed counterfactual causal discovery method and evaluate these against state of the art observational temporal causal discovery methods across 3396 causal scenes extracted from a real world driving dataset. We find that the proposed method significantly outperforms the state of the art on the proposed task quantitatively and can offer additional insights by comparing the outcome of an alternate series of decisions in a way that observational and interventional approaches cannot. | ['Lars Kunze', 'Rhys Howard'] | 2023-06-06 | null | null | null | null | ['causal-discovery'] | ['knowledge-base'] | [ 4.06838208e-01 2.33173132e-01 -7.22072899e-01 -2.91902602e-01
-3.07863593e-01 -3.61638397e-01 1.30521929e+00 8.06870088e-02
-4.76061851e-01 1.18855047e+00 6.12379789e-01 -8.06385994e-01
-6.33244634e-01 -7.85693944e-01 -1.02560329e+00 -6.40129268e-01
-4.14319128e-01 3.82362038e-01 3.02350938e-01 -2.35813349e-01
1.69872820e-01 4.26213682e-01 -1.89985645e+00 -2.24753112e-01
8.07658255e-01 3.67075086e-01 -7.68305436e-02 3.01158637e-01
2.91355103e-01 8.85112822e-01 -1.65589094e-01 -3.57659757e-01
2.42035747e-01 -2.96775341e-01 -4.52969640e-01 -3.84848714e-01
5.18565416e-01 -3.53529304e-01 -7.52001822e-01 7.78425694e-01
2.32765675e-01 1.76736876e-01 7.12656021e-01 -1.90374660e+00
-2.51933694e-01 5.75043559e-01 -5.51102400e-01 4.14417505e-01
2.42108807e-01 3.36187154e-01 9.20117915e-01 -1.45741224e-01
6.87542856e-01 1.38006425e+00 3.82639319e-01 2.39076063e-01
-1.31765151e+00 -1.09789765e+00 4.73410964e-01 5.85948050e-01
-7.46413291e-01 -5.65094054e-01 8.53592455e-01 -5.88684678e-01
6.36111379e-01 2.29104981e-01 4.87266123e-01 1.49374330e+00
3.24270993e-01 4.49832350e-01 1.29600382e+00 5.82588874e-02
3.05122018e-01 -3.20384473e-01 1.70520246e-01 5.95784545e-01
5.24782896e-01 1.07383156e+00 -7.66816914e-01 -2.78792322e-01
1.79398909e-01 -8.83262604e-02 7.94749781e-02 -4.07422543e-01
-1.41296113e+00 9.52431679e-01 2.00800732e-01 -1.17439471e-01
-6.26636744e-01 4.61850941e-01 5.04785061e-01 3.11264455e-01
2.67604411e-01 1.74799964e-01 -3.71929139e-01 -5.03989011e-02
-6.45434082e-01 7.80042648e-01 5.70096791e-01 5.23568392e-01
4.02915806e-01 -2.34638944e-01 -2.22440347e-01 1.04626395e-01
2.50637889e-01 6.42173946e-01 -1.88172534e-02 -8.77565205e-01
4.66276467e-01 3.86637032e-01 4.44831580e-01 -9.16622460e-01
-3.77638847e-01 -3.53193209e-02 -5.27157247e-01 3.17154497e-01
7.23506331e-01 -4.16068077e-01 -6.80486619e-01 2.04763818e+00
6.41081810e-01 6.64565980e-01 -7.83186033e-02 8.43163371e-01
4.32502478e-01 4.12624657e-01 5.07372558e-01 -6.56548858e-01
1.09753931e+00 -4.12810415e-01 -1.01125157e+00 -1.31478518e-01
4.83486980e-01 -4.62648928e-01 6.12164378e-01 2.68065631e-02
-5.28586447e-01 -3.45708907e-01 -9.71894801e-01 3.70532751e-01
-3.47090989e-01 -3.40160877e-01 9.77809608e-01 5.81055701e-01
-5.19580245e-01 2.30967730e-01 -5.56993186e-01 -3.13461810e-01
4.55907166e-01 1.99393108e-01 -3.10443640e-01 7.19667822e-02
-1.56550598e+00 1.01750875e+00 1.36651501e-01 -2.46094987e-02
-1.31941366e+00 -1.24138784e+00 -6.83083713e-01 -3.08620304e-01
1.00964284e+00 -6.43380940e-01 1.22525704e+00 -5.02441585e-01
-1.10381544e+00 3.19139659e-01 -4.27631944e-01 -8.14444661e-01
9.48256075e-01 -2.70926625e-01 -7.29107976e-01 -2.84454793e-01
4.58764911e-01 4.44582403e-01 6.42083168e-01 -1.19137192e+00
-1.03404355e+00 -3.08638602e-01 2.54113346e-01 -2.70879921e-02
2.97101170e-01 -1.49733767e-01 1.63258031e-01 -4.28814411e-01
-5.18665731e-01 -1.19018638e+00 -4.09208655e-01 -2.47320458e-01
-5.13852656e-01 -5.49728870e-01 1.06845868e+00 -1.65610880e-01
1.10169601e+00 -1.92200530e+00 -1.62331522e-01 -4.53712652e-03
2.13075817e-01 -4.76649404e-02 6.27767593e-02 5.75400472e-01
-2.79528141e-01 -8.24604407e-02 -9.35062841e-02 1.42506137e-01
4.48913909e-02 2.52142876e-01 -5.48331082e-01 8.87379527e-01
1.88803583e-01 8.80019546e-01 -1.30296850e+00 -4.52039182e-01
4.64860767e-01 4.01245467e-02 -4.10506159e-01 -9.53940973e-02
-4.38952416e-01 5.87250710e-01 -7.18420446e-01 -3.85259539e-02
3.50161463e-01 1.99703887e-01 3.36393565e-01 1.45700872e-01
-6.18434250e-01 6.60414338e-01 -1.17154658e+00 1.15247869e+00
-2.94799745e-01 7.96030104e-01 -3.81626636e-01 -1.21228838e+00
5.53941309e-01 5.84195673e-01 6.74674451e-01 -8.56578708e-01
-2.95034982e-02 -1.14183865e-01 3.79328012e-01 -7.90403903e-01
1.28930494e-01 -2.29453132e-01 -1.48211449e-01 5.59540689e-01
-5.31559289e-01 3.27703416e-01 3.18562031e-01 1.98119849e-01
1.37096035e+00 9.15360227e-02 2.58159190e-01 -3.93350661e-01
2.14129329e-01 4.44175392e-01 5.80771506e-01 9.41486716e-01
-2.98935115e-01 -3.33416551e-01 5.41949987e-01 -5.91681302e-01
-9.54640150e-01 -1.18376660e+00 -9.51678008e-02 8.02214205e-01
2.70256191e-01 2.65159328e-02 -2.42968157e-01 -8.42032790e-01
3.87271523e-01 1.25612760e+00 -1.12895930e+00 -2.58289635e-01
-5.85281253e-01 -9.09849107e-01 3.29434276e-01 2.52994001e-01
5.38779616e-01 -6.99806869e-01 -1.07378459e+00 1.58718541e-01
-3.12174648e-01 -1.02444744e+00 -5.80927767e-02 -1.56694084e-01
-4.27982152e-01 -1.42557299e+00 -2.83624083e-02 9.16820671e-03
3.34904134e-01 4.55459177e-01 1.01652145e+00 -2.36441925e-01
-4.02146131e-02 1.22263588e-01 -1.15473885e-02 -9.05668736e-01
-4.49810386e-01 -3.57416600e-01 3.10613602e-01 1.37403667e-01
5.45615554e-01 -6.91145122e-01 -6.96337163e-01 3.76083344e-01
-5.01623034e-01 1.78577393e-01 4.37038511e-01 6.14583135e-01
2.20716700e-01 5.15684187e-01 1.01630569e+00 -7.94646144e-01
3.94684613e-01 -8.58071029e-01 -8.59971523e-01 -6.20673187e-02
-7.43247390e-01 2.33483136e-01 3.19345444e-01 -5.60225487e-01
-1.46781898e+00 1.25560105e-01 4.43635732e-01 -2.00269707e-02
-2.75244385e-01 4.21689868e-01 1.55069446e-02 4.70966935e-01
5.23325205e-01 -3.36001605e-01 4.92070466e-02 -6.29726425e-02
6.50767803e-01 2.63705850e-01 4.88582522e-01 -4.82696623e-01
8.42683911e-01 9.44644570e-01 4.38512176e-01 -5.46797574e-01
-8.13247144e-01 -3.63427788e-01 -4.98595178e-01 -6.16843760e-01
8.71717334e-01 -9.33358133e-01 -1.18477654e+00 -1.41586244e-01
-1.06553209e+00 -4.59719896e-01 -2.84867704e-01 7.40339637e-01
-6.44428611e-01 -1.18907161e-01 2.34788731e-01 -1.10645330e+00
5.80980957e-01 -9.06620800e-01 7.97129691e-01 -1.30146459e-01
-4.14691985e-01 -8.51174712e-01 3.65321755e-01 3.79499078e-01
1.57020278e-02 6.51607215e-01 8.83919895e-01 -2.10071445e-01
-5.28707027e-01 -1.67881295e-01 -3.15367430e-01 -5.60245812e-01
2.96357155e-01 1.47722065e-02 -7.71013558e-01 1.92413300e-01
-2.73762941e-01 2.39352167e-01 8.33567500e-01 6.86527371e-01
6.56753540e-01 -4.28509206e-01 -7.91832030e-01 -2.99847990e-01
1.12423062e+00 3.70268553e-01 4.44928527e-01 3.17026764e-01
4.83630687e-01 1.09523356e+00 9.14376616e-01 3.52477282e-01
7.89931118e-01 8.34613740e-01 6.76134884e-01 -1.78994447e-01
-1.38518915e-01 -5.01140952e-01 2.41108790e-01 -2.71213442e-01
-2.63201058e-01 -1.34013161e-01 -8.41566801e-01 1.02804160e+00
-2.31182098e+00 -1.47190130e+00 -9.39089775e-01 2.27629352e+00
5.43896198e-01 2.42984697e-01 5.61481595e-01 5.68083040e-02
4.94544983e-01 5.10018952e-02 -6.10746622e-01 -3.59646291e-01
6.60685077e-02 -3.00101370e-01 1.08479798e+00 3.90630335e-01
-1.06490421e+00 4.95996922e-01 6.63562727e+00 4.49339777e-01
-8.16098511e-01 2.69384801e-01 4.75042254e-01 -3.41744423e-01
-2.29988098e-01 1.50291651e-01 -6.63180351e-01 4.90333915e-01
1.34536469e+00 -5.33894658e-01 2.56488234e-01 2.04054892e-01
1.08100665e+00 -4.33216959e-01 -1.43303394e+00 4.17207956e-01
-5.55046201e-01 -1.33318162e+00 -2.25036442e-01 2.76225001e-01
8.74014497e-01 1.44853488e-01 4.60675992e-02 5.32481521e-02
1.04189336e+00 -9.70433712e-01 9.00295675e-01 6.34402096e-01
4.55831468e-01 -5.59382021e-01 4.91988182e-01 3.88179004e-01
-8.71321261e-01 -3.63982081e-01 2.37031132e-01 -7.00611591e-01
4.36840624e-01 6.56774759e-01 -8.12996387e-01 6.87817097e-01
6.56635702e-01 6.50017679e-01 -2.33629271e-01 7.04795122e-01
-4.13840383e-01 8.81473124e-01 -3.03163230e-01 -1.34215027e-01
3.34943861e-01 -9.95724872e-02 7.87889242e-01 8.29592407e-01
8.56220275e-02 1.28110617e-01 -1.01413764e-01 7.62666821e-01
2.73894966e-01 -3.26874644e-01 -1.26923490e+00 2.98864301e-02
4.26614553e-01 5.77993870e-01 -4.06062365e-01 -3.87642324e-01
-6.86162353e-01 1.54942319e-01 3.95707320e-03 4.69076365e-01
-1.13263309e+00 4.61307019e-01 9.27798092e-01 2.89421529e-01
1.45537616e-03 -2.09548458e-01 -3.91230077e-01 -6.06630266e-01
2.31341983e-04 -7.22601831e-01 5.02246797e-01 -5.05100727e-01
-1.18998027e+00 -2.56605595e-01 5.95453799e-01 -1.06326437e+00
-2.82109380e-01 -1.56557605e-01 -5.42100549e-01 6.06784642e-01
-1.60996783e+00 -9.76028979e-01 1.25087216e-01 4.72529143e-01
5.70840061e-01 1.89360604e-01 3.50582987e-01 3.92508864e-01
-5.48732996e-01 1.98523495e-02 -2.06527174e-01 -3.47797126e-01
6.55854285e-01 -1.02001965e+00 3.93500090e-01 9.90822017e-01
-1.36596337e-01 3.77665758e-01 1.31015503e+00 -1.02598608e+00
-1.51264811e+00 -1.24032307e+00 1.08997846e+00 -6.18854344e-01
1.15554380e+00 -2.30714038e-01 -4.49680358e-01 5.93009233e-01
4.43423688e-02 -1.83103979e-01 2.57692277e-01 4.66132551e-01
-2.44533837e-01 -5.35554662e-02 -8.22276175e-01 1.05951345e+00
1.41652656e+00 -2.98424989e-01 -8.07274878e-01 3.43390822e-01
7.52542794e-01 1.50983885e-01 -3.47824097e-01 5.33028245e-01
7.52909541e-01 -8.27461064e-01 9.26439285e-01 -9.70641792e-01
5.53158760e-01 -5.50307870e-01 5.38662970e-02 -1.27968931e+00
-4.40433808e-02 -7.49328911e-01 1.52853861e-01 9.71351564e-01
5.39594710e-01 -9.03490663e-01 5.34713149e-01 6.01607203e-01
1.38212487e-01 -1.99928492e-01 -1.30153728e+00 -8.53604853e-01
-1.74151715e-02 -7.59578407e-01 7.45580018e-01 1.04464960e+00
-9.72297117e-02 4.21750695e-01 -4.35919434e-01 4.78779554e-01
9.82087135e-01 7.42966086e-02 1.01224720e+00 -1.40286899e+00
4.11812253e-02 -3.60494494e-01 -2.59815305e-01 -5.70687056e-01
4.84494835e-01 -3.76409560e-01 9.27764922e-02 -1.14377260e+00
2.74199247e-01 -6.56402469e-01 -1.86144397e-01 2.90587842e-01
-1.46942407e-01 -4.57359590e-02 -1.81858465e-01 -1.00865833e-01
-2.42734253e-01 5.22915184e-01 1.10690582e+00 -2.64971018e-01
-2.00877473e-01 4.21680585e-02 -4.92854774e-01 7.25848854e-01
6.15382493e-01 -6.59081638e-01 -7.32345700e-01 7.51989558e-02
5.50260365e-01 4.05715644e-01 9.99078155e-01 -5.47619283e-01
1.43857777e-01 -6.24859929e-01 -2.83703327e-01 -5.38303494e-01
-7.83637588e-05 -9.57555234e-01 6.56293273e-01 6.06688261e-01
-5.08627295e-01 4.52742539e-02 2.14300990e-01 1.20735753e+00
6.93479106e-02 3.38205993e-01 1.30574629e-01 2.55843639e-01
-7.98711479e-01 2.03245893e-01 -5.34536600e-01 -1.81833625e-01
1.28546369e+00 7.10369572e-02 -5.98616719e-01 -3.38272303e-01
-3.40570778e-01 5.90893924e-01 1.36854172e-01 7.76620626e-01
1.84891120e-01 -1.29217398e+00 -8.70591283e-01 -2.46354550e-01
2.13307321e-01 -5.77227950e-01 3.35734993e-01 1.35237527e+00
2.87824541e-01 6.70020103e-01 -3.40284631e-02 -3.67753148e-01
-1.04831159e+00 9.17377889e-01 2.02011183e-01 -2.47195125e-01
-7.43487358e-01 1.76690072e-01 6.83689594e-01 -1.43322393e-01
-6.75289184e-02 -2.07469165e-02 -3.07784349e-01 1.41604111e-01
3.48325074e-01 6.68373466e-01 -7.61153027e-02 -5.65912247e-01
-3.26608956e-01 -4.02634218e-02 2.10208014e-01 -3.08857799e-01
1.26949501e+00 -3.43285590e-01 1.98556006e-01 6.84987605e-01
6.59254611e-01 -1.08914837e-01 -1.48990452e+00 -4.09592576e-02
2.02363059e-02 -4.75016326e-01 3.17107350e-01 -9.68594313e-01
-5.77199936e-01 3.41522068e-01 6.22865379e-01 4.32066500e-01
7.93276489e-01 -1.80773865e-02 5.19455492e-01 3.51488404e-02
4.36475039e-01 -8.51448119e-01 -5.56593120e-01 -9.23493877e-02
7.00260460e-01 -1.52097261e+00 2.38540359e-02 -5.59130669e-01
-2.67757207e-01 4.83727604e-01 4.23473418e-02 -9.61273685e-02
7.84338772e-01 1.34308800e-01 -2.43613169e-01 -5.65756619e-01
-1.19936192e+00 -4.33827698e-01 2.61927247e-01 5.48020720e-01
2.62659192e-01 6.63411558e-01 -6.54028952e-01 -1.27718346e-02
-2.37955615e-01 1.06551331e-02 4.91979569e-01 5.26102781e-01
2.13949561e-01 -7.91507959e-01 -3.06147039e-01 6.62824929e-01
-3.30804199e-01 1.24698505e-01 -2.70891726e-01 1.22942650e+00
2.57227480e-01 1.53122294e+00 1.56438977e-01 -7.25264251e-02
5.58526695e-01 -1.53811470e-01 2.46445611e-01 -1.46151736e-01
6.98289648e-02 -4.14807588e-01 6.24094486e-01 -8.17978442e-01
-9.89414454e-01 -1.39235699e+00 -9.36244726e-01 -4.85583425e-01
-1.03596859e-01 -7.19744759e-03 4.72206950e-01 1.41071332e+00
2.12030396e-01 8.59346628e-01 6.59713209e-01 -5.83884358e-01
-2.42554396e-01 -5.84753633e-01 -1.61256626e-01 4.54192758e-01
7.37880826e-01 -1.42559230e+00 -3.82687122e-01 1.79524735e-01] | [7.843807220458984, 5.303234100341797] |
d005d65f-f029-40b0-b848-92957582087f | the-cropandweed-dataset-a-multi-modal | null | null | https://openaccess.thecvf.com/content/WACV2023/html/Steininger_The_CropAndWeed_Dataset_A_Multi-Modal_Learning_Approach_for_Efficient_Crop_WACV_2023_paper.html | https://openaccess.thecvf.com/content/WACV2023/papers/Steininger_The_CropAndWeed_Dataset_A_Multi-Modal_Learning_Approach_for_Efficient_Crop_WACV_2023_paper.pdf | The CropAndWeed Dataset: A Multi-Modal Learning Approach for Efficient Crop and Weed Manipulation | Precision Agriculture and especially the application of automated weed intervention represents an increasingly essential research area, as sustainability and efficiency considerations are becoming more and more relevant. While the potentials of Convolutional Neural Networks for detection, classification and segmentation tasks have successfully been demonstrated in other application areas, this relatively new field currently lacks the required quantity and quality of training data for such a highly data-driven approach. Therefore, we propose a novel large-scale image dataset specializing in the fine-grained identification of 74 relevant crop and weed species with a strong emphasis on data variability. We provide annotations of labeled bounding boxes, semantic masks and stem positions for about 112k instances in more than 8k high-resolution images of both real-world agricultural sites and specifically cultivated outdoor plots of rare weed types. Additionally, each sample is enriched with an extensive set of meta-annotations regarding environmental conditions and recording parameters. We furthermore conduct benchmark experiments for multiple learning tasks on different variants of the dataset to demonstrate its versatility and provide examples of useful mapping schemes for tailoring the annotated data to the requirements of specific applications. In the course of the evaluation, we furthermore demonstrate how incorporating multiple species of weeds into the learning process increases the accuracy of crop detection. Overall, the evaluation clearly demonstrates that our dataset represents an essential step towards overcoming the data gap and promoting further research in the area of Precision Agriculture. | ['Verena Widhalm', 'Julia Simon', 'Gerardus Croonen', 'Andreas Trondl', 'Daniel Steininger'] | 2023-01-06 | null | null | null | winter-conference-on-applications-of-computer-2 | ['plant-phenotyping', 'fine-grained-image-classification', 'crop-classification'] | ['computer-vision', 'computer-vision', 'miscellaneous'] | [ 6.41694546e-01 -1.64757669e-01 -1.89875469e-01 -4.08107996e-01
-3.80421430e-01 -1.00150979e+00 3.10132623e-01 8.32566559e-01
-3.12119603e-01 5.78610182e-01 -3.63751262e-01 -6.19866848e-01
-4.51178700e-01 -9.11238372e-01 -6.59950972e-01 -7.29940832e-01
-4.76722330e-01 3.28894675e-01 2.73902714e-01 -4.13482994e-01
3.76961706e-03 8.62921834e-01 -1.70942688e+00 2.22895712e-01
7.33936608e-01 9.03977394e-01 8.66220534e-01 6.43685997e-01
4.41032313e-02 5.18313386e-02 -6.69215322e-01 -9.13576223e-03
3.64859074e-01 -8.28173885e-04 -4.04260069e-01 5.04207909e-02
4.94865805e-01 -2.28837475e-01 4.03687060e-01 1.10325813e+00
5.77892184e-01 -3.48635703e-01 4.38059002e-01 -9.43261683e-01
-3.67433548e-01 5.60114682e-01 -9.44711268e-01 1.05288573e-01
-3.09045881e-01 1.38076946e-01 7.92888343e-01 -3.30336571e-01
5.40125251e-01 7.77552843e-01 8.41289937e-01 -1.66523337e-01
-1.32922232e+00 -7.12936759e-01 3.89053673e-01 -1.18442118e-01
-1.20101273e+00 -2.71208555e-01 2.13827252e-01 -5.33479333e-01
6.92765117e-01 2.80154943e-01 7.27278709e-01 6.79988384e-01
5.88105395e-02 5.64768910e-01 7.90123403e-01 -4.37775761e-01
2.35195026e-01 5.99958748e-03 1.45593539e-01 5.09190679e-01
7.15253413e-01 2.11026907e-01 -9.62283611e-02 5.81627619e-03
9.02915716e-01 -1.56709269e-01 -9.99232978e-02 -7.74390817e-01
-1.26489568e+00 7.06143737e-01 7.40596116e-01 3.56189072e-01
-4.89683837e-01 -7.20538124e-02 6.90627515e-01 -3.05716060e-02
5.47098517e-01 8.47814322e-01 -9.28996205e-01 2.89556473e-01
-1.07052386e+00 1.53562889e-01 6.68816030e-01 1.16484797e+00
7.27314532e-01 -8.04251805e-02 -1.49878368e-01 7.13917077e-01
-1.32739022e-01 7.04968333e-01 -1.65188927e-02 -8.28004301e-01
4.28077914e-02 7.87000656e-01 4.04467374e-01 -1.15062809e+00
-6.81319535e-01 -5.11068165e-01 -8.86289358e-01 3.27682048e-01
5.83972991e-01 -2.05711782e-01 -1.14710295e+00 1.40122616e+00
1.43236876e-01 -7.86551163e-02 -1.63701791e-02 6.11575186e-01
7.08903551e-01 5.19727230e-01 3.92255187e-01 -5.60052693e-02
1.76749861e+00 -4.91557479e-01 -5.99196732e-01 -5.00351846e-01
5.78253925e-01 -7.79655397e-01 8.04406285e-01 1.54446587e-01
-5.39264083e-01 -4.96763468e-01 -1.17971873e+00 3.33995074e-01
-9.32979584e-01 7.48034716e-01 1.10802710e+00 6.95723712e-01
-7.02968299e-01 5.20150721e-01 -8.71322513e-01 -7.88953364e-01
8.86154056e-01 4.00077045e-01 -4.26861554e-01 4.21787649e-02
-8.41168404e-01 1.03327155e+00 7.20424354e-01 6.57478988e-01
-7.65820622e-01 -8.48541796e-01 -7.91188717e-01 1.51071176e-01
5.47985077e-01 3.77353393e-02 9.82818723e-01 -6.79295540e-01
-1.17023420e+00 1.12788308e+00 2.04338342e-01 -5.60566127e-01
2.45788381e-01 -2.27190763e-01 -3.02922487e-01 -1.17767945e-01
3.56108755e-01 8.26546490e-01 5.54931402e-01 -1.28818274e+00
-1.17252588e+00 -4.78316486e-01 1.94391176e-01 -1.44554809e-01
-3.09166819e-01 1.19232431e-01 -8.40093121e-02 -5.10610163e-01
8.85256156e-02 -8.30249488e-01 -5.38566172e-01 2.55603254e-01
-2.02972740e-01 5.94417214e-01 7.65034974e-01 -6.48119986e-01
9.13177550e-01 -2.11106086e+00 -2.83706002e-02 5.97019540e-03
-1.34817481e-01 6.32745564e-01 -3.88360977e-01 2.10538194e-01
-1.13612071e-01 1.87459737e-01 -4.71407652e-01 3.71950984e-01
-2.03515127e-01 2.70680457e-01 -1.94803983e-01 4.41623569e-01
5.74205399e-01 8.21024895e-01 -9.77830887e-01 -1.47017539e-01
6.07879639e-01 3.65801662e-01 -5.37599288e-02 -3.24236229e-02
-3.68509442e-01 2.73390889e-01 -5.05768359e-01 9.53584313e-01
9.61518764e-01 -1.48223817e-01 2.44882956e-01 -3.19567919e-01
-5.03124833e-01 -4.41425890e-01 -1.08609414e+00 1.46497321e+00
-4.14761335e-01 6.80688262e-01 3.86030823e-01 -1.08043802e+00
1.05166316e+00 1.03416637e-01 3.15861732e-01 -3.79846990e-01
3.78015190e-02 3.85566235e-01 2.22381338e-01 -3.59488755e-01
8.11890185e-01 3.99138927e-01 6.41649440e-02 -1.11156769e-01
7.67574683e-02 -3.46508235e-01 5.01158834e-01 -3.45832407e-01
8.20251346e-01 4.77333337e-01 6.20015502e-01 -6.49071157e-01
8.99991170e-02 6.65924847e-01 2.12460831e-01 6.58483446e-01
-4.41342264e-01 3.78578454e-01 4.80285078e-01 -5.26133537e-01
-1.03004801e+00 -4.65984255e-01 -6.03250444e-01 1.29138029e+00
1.07336283e-01 -7.16731548e-02 -5.25769234e-01 -4.52361405e-01
2.91019857e-01 4.82513219e-01 -8.12173307e-01 7.56766424e-02
-2.78834611e-01 -1.43218279e+00 6.29839778e-01 6.29656494e-01
6.84476614e-01 -1.34961271e+00 -1.33348405e+00 1.96432605e-01
1.05982393e-01 -1.32568634e+00 3.84066164e-01 1.00968409e+00
-6.43442094e-01 -1.12916100e+00 -7.66253233e-01 -8.94695222e-01
5.83816171e-01 4.72363949e-01 1.16144896e+00 -2.21987039e-01
-8.60253572e-01 -1.91818878e-01 -5.35400152e-01 -8.70822012e-01
-4.46694523e-01 6.32199585e-01 -5.48501849e-01 -5.28536379e-01
5.90811968e-01 -3.48379582e-01 -4.88738894e-01 2.62545675e-01
-8.09641123e-01 -6.04600273e-03 7.15326369e-01 8.13664973e-01
6.21154249e-01 1.96316093e-01 5.31539977e-01 -8.05519402e-01
1.96296498e-01 -2.34685630e-01 -1.24165535e+00 5.43876708e-01
-7.60851502e-02 -2.45782390e-01 2.49690086e-01 -1.54118463e-01
-8.56039166e-01 5.56195438e-01 2.83898711e-01 2.68156052e-01
-7.16526628e-01 8.18169355e-01 -3.46513957e-01 -1.43771321e-01
9.56734538e-01 -3.04335654e-01 -1.96650594e-01 -2.37243682e-01
3.15844357e-01 6.19420648e-01 6.63476408e-01 -3.45546603e-01
6.69376373e-01 3.69807750e-01 2.15498492e-01 -1.11830199e+00
-7.92113900e-01 -5.58398306e-01 -9.45274651e-01 1.58233732e-01
7.95885623e-01 -8.21172655e-01 -4.97996092e-01 5.66953897e-01
-1.00557935e+00 -6.20384634e-01 -3.35944265e-01 2.90844947e-01
-2.34958842e-01 -2.01256294e-02 -9.04527828e-02 -6.76432252e-01
-1.86297148e-01 -1.32254946e+00 1.45562708e+00 2.38030627e-01
-2.57760007e-02 -7.64216542e-01 -4.04401511e-01 -1.67899549e-01
4.26339269e-01 6.02218568e-01 9.16175246e-01 -4.47885603e-01
-3.16477656e-01 -6.35846019e-01 -7.94678092e-01 1.07619084e-01
6.36275470e-01 4.14885849e-01 -1.18843317e+00 -1.53511912e-01
-5.37949860e-01 -2.13017061e-01 8.82038951e-01 8.21246088e-01
1.11739445e+00 5.94639778e-01 -6.20884895e-01 6.19678259e-01
1.62070906e+00 2.41628155e-01 3.25046062e-01 6.06279790e-01
4.81147379e-01 9.95546639e-01 1.10252988e+00 4.05925840e-01
-1.47336274e-01 5.84721684e-01 9.77417886e-01 -6.43764079e-01
2.09871441e-01 2.98668116e-01 -1.88741803e-01 -3.22909713e-01
-2.58034259e-01 -3.77074689e-01 -1.10796022e+00 7.49004185e-01
-1.73177028e+00 -8.22346032e-01 -1.20282933e-01 2.22580743e+00
6.64719522e-01 -1.37364879e-01 -1.80507466e-01 1.99825749e-01
9.40028608e-01 2.08859757e-01 -7.34841347e-01 -2.64582373e-02
-4.24946666e-01 4.24897403e-01 1.30647290e+00 1.97601795e-01
-1.69235349e+00 1.35440946e+00 6.49517155e+00 6.61501825e-01
-1.10982585e+00 -5.26127815e-01 5.93516290e-01 3.97339880e-01
3.70117068e-01 -7.47867748e-02 -9.20874953e-01 -7.88450837e-02
6.47609353e-01 3.21082801e-01 2.11986020e-01 7.85983145e-01
2.70737529e-01 -4.42447871e-01 -8.05376649e-01 3.26333970e-01
-4.92219418e-01 -1.22184467e+00 -2.94800639e-01 3.21232416e-02
5.59689224e-01 3.30512039e-02 -1.29518345e-01 -5.62862121e-02
5.55501699e-01 -1.10750139e+00 4.91042703e-01 -1.53338328e-01
9.42840159e-01 -4.58786666e-01 9.67383385e-01 6.14524260e-02
-1.32180452e+00 -1.18694790e-01 -6.25913501e-01 2.26830393e-02
-1.70734942e-01 5.11547685e-01 -6.89421058e-01 8.41071784e-01
9.69643891e-01 7.78513908e-01 -9.58020091e-01 1.11908722e+00
-5.34236431e-02 4.92075890e-01 -5.10655284e-01 1.79854840e-01
3.91557485e-01 -3.80008928e-02 3.47750857e-02 1.31128347e+00
4.08112347e-01 -2.60019839e-01 9.94324908e-02 5.68391144e-01
3.03874254e-01 1.07556924e-01 -5.41387498e-01 -2.63349205e-01
5.67069411e-01 1.59563565e+00 -1.36641836e+00 -5.33712795e-03
-1.21447302e-01 7.18063712e-01 -3.70262116e-02 2.05000818e-01
-6.06176615e-01 -4.34854001e-01 5.87079108e-01 -1.66543171e-01
6.05795562e-01 -1.95335522e-01 -3.97143185e-01 -5.66865504e-01
-9.88001004e-02 -8.89232635e-01 3.38090025e-02 -6.24090791e-01
-8.99612963e-01 3.72054487e-01 1.65506616e-01 -9.12616432e-01
7.63771459e-02 -1.23276150e+00 -1.81727961e-01 1.11341202e+00
-1.77256966e+00 -1.48316073e+00 -1.06943989e+00 -1.90622032e-01
4.25443649e-01 -2.18698740e-01 1.57252789e+00 1.61397412e-01
-5.78900218e-01 1.30061552e-01 3.33282262e-01 -1.08148620e-01
7.60905564e-01 -1.09605491e+00 5.80376208e-01 7.42596388e-01
-1.37114421e-01 1.87381327e-01 6.75752521e-01 -5.24812460e-01
-1.08722663e+00 -1.40399015e+00 3.73645931e-01 -1.33938581e-01
7.69405663e-01 -3.09222430e-01 -7.33240664e-01 6.26893580e-01
-2.42801029e-02 1.36070430e-01 5.58151603e-01 9.20672417e-02
1.07312456e-01 -2.11042836e-01 -1.18953192e+00 2.96626210e-01
9.09650147e-01 -9.73757729e-02 2.63300925e-01 5.68823159e-01
5.41322589e-01 -5.59685111e-01 -8.09279859e-01 8.08712304e-01
6.34130239e-01 -7.01980054e-01 9.17301655e-01 -5.12175202e-01
5.53354204e-01 -3.58576208e-01 -4.72823322e-01 -1.37304592e+00
-5.15464664e-01 -2.36201152e-01 4.55928564e-01 1.29319513e+00
3.47749710e-01 -3.79615605e-01 7.89313018e-01 4.98642176e-02
-1.91461682e-01 -3.48256230e-01 -2.29610741e-01 -6.06925786e-01
8.74044076e-02 -2.05134898e-01 8.88968647e-01 8.05648148e-01
-5.69165170e-01 -2.27568820e-01 -7.38925636e-02 8.02440882e-01
1.53522000e-01 3.25440615e-01 7.67521560e-01 -1.45394480e+00
1.12715684e-01 -5.68215072e-01 -6.57842159e-01 -5.02224445e-01
5.14226593e-02 -6.12312913e-01 1.96743265e-01 -1.51116002e+00
-7.41193295e-02 -6.32285953e-01 5.30806221e-02 7.50737488e-01
-3.52841735e-01 2.21899539e-01 -1.13277085e-01 -1.97081491e-01
2.54028052e-01 -5.92142157e-02 1.03468573e+00 -2.14304805e-01
-3.56579453e-01 2.60814011e-01 -7.05910683e-01 7.09453046e-01
9.64786172e-01 -1.24744140e-01 -3.24728638e-01 -7.35545933e-01
-3.19100507e-02 -3.88444841e-01 5.72085798e-01 -8.16966176e-01
-2.67023861e-01 -3.30843687e-01 5.17095625e-01 -7.21101165e-01
1.19991258e-01 -1.18993068e+00 -2.48153452e-02 4.94550556e-01
-3.03000897e-01 -2.62874573e-01 8.84292841e-01 3.31109315e-01
3.13489623e-02 -3.20445389e-01 6.81716800e-01 -2.01931894e-01
-1.24175286e+00 2.38216594e-01 -1.81354582e-01 -3.40269089e-01
1.27229226e+00 -2.67174423e-01 -3.29174638e-01 2.38580167e-01
-6.07340097e-01 3.52120459e-01 4.81518924e-01 4.71830338e-01
-8.37123096e-02 -5.89970708e-01 -7.60315061e-01 3.31063628e-01
6.15033865e-01 3.56948137e-01 -1.51848763e-01 3.34302902e-01
-1.05554068e+00 4.97814208e-01 -6.84557080e-01 -8.07037890e-01
-1.23232901e+00 3.44620675e-01 3.41172814e-01 -2.44115517e-01
-4.63005662e-01 6.84886634e-01 1.96677178e-01 -4.76244122e-01
2.93053627e-01 -5.90974331e-01 -5.18889546e-01 4.44604963e-01
2.44239807e-01 1.23589568e-01 3.35034758e-01 -5.45485854e-01
-3.00095975e-01 3.13946813e-01 1.36278003e-01 3.47161859e-01
1.61691332e+00 2.93171257e-01 -1.23994939e-01 2.75258183e-01
4.38129127e-01 -4.10746008e-01 -1.32627594e+00 5.13772070e-02
4.58587021e-01 -4.38058347e-01 7.69523755e-02 -1.13928556e+00
-1.11799967e+00 8.69677484e-01 1.06277299e+00 3.16666722e-01
1.30797362e+00 -2.83365607e-01 -1.15915760e-01 5.38911462e-01
4.04022276e-01 -7.57368028e-01 -6.68130398e-01 2.94778585e-01
8.03491414e-01 -1.62131798e+00 1.88067243e-01 -6.84594393e-01
-3.36796552e-01 1.09240234e+00 4.58543569e-01 6.30562454e-02
5.21403372e-01 6.70142293e-01 3.38634402e-01 -2.49687895e-01
-4.48477343e-02 -4.85651791e-01 -1.82037190e-01 1.16011143e+00
6.46640480e-01 3.67140174e-01 1.97524466e-02 2.99920082e-01
-1.80877596e-01 5.96364625e-02 2.85792470e-01 1.17304718e+00
-5.76378584e-01 -1.02013361e+00 -3.59291971e-01 6.05973065e-01
-1.75498307e-01 -8.87932628e-02 -4.57721412e-01 1.09879410e+00
2.77968585e-01 6.49922907e-01 -1.08400382e-01 1.32675633e-01
5.09583414e-01 -3.14343303e-01 4.31377321e-01 -9.21378374e-01
-6.35096312e-01 2.16456410e-03 -1.93691079e-03 -2.07855314e-01
-7.07698107e-01 -6.83545232e-01 -7.09147751e-01 -9.00810361e-02
-7.53375709e-01 -2.52630323e-01 9.66979563e-01 5.99902630e-01
2.75338858e-01 8.93464684e-01 1.57655627e-01 -1.16941464e+00
-3.69791627e-01 -9.88005459e-01 -9.04105544e-01 -1.41486347e-01
2.87730545e-01 -7.08051920e-01 7.05634505e-02 1.87437847e-01] | [9.126935958862305, -1.5576788187026978] |
48360c01-716d-4afb-b46b-9b56bf00e85b | equitability-analysis-of-the-maximal | 1301.6314 | null | http://arxiv.org/abs/1301.6314v2 | http://arxiv.org/pdf/1301.6314v2.pdf | Equitability Analysis of the Maximal Information Coefficient, with Comparisons | A measure of dependence is said to be equitable if it gives similar scores to
equally noisy relationships of different types. Equitability is important in
data exploration when the goal is to identify a relatively small set of
strongest associations within a dataset as opposed to finding as many non-zero
associations as possible, which often are too many to sift through. Thus an
equitable statistic, such as the maximal information coefficient (MIC), can be
useful for analyzing high-dimensional data sets. Here, we explore both
equitability and the properties of MIC, and discuss several aspects of the
theory and practice of MIC. We begin by presenting an intuition behind the
equitability of MIC through the exploration of the maximization and
normalization steps in its definition. We then examine the speed and optimality
of the approximation algorithm used to compute MIC, and suggest some directions
for improving both. Finally, we demonstrate in a range of noise models and
sample sizes that MIC is more equitable than natural alternatives, such as
mutual information estimation and distance correlation. | ['Michael Mitzenmacher', 'Yakir Reshef', 'David Reshef', 'Pardis Sabeti'] | 2013-01-27 | null | null | null | null | ['mutual-information-estimation'] | ['methodology'] | [ 1.17260844e-01 -7.89518207e-02 -1.71254486e-01 -4.56440985e-01
-6.01919830e-01 -5.48855543e-01 4.13289398e-01 6.50545597e-01
-5.82244456e-01 1.05315042e+00 3.40870917e-01 -3.68550122e-01
-1.02969217e+00 -7.26243556e-01 -2.89366096e-01 -6.95279896e-01
-8.39292943e-01 6.73210263e-01 -4.74973992e-02 1.11034095e-01
2.96804219e-01 3.37353081e-01 -1.54276264e+00 -2.16922313e-01
8.36712241e-01 6.99856043e-01 1.42465055e-01 4.94539797e-01
1.15585163e-01 2.94976294e-01 -5.33775568e-01 -4.84062552e-01
7.00281188e-02 -6.49545789e-01 -8.73349190e-01 -3.76589507e-01
-8.50111619e-02 1.76199958e-01 1.94130391e-01 1.14628196e+00
4.44245458e-01 1.64646402e-01 1.09202921e+00 -1.33756709e+00
-2.88105339e-01 7.89547503e-01 -7.14038014e-01 4.52182800e-01
2.12134689e-01 -1.29352257e-01 1.38630867e+00 -7.80070007e-01
5.86905718e-01 1.27487075e+00 7.86785483e-01 -6.86888099e-02
-1.68456769e+00 -4.82556552e-01 -1.57640219e-01 -9.79215354e-02
-1.59245420e+00 -1.50507376e-01 2.52541065e-01 -3.89779091e-01
5.56651950e-01 7.96307027e-01 3.96686375e-01 5.35214782e-01
7.20032975e-02 4.52513218e-01 1.08197200e+00 -2.85009801e-01
3.23440492e-01 5.79622202e-02 3.73182267e-01 9.94600430e-02
8.00611556e-01 -4.04253276e-03 -5.04358232e-01 -6.89879060e-01
6.87712133e-01 -2.66930670e-01 -2.56289274e-01 -4.75911558e-01
-1.29862261e+00 8.76637995e-01 2.52522558e-01 3.31035137e-01
-2.43959948e-01 -7.95790479e-02 2.86203206e-01 4.13308471e-01
3.10069352e-01 1.10357368e+00 -3.34293723e-01 -2.60024220e-01
-6.06001019e-01 3.71088743e-01 8.86330903e-01 5.03580093e-01
6.81622624e-01 -7.00570941e-01 9.57581252e-02 1.03756213e+00
1.36624366e-01 2.29087651e-01 1.10243805e-01 -8.61053765e-01
2.35937402e-01 2.64281243e-01 8.99245888e-02 -1.19722831e+00
-6.34880304e-01 -3.35083932e-01 -8.99992883e-01 -2.56996248e-02
6.65853083e-01 -1.98212370e-01 -2.90947199e-01 2.31533575e+00
2.60522097e-01 1.13654463e-02 -8.98243859e-02 7.17057407e-01
3.26048255e-01 3.98934215e-01 1.16622463e-01 -6.41952634e-01
1.01399493e+00 2.64305510e-02 -6.98526561e-01 2.18678103e-03
8.41881096e-01 -7.63558865e-01 8.95656705e-01 1.39374211e-01
-1.00499582e+00 -6.64397329e-02 -8.48071992e-01 1.83896556e-01
-2.31991112e-01 -6.19807184e-01 9.06940937e-01 6.38237298e-01
-8.12798381e-01 7.39193738e-01 -7.13875651e-01 -4.66751158e-01
2.24780858e-01 6.47465527e-01 -3.22602361e-01 1.83139890e-01
-1.07356393e+00 8.18008721e-01 3.18904549e-01 -2.18718112e-01
-1.24482319e-01 -5.03754079e-01 -5.30978203e-01 2.71146987e-02
3.01990479e-01 -4.50605661e-01 5.33179283e-01 -5.10161400e-01
-4.15101945e-01 6.82062089e-01 -2.52600551e-01 -1.48983479e-01
1.73584417e-01 -7.33897649e-03 -3.32405895e-01 -3.14982891e-01
2.80222893e-01 4.03631866e-01 -1.39531299e-01 -1.10709965e+00
-4.99929637e-01 -5.07723153e-01 -1.08204469e-01 3.53444517e-01
-4.17352855e-01 1.75525486e-01 -2.43507758e-01 -6.49494827e-01
3.65074903e-01 -8.82389843e-01 -4.74382937e-01 3.48635321e-03
-5.78044236e-01 -1.60354063e-01 4.12614465e-01 -2.09481955e-01
1.47428048e+00 -2.03716660e+00 2.29359686e-01 9.43319976e-01
3.67007643e-01 -2.84682423e-01 -1.92746878e-01 5.85625231e-01
-3.00052822e-01 3.40296119e-01 -4.37543839e-01 1.78053409e-01
-2.89230555e-01 1.94957152e-01 7.64318258e-02 7.36226380e-01
1.56914517e-01 6.55017316e-01 -1.05746818e+00 -4.40140784e-01
-1.50660202e-01 2.22988576e-01 -4.85267133e-01 7.93460757e-02
1.84595793e-01 1.51169688e-01 -3.32528651e-01 5.25484324e-01
6.24929845e-01 -6.30437672e-01 5.97021043e-01 -6.53130934e-02
-2.10079268e-01 4.23830569e-01 -1.48475838e+00 9.58429217e-01
-5.38695641e-02 7.39588916e-01 -2.67289102e-01 -9.92133737e-01
1.06356919e+00 6.37461841e-02 6.90363228e-01 -4.78251606e-01
9.95062143e-02 3.60478073e-01 4.01049793e-01 -3.83075446e-01
2.41364643e-01 -1.91909581e-01 -6.75016418e-02 1.08513534e+00
-5.25093079e-01 5.24326377e-02 2.97153443e-01 2.17960387e-01
1.29909909e+00 -5.38622439e-01 6.04982674e-01 -8.56629133e-01
-6.43330887e-02 -1.65102720e-01 5.09211361e-01 8.65810931e-01
4.74456400e-02 4.21859890e-01 8.18724990e-01 -1.82593539e-01
-1.14054215e+00 -1.23872566e+00 -8.18735063e-01 5.88171184e-01
2.65435785e-01 -3.91204536e-01 -3.07878375e-01 -3.90603274e-01
2.44380817e-01 3.96715045e-01 -9.34887767e-01 6.14941418e-02
-2.13258103e-01 -1.57156301e+00 4.03064460e-01 3.77809256e-01
4.57716882e-02 -4.89746451e-01 -4.71115768e-01 -7.71606788e-02
-2.47023404e-01 -4.11337852e-01 -1.48588598e-01 6.52444482e-01
-8.83356035e-01 -1.19038618e+00 -3.84197801e-01 -5.18823028e-01
6.70704663e-01 2.92036593e-01 1.16321790e+00 -6.47214726e-02
-1.35854855e-01 -3.96029912e-02 -1.85948208e-01 -2.43793964e-01
-2.03513697e-01 -2.68515795e-01 4.60060775e-01 -2.60251135e-01
7.94161618e-01 -7.88021266e-01 -3.65638286e-01 5.19468069e-01
-8.75709236e-01 -3.39926869e-01 5.22004366e-01 9.50901270e-01
6.95345342e-01 1.89491138e-01 6.53692961e-01 -9.39689517e-01
1.08219540e+00 -9.45985496e-01 -3.00608933e-01 2.53097177e-01
-7.61753500e-01 3.58051926e-01 2.33020857e-01 -5.20414531e-01
-5.29564381e-01 -3.05376232e-01 5.16379662e-02 1.83957338e-01
2.55902231e-01 7.29555130e-01 -8.33309069e-02 1.30023316e-01
8.97147000e-01 -3.24682415e-01 1.66226119e-01 -4.51035917e-01
2.00840399e-01 7.76499331e-01 3.29574525e-01 -6.07310534e-01
4.52541202e-01 4.27061021e-01 3.52277130e-01 -8.73673081e-01
-5.42055786e-01 -4.48045015e-01 -4.83850867e-01 4.50476110e-02
4.36382234e-01 -4.49910015e-01 -9.04115260e-01 -3.60009517e-03
-7.72468150e-01 -4.66803499e-02 -4.23043191e-01 6.52399182e-01
-5.86859941e-01 1.97139680e-01 -2.23288521e-01 -8.66500854e-01
2.47898281e-01 -1.08116269e+00 6.48834169e-01 -3.97114381e-02
-1.03497934e+00 -1.13564563e+00 5.14013588e-01 -1.70064345e-01
9.55290869e-02 3.68997514e-01 1.11324275e+00 -9.16743875e-01
-1.22293770e-01 -9.29008946e-02 -2.90696979e-01 -2.27777541e-01
5.97635865e-01 -6.73006997e-02 -6.58687532e-01 -2.92484224e-01
-1.50028393e-01 -3.38086277e-01 7.66236961e-01 6.20430052e-01
1.13044167e+00 -2.55611718e-01 -6.06467664e-01 5.64527690e-01
1.34701777e+00 2.96909779e-01 4.69068915e-01 2.95919985e-01
3.37229311e-01 1.04819143e+00 8.04098904e-01 5.85780680e-01
1.32325545e-01 7.15495110e-01 1.83047473e-01 -1.43409669e-01
3.08237016e-01 -4.35090661e-02 -3.68398547e-01 7.16948509e-01
-6.08822377e-03 -3.26061994e-01 -9.24618125e-01 6.08617365e-01
-1.85303116e+00 -8.97805691e-01 -3.02635014e-01 2.66093612e+00
9.88945663e-01 6.61505163e-02 3.96742761e-01 2.95670420e-01
7.90632248e-01 -2.07534164e-01 -6.25555396e-01 -4.35529828e-01
-5.17128825e-01 -5.24647944e-02 3.82877469e-01 6.99264884e-01
-9.32977498e-01 2.60817558e-01 7.79506540e+00 6.55061603e-01
-6.04713261e-01 -3.10069770e-01 1.21013653e+00 -1.18372738e-01
-6.55590832e-01 2.69459514e-03 -4.06753123e-01 5.14214516e-01
9.19588387e-01 -3.62875074e-01 1.05472639e-01 5.12216151e-01
2.84128100e-01 -4.64253753e-01 -1.35176933e+00 8.67896497e-01
-4.62058514e-01 -1.17438865e+00 -2.84377098e-01 3.28158885e-01
7.13377655e-01 -5.96891232e-02 9.89533588e-02 -5.99449098e-01
6.52660489e-01 -1.31206477e+00 1.69423118e-01 3.44782710e-01
7.24318743e-01 -1.01111007e+00 6.85744226e-01 1.40740156e-01
-9.79258120e-01 -4.15799320e-02 -4.84976232e-01 -2.35100463e-01
-2.76149753e-02 1.02927232e+00 -7.24683464e-01 1.54430956e-01
7.58711278e-01 5.66178679e-01 -3.03378880e-01 1.21046770e+00
3.62063318e-01 2.31480837e-01 -8.61038625e-01 -3.83914143e-01
3.68083268e-02 -2.36741662e-01 4.80906397e-01 1.23323250e+00
3.83530885e-01 8.41548964e-02 -2.64944106e-01 7.79757738e-01
1.40758455e-01 2.46604547e-01 -8.14283729e-01 -3.64731178e-02
9.91378665e-01 7.40255713e-01 -8.14361989e-01 2.73110233e-02
-3.34871829e-01 4.59287405e-01 4.48770434e-01 2.14921936e-01
-3.35366189e-01 -3.95171553e-01 9.92698252e-01 8.93957615e-02
-8.44552666e-02 -1.04101837e-01 -6.64025784e-01 -6.66061997e-01
-1.45659074e-01 -9.49531317e-01 8.29594016e-01 -3.51672351e-01
-1.50810897e+00 3.67147118e-01 2.67549902e-01 -1.31517100e+00
-3.21618348e-01 -4.36655462e-01 -3.46861899e-01 1.00563025e+00
-8.09338748e-01 -1.85096890e-01 8.55360106e-02 2.09215671e-01
-2.56996253e-03 2.89913416e-01 7.94119596e-01 7.09727593e-03
-3.95582259e-01 6.48463130e-01 5.87371409e-01 -1.75622866e-01
6.14933074e-01 -1.17670739e+00 3.40817720e-01 6.59459591e-01
8.79898965e-02 1.08673942e+00 1.07724035e+00 -8.73840690e-01
-1.09001088e+00 -6.17807269e-01 8.75700712e-01 -3.41938049e-01
8.13557744e-01 -7.22737387e-02 -8.95939052e-01 4.46946830e-01
-2.81767994e-01 -3.47699493e-01 1.11410832e+00 7.17392385e-01
-3.52285355e-01 -5.20452671e-03 -1.14989793e+00 9.06767905e-01
1.12343252e+00 -1.07629329e-01 -1.91217631e-01 2.93536514e-01
2.32406244e-01 2.53123015e-01 -1.14587939e+00 5.89923322e-01
8.42008770e-01 -1.17201710e+00 9.82482314e-01 -6.61832035e-01
1.98268786e-01 -2.36990198e-01 -4.39763933e-01 -1.26225269e+00
-6.15267038e-01 -5.44055283e-01 3.03590000e-01 8.66287351e-01
6.95066571e-01 -6.00370407e-01 6.67865217e-01 8.08467090e-01
4.94883865e-01 -1.05569518e+00 -8.41573596e-01 -1.07326138e+00
3.28937508e-02 -4.71591175e-01 6.26637042e-01 1.29007757e+00
5.12728393e-01 3.07892174e-01 -3.51343632e-01 5.50543249e-04
9.14289474e-01 1.67838156e-01 6.59229815e-01 -1.44533336e+00
-5.44762075e-01 -5.03075004e-01 -5.79181910e-01 -9.87266302e-01
-4.28512216e-01 -7.04575181e-01 -5.33415228e-02 -1.06866717e+00
7.22291529e-01 -1.02597821e+00 -3.70678306e-01 1.92744821e-01
-3.69219154e-01 2.87347585e-01 -2.82447368e-01 4.66963321e-01
-4.18738276e-01 6.33504838e-02 9.84127581e-01 1.86734691e-01
-4.52579081e-01 -3.40883769e-02 -1.17474711e+00 5.89786112e-01
7.08882153e-01 -5.59014678e-01 -5.89235783e-01 -1.54391035e-01
6.30245984e-01 -3.35777848e-04 -6.83401898e-02 -5.41568398e-01
1.61658123e-01 -4.65413481e-01 5.00899017e-01 -3.96903247e-01
3.57489944e-01 -5.18703938e-01 5.83043754e-01 4.46245223e-01
-6.13714874e-01 3.01490813e-01 -1.22943018e-02 4.26797241e-01
-5.11025935e-02 -2.79437989e-01 8.14261436e-01 -4.17447463e-02
-2.45966569e-01 7.89759010e-02 -1.85098171e-01 1.66102365e-01
9.89780784e-01 -1.53365642e-01 -4.04146105e-01 -5.16721725e-01
-5.91466069e-01 2.93871731e-01 5.63796341e-01 1.31644949e-01
5.90749741e-01 -1.23386431e+00 -7.98578441e-01 8.08009356e-02
1.54691875e-01 -3.88923645e-01 -2.01132834e-01 1.02653885e+00
-2.27590337e-01 2.13659972e-01 2.64978013e-03 -5.96611440e-01
-1.46332109e+00 4.27296907e-01 9.88309234e-02 -2.41431564e-01
-2.64186889e-01 9.22377527e-01 4.55472916e-01 -9.45227891e-02
1.72896117e-01 1.04307562e-01 -1.09143451e-01 1.23889662e-01
7.81056821e-01 5.36631882e-01 -2.12604448e-01 -4.79286164e-01
-7.10582674e-01 5.26243687e-01 3.15295644e-02 -2.09498167e-01
1.34682310e+00 -2.56750166e-01 -5.95453560e-01 5.79295754e-01
1.27008855e+00 -2.65549086e-02 -8.63774419e-01 -6.66103438e-02
3.43135089e-01 -7.57623971e-01 -2.12405160e-01 -5.35823882e-01
-5.36229908e-01 3.46305549e-01 4.80191946e-01 6.76317155e-01
1.10269666e+00 3.51479143e-01 8.38617384e-02 5.23845255e-01
1.84359267e-01 -8.06013107e-01 -1.23505123e-01 8.27203020e-02
5.31784892e-01 -1.24554741e+00 2.97252387e-01 -5.41122079e-01
-4.21243697e-01 8.80404770e-01 2.66117036e-01 1.89835295e-01
9.01992798e-01 6.12752140e-01 -1.03193231e-01 -2.75676280e-01
-7.45903730e-01 -2.48316601e-01 1.10822462e-01 6.31060779e-01
6.40280545e-01 1.65271610e-01 -8.88537347e-01 6.64652362e-02
-3.32034737e-01 -5.11870801e-01 3.04970831e-01 7.83132017e-01
-4.74663526e-01 -1.18293202e+00 -3.78285617e-01 1.02959096e+00
-5.52597880e-01 -2.60177046e-01 -6.06837809e-01 6.75882518e-01
-7.00504035e-02 1.05299520e+00 3.92854750e-01 -5.14352977e-01
-6.55215904e-02 -3.22759688e-01 2.41080165e-01 -3.82168591e-01
-6.26581386e-02 1.02503777e-01 3.95290315e-01 -4.35319006e-01
-5.06830633e-01 -1.00793815e+00 -9.03229177e-01 -5.62932670e-01
-4.19591069e-01 4.99186248e-01 4.21826571e-01 9.34043586e-01
1.35546848e-01 -2.52169259e-02 6.76002443e-01 -1.91334918e-01
-2.52667874e-01 -7.05127239e-01 -8.66766274e-01 4.71745908e-01
2.75178581e-01 -6.65565491e-01 -5.39118707e-01 -3.80665779e-01] | [7.628860950469971, 4.672356128692627] |
9cf2db4f-8ea2-43b9-bd19-0a8ff83f490a | dynamic-semantic-graph-construction-and | 2105.11776 | null | https://arxiv.org/abs/2105.11776v1 | https://arxiv.org/pdf/2105.11776v1.pdf | Dynamic Semantic Graph Construction and Reasoning for Explainable Multi-hop Science Question Answering | Knowledge retrieval and reasoning are two key stages in multi-hop question answering (QA) at web scale. Existing approaches suffer from low confidence when retrieving evidence facts to fill the knowledge gap and lack transparent reasoning process. In this paper, we propose a new framework to exploit more valid facts while obtaining explainability for multi-hop QA by dynamically constructing a semantic graph and reasoning over it. We employ Abstract Meaning Representation (AMR) as semantic graph representation. Our framework contains three new ideas: (a) {\tt AMR-SG}, an AMR-based Semantic Graph, constructed by candidate fact AMRs to uncover any hop relations among question, answer and multiple facts. (b) A novel path-based fact analytics approach exploiting {\tt AMR-SG} to extract active facts from a large fact pool to answer questions. (c) A fact-level relation modeling leveraging graph convolution network (GCN) to guide the reasoning process. Results on two scientific multi-hop QA datasets show that we can surpass recent approaches including those using additional knowledge graphs while maintaining high explainability on OpenBookQA and achieve a new state-of-the-art result on ARC-Challenge in a computationally practicable setting. | ['Wai Lam', 'Deng Cai', 'Huihui Zhang', 'Weiwen Xu'] | 2021-05-25 | null | https://aclanthology.org/2021.findings-acl.90 | https://aclanthology.org/2021.findings-acl.90.pdf | findings-acl-2021-8 | ['multi-hop-question-answering', 'science-question-answering'] | ['knowledge-base', 'miscellaneous'] | [-0.14973214 0.75143665 -0.30149794 -0.27714568 -1.1452833 -0.771951
0.40973285 0.6298919 0.06550033 0.9194366 0.48534396 -0.8120061
-0.89833635 -1.4470533 -1.1659086 0.10794199 -0.10998552 0.7947577
0.731428 -0.6952317 0.21023792 0.17912154 -1.3150518 0.67096126
1.3165418 1.0633367 -0.09192512 0.73450124 -0.8298721 1.3550189
-0.4487351 -0.92406243 -0.1049577 -0.259523 -2.0438166 -0.40077534
0.52006924 0.1355476 -0.52577394 1.0603137 -0.06718107 0.20101401
0.20208302 -1.1753255 -1.171272 1.0201321 -0.09227076 0.6690619
0.96665245 -0.03718129 1.5859262 -0.4502867 1.0845513 1.289055
0.4538701 0.28744927 -0.64530927 -0.06733804 0.24372156 0.96534187
-1.1131418 0.21073598 0.6005104 0.08184448 1.2708867 0.64894265
0.7536467 0.7487392 -0.03157444 0.6345838 0.99643165 -0.2561316
0.29607543 -0.21200238 0.7643665 1.3199958 0.3422505 -0.5240423
-0.96784556 -0.33806568 0.4505764 -0.2549351 -0.3801364 0.02468121
-0.9968492 0.92087066 0.90623504 0.25930226 -0.5615999 0.34641215
0.05564119 0.4967587 0.1971079 0.6882063 -0.6918398 0.02109619
-0.43468127 0.362224 1.1750488 1.0072132 0.85708183 -0.59438336
-0.42839587 0.41728526 0.16026944 0.14484172 0.13064365 -1.0172529
0.8215835 1.3244085 -0.16832234 -1.0934832 -0.46743008 -0.4414093
-0.30389214 -0.4968514 0.35056543 0.31545517 -0.8037851 1.3261511
0.7491996 0.33113968 0.38280436 0.92531556 1.4226989 0.38716015
0.05642778 0.37998596 1.9991095 -1.1541127 -0.7065375 -0.16501948
0.6750294 -0.15144342 1.3301022 0.37095264 -1.1544042 -0.10023373
-1.0352187 -0.68219167 -0.9647893 -0.48027453 1.3438288 0.30451298
-1.1712663 0.44108787 -0.19597368 -0.2896802 0.7336545 0.10649201
-0.37940517 -0.76315707 -1.7180796 0.74474937 0.6777321 -0.11021455
-0.8203392 -0.9398179 -0.9014469 0.54297125 1.315242 -1.3992692
0.9149972 -0.03868224 -1.1463159 0.49479544 -0.05136729 -0.6413227
-0.02148833 -0.16517907 -0.85377365 0.7480531 0.20241982 0.2955267
0.3204804 -1.1138563 -0.4824473 -0.48134235 0.8600974 0.08936272
0.05231866 -0.45491952 -0.5818284 -0.12254692 0.37101984 -0.38512108
-0.01714462 -0.28060773 -0.5670256 -0.7714209 0.55927336 -0.8961475
1.2297527 -1.2112048 0.11350095 0.359524 0.80957747 0.07146507
-0.19862469 0.6652212 0.24270849 0.17813002 -0.24687351 0.4729068
0.01798879 0.6152117 -0.5700381 -0.19469988 0.29169956 1.5405613
-1.3814415 -0.64953214 -0.25866988 0.15082984 -0.72751164 -0.11506127
-1.0062656 0.08084009 -0.9775344 0.97814286 0.512101 -1.0313581
0.24694118 -0.27838087 0.5207422 0.7382981 -0.90232444 2.161402
-0.39904305 0.29135758 -0.19271968 -0.96715933 0.68394876 0.1514012
0.14458604 -0.7492547 -0.26367387 0.15850326 -0.20787244 -0.795553
0.59572965 0.26431742 0.09773567 0.19275086 0.42087078 -0.09561683
0.4822996 1.0045068 1.680186 -0.03886112 0.11298572 -0.2848447
0.7226875 0.55089164 -0.06422332 0.8527604 0.26838258 0.03140702
0.8300235 -0.6941235 -0.23993325 -1.1853701 0.2434548 0.78438854
0.34259045 -0.8376139 -0.49966717 -1.3344216 0.16725783 1.0191952
-0.5764933 -0.18269452 -0.42772064 -0.26906297 0.63094866 0.34981334
0.72293603 -0.90101916 -0.15685001 0.06503084 -0.77588546 -1.3760631
0.07320591 -0.26869714 -0.82092553 -1.6427624 0.03281253 -0.29253173
0.5255104 0.32117942 1.7350708 0.5280283 -0.3471811 0.8840632
-0.46267867 -0.32698193 0.02454183 -0.06685232 -0.5634649 -0.36126837
0.44328558 -0.633292 -0.95813775 -0.00885407 -1.0973253 -0.21287428
0.46735665 0.47071248 0.7523486 0.27751997 0.71863765 -1.1006241
0.9278887 -1.0270426 -0.4673919 0.8008374 -1.039584 0.4650561
0.59684247 0.16876438 -1.1830405 -0.825872 -0.08594937 -0.02559299
0.15557447 0.9591367 0.04833295 0.02200459 0.7587809 0.02925546
-0.47711873 -0.3783912 1.1675457 0.19234282 0.6721158 -0.96340185
0.75866675 0.7512444 0.42104495 -0.29139405 -1.5526179 -0.62445873
-0.13012546 -0.05309446 0.9252794 -0.7616759 -1.397489 -0.5911125
-1.1347078 0.09302238 -0.4928421 0.22541602 -0.3349448 0.48415774
-0.5287787 -0.41828817 -0.6879237 -0.52454436 1.1107342 0.39976156
0.10705212 -1.1404424 0.02333211 1.3721722 0.2455835 0.1866238
1.3021075 -0.7976344 -1.2543716 -0.10429542 -0.5802356 -0.27160135
-0.19911942 -0.57478595 -0.7058283 0.244373 -0.36549032 -0.41154885
0.92503643 -0.02538398 1.5847645 -0.7256475 -0.29616764 0.2499386
1.5078884 -0.38375556 0.66205746 0.3570355 0.7266628 0.5163235
0.4098614 0.02177449 1.1596395 0.20867732 0.65185535 0.27487212
-0.51070815 -0.5821683 -0.1170155 0.76819813 -0.2605399 -0.44149467
-1.1016603 0.735541 -1.9732877 -0.8606648 -0.6864031 1.6964414
0.7965735 0.12886311 -0.22365628 -0.1304938 0.24471018 0.05537111
-0.45213494 -0.11822592 -0.25932318 0.69708383 0.3071002 0.69759935
-0.3760615 1.087556 5.333355 0.8939157 -0.1843874 0.21454282
0.10828017 0.20436017 -1.074715 0.5115308 -0.49445364 -0.05622837
0.8253874 -0.24130271 0.72714126 0.67724246 -0.5758142 -0.23607957
-0.8353023 0.6583369 0.0477433 -2.133153 0.6651495 -0.2322364
0.61519426 -0.09324615 -0.38059074 0.49410897 0.8428333 -1.0684183
0.16944204 0.8998656 0.5154883 -0.5289421 0.6694103 0.1224748
-1.3721448 -0.11968739 -0.3794514 0.12787926 0.05113216 0.7923163
-0.87832403 1.8372154 0.8925115 0.41735366 -0.7994351 0.59924555
-0.6503412 0.8129591 -0.21190701 -0.11970647 0.46108955 0.06004468
0.5331891 0.81629324 0.22184768 0.36630034 -0.1842604 1.1333627
-0.5618535 -0.0262952 -0.41128778 -0.31363726 0.3697297 1.3809214
-0.79769707 -0.5757041 -0.49292812 0.9176349 0.883536 0.43682355
-0.7233377 -0.32391417 0.3414407 -0.05560342 0.22358647 -0.08323631
-0.03466618 -1.2533872 0.34376222 -0.67609537 1.4265815 -1.1538616
-1.4400178 0.55598724 -0.01749063 -0.43559954 0.11896708 -0.6160877
-0.4677949 0.7951354 -2.0940185 -1.2922382 -0.64682573 1.0887108
0.34362978 0.14078495 0.7501596 0.06040217 -0.03942454 0.12400461
-0.64916676 -0.20195843 0.0101222 -1.6078686 0.24278072 0.63469994
0.7348446 0.7649004 0.35986277 -0.912366 -1.923771 -0.86512387
0.85281897 -0.94412315 0.8843649 -0.04137063 -1.1749877 0.81637794
0.19489206 0.1999779 0.6591315 0.5157106 -0.7187733 0.06472544
-1.1530672 0.59669256 1.5230105 -0.6218374 -1.0980681 0.69916654
1.4869331 -0.37935126 -1.1979159 0.29658136 0.0695421 -0.73408216
1.2033556 -1.2015246 0.61131257 -0.4563152 -0.08712877 -0.98006874
-0.09125136 -0.76045924 -1.0090555 0.80801743 0.58873 -0.7241211
0.7993794 0.42898923 -0.2269828 -0.96165603 -1.0991361 -0.7109396
-0.19111097 -0.5554503 1.0274067 1.0533059 0.21955 0.69308436
0.2171174 0.7257682 0.5820815 0.5084575 0.34189066 -1.3299605
-0.24781881 -0.20026866 -0.15865017 -0.93181026 -0.01038773 -1.5090655
-0.6547628 -2.513633 0.13545628 -0.17362353 -0.3547452 0.4671256
-0.4653285 -0.43346533 -0.23589467 -0.08608034 -1.1526992 0.44837666
1.5510483 -0.14895228 0.37352568 -0.47640815 -0.99802816 0.44974026
0.48930165 -0.45315167 -0.91008717 -0.48733187 1.1050739 0.34967425
0.7767783 -0.7114183 0.53068644 -0.12650892 0.0539625 -0.5290271
0.14791225 -0.64535755 -0.17171283 0.27448493 -0.2983184 0.03529194
0.1640003 1.0662235 -0.31539574 -0.04445224 -0.22178215 -0.46040782
-0.93696856 0.40594882 0.25259432 0.5373433 0.48553875 0.29116932
-1.1278409 -0.47084662 -0.584994 0.6825587 -0.46286827 0.40379322
0.74614316 -1.149467 -0.61203504 -0.4446535 0.41740972 0.23538989
0.7480106 0.71770054 -0.5642048 0.7776051 0.1907847 -0.11561434
-0.85006857 0.8496591 0.07001772 -0.9593574 -0.8053392 0.94611895
-0.3229493 -0.57660705 -0.27991712 -0.43655357 -0.23716651 -0.3323197
0.3639011 0.5812743 0.24304564 0.17342429 -0.38531032 0.19974512
0.07252914 0.16964236 1.3326409 -0.14818731 -0.38847837 -0.03604404
0.7427081 0.16059744 -0.6020983 -0.43221796 0.34260038 -0.41650167
-0.09549888 -1.3662285 -0.7630625 0.55018914 -0.05153931 0.7187751
0.9241715 0.7681672 1.1645374 0.89083654 0.4345219 -1.0021518
0.36975574 0.37612113 1.1769546 -1.14468 -0.01310966 -0.94047594
-0.50834435 1.1720251 0.50890905 0.22164516 0.321127 -0.43801114
-0.29630786 -1.4857633 -0.85066724 -0.6603064 0.7391387 0.46651554
-0.05886689 0.01558647 -0.22765638 0.85757947 -0.47832906 0.05554678
0.23967361 0.7764597 -0.43432623 -0.8125444 0.04542427 0.59225965
-0.303819 -0.36051527 -0.5018254 0.8239834 -0.09338094 1.2403693
-0.36774066 -0.25025445 0.39552653 0.43024185 0.57184494 -0.41001564
-0.48822567 -0.98123145 0.6096832 -1.0313746 -0.2736983 0.07749187
-1.6786047 -0.3159197 -0.18405522 0.45026675 0.48669615 1.1577652
0.83998257 0.9059586 -0.05259649 0.7192017 -0.21684314 -0.6483852
-0.39172515 0.4860584 0.15054695 -0.5398127 -0.03682924 -0.28296754] | [10.468173027038574, 7.892541885375977] |
b888ee7e-9d75-4291-86a6-575e141809aa | animal-kingdom-a-large-and-diverse-dataset | 2204.08129 | null | https://arxiv.org/abs/2204.08129v2 | https://arxiv.org/pdf/2204.08129v2.pdf | Animal Kingdom: A Large and Diverse Dataset for Animal Behavior Understanding | Understanding animals' behaviors is significant for a wide range of applications. However, existing animal behavior datasets have limitations in multiple aspects, including limited numbers of animal classes, data samples and provided tasks, and also limited variations in environmental conditions and viewpoints. To address these limitations, we create a large and diverse dataset, Animal Kingdom, that provides multiple annotated tasks to enable a more thorough understanding of natural animal behaviors. The wild animal footages used in our dataset record different times of the day in extensive range of environments containing variations in backgrounds, viewpoints, illumination and weather conditions. More specifically, our dataset contains 50 hours of annotated videos to localize relevant animal behavior segments in long videos for the video grounding task, 30K video sequences for the fine-grained multi-label action recognition task, and 33K frames for the pose estimation task, which correspond to a diverse range of animals with 850 species across 6 major animal classes. Such a challenging and comprehensive dataset shall be able to facilitate the community to develop, adapt, and evaluate various types of advanced methods for animal behavior analysis. Moreover, we propose a Collaborative Action Recognition (CARe) model that learns general and specific features for action recognition with unseen new animals. This method achieves promising performance in our experiments. Our dataset can be found at https://sutdcv.github.io/Animal-Kingdom. | ['Jun Liu', 'Si Yong Yeo', 'Yun Ni', 'Qichen Zheng', 'Kian Eng Ong', 'Xun Long Ng'] | 2022-04-18 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Ng_Animal_Kingdom_A_Large_and_Diverse_Dataset_for_Animal_Behavior_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Ng_Animal_Kingdom_A_Large_and_Diverse_Dataset_for_Animal_Behavior_CVPR_2022_paper.pdf | cvpr-2022-1 | ['video-grounding', 'animal-pose-estimation', 'animal-action-recognition'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 1.47762254e-01 -7.84712791e-01 -9.85144749e-02 -5.97245097e-01
-3.06724072e-01 -6.73681140e-01 2.04718083e-01 -2.06372932e-01
-4.60884243e-01 7.10671186e-01 7.04780892e-02 5.07675588e-01
2.70396173e-02 -4.52598363e-01 -8.13915014e-01 -8.39817524e-01
-4.90409255e-01 1.40505910e-01 5.16536176e-01 -1.62398573e-02
8.39337409e-02 6.01735771e-01 -1.73627412e+00 4.49462682e-01
2.78924435e-01 8.30577850e-01 4.57264841e-01 7.66590118e-01
3.93083274e-01 8.13703418e-01 -5.04475832e-01 -2.57241100e-01
2.71215737e-01 -4.34325516e-01 -7.36078143e-01 2.64044672e-01
5.72274804e-01 -6.48390234e-01 -3.58385801e-01 9.46015179e-01
1.66694313e-01 4.28066671e-01 5.21995246e-01 -1.65977502e+00
-5.96351206e-01 2.65109479e-01 -5.59460104e-01 5.34207761e-01
4.22433197e-01 4.46630627e-01 7.16242969e-01 -3.27423692e-01
6.56868517e-01 1.09054744e+00 6.59634531e-01 6.50000453e-01
-9.85650063e-01 -9.45760608e-01 2.16690361e-01 6.05647087e-01
-1.16075516e+00 -5.13503850e-01 5.40792584e-01 -6.42954826e-01
7.62314260e-01 1.13067003e-02 7.78883159e-01 1.42082143e+00
1.82123244e-01 7.91500270e-01 9.22605276e-01 3.39694858e-01
1.07461683e-01 -2.55141884e-01 -9.04361904e-02 7.73400903e-01
2.29442269e-01 -1.96311064e-02 -4.82621968e-01 -2.39719406e-01
7.09925175e-01 6.19276643e-01 -2.77312398e-01 -7.57099986e-01
-1.56837249e+00 5.72957695e-01 4.34761852e-01 -4.44667339e-02
-3.24036151e-01 2.44150952e-01 5.77042878e-01 2.75094122e-01
-1.68735739e-02 2.08847478e-01 -8.50950241e-01 -2.18557835e-01
-2.58058757e-01 4.49553877e-01 6.65833712e-01 1.11278713e+00
7.11903334e-01 -8.19847211e-02 7.61342468e-04 1.05459344e+00
4.88435954e-01 8.58410835e-01 5.54076016e-01 -1.25413418e+00
3.65419000e-01 4.36757922e-01 6.55399933e-02 -1.11534500e+00
-4.42781657e-01 -2.07849834e-02 -6.91068649e-01 -6.00482933e-02
3.32633108e-01 2.35178825e-02 -7.56590426e-01 1.71983635e+00
7.07504690e-01 4.16653186e-01 2.52895225e-02 8.94579947e-01
1.08691680e+00 6.36362076e-01 2.50130773e-01 2.97432840e-02
1.51619339e+00 -1.25505221e+00 -4.71350104e-01 -3.73511672e-01
7.11481452e-01 -5.12123227e-01 7.13598847e-01 3.05368900e-01
-4.25129443e-01 -5.32830775e-01 -8.04131508e-01 1.83644384e-01
-5.03008604e-01 2.65248567e-01 7.17977941e-01 1.20777756e-01
-6.46824181e-01 4.12176818e-01 -9.41314936e-01 -6.44461572e-01
7.19867527e-01 2.83308774e-01 -8.31020415e-01 -2.33576015e-01
-7.74063289e-01 6.41867578e-01 3.10749412e-01 2.64166534e-01
-1.38291681e+00 -3.38156134e-01 -9.35782373e-01 -4.24919844e-01
5.88254809e-01 -3.84117931e-01 1.20797825e+00 -8.94938946e-01
-1.13599122e+00 9.95357871e-01 3.75523686e-01 -5.25668800e-01
3.02825063e-01 -4.67276156e-01 -3.60435694e-01 2.47685775e-01
2.43515849e-01 8.04781914e-01 5.79748094e-01 -8.97622108e-01
-6.42781198e-01 -6.09692156e-01 6.77807629e-02 8.30976963e-02
-2.58061662e-03 2.13632062e-01 -4.28679317e-01 -8.42534602e-01
-2.37423152e-01 -1.35583806e+00 -2.24461168e-01 3.15016538e-01
2.48758823e-01 1.07352413e-01 1.01582241e+00 -5.99562764e-01
7.90571213e-01 -2.31977344e+00 3.41223806e-01 -4.92681772e-01
-1.82533786e-02 1.37277439e-01 -4.15050268e-01 4.04677063e-01
1.05394691e-01 -2.29868636e-01 -2.35389933e-01 -2.33045258e-02
-2.38248512e-01 4.96723533e-01 7.46336132e-02 8.32539737e-01
-1.88758690e-02 7.44888961e-01 -8.71571958e-01 -6.56289518e-01
3.45144302e-01 1.11686252e-01 -7.06894517e-01 4.96343106e-01
-1.28909886e-01 7.18087852e-01 -7.12627590e-01 1.06377888e+00
4.96794999e-01 -1.11485429e-01 3.00984993e-03 -1.54164463e-01
1.57381490e-01 -5.22544086e-01 -1.04790294e+00 1.76349175e+00
-1.02756575e-01 3.39351475e-01 1.42297149e-01 -1.12671053e+00
5.15494347e-01 7.74797797e-02 9.01271403e-01 -3.16285610e-01
2.60059148e-01 3.22849341e-02 -3.27331237e-02 -9.32507575e-01
1.51888400e-01 3.17343861e-01 -2.51387954e-01 7.32774287e-02
4.14504826e-01 2.62249298e-02 4.84828055e-01 -1.38193861e-01
1.40537214e+00 5.27999043e-01 5.16859353e-01 -3.38975750e-02
4.37761933e-01 7.85738304e-02 9.64373350e-01 4.12655145e-01
-9.31117296e-01 2.48999387e-01 1.47337496e-01 -7.12899506e-01
-7.73160994e-01 -7.08411753e-01 -5.12075126e-02 1.52034140e+00
4.11116749e-01 -1.85950935e-01 -5.28610289e-01 -7.64838040e-01
3.88493687e-02 4.53767702e-02 -7.20831275e-01 -1.49729058e-01
-6.84948504e-01 -8.06346536e-01 6.08390450e-01 7.44007170e-01
9.24790382e-01 -1.46561384e+00 -7.33616471e-01 6.38458952e-02
-4.30841118e-01 -1.44680989e+00 -3.87258142e-01 1.21877439e-01
-6.84504747e-01 -1.41231763e+00 -4.49594826e-01 -7.80789256e-01
4.60029274e-01 5.08906066e-01 8.00626636e-01 -5.99992089e-02
-5.36135912e-01 3.53114009e-01 -6.09307826e-01 -3.04092288e-01
-2.05591545e-01 -3.37596774e-01 2.84099728e-01 -8.88554603e-02
3.65021646e-01 -4.02496547e-01 -6.04853928e-01 9.02937353e-01
-9.25895751e-01 -2.18639478e-01 6.01668239e-01 5.97599387e-01
6.68772697e-01 -2.45396510e-01 2.96483159e-01 -5.45504808e-01
-2.75465965e-01 -9.09051597e-01 -6.65070355e-01 2.37044677e-01
2.93742001e-01 -2.63158053e-01 6.93268716e-01 -6.40362680e-01
-7.72204578e-01 3.13607067e-01 -2.83473078e-02 -2.55939037e-01
-8.08266938e-01 -7.75581179e-03 -1.97412282e-01 -1.75712064e-01
5.00332832e-01 9.27887857e-02 -1.22127421e-01 -5.25673151e-01
-3.14931609e-02 5.44236720e-01 5.71278155e-01 -4.78170604e-01
6.76109672e-01 3.90690416e-01 2.62553580e-02 -7.48206496e-01
-7.88181663e-01 -6.61547661e-01 -8.29322040e-01 -5.67671359e-01
1.23823965e+00 -1.14041543e+00 -6.57470822e-01 9.52927053e-01
-6.17293894e-01 -4.74411517e-01 2.35658869e-01 8.65319669e-01
-7.41328597e-01 3.78815800e-01 -6.16080225e-01 -4.17264283e-01
-2.56570876e-02 -1.23673677e+00 1.18715286e+00 1.25823289e-01
-1.54803529e-01 -5.82464159e-01 1.17449157e-01 9.75985765e-01
1.49759546e-01 5.70043504e-01 5.47136128e-01 -7.00467288e-01
-7.61133075e-01 -1.23029202e-01 1.01535909e-01 3.92910630e-01
2.24148378e-01 2.99032778e-01 -4.96962279e-01 -3.65516514e-01
-2.20739886e-01 -9.48953748e-01 6.01181626e-01 3.06962699e-01
1.34639776e+00 -1.17956012e-01 -4.02443975e-01 6.83482528e-01
1.23334968e+00 4.65107650e-01 2.97463149e-01 2.67444968e-01
7.39774704e-01 6.11666799e-01 1.00780594e+00 5.46845198e-01
2.66490936e-01 8.32342088e-01 6.01722896e-01 3.27892184e-01
8.65074024e-02 2.92492080e-02 5.47847092e-01 6.58343256e-01
-2.75051326e-01 -3.95586073e-01 -9.56167638e-01 4.57172573e-01
-1.92654967e+00 -1.26738787e+00 -3.25044319e-02 1.93592334e+00
4.47633117e-01 -3.90306294e-01 3.30932826e-01 -3.34440202e-01
6.43041432e-01 3.21907848e-01 -7.10875332e-01 1.43723339e-01
7.48801529e-02 -5.36720514e-01 5.73010921e-01 -3.87681305e-01
-1.59636724e+00 8.41319382e-01 6.47868586e+00 6.03965998e-01
-8.09193313e-01 -7.54012614e-02 1.70913160e-01 -1.66002929e-01
5.99377871e-01 -4.68886048e-01 -8.56222570e-01 5.74840486e-01
8.12870145e-01 2.43251428e-01 4.25882787e-01 1.25440490e+00
5.41723482e-02 -1.09870553e-01 -1.05097318e+00 1.04295540e+00
3.48133296e-01 -8.09210539e-01 -1.56696990e-01 1.34031018e-02
6.78890049e-01 3.97468358e-01 -3.06814879e-01 4.48604912e-01
4.01033491e-01 -7.12000251e-01 5.08330107e-01 4.75281328e-01
2.62614340e-01 -4.09954876e-01 8.06243300e-01 5.21656930e-01
-1.35060644e+00 -2.28870302e-01 -3.38449329e-01 2.49248426e-02
1.51165826e-02 -1.74576327e-01 -2.34573424e-01 2.86730021e-01
1.35062611e+00 1.23773551e+00 -7.65847743e-01 1.14008760e+00
1.17587171e-01 4.48399246e-01 -3.79194736e-01 -9.65236798e-02
1.80474684e-01 -3.10952932e-01 4.29196149e-01 8.97934794e-01
1.87670231e-01 3.10039520e-01 5.32190263e-01 2.15041205e-01
-8.91391486e-02 1.24433704e-01 -8.42184842e-01 -2.88377076e-01
1.96982279e-01 1.29491866e+00 -8.72731805e-01 -2.26362228e-01
-7.05819905e-01 9.26872909e-01 3.35334688e-01 1.46160364e-01
-1.33232677e+00 9.99630764e-02 1.15187657e+00 -2.04521075e-01
4.94008631e-01 -2.79335201e-01 5.53934872e-01 -1.47279251e+00
7.30773574e-03 -1.10863769e+00 4.73971933e-01 -6.51557922e-01
-1.18810976e+00 4.03611511e-01 4.60073352e-01 -1.64436769e+00
1.97306406e-02 -7.06616044e-01 -2.49531139e-02 -1.06091000e-01
-1.06803524e+00 -1.27028215e+00 -6.24809921e-01 6.23335600e-01
8.96096051e-01 -3.38708550e-01 7.15035021e-01 5.45262694e-01
-6.83607817e-01 5.46829328e-02 1.78256691e-01 3.06418955e-01
8.58026206e-01 -6.87638283e-01 5.83816096e-02 7.79742897e-01
1.08776122e-01 3.54340166e-01 6.57370746e-01 -5.52094817e-01
-1.58187616e+00 -1.39376998e+00 9.46151987e-02 -5.84197342e-01
7.69640863e-01 -3.32450658e-01 -7.57950366e-01 9.26620543e-01
-2.59744227e-01 4.34375852e-01 1.00041497e+00 -3.28474700e-01
-1.23856038e-01 -3.39767300e-02 -1.09058022e+00 3.51738840e-01
1.37265098e+00 -1.35179788e-01 -5.30610979e-01 4.83849943e-01
3.93285125e-01 -3.58273208e-01 -9.19362962e-01 7.17684984e-01
8.57764482e-01 -7.61490524e-01 8.74873579e-01 -7.86099076e-01
6.85704768e-01 -6.38125241e-01 -5.53140402e-01 -1.06823647e+00
-4.53485876e-01 1.51023999e-01 -2.39816289e-02 1.04860079e+00
-4.07712720e-03 -4.93220896e-01 5.38762212e-01 2.17210084e-01
-2.15157688e-01 -6.61678374e-01 -5.26791632e-01 -7.68601358e-01
-3.19503516e-01 -1.25477269e-01 5.53027630e-01 8.02159309e-01
-5.79082608e-01 -2.52707805e-02 -7.91149855e-01 2.36433715e-01
6.00706697e-01 4.82235551e-01 1.19979644e+00 -1.04566634e+00
-2.14361802e-01 3.72236408e-02 -1.06888747e+00 -1.03405166e+00
2.50619739e-01 -5.30273080e-01 2.88457692e-01 -1.26289630e+00
4.96988744e-01 2.79556457e-02 -1.85496300e-01 7.58948207e-01
2.57455371e-02 5.67452550e-01 -6.76326752e-02 2.90897429e-01
-1.05961335e+00 5.32337010e-01 1.15548456e+00 -7.41183609e-02
1.77814946e-01 2.47408748e-02 -1.94348007e-01 1.06780922e+00
7.48642027e-01 -5.30518889e-01 -3.69708627e-01 -5.63920856e-01
-1.96628079e-01 1.00367472e-01 6.53157651e-01 -1.24530888e+00
-1.76821515e-01 -7.64713347e-01 4.67113763e-01 -5.99406779e-01
4.17165339e-01 -1.12560618e+00 5.57073474e-01 5.66067934e-01
-3.43731403e-01 2.13552825e-03 8.45282152e-02 8.93246770e-01
-2.58620322e-01 2.42946092e-02 8.94544482e-01 -4.71988410e-01
-1.38261735e+00 7.07187176e-01 -5.25181651e-01 1.12771936e-01
1.71716523e+00 -2.70899981e-01 -3.23041946e-01 1.61286265e-01
-6.60908043e-01 4.35271323e-01 6.27811491e-01 8.07412624e-01
3.72867197e-01 -1.39003146e+00 -6.00711107e-01 2.45677382e-01
6.95091009e-01 -1.45942435e-01 6.48802936e-01 8.55462849e-01
-9.08657908e-01 1.01024345e-01 -9.47126269e-01 -5.61104178e-01
-1.73322403e+00 5.67730367e-01 2.25846007e-01 1.96136415e-01
-5.20462096e-01 7.90349126e-01 3.82002652e-01 -6.29832447e-01
8.68934914e-02 -5.32651305e-01 -4.38273609e-01 5.96214086e-02
5.17330587e-01 3.65128964e-01 -3.01096350e-01 -1.17797029e+00
-6.65932059e-01 6.56029046e-01 7.90218189e-02 5.87114513e-01
1.56609154e+00 -2.13312462e-01 7.41392225e-02 5.35364747e-01
1.32300866e+00 -3.36689502e-01 -1.33405364e+00 -1.30954280e-01
-3.08870643e-01 -6.06616199e-01 -4.52301592e-01 -4.54462975e-01
-1.17617095e+00 5.45882285e-01 6.02456152e-01 -1.74397483e-01
1.05463517e+00 7.28703365e-02 8.43228519e-01 7.14917660e-01
8.28510761e-01 -1.15655470e+00 1.63068816e-01 4.03823286e-01
1.02284825e+00 -1.59752369e+00 1.79776236e-01 -2.85495520e-01
-9.46007729e-01 7.81167686e-01 1.06058323e+00 -5.48443124e-02
5.12702584e-01 2.04230443e-01 1.33497104e-01 -2.24169761e-01
-8.35849762e-01 -1.39735222e-01 -6.02081977e-02 6.61415875e-01
2.16301069e-01 1.05805695e-01 2.79011335e-02 2.66470313e-01
-6.71320632e-02 -1.09007265e-02 2.71604687e-01 1.15860665e+00
-4.90759045e-01 -6.80498183e-01 -1.72377765e-01 3.71602863e-01
-5.32392561e-01 4.38489348e-01 -3.13399404e-01 6.25180542e-01
4.14693832e-01 7.77772248e-01 -3.66745979e-01 -3.11793059e-01
3.10622007e-01 -8.66012052e-02 3.39244455e-01 -6.23654425e-01
-2.88401335e-01 -1.59323052e-01 3.19817550e-02 -8.97160053e-01
-1.08958161e+00 -9.11295950e-01 -1.06041443e+00 -1.94935724e-01
-1.15472570e-01 -1.04933925e-01 3.11828554e-01 8.88918042e-01
2.05388412e-01 3.17969412e-01 5.12965739e-01 -9.47227120e-01
-2.60201156e-01 -1.02740169e+00 -7.06526041e-01 7.12399423e-01
2.45469809e-01 -9.91765141e-01 -3.69238794e-01 5.36343634e-01] | [7.72604513168335, -0.7765602469444275] |
82f1c5d2-5938-4662-96ea-e11305440ce8 | can-foundation-models-wrangle-your-data | 2205.09911 | null | https://arxiv.org/abs/2205.09911v2 | https://arxiv.org/pdf/2205.09911v2.pdf | Can Foundation Models Wrangle Your Data? | Foundation Models (FMs) are models trained on large corpora of data that, at very large scale, can generalize to new tasks without any task-specific finetuning. As these models continue to grow in size, innovations continue to push the boundaries of what these models can do on language and image tasks. This paper aims to understand an underexplored area of FMs: classical data tasks like cleaning and integration. As a proof-of-concept, we cast five data cleaning and integration tasks as prompting tasks and evaluate the performance of FMs on these tasks. We find that large FMs generalize and achieve SoTA performance on data cleaning and integration tasks, even though they are not trained for these data tasks. We identify specific research challenges and opportunities that these models present, including challenges with private and domain specific data, and opportunities to make data management systems more accessible to non-experts. We make our code and experiments publicly available at: https://github.com/HazyResearch/fm_data_tasks. | ['Christopher Ré', 'Simran Arora', 'Laurel Orr', 'Ines Chami', 'Avanika Narayan'] | 2022-05-20 | null | null | null | null | ['entity-resolution'] | ['natural-language-processing'] | [-1.78388894e-01 1.45596504e-01 -1.22995839e-01 -7.77692795e-01
-8.82879674e-01 -5.61268926e-01 6.07798040e-01 1.18074335e-01
-3.89511466e-01 5.53363860e-01 2.15018049e-01 -4.61608320e-01
7.69567639e-02 -4.39142287e-01 -9.30851758e-01 -3.18204373e-01
-1.01946080e-02 5.39505601e-01 5.09081874e-03 -1.99912921e-01
2.17656016e-01 3.64222884e-01 -1.66360056e+00 5.68583786e-01
9.56877232e-01 6.83898509e-01 3.01135093e-01 5.67741156e-01
-1.47697002e-01 6.24326468e-01 -6.49703741e-01 -3.89258325e-01
5.14429569e-01 3.47633034e-01 -1.12425053e+00 -3.00958436e-02
9.89666700e-01 -5.22696450e-02 -1.16080791e-01 9.52191412e-01
3.24767023e-01 1.19480237e-01 3.08307767e-01 -1.59659612e+00
-1.39894617e+00 5.11273265e-01 -3.20747882e-01 1.39243409e-01
-2.11310089e-02 4.32562917e-01 8.05561185e-01 -1.01680374e+00
8.49536300e-01 1.24213707e+00 9.05298233e-01 7.34269321e-01
-1.24872601e+00 -1.01157236e+00 2.52548367e-01 1.54162481e-01
-1.11942744e+00 -9.57427859e-01 1.13425218e-01 -7.32979774e-01
1.08469701e+00 2.82525957e-01 1.71523646e-03 1.22357726e+00
1.82629779e-01 8.29130054e-01 9.89005864e-01 -3.85052234e-01
1.56160101e-01 3.24281633e-01 5.73244929e-01 5.36055088e-01
6.53904021e-01 -1.95276141e-01 -6.43342912e-01 1.07540883e-01
4.03876156e-01 1.33866280e-01 -1.26491254e-02 -2.75912732e-01
-1.17407453e+00 6.48790181e-01 3.50460917e-01 3.28539819e-01
-2.29664996e-01 2.37244666e-02 4.58065778e-01 5.45187771e-01
7.95209706e-01 8.95461977e-01 -9.31304157e-01 1.04496539e-01
-9.46132779e-01 4.70419884e-01 7.90489793e-01 1.43316483e+00
1.04743028e+00 -2.64068574e-01 -2.83610195e-01 8.32846701e-01
3.69015448e-02 4.84064728e-01 3.42554241e-01 -1.17321324e+00
6.49125814e-01 6.10615432e-01 2.25337386e-01 -5.23904443e-01
-4.68668371e-01 -3.62520099e-01 -8.30863535e-01 2.09298685e-01
2.55065769e-01 -1.52242675e-01 -1.41779685e+00 1.72234285e+00
2.59114474e-01 1.97276250e-02 -1.19296886e-01 3.13649863e-01
9.38945651e-01 3.23827028e-01 4.19748724e-01 7.21893013e-02
1.41437078e+00 -1.19408095e+00 -9.53716755e-01 -5.53311825e-01
8.43116462e-01 -9.54612613e-01 1.44930816e+00 4.23719496e-01
-1.14149833e+00 -8.57461572e-01 -8.07790697e-01 -5.55247247e-01
-1.01033497e+00 -9.75267142e-02 7.70242870e-01 3.58357072e-01
-1.22612858e+00 6.15321338e-01 -1.05947244e+00 -7.25057244e-01
7.92933047e-01 1.25170678e-01 -5.43652475e-01 -1.04829066e-01
-8.58833373e-01 1.05661809e+00 3.39798659e-01 -8.20192918e-02
-1.03838146e+00 -1.05690336e+00 -7.81922400e-01 -1.28746554e-01
5.34409523e-01 -8.38647664e-01 1.53589857e+00 -5.21098137e-01
-6.11692190e-01 1.06238413e+00 -4.27610308e-01 -5.32395780e-01
1.09140910e-01 -6.71948373e-01 -6.80826604e-01 -3.74442548e-01
5.27791262e-01 7.00309813e-01 7.53154755e-01 -1.37315142e+00
-8.19443583e-01 -3.97706628e-01 2.32550483e-02 -1.38291329e-01
-4.25448328e-01 1.99662298e-01 -4.61708099e-01 -4.81219500e-01
-3.68324459e-01 -7.51506090e-01 -3.45407188e-01 -1.08578213e-01
-4.64729428e-01 -2.88895190e-01 6.78181112e-01 -7.51136184e-01
1.29890990e+00 -2.28743124e+00 -1.09801680e-01 -3.97186786e-01
2.85444796e-01 6.53757930e-01 -3.94453943e-01 5.28332651e-01
-4.40359339e-02 6.87784016e-01 -2.08229452e-01 -9.17587757e-01
4.79184091e-03 3.08865607e-01 -3.18756461e-01 1.26184925e-01
3.79581839e-01 1.00457609e+00 -7.91008174e-01 -2.24426478e-01
1.11195259e-01 2.11063713e-01 -5.94482780e-01 4.58640933e-01
-4.76181686e-01 4.43964988e-01 -2.78472036e-01 6.98360384e-01
9.04447615e-01 -3.57707202e-01 -3.10251623e-01 -3.34406078e-01
-2.93433905e-01 2.37463608e-01 -1.06450522e+00 1.99465215e+00
-3.75789553e-01 5.88925242e-01 4.95631248e-01 -9.74190176e-01
6.39429629e-01 1.26226768e-01 9.45026428e-02 -6.34075880e-01
-2.50644028e-01 1.11877173e-01 -2.90593177e-01 -8.98179650e-01
6.79735720e-01 1.86075956e-01 -9.81696025e-02 2.59544224e-01
3.33985746e-01 -1.24974489e-01 5.20589828e-01 2.27777556e-01
1.21226323e+00 -1.23884771e-02 2.86358654e-01 -3.76128823e-01
1.98663905e-01 4.17559206e-01 4.98572528e-01 1.11832964e+00
-1.63019240e-01 6.25260055e-01 -3.93791273e-02 -7.11162269e-01
-1.00494695e+00 -1.01872194e+00 -2.57944256e-01 1.54172957e+00
-2.61907727e-01 -7.22769022e-01 -7.36628711e-01 -7.20108390e-01
2.59641767e-01 7.97336876e-01 -7.65493214e-01 -3.70330103e-02
-4.57460344e-01 -7.19183266e-01 4.23716545e-01 4.84691590e-01
4.47204769e-01 -1.24490368e+00 -1.59913808e-01 -4.42397334e-02
-1.83650434e-01 -1.25451171e+00 -3.32736731e-01 2.36348704e-01
-8.73824060e-01 -1.13759303e+00 -1.81399807e-02 -6.65815532e-01
5.77056825e-01 6.08549893e-01 1.53068841e+00 2.13830292e-01
-5.01675367e-01 4.36919808e-01 -4.46996540e-01 -1.15735722e+00
-4.79058564e-01 5.01672566e-01 -3.62748355e-02 -4.45244163e-01
7.77933896e-01 -5.27829766e-01 -5.45181930e-01 3.04304183e-01
-1.20239198e+00 5.40479757e-02 6.20673180e-01 4.76372004e-01
6.71186566e-01 -2.64419585e-01 7.31807172e-01 -1.35272169e+00
6.97227776e-01 -8.34354460e-01 -3.58425587e-01 4.66848224e-01
-9.82291162e-01 -9.71971154e-02 4.02709663e-01 -2.60972846e-02
-1.07744873e+00 -1.74779788e-01 -1.60218641e-01 -2.16118366e-01
-5.93012929e-01 4.33772117e-01 -2.56042294e-02 2.73457915e-01
9.50795352e-01 -2.90188849e-01 3.40916147e-03 -1.14123166e+00
6.94424152e-01 8.95313978e-01 6.77310944e-01 -5.71134329e-01
9.18976307e-01 7.19688952e-01 -4.14087296e-01 -6.16634190e-01
-1.34480107e+00 -6.62563562e-01 -6.93327129e-01 4.92527753e-01
1.10802531e+00 -1.28390718e+00 -1.65636629e-01 3.16864699e-01
-1.06614912e+00 -4.49179292e-01 -5.49595594e-01 -1.08531499e-02
-3.12619925e-01 -5.24977548e-03 -3.02302897e-01 -3.95505011e-01
-4.98804778e-01 -9.54617858e-01 1.18132150e+00 2.74085313e-01
-2.88160235e-01 -9.15617168e-01 2.78499089e-02 7.81786501e-01
7.41050422e-01 1.10639110e-01 6.74272537e-01 -8.52511108e-01
-7.12935448e-01 -9.42530856e-02 -2.39023089e-01 8.06196630e-01
2.13305086e-01 1.03126436e-01 -1.17358148e+00 -3.76280725e-01
1.27790952e-02 -2.89333194e-01 1.06190681e+00 3.13183278e-01
1.55475211e+00 -2.40378112e-01 -3.41999680e-01 9.27071750e-01
1.16261446e+00 -5.22618145e-02 4.00194913e-01 5.27419329e-01
7.23849714e-01 3.65731537e-01 7.78966486e-01 1.48969159e-01
5.73965311e-01 2.88942069e-01 4.61746901e-01 -2.87615210e-01
-4.16334987e-01 -4.79211956e-02 1.90107495e-01 7.73383200e-01
-2.55099982e-01 -8.20435360e-02 -1.10208750e+00 9.42534566e-01
-2.02208638e+00 -8.62080336e-01 -2.82057792e-01 2.07913375e+00
9.72876012e-01 -6.57889917e-02 -2.01722085e-01 -5.33747613e-01
4.89940226e-01 3.88211980e-02 -6.01037681e-01 -4.21981603e-01
-6.71707988e-02 3.32053959e-01 5.37754595e-01 3.64548206e-01
-1.35982895e+00 1.14002061e+00 6.62013340e+00 6.48080885e-01
-7.59037912e-01 5.20320237e-01 4.43335712e-01 -2.37896666e-01
-2.38101095e-01 1.30050227e-01 -1.17725241e+00 3.09108317e-01
1.08773601e+00 -1.04499891e-01 6.56564474e-01 8.97202730e-01
1.39313266e-01 2.23301146e-02 -1.40396571e+00 7.04136372e-01
-1.49585098e-01 -1.56978023e+00 7.04461196e-03 1.36292130e-01
9.30595875e-01 7.91253865e-01 1.04360908e-01 6.49822712e-01
7.20856845e-01 -1.25670612e+00 6.18487418e-01 5.94429016e-01
5.94546318e-01 -1.46964729e-01 6.73024774e-01 2.14875832e-01
-7.49940455e-01 -2.15391934e-01 -6.64792657e-01 -1.19253129e-01
1.30581707e-01 6.54379666e-01 -7.73961902e-01 6.63170815e-01
1.24034286e+00 8.07059586e-01 -1.18336082e+00 1.03072476e+00
-7.98420608e-02 5.77246666e-01 -1.09844670e-01 6.70860171e-01
-1.15180433e-01 2.84658521e-02 2.92715013e-01 1.51750660e+00
2.35378027e-01 -2.90596515e-01 1.23053677e-01 6.70452952e-01
-3.26421022e-01 -1.85309187e-01 -7.21713424e-01 -2.08490610e-01
7.18170106e-01 1.37358928e+00 -2.10179970e-01 -4.40859795e-01
-6.34678006e-01 4.93608713e-01 6.46286726e-01 4.81226236e-01
-5.19224226e-01 -1.69336259e-01 1.20475686e+00 2.86413401e-01
1.43297151e-01 -2.17602551e-01 -6.23918653e-01 -1.44159460e+00
5.28207682e-02 -1.06793058e+00 5.88557899e-01 -7.72652805e-01
-1.66247439e+00 3.25816184e-01 2.24856347e-01 -8.26477647e-01
1.96891092e-02 -5.52251518e-01 -4.04061317e-01 6.62185967e-01
-1.67330575e+00 -1.34211063e+00 -6.68029785e-01 5.99594355e-01
5.90055704e-01 -1.69308364e-01 9.89120483e-01 4.72564846e-01
-7.11362004e-01 4.30775553e-01 3.61791283e-01 1.62512854e-01
1.38529849e+00 -1.26602960e+00 9.76645172e-01 9.33176219e-01
1.30530894e-01 1.14546406e+00 8.23797405e-01 -6.42443419e-01
-1.28638148e+00 -1.45145380e+00 9.78176296e-01 -1.15305436e+00
6.44314647e-01 -7.86249697e-01 -1.07330048e+00 1.33473897e+00
4.83115971e-01 9.03636962e-02 6.90363765e-01 7.59645700e-01
-6.56827986e-01 -3.65447789e-01 -1.08342993e+00 3.78818363e-01
1.54599655e+00 -4.76445436e-01 -7.04580665e-01 7.36954689e-01
8.94034505e-01 -3.36110502e-01 -1.01041174e+00 5.20330489e-01
2.98341244e-01 -9.36838508e-01 1.11654198e+00 -1.08564508e+00
4.52498019e-01 -2.75360048e-01 -1.57319427e-01 -1.36271846e+00
-4.74584907e-01 -7.36308038e-01 -8.10847282e-02 1.52024734e+00
5.38460791e-01 -6.64559066e-01 3.54055524e-01 9.13572788e-01
-5.66229463e-01 -3.25559437e-01 -6.56172633e-01 -9.97563541e-01
2.90596306e-01 -5.08625150e-01 8.84189069e-01 1.19761133e+00
-4.69170749e-01 3.89379084e-01 -1.82873145e-01 1.86786860e-01
6.49223089e-01 1.24349058e-01 1.28193974e+00 -1.38575208e+00
-1.86086390e-02 -1.93528980e-01 3.85039411e-02 -5.70954800e-01
-1.05818719e-01 -1.01699436e+00 -1.65743858e-01 -1.85042119e+00
1.68237001e-01 -5.13627112e-01 -4.38385010e-01 9.68369722e-01
-2.71060705e-01 2.00075835e-01 2.32153490e-01 3.98206025e-01
-9.27182615e-01 1.96759239e-01 1.07691050e+00 -1.00005776e-01
-7.89400935e-02 -1.25304997e-01 -1.38293600e+00 6.81234002e-01
9.94163990e-01 -5.63816488e-01 -4.40268159e-01 -1.03146577e+00
1.22834362e-01 -6.89915121e-01 2.25569159e-01 -1.04787278e+00
2.57656515e-01 -3.87208998e-01 2.78844833e-01 -3.99611771e-01
6.48045763e-02 -7.19217777e-01 3.69668454e-01 -4.15804014e-02
-2.87107140e-01 1.78754762e-01 4.51405525e-01 2.89065927e-01
-1.68411180e-01 -1.09645151e-01 6.16066575e-01 -4.23988670e-01
-8.03044856e-01 4.11781818e-01 1.04971789e-01 3.11869383e-01
9.77557659e-01 1.16676234e-01 -9.82218683e-01 -1.05787098e-01
-9.31493044e-01 5.41688323e-01 5.40413022e-01 8.49649072e-01
9.07928720e-02 -1.00510383e+00 -8.23207080e-01 2.51778066e-01
3.68079752e-01 5.50388634e-01 2.15675592e-01 6.88846052e-01
-1.93415761e-01 5.20551085e-01 -2.20496565e-01 -4.20289487e-01
-1.17933178e+00 8.51225674e-01 -2.02855486e-02 -2.39927441e-01
-4.03477550e-01 8.82600009e-01 4.02380288e-01 -7.35404372e-01
3.17243218e-01 -5.55621743e-01 1.24324389e-01 1.37979656e-01
7.43922293e-01 2.78061330e-01 3.85672271e-01 -9.62547958e-02
-2.59071529e-01 1.44932821e-01 -4.91300434e-01 4.68050390e-01
1.78946543e+00 -2.62716413e-01 -2.42523909e-01 3.25076818e-01
9.82789278e-01 -1.02779403e-01 -1.21498477e+00 -3.59932870e-01
1.74692616e-01 -4.62412655e-01 -2.23123834e-01 -1.19102967e+00
-9.29338932e-01 8.72362137e-01 4.64541018e-01 4.34534997e-01
1.11601424e+00 3.14378768e-01 5.58648050e-01 3.94411832e-01
1.36263251e-01 -1.26190412e+00 -3.07414234e-01 5.16748011e-01
1.17814541e+00 -1.53954160e+00 8.99293646e-02 -2.86831051e-01
-5.40665150e-01 6.67535484e-01 8.19600582e-01 3.75999659e-02
8.65028560e-01 4.36011672e-01 1.52198717e-01 -4.25212741e-01
-9.68090534e-01 -2.85181284e-01 1.43466085e-01 9.76921141e-01
4.18677688e-01 5.83527200e-02 -1.27481539e-02 6.84766054e-01
-2.77252436e-01 3.23537380e-01 3.77781034e-01 9.74958241e-01
-4.11295921e-01 -1.34308982e+00 -2.77844340e-01 7.03416228e-01
-4.57153022e-01 -3.51256460e-01 -2.52290964e-01 9.52442944e-01
4.88961339e-01 1.06539500e+00 3.81046571e-02 -2.28099763e-01
6.69607759e-01 3.42365891e-01 9.75203142e-02 -1.11994278e+00
-6.81636810e-01 -4.71671283e-01 2.66297907e-01 -8.21158588e-01
-3.37256134e-01 -7.96452761e-01 -9.33194399e-01 -4.74731237e-01
1.03172865e-02 7.27623552e-02 8.91225636e-01 7.56238043e-01
1.00028741e+00 4.32306170e-01 3.10341343e-02 -6.84426010e-01
-6.27029419e-01 -1.07568061e+00 -1.31221160e-01 7.17984140e-01
5.21562338e-01 -3.57295185e-01 -2.49164566e-01 4.59075868e-01] | [9.513955116271973, 8.45335578918457] |
bbd996ec-3426-46ea-a608-893729f7cdee | zo-grof-a-comprehensive-corpus-for-offensive | null | null | https://aclanthology.org/2022.woah-1.5 | https://aclanthology.org/2022.woah-1.5.pdf | “Zo Grof !”: A Comprehensive Corpus for Offensive and Abusive Language in Dutch | This paper presents a comprehensive corpus for the study of socially unacceptable language in Dutch. The corpus extends and revise an existing resource with more data and introduces a new annotation dimension for offensive language, making it a unique resource in the Dutch language panorama. Each language phenomenon (abusive and offensive language) in the corpus has been annotated with a multi-layer annotation scheme modelling the explicitness and the target(s) of the message. We have conducted a new set of experiments with different classification algorithms on all annotation dimensions. Monolingual Pre-Trained Language Models prove as the best systems, obtaining a macro-average F1 of 0.828 for binary classification of offensive language, and 0.579 for the targets of offensive messages. Furthermore, the best system obtains a macro-average F1 of 0.667 for distinguishing between abusive and offensive messages. | ['Tommaso Caselli', 'Zhenja Gnezdilov', 'Robin Van Der Noord', 'Victor Zwart', 'Ward Ruitenbeek'] | null | null | null | null | naacl-woah-2022-7 | ['abusive-language'] | ['natural-language-processing'] | [ 9.03251581e-03 2.71185040e-01 -4.12688911e-01 -2.99205035e-01
-7.33552873e-01 -7.80891180e-01 1.04527009e+00 2.99093425e-01
-7.76516616e-01 7.82154143e-01 4.34515566e-01 -3.77752364e-01
-1.24715343e-01 -2.87407368e-01 -3.89800631e-02 -5.63808799e-01
1.60471722e-01 7.37648606e-01 -5.56583293e-02 -6.55091345e-01
2.97446281e-01 4.31336761e-01 -1.07970858e+00 6.78116083e-01
7.59153783e-01 6.50899589e-01 -1.64275035e-01 5.74589729e-01
-2.69725889e-01 1.11753511e+00 -1.13235343e+00 -1.07505500e+00
5.63604981e-02 -4.72021639e-01 -9.57267344e-01 5.40289320e-02
4.32299346e-01 -1.34696746e-02 9.66454893e-02 1.14161861e+00
3.67209464e-01 -3.38350415e-01 8.00235510e-01 -9.04777467e-01
-6.63145304e-01 1.02108264e+00 -5.20082831e-01 2.31411800e-01
6.25073850e-01 -1.21548325e-01 1.17619479e+00 -5.45022190e-01
9.50341463e-01 1.66205740e+00 4.14587587e-01 6.56359732e-01
-1.28709614e+00 -9.04822230e-01 -2.50466447e-03 -4.74217236e-02
-1.37956619e+00 -4.86764312e-01 6.40541077e-01 -1.02114141e+00
8.91870797e-01 6.00983083e-01 6.29844427e-01 1.56514835e+00
1.51251599e-01 4.66036201e-01 1.29874694e+00 -5.59980929e-01
-2.63600290e-01 7.92648911e-01 2.43787780e-01 3.07910383e-01
3.23615432e-01 -4.33187224e-02 -3.81704658e-01 -4.51916665e-01
-6.04982674e-02 -7.76920557e-01 1.06712736e-01 2.98928648e-01
-9.29935575e-01 1.17563641e+00 -3.25574070e-01 1.04755747e+00
3.03731244e-02 -3.31719667e-01 9.92855012e-01 5.57479084e-01
8.01020920e-01 5.88434935e-01 -3.75490725e-01 -1.98458746e-01
-7.83101201e-01 7.75679871e-02 1.13049853e+00 6.29548132e-01
1.96457669e-01 1.21994197e-01 -1.30656630e-01 1.29539847e+00
1.92651093e-01 5.53562760e-01 5.50146222e-01 -6.52023315e-01
3.96742433e-01 6.91896319e-01 1.15763946e-02 -1.07379353e+00
-5.44462085e-01 -4.24379736e-01 -3.65728259e-01 -1.14492938e-01
5.43045700e-01 -1.78237826e-01 -2.93961436e-01 1.66819191e+00
9.60943401e-02 -6.27126276e-01 2.78871924e-01 4.79746640e-01
7.01766968e-01 8.29682827e-01 4.26951408e-01 -6.17493451e-01
1.28613138e+00 -5.71111143e-01 -8.91587138e-01 6.36911169e-02
1.06022227e+00 -1.23585427e+00 1.19146204e+00 6.77948415e-01
-9.70511794e-01 -1.76750839e-01 -7.59711802e-01 -7.41609633e-02
-5.70941448e-01 8.07755627e-03 4.09061641e-01 1.19078481e+00
-7.92351663e-01 -3.38842385e-02 1.46859542e-01 -3.57560754e-01
-2.06322633e-02 4.84497771e-02 -4.11720604e-01 5.17065525e-01
-1.41196346e+00 1.21742737e+00 4.29305017e-01 -1.78396642e-01
-6.32108271e-01 -3.50756466e-01 -7.55088747e-01 -4.79763150e-01
2.97987461e-01 4.57039714e-01 9.59118426e-01 -1.66231346e+00
-1.44641519e+00 1.80946887e+00 4.49055344e-01 -7.12780282e-02
8.10655296e-01 -2.38284141e-01 -1.01230407e+00 -9.83260572e-02
3.81068558e-01 2.13061109e-01 6.17012262e-01 -1.30306888e+00
-6.10564649e-01 -2.87688494e-01 1.97289377e-01 -1.15703769e-01
-6.44335330e-01 1.05129910e+00 3.80014271e-01 -8.81228089e-01
-4.33038682e-01 -7.92582095e-01 3.41846615e-01 -5.21625578e-01
-6.32404029e-01 -4.90426451e-01 4.34239537e-01 -9.24604952e-01
1.66652870e+00 -2.48099566e+00 1.92848578e-01 3.04889500e-01
2.38923296e-01 3.38614881e-01 9.11076814e-02 6.39960468e-01
-1.53686926e-01 4.35064912e-01 -1.62870631e-01 -6.70188218e-02
3.42475086e-01 3.68738204e-01 -2.53890961e-01 6.87978029e-01
-1.42623544e-01 3.95878464e-01 -8.73005807e-01 -7.19428778e-01
-1.12519801e-01 3.08319896e-01 -3.68534207e-01 -1.02494270e-01
2.51356900e-01 3.19173276e-01 -2.59134978e-01 4.80290830e-01
4.97673422e-01 6.27473295e-01 3.91396284e-01 4.39597875e-01
-4.57678795e-01 3.95949483e-01 -6.52368128e-01 9.84209239e-01
-6.04086339e-01 7.32564211e-01 2.57554054e-01 -6.72758520e-01
1.05052340e+00 4.40265775e-01 2.26564676e-01 -5.42273521e-01
4.24839348e-01 6.41354740e-01 6.69730902e-01 -6.78878605e-01
6.42178357e-01 -5.87442517e-01 -7.42287576e-01 3.95159453e-01
1.50473984e-02 -1.80346817e-01 4.59831983e-01 1.94790646e-01
5.69113612e-01 -1.44331649e-01 6.64454758e-01 -9.26775753e-01
1.23011243e+00 -1.33188128e-01 3.47706974e-01 4.29351717e-01
-3.74082029e-01 9.21888370e-03 1.09071219e+00 -5.44212103e-01
-7.56403089e-01 -8.91573250e-01 -5.77848852e-01 1.64518535e+00
-2.06265166e-01 -2.78989673e-01 -8.43218386e-01 -8.15540552e-01
-1.80113196e-01 1.28746188e+00 -5.17260313e-01 -5.56630678e-02
-5.18991590e-01 -6.29546046e-01 1.02978277e+00 -2.57539690e-01
1.25180945e-01 -1.19511545e+00 -3.11453819e-01 -2.42971326e-03
-2.89741844e-01 -1.30039847e+00 -4.64130826e-02 -4.78940494e-02
-7.54319876e-02 -1.06147110e+00 -2.28229478e-01 -7.59862185e-01
1.64848119e-01 -6.24199927e-01 9.24183547e-01 2.62814730e-01
6.88711181e-02 -1.27687722e-01 -4.27601814e-01 -5.70723057e-01
-1.29539287e+00 2.23092049e-01 -9.51727945e-03 1.99110299e-01
5.05864501e-01 -1.48160225e-02 4.74805683e-01 -2.35824008e-02
-8.23811293e-01 -4.05926883e-01 -2.31311973e-02 4.45888728e-01
-2.48123795e-01 -4.26542342e-01 3.69499028e-01 -1.27864099e+00
1.03357708e+00 -5.26521146e-01 -5.33305764e-01 3.39134187e-02
-1.53241292e-01 -5.32512307e-01 7.68378437e-01 -5.19069433e-01
-1.05996466e+00 -3.72016579e-01 -5.28161168e-01 5.02287328e-01
-4.01405573e-01 1.62024736e-01 -1.00598782e-01 1.30416036e-01
7.80818701e-01 -2.14663520e-01 -3.47426981e-01 -4.53818262e-01
7.07991272e-02 1.17850184e+00 1.34778246e-01 -5.22800386e-01
5.97456396e-01 2.01427624e-01 -3.23807448e-01 -1.30483317e+00
-1.09568799e+00 -5.51577866e-01 -9.26191747e-01 -4.92744952e-01
9.42174375e-01 -6.70140207e-01 -6.23479486e-01 5.02797663e-01
-1.30636466e+00 -2.23172992e-01 -1.06921144e-01 4.78243321e-01
-2.97327578e-01 3.55501115e-01 -5.64926326e-01 -9.54776287e-01
-1.36387140e-01 -9.49207842e-01 9.07724917e-01 -7.95272052e-01
-9.50876832e-01 -1.16649401e+00 3.29284906e-01 4.74719286e-01
5.76954409e-02 4.82647151e-01 1.27724850e+00 -1.33191049e+00
6.67023301e-01 -3.15660745e-01 1.97167248e-02 5.43350279e-01
-6.99734762e-02 1.88187584e-01 -9.21140254e-01 -2.10470762e-02
2.49268562e-01 -6.02896035e-01 4.99842584e-01 -3.33487660e-01
5.13917446e-01 -5.02247095e-01 7.14629889e-02 4.71520573e-02
9.87603009e-01 2.68144727e-01 5.21983862e-01 5.03182292e-01
4.66181070e-01 1.07406068e+00 4.27687436e-01 3.80674809e-01
9.48742479e-02 8.16377163e-01 3.62922475e-02 1.16446286e-01
2.82657873e-02 4.19411436e-02 9.53748584e-01 8.07485580e-01
6.62599951e-02 -3.14698875e-01 -1.12046981e+00 5.08471370e-01
-1.31617928e+00 -9.85945284e-01 -5.82752585e-01 1.88384914e+00
9.23640251e-01 1.94343776e-01 3.76354426e-01 4.24041152e-01
8.31711233e-01 3.53945553e-01 3.34790379e-01 -1.35307610e+00
-3.00368875e-01 6.73580635e-03 7.53338709e-02 1.18289483e+00
-1.03414762e+00 1.37435865e+00 6.66760397e+00 1.13700616e+00
-1.10535979e+00 3.47196907e-01 5.04450262e-01 -4.87174131e-02
-2.35587656e-01 -3.60537261e-01 -8.86261404e-01 5.35692811e-01
1.17842937e+00 -2.13546187e-01 1.08531408e-01 7.52941489e-01
2.90928721e-01 -9.76794958e-02 -8.50268185e-01 5.58232307e-01
6.15816534e-01 -6.49789393e-01 2.03383178e-01 2.29200721e-01
6.00042641e-01 -1.26150809e-02 -6.99771494e-02 3.33636433e-01
8.28978792e-02 -8.63657832e-01 1.31989038e+00 6.81502223e-02
8.03517580e-01 -8.79185617e-01 1.01019132e+00 5.58950543e-01
-3.03501070e-01 -1.44393474e-01 -2.68682122e-01 -5.31491220e-01
1.14212893e-01 3.76871079e-01 -3.55381191e-01 4.56928723e-02
5.02935529e-01 5.92024088e-01 -6.78709865e-01 9.38061327e-02
-4.86875474e-01 7.15943813e-01 -1.20719165e-01 -4.23671603e-01
5.41408420e-01 -3.25170964e-01 9.27642465e-01 1.75254035e+00
-7.38081634e-02 -8.49326104e-02 3.44364583e-01 3.02230299e-01
8.06992203e-02 8.79122496e-01 -7.72295356e-01 -2.33190760e-01
-1.43457428e-02 1.30255675e+00 -6.52106166e-01 -4.82927084e-01
-2.98094571e-01 8.32999766e-01 2.87695706e-01 -4.94463257e-02
-7.07231104e-01 -4.16535586e-01 4.54754323e-01 2.98844606e-01
-3.68953586e-01 -5.06044440e-02 -2.82938570e-01 -8.47257912e-01
-2.71594316e-01 -9.97700095e-01 5.71469128e-01 -2.74585962e-01
-1.34618485e+00 9.51360345e-01 2.13384509e-01 -8.19636524e-01
-3.05215448e-01 -8.71155143e-01 -1.46001130e-01 6.10666811e-01
-1.10787356e+00 -1.25972176e+00 3.63585263e-01 3.14474434e-01
3.34239662e-01 -5.84076583e-01 9.97424304e-01 3.90334368e-01
-5.17719805e-01 5.81544280e-01 -6.86146617e-02 1.52813226e-01
7.65072107e-01 -9.21051621e-01 -3.41388226e-01 5.95282912e-01
1.01841791e-02 3.68905127e-01 7.29376674e-01 -6.11195922e-01
-3.52205038e-01 -6.24419570e-01 1.87529576e+00 -5.94099343e-01
1.48345506e+00 -7.07314372e-01 -6.72133565e-01 6.77995861e-01
5.32209992e-01 -7.99086273e-01 1.22260606e+00 1.86868787e-01
-5.91736019e-01 2.15009376e-01 -1.36885428e+00 6.31395519e-01
1.03692460e+00 -4.87997800e-01 -6.64007366e-01 7.35359967e-01
5.50263524e-01 9.10021178e-03 -8.44081938e-01 -1.28201947e-01
5.80079794e-01 -1.08515620e+00 4.89973515e-01 -7.66740680e-01
6.57293856e-01 2.25118652e-01 -2.88098007e-01 -9.91049230e-01
-1.68934450e-01 -8.32708836e-01 6.71052456e-01 1.38590980e+00
7.76267946e-01 -8.47718775e-01 3.69141586e-02 1.73134893e-01
-1.23147003e-01 -2.35980868e-01 -1.17353320e+00 -7.44723916e-01
5.77012002e-01 -6.46434844e-01 1.12689354e-01 1.36547089e+00
4.21504050e-01 5.63143492e-01 -3.80862087e-01 -4.03505772e-01
2.96317935e-01 -1.60379559e-01 4.96724129e-01 -1.32820654e+00
7.63206407e-02 -8.17215860e-01 -5.45565188e-01 -2.13494837e-01
9.26392972e-01 -1.26880884e+00 -3.29659462e-01 -9.28159595e-01
2.02153638e-01 -6.99547008e-02 2.03716576e-01 3.02218556e-01
2.89034963e-01 6.98273718e-01 2.82776505e-01 4.97885868e-02
-3.39115113e-01 5.92643730e-02 7.75743425e-01 7.16893300e-02
-2.26207778e-01 -1.91144824e-01 -8.17775190e-01 1.31790721e+00
9.65290964e-01 -6.58885658e-01 -5.67888021e-02 -3.68508279e-01
4.61052269e-01 -4.39331889e-01 -4.20629568e-02 -4.79402095e-01
-3.85796845e-01 -3.01195949e-01 3.87232266e-02 -1.75261602e-01
2.92760998e-01 -8.81167233e-01 -1.23484612e-01 9.22089875e-01
-7.16221392e-01 -2.88420711e-02 1.89266965e-01 -4.65377495e-02
-2.69281536e-01 -6.57338560e-01 1.07594991e+00 -4.05970253e-02
-5.07483065e-01 -4.49370950e-01 -1.05917490e+00 4.64901596e-01
1.29560578e+00 -1.53480023e-01 -7.29997382e-02 -3.12714696e-01
-8.20228636e-01 -2.05960765e-01 2.93998837e-01 6.50330484e-01
5.31293526e-02 -9.30529416e-01 -1.09009850e+00 7.80108646e-02
3.16651285e-01 -1.20515358e+00 -2.05430895e-01 8.08622599e-01
-7.32971430e-01 2.92695045e-01 -1.73110649e-01 -1.70706093e-01
-1.58781230e+00 2.71912575e-01 1.98241010e-01 -1.97108805e-01
-1.15536973e-01 5.59693217e-01 4.03357111e-02 -7.29959786e-01
1.29275555e-02 2.94121683e-01 -6.82068646e-01 6.12952530e-01
4.57064688e-01 6.02004588e-01 -2.89611965e-01 -1.84747159e+00
-3.99525583e-01 4.23100621e-01 -1.03558049e-01 -2.85941243e-01
1.07979560e+00 -2.82221939e-02 -8.60781908e-01 8.55061293e-01
1.35100651e+00 8.03801715e-01 3.80604714e-02 1.94362238e-01
4.12516385e-01 -5.42665899e-01 -2.15869546e-01 -9.95356858e-01
-5.26734650e-01 6.47991955e-01 -1.58363022e-02 6.82695329e-01
5.90451300e-01 5.90193495e-02 4.00574565e-01 2.20029056e-01
1.21050738e-01 -1.48565221e+00 -1.57663703e-01 8.76528740e-01
1.33076632e+00 -9.03167248e-01 3.31818201e-02 -7.51501501e-01
-9.71334100e-01 1.03068078e+00 3.51064831e-01 -1.05350189e-01
5.68899810e-01 3.39552641e-01 6.80522859e-01 -3.61631125e-01
-3.66243213e-01 4.75728661e-02 3.20402205e-01 4.98427093e-01
9.46734607e-01 4.11850810e-01 -1.45887756e+00 6.41316473e-01
-8.67424190e-01 -6.94149733e-01 6.79185450e-01 3.05904835e-01
-3.78632754e-01 -1.06740057e+00 -3.43352258e-01 6.27141595e-02
-1.07280374e+00 -7.11560622e-02 -1.44593811e+00 1.35144937e+00
4.21651453e-01 1.06506300e+00 1.63510069e-01 -2.73193032e-01
3.86722267e-01 7.61147067e-02 1.18158907e-01 -7.44839609e-01
-1.34485590e+00 2.61526793e-01 8.33875895e-01 -2.64776915e-01
-4.87632126e-01 -6.02291167e-01 -1.01519608e+00 -4.46347505e-01
-1.82922170e-01 3.93144935e-01 5.44558644e-01 1.00989604e+00
-3.99313986e-01 6.82309717e-02 6.41483843e-01 -3.90155494e-01
-5.01883566e-01 -9.99196231e-01 -8.15808892e-01 5.12461960e-01
7.31843635e-02 -3.65048259e-01 -6.40443206e-01 -1.57995403e-01] | [8.800089836120605, 10.54399299621582] |
80a751aa-58a2-4c91-9b21-feb5e8dd824a | low-latency-time-domain-multichannel-speech | 2204.05609 | null | https://arxiv.org/abs/2204.05609v1 | https://arxiv.org/pdf/2204.05609v1.pdf | Low Latency Time Domain Multichannel Speech and Music Source Separation | The Goal is to obtain a simple multichannel source separation with very low latency. Applications can be teleconferencing, hearing aids, augmented reality, or selective active noise cancellation. These real time applications need a very low latency, usually less than about 6 ms, and low complexity, because they usually run on small portable devices. For that we don't need the best separation, but "useful" separation, and not just on speech, but also music and noise. Usual frequency domain approaches have higher latency and complexity. Hence we introduce a novel probabilistic optimization method which we call "Random Directions", which can overcome local minima, applied to a simple time domain unmixing structure, and which is scalable for low complexity. Then it is compared to frequency domain approaches on separating speech and music sources, and using 3D microphone setups. | ['Gerald Schuller'] | 2022-04-12 | null | null | null | null | ['music-source-separation'] | ['music'] | [ 3.08859020e-01 -3.91054749e-01 9.32835117e-02 9.96342674e-02
-1.24576461e+00 -6.67456448e-01 3.11945945e-01 -1.96478054e-01
-4.39393461e-01 6.93685710e-01 1.97681621e-01 -1.28482744e-01
-5.98382771e-01 -3.27585280e-01 -1.20785095e-01 -9.08775449e-01
-6.76386505e-02 3.12479615e-01 3.49890053e-01 -1.17436079e-02
1.32396936e-01 3.21964771e-01 -1.63385975e+00 5.20926761e-03
7.20490098e-01 8.76338840e-01 4.94194597e-01 1.12428868e+00
-4.02070343e-01 4.67406571e-01 -7.98445165e-01 1.57198817e-01
2.80317180e-02 -3.43807429e-01 -2.45836541e-01 -2.65939236e-01
8.36653635e-02 -1.12430863e-01 6.91387951e-02 1.14415467e+00
1.16494632e+00 3.33468825e-01 3.48437339e-01 -1.06412971e+00
8.32355693e-02 8.52773547e-01 -6.39448583e-01 1.92414269e-01
7.72488058e-01 -5.69499075e-01 6.74567223e-01 -6.71729326e-01
2.78961658e-01 9.94689941e-01 8.07958603e-01 5.94627261e-01
-1.20180941e+00 -7.65933812e-01 6.32881448e-02 1.24290064e-01
-1.51708305e+00 -1.07140839e+00 8.04097652e-01 -1.28535941e-01
8.03007364e-01 7.08404541e-01 3.09633195e-01 1.02217436e+00
-2.87949204e-01 7.02443361e-01 1.06495714e+00 -5.13657391e-01
3.67511451e-01 2.57350981e-01 1.67466775e-02 9.35368687e-02
1.03878388e-02 -1.40766963e-01 -7.91675687e-01 -4.31838989e-01
4.61098522e-01 -2.59705335e-01 -7.21107900e-01 -1.11344710e-01
-1.43873501e+00 3.39853138e-01 -3.61174881e-01 7.59018302e-01
-1.12386070e-01 6.76634312e-02 5.63252270e-02 4.43235040e-01
4.31843549e-01 2.49553666e-01 -5.41201949e-01 -6.13335967e-01
-1.05553472e+00 -3.40202563e-02 1.14997411e+00 7.86322355e-01
3.05632710e-01 6.58750162e-02 3.51537257e-01 1.10984576e+00
4.84580874e-01 8.82167459e-01 6.43059909e-01 -8.76788199e-01
4.07917619e-01 -2.43629277e-01 2.41982311e-01 -6.95923984e-01
-5.80779135e-01 -4.43291545e-01 -9.46081221e-01 1.41833410e-01
4.07363713e-01 -5.05645394e-01 -6.00877345e-01 1.70718265e+00
2.96784282e-01 7.59908915e-01 3.84045625e-03 9.46121633e-01
7.15982497e-01 7.86913514e-01 -5.99249125e-01 -8.15148771e-01
1.12697589e+00 -6.95262253e-01 -1.09081459e+00 -1.91014007e-01
1.91517681e-01 -1.35501909e+00 6.64548278e-01 9.41296160e-01
-1.21414435e+00 -2.43508369e-01 -9.87918973e-01 2.14386716e-01
-1.32822409e-01 -1.09806575e-01 6.38524175e-01 1.21010232e+00
-1.18261981e+00 5.40187716e-01 -7.89699256e-01 -1.68572560e-01
-3.68875325e-01 6.47695780e-01 -1.70639768e-01 2.53238350e-01
-7.55414665e-01 5.71317196e-01 -2.46028751e-01 1.98040649e-01
-3.66001427e-01 -5.91902435e-01 -6.13038599e-01 6.76855743e-02
3.15092385e-01 -4.59364414e-01 1.26628244e+00 -8.91245902e-01
-2.15714335e+00 4.64756846e-01 -6.29730999e-01 -1.87631130e-01
2.12163404e-01 -4.82764721e-01 -8.42122734e-01 -4.33678254e-02
-1.57529879e-02 -1.00704797e-01 1.13925564e+00 -1.02515626e+00
-4.31735426e-01 -1.65707707e-01 -3.15504104e-01 2.41689354e-01
-3.17814559e-01 5.29282987e-01 -3.69138271e-01 -7.17199028e-01
6.04123652e-01 -7.50888050e-01 -2.65386969e-01 -3.68777931e-01
-3.25573772e-01 3.12319070e-01 6.93286955e-01 -6.70009494e-01
1.25634551e+00 -2.45078564e+00 3.16923052e-01 4.55469847e-01
-1.62152156e-01 1.47885725e-01 -7.94978216e-02 2.88940042e-01
-1.72152877e-01 -2.20967829e-01 -2.69444615e-01 -5.46007276e-01
1.25643136e-02 -1.86663017e-01 -2.38061473e-01 5.64999878e-01
-4.02107596e-01 1.73410982e-01 -7.65399575e-01 -1.38789296e-01
3.29148650e-01 5.46431243e-01 -5.50434649e-01 -2.37266384e-02
3.46523404e-01 4.80946690e-01 1.28282961e-02 5.58970869e-01
9.19579029e-01 8.36564079e-02 2.98739791e-01 3.28042731e-02
-3.49748611e-01 4.15879548e-01 -2.12005377e+00 1.89428174e+00
-8.30021918e-01 8.26749921e-01 8.61406982e-01 -9.64905322e-01
9.60027218e-01 8.41129601e-01 5.41498959e-01 -3.18995804e-01
2.15312675e-01 7.09371090e-01 -2.15893745e-01 -3.60274374e-01
3.08735996e-01 -2.38339111e-01 2.03496963e-01 5.32975316e-01
6.58150688e-02 -1.24521770e-01 -1.41137213e-01 -3.77965234e-02
1.00445437e+00 -1.51025921e-01 3.41016710e-01 -3.40010673e-01
6.73460126e-01 -5.46852231e-01 4.26948547e-01 6.15651548e-01
6.58876076e-02 8.36316466e-01 1.20952897e-01 4.93600726e-01
-3.05864751e-01 -1.06895840e+00 3.24522555e-02 1.04683936e+00
3.04925263e-01 -4.91082877e-01 -5.13902783e-01 -1.36912568e-02
-4.01923835e-01 3.09382737e-01 1.55033052e-01 2.24887177e-01
-6.48720741e-01 -7.97921181e-01 5.03406763e-01 1.79136485e-01
1.50555417e-01 -5.52751601e-01 -1.37339249e-01 4.78486717e-01
-3.86737525e-01 -1.16942048e+00 -5.72693467e-01 4.50792223e-01
-1.00474989e+00 -5.57197392e-01 -1.02923644e+00 -8.88164699e-01
1.31404310e-01 6.25667989e-01 8.59231174e-01 -3.88922811e-01
-1.00199975e-01 5.40796041e-01 -1.66609704e-01 -4.85353738e-01
8.79452303e-02 -5.86568490e-02 4.55283284e-01 2.41102815e-01
2.05751568e-01 -1.23306680e+00 -4.44265366e-01 3.62497598e-01
-5.16691446e-01 -4.70718384e-01 4.56321836e-01 5.03398180e-01
2.63775617e-01 3.48114133e-01 5.77388465e-01 -4.29696769e-01
6.67169273e-01 -2.51062721e-01 -6.28656507e-01 -2.60965731e-02
-2.06552535e-01 -1.97344795e-01 6.13092303e-01 -7.49982536e-01
-1.21098590e+00 1.82275653e-01 -3.48834187e-01 -2.57564038e-01
-3.02385926e-01 2.55067110e-01 -4.45197344e-01 -2.32536316e-01
7.27247953e-01 7.35254660e-02 -2.20886007e-01 -1.04413664e+00
1.96867406e-01 1.14521229e+00 5.28118670e-01 -3.02365631e-01
6.46417916e-01 6.03264809e-01 -2.09138036e-01 -1.37679899e+00
-2.13124260e-01 -1.04833591e+00 -2.55582362e-01 1.06478810e-01
4.09417838e-01 -8.74420702e-01 -6.41331971e-01 5.50932765e-01
-1.28507805e+00 -1.13090150e-01 -5.86211234e-02 1.14902890e+00
-4.74124223e-01 4.62693691e-01 -4.29287821e-01 -1.32860327e+00
-7.97894672e-02 -7.82373428e-01 1.07199490e+00 2.23494500e-01
-2.68856525e-01 -8.72905731e-01 3.50920707e-01 1.96723178e-01
4.67933089e-01 -2.85496384e-01 2.55481154e-01 -2.91475266e-01
-6.21393919e-01 -9.51809362e-02 3.83915263e-03 3.42605084e-01
3.22891206e-01 -1.82083949e-01 -1.32145250e+00 -9.48595628e-02
4.88587320e-01 3.15185457e-01 5.35640180e-01 7.68014371e-01
7.28349745e-01 -4.18524370e-02 -5.11226237e-01 6.70906782e-01
1.33051240e+00 4.60335672e-01 5.29931426e-01 -5.63472658e-02
3.33730996e-01 4.79133338e-01 4.27757919e-01 4.71990764e-01
-1.38767913e-01 8.66416991e-01 -4.23691273e-02 -9.65093728e-03
-2.74261869e-02 2.77638674e-01 5.50064862e-01 1.48792052e+00
-1.89706296e-01 -3.56775999e-01 -5.98379254e-01 5.84620774e-01
-1.89826989e+00 -9.76550341e-01 -3.93085808e-01 2.51194477e+00
6.58219635e-01 -1.05730705e-01 7.86212906e-02 6.41573250e-01
7.20379412e-01 1.04323246e-01 -1.10580288e-01 -2.47151360e-01
-8.25297460e-02 6.14639997e-01 5.61319709e-01 9.21480954e-01
-1.09088218e+00 3.77654225e-01 6.99022007e+00 1.15046501e+00
-1.23397911e+00 4.44951385e-01 -2.24184453e-01 -4.10941601e-01
-2.16704294e-01 -1.98863313e-01 -5.55501759e-01 4.99480397e-01
1.06525600e+00 -3.11478251e-03 7.01362133e-01 6.50246859e-01
3.71003836e-01 -3.24581116e-01 -9.28179741e-01 1.76302421e+00
7.39319772e-02 -8.03228378e-01 -6.85437322e-01 -6.90370351e-02
2.98739940e-01 -2.03427628e-01 -1.80254117e-01 -1.93459556e-01
-1.58907995e-01 -9.06086445e-01 4.73283470e-01 3.40129465e-01
5.04950285e-01 -7.77839422e-01 6.02007031e-01 6.56167746e-01
-1.32382965e+00 4.63516824e-02 -2.81778187e-01 -9.26836133e-02
5.44334888e-01 1.04045951e+00 -2.65343785e-01 4.63018894e-01
8.14580977e-01 3.10620666e-01 3.66127819e-01 1.41126919e+00
8.57831016e-02 4.98532385e-01 -8.47514093e-01 -1.72472686e-01
-2.11874977e-01 -2.33206406e-01 9.72144067e-01 1.27878869e+00
9.33574319e-01 1.94734290e-01 -1.59451202e-01 2.81419810e-02
3.88974458e-01 2.16507062e-01 -4.98398632e-01 1.97098956e-01
3.38313639e-01 1.06634867e+00 -9.42319095e-01 -3.11649255e-02
-3.54553342e-01 1.06601357e+00 -6.31525934e-01 4.61612195e-01
-8.43721151e-01 -9.42949176e-01 7.01463044e-01 -3.49842496e-02
1.75086245e-01 -6.11341059e-01 -2.30727702e-01 -1.29941952e+00
1.92519948e-01 -1.03175664e+00 -2.20371708e-02 -5.83649635e-01
-8.35912526e-01 4.61470395e-01 -2.06577361e-01 -1.50637269e+00
-3.11004072e-01 -6.11534953e-01 -3.67607713e-01 8.46255660e-01
-1.31435084e+00 -2.84482181e-01 -8.59989151e-02 1.05644166e+00
4.32471961e-01 -2.00022295e-01 9.55454350e-01 1.02388036e+00
-2.46157065e-01 4.83064085e-01 5.01465678e-01 -4.22719777e-01
8.75466406e-01 -1.25240791e+00 -1.10823818e-01 7.27402568e-01
4.66375083e-01 6.28831089e-01 8.59338224e-01 -1.20338075e-01
-1.42550635e+00 -3.84752065e-01 1.15013039e+00 -3.00167829e-01
5.59775710e-01 -5.65295279e-01 -5.88209271e-01 1.12853153e-02
1.60979539e-01 -3.82214338e-01 9.97983158e-01 4.36749369e-01
4.75605130e-02 -3.20034057e-01 -1.03099859e+00 4.39000726e-01
1.14098692e+00 -6.78984642e-01 -5.16823947e-01 5.32328904e-01
5.43978095e-01 -2.56843060e-01 -3.89085680e-01 1.27629220e-01
7.14595258e-01 -1.11352432e+00 1.20742571e+00 1.26447007e-01
-4.39739645e-01 -5.22531986e-01 -3.69651794e-01 -1.16340864e+00
-1.59523234e-01 -1.48256111e+00 -1.48937240e-01 1.38109565e+00
5.72861314e-01 -8.61189663e-01 6.63460016e-01 6.61736131e-02
8.70859623e-02 -3.29133607e-02 -1.27312934e+00 -9.75567937e-01
-6.99442923e-01 -1.01699901e+00 5.26114643e-01 8.31000090e-01
1.00484990e-01 4.70287651e-01 -7.92924106e-01 4.63414371e-01
6.98225796e-01 7.23446235e-02 5.82099319e-01 -1.26127040e+00
-8.39449167e-01 -3.12392831e-01 -3.41986835e-01 -1.50534570e+00
-2.33961985e-01 -3.65920484e-01 2.65272975e-01 -1.21788085e+00
-3.10228944e-01 -5.35922825e-01 -3.73202264e-01 -2.15215340e-01
1.56868383e-01 3.52464646e-01 -1.17802925e-01 5.53944856e-02
-3.87887508e-01 3.18061471e-01 6.19269848e-01 -6.77722925e-03
-6.95405483e-01 6.30515873e-01 -5.82877874e-01 1.04351997e+00
5.18616080e-01 -5.66280544e-01 -5.72419643e-01 -5.97321093e-01
3.45255315e-01 4.02199417e-01 7.53907114e-02 -1.19180810e+00
5.58286786e-01 1.14733197e-01 -2.71335058e-02 -3.44289809e-01
8.53058577e-01 -1.06025577e+00 2.93555826e-01 1.45186000e-02
7.57145435e-02 -5.00276029e-01 1.24693498e-01 5.77899814e-01
-5.08992970e-01 -4.26182538e-01 5.59741259e-01 1.34208193e-02
-2.61666000e-01 -1.02535181e-01 -7.10922956e-01 -2.78488547e-01
6.43654823e-01 -3.01211804e-01 2.20611721e-01 -8.77710104e-01
-1.19779444e+00 -2.32117549e-01 -6.51279613e-02 2.59382010e-01
4.00664568e-01 -1.20945716e+00 -3.86947840e-01 2.32415140e-01
-6.08262956e-01 -1.97910771e-01 3.78155172e-01 1.02584183e+00
-2.95501918e-01 3.93771440e-01 2.38097817e-01 -7.69909263e-01
-1.63572645e+00 3.27529997e-01 2.42058113e-01 7.74670094e-02
-3.06211203e-01 1.13297558e+00 -5.50365448e-03 -5.01080394e-01
5.64832926e-01 -3.89951110e-01 -2.86851436e-01 2.24446163e-01
7.92880952e-01 7.53879845e-01 2.45869771e-01 -3.99400324e-01
-7.08150744e-01 1.10159922e+00 6.53898358e-01 -8.29470634e-01
1.06643760e+00 -4.30626929e-01 -3.14202726e-01 6.27741337e-01
1.14620888e+00 9.88051116e-01 -5.65694869e-01 -1.11049362e-01
6.74030483e-02 -6.60327911e-01 2.70525105e-02 -3.80378157e-01
-7.54606187e-01 9.15728211e-01 9.34800327e-01 6.37248278e-01
1.60251081e+00 -1.65881246e-01 6.97287679e-01 5.57745039e-01
6.58608615e-01 -9.11552012e-01 -3.03663611e-01 4.14710909e-01
6.29901826e-01 -8.40692699e-01 -1.62146315e-01 -6.28497243e-01
-8.37853327e-02 1.04574955e+00 -2.63234302e-02 -3.54693197e-02
1.09561515e+00 8.69119763e-01 2.94359624e-01 3.56881529e-01
-5.44957101e-01 -4.00699556e-01 2.11968184e-01 8.09983790e-01
5.58316350e-01 -3.40427086e-02 -1.88140541e-01 7.95281053e-01
-1.97922692e-01 1.54397143e-02 2.27755591e-01 7.50284493e-01
-5.93378901e-01 -1.43047476e+00 -8.40658963e-01 1.85790852e-01
-8.58985305e-01 -2.01444164e-01 -1.95389748e-01 1.99518487e-01
1.87991649e-01 1.43627346e+00 -1.38209924e-01 -3.41235787e-01
3.86459529e-01 7.19135329e-02 4.24141318e-01 -5.48654974e-01
-4.15065646e-01 8.47399712e-01 2.34583318e-01 -6.88448787e-01
-8.27160418e-01 -5.27699769e-01 -8.14147592e-01 -2.59316415e-01
-6.77484453e-01 4.69385415e-01 8.70240927e-01 7.87951827e-01
2.53311366e-01 4.61613268e-01 6.69046521e-01 -1.03073311e+00
-1.13991074e-01 -8.62154484e-01 -8.92200589e-01 -3.05370271e-01
6.70276225e-01 -3.46683323e-01 -6.06360555e-01 -2.17620894e-01] | [15.172436714172363, 5.653975486755371] |
3d9b335c-2c52-4627-b0e5-7910dad5a22d | from-hero-to-z-eroe-a-benchmark-of-low-level | null | null | https://aclanthology.org/2020.aacl-main.79 | https://aclanthology.org/2020.aacl-main.79.pdf | From Hero to Z\'eroe: A Benchmark of Low-Level Adversarial Attacks | Adversarial attacks are label-preserving modifications to inputs of machine learning classifiers designed to fool machines but not humans. Natural Language Processing (NLP) has mostly focused on high-level attack scenarios such as paraphrasing input texts. We argue that these are less realistic in typical application scenarios such as in social media, and instead focus on low-level attacks on the character-level. Guided by human cognitive abilities and human robustness, we propose the first large-scale catalogue and benchmark of low-level adversarial attacks, which we dub Z{\'e}roe, encompassing nine different attack modes including visual and phonetic adversaries. We show that RoBERTa, NLP{'}s current workhorse, fails on our attacks. Our dataset provides a benchmark for testing robustness of future more human-like NLP models. | ['Yannik Benz', 'Steffen Eger'] | 2020-12-01 | null | null | null | asian-chapter-of-the-association-for | ['toxic-comment-classification'] | ['natural-language-processing'] | [ 5.80366910e-01 2.97756612e-01 1.58320606e-01 -4.51247357e-02
-8.83823752e-01 -1.41517174e+00 1.09631288e+00 8.85705575e-02
-4.05385524e-01 5.75428367e-01 -8.28087423e-03 -7.98637986e-01
1.67079210e-01 -6.86050951e-01 -8.77117038e-01 -3.35508466e-01
1.38828650e-01 4.20116931e-01 2.69198567e-01 -2.69545585e-01
2.08538473e-01 8.56229663e-01 -8.28469992e-01 6.73812330e-01
3.76167655e-01 2.26643011e-01 -7.32072413e-01 1.19818866e+00
3.35638076e-01 9.17911112e-01 -1.01664829e+00 -1.28072298e+00
5.27267814e-01 -2.42961869e-01 -1.10608673e+00 -4.59722251e-01
8.48114371e-01 -7.47911409e-02 -6.75477505e-01 1.21044374e+00
7.17592239e-01 -6.57050163e-02 6.72709703e-01 -1.61911809e+00
-8.53777707e-01 8.15652788e-01 -1.80219620e-01 1.87449649e-01
6.04402125e-01 8.86226535e-01 7.83046961e-01 -5.16578317e-01
6.57903016e-01 1.65349662e+00 8.80490243e-01 1.10778725e+00
-1.40643919e+00 -7.79068947e-01 -1.80090532e-01 7.14671761e-02
-1.03958678e+00 -4.78617251e-01 6.83627725e-01 -3.49496126e-01
8.32456887e-01 5.84348202e-01 -1.64383650e-01 1.98272705e+00
5.06905735e-01 6.47611558e-01 1.38562059e+00 -4.53091860e-01
1.17910311e-01 1.97566032e-01 3.06000765e-02 5.91525316e-01
1.26030684e-01 3.53896350e-01 -2.18788639e-01 -7.53619790e-01
2.92514026e-01 -4.27305222e-01 -1.92379862e-01 -2.14522853e-02
-1.38238668e+00 9.87601638e-01 3.04995924e-01 1.12914592e-01
1.03606418e-01 3.36629361e-01 6.35644257e-01 5.79342067e-01
4.95775081e-02 9.98611212e-01 -3.43893886e-01 1.07495368e-01
-4.87075359e-01 6.06393218e-01 1.06554413e+00 9.37536418e-01
1.64907873e-01 -4.24401052e-02 -3.48037332e-01 5.41625202e-01
-8.93390849e-02 6.36729896e-01 2.85069883e-01 -9.66306269e-01
6.05431080e-01 3.62016708e-02 1.25891462e-01 -1.04991984e+00
-3.94864947e-01 -1.11308426e-01 -6.39376700e-01 3.61174315e-01
8.03676128e-01 -3.78906637e-01 -6.72029555e-01 1.79036033e+00
-5.46629280e-02 7.45512769e-02 3.05534631e-01 5.10511398e-01
5.35712600e-01 5.06877542e-01 3.88088375e-01 9.37797353e-02
1.31722009e+00 -6.31136894e-01 -3.32918614e-01 -3.95003915e-01
6.74419880e-01 -8.28963518e-01 1.53725898e+00 4.57437068e-01
-1.12347746e+00 -3.79757255e-01 -1.05070448e+00 -1.82305425e-02
-7.46402562e-01 -5.51532209e-01 5.06966054e-01 1.40294659e+00
-6.85464323e-01 7.42818594e-01 -3.73176366e-01 -3.54150951e-01
6.27669871e-01 2.14618728e-01 -5.44945776e-01 4.07058187e-02
-1.71976113e+00 1.21734774e+00 1.94011822e-01 -3.06990862e-01
-1.03588581e+00 -6.90397322e-01 -5.78642428e-01 -1.93606213e-01
2.40590855e-01 -5.79333186e-01 1.33460629e+00 -8.90926600e-01
-1.40097809e+00 1.38837969e+00 3.32057387e-01 -9.37854886e-01
1.06375039e+00 -1.65680885e-01 -7.12047219e-01 1.33409962e-01
-2.83156842e-01 6.37475967e-01 9.46138859e-01 -1.30195665e+00
-8.60250369e-02 -9.87783819e-02 4.98718560e-01 -1.11337386e-01
-3.36282611e-01 6.29344463e-01 1.16447262e-01 -9.72413599e-01
-7.48898506e-01 -1.16885579e+00 -1.06773168e-01 6.83576241e-02
-8.00267279e-01 2.37474255e-02 9.40631509e-01 -4.78323907e-01
9.75762784e-01 -2.03027678e+00 -1.78071067e-01 1.94000423e-01
1.26563966e-01 7.58850098e-01 -3.01249683e-01 4.94247705e-01
-2.79603660e-01 7.95018971e-01 -3.72885466e-01 -1.61960691e-01
4.27926660e-01 6.11550640e-03 -9.00035322e-01 6.73733652e-01
2.32760429e-01 1.20212698e+00 -8.10748160e-01 -4.19146627e-01
-4.66119349e-02 1.49340257e-01 -6.89093411e-01 5.04694097e-02
-3.91288102e-01 2.59043455e-01 -1.51896819e-01 4.45511401e-01
3.54067534e-01 1.27537176e-01 -2.53136933e-01 2.21570179e-01
5.05509555e-01 4.96124700e-02 -6.06584489e-01 1.26055706e+00
-4.28319842e-01 8.66269648e-01 -2.10448932e-02 -5.54462910e-01
6.40591919e-01 1.62391379e-01 -3.55998933e-01 -2.66119123e-01
1.31022677e-01 -4.07848619e-02 2.93613881e-01 -3.70936692e-01
2.60361284e-01 -3.00259769e-01 -7.15024054e-01 6.07487082e-01
-1.33915290e-01 -5.48121691e-01 -2.58740991e-01 5.83406210e-01
1.62056804e+00 -1.06620975e-01 2.27754161e-01 -2.79878676e-01
5.60623586e-01 -1.12406678e-01 1.51188329e-01 1.48501146e+00
-7.17718959e-01 6.71780765e-01 7.37597048e-01 -2.48025775e-01
-1.41961646e+00 -1.20965683e+00 -1.08505048e-01 1.18787730e+00
-1.46440208e-01 -2.35266209e-01 -1.20067334e+00 -1.19524002e+00
1.56810224e-01 1.10979581e+00 -5.79860151e-01 -5.40203750e-01
-7.29844034e-01 -5.41450799e-01 1.73656154e+00 3.40180695e-01
4.56441164e-01 -1.41415906e+00 -2.64747858e-01 -4.03149165e-02
1.60104260e-01 -1.21365309e+00 -4.40199554e-01 5.44825643e-02
-1.86601594e-01 -1.19360864e+00 -3.96831095e-01 -6.31746173e-01
3.78829062e-01 -3.77182305e-01 1.22788525e+00 1.53700396e-01
-4.18124139e-01 4.69753742e-01 -3.77324253e-01 -7.53561914e-01
-1.32468736e+00 1.09866783e-01 2.38032699e-01 -1.93923041e-01
3.87876630e-01 -5.60765684e-01 -1.06253706e-01 5.38980484e-01
-1.09549391e+00 -3.56771648e-01 3.11899662e-01 7.09225893e-01
-9.91651416e-02 2.21411493e-02 5.47485888e-01 -1.59518421e+00
1.05556846e+00 -3.67811114e-01 -4.04597729e-01 3.03642511e-01
-4.91917878e-02 -2.39509180e-01 1.29204750e+00 -8.64032924e-01
-7.36989141e-01 -1.30323753e-01 -4.17143345e-01 -4.86114591e-01
-6.42551422e-01 -8.03841278e-02 -5.05768716e-01 -4.01798844e-01
1.39164889e+00 4.21557687e-02 -4.62060720e-01 -2.28047311e-01
6.19495511e-01 6.15461111e-01 9.89559948e-01 -9.16594923e-01
1.58568013e+00 2.06294075e-01 -1.15302382e-02 -7.64772356e-01
-6.15141749e-01 3.25457603e-01 -4.62920964e-01 -2.66218260e-02
7.05037892e-01 -4.28747088e-01 -1.15985107e+00 5.84436953e-01
-1.43314183e+00 -4.54601526e-01 -2.80167043e-01 -5.98006770e-02
-7.26965964e-01 5.25917530e-01 -8.58064175e-01 -5.79518616e-01
-2.43834779e-01 -8.67537141e-01 6.09490216e-01 -2.44741470e-01
-5.87521017e-01 -8.88078749e-01 3.85912843e-02 5.79603493e-01
3.69759262e-01 5.93446434e-01 1.28242731e+00 -1.39826596e+00
-1.65931538e-01 -5.93635321e-01 -1.54340267e-03 3.67212206e-01
-3.84287775e-01 7.61935562e-02 -1.21595550e+00 -4.62323874e-01
5.50520979e-02 -7.99139917e-01 5.46247721e-01 -4.98769224e-01
1.23580754e+00 -8.63475621e-01 -2.26799905e-01 4.96791869e-01
9.47050214e-01 -5.02578169e-02 7.42532492e-01 2.12642044e-01
7.76648164e-01 6.85246944e-01 1.66523039e-01 -1.40666086e-02
-3.42361659e-01 5.79895914e-01 4.21580762e-01 5.79615422e-02
6.39092773e-02 -5.89891613e-01 6.22433424e-01 6.06081448e-02
4.65013295e-01 -8.46928179e-01 -1.20620704e+00 2.20933706e-01
-1.53105640e+00 -1.21962178e+00 1.31441176e-01 1.74818575e+00
9.89791095e-01 7.85899043e-01 9.57384259e-02 3.48073900e-01
9.43991601e-01 2.38937736e-01 -6.11796796e-01 -1.03596807e+00
-2.11440504e-01 2.95182437e-01 6.06989920e-01 5.91744065e-01
-1.41145730e+00 1.29989421e+00 7.07518339e+00 9.98971403e-01
-7.22606361e-01 1.57799393e-01 5.43178022e-01 -1.77483588e-01
-3.23095083e-01 -1.14221551e-01 -5.99419296e-01 6.37539506e-01
1.25623274e+00 -2.84976959e-01 5.83705246e-01 7.58233964e-01
-2.47967988e-01 6.73077345e-01 -1.40509725e+00 5.26916087e-01
2.19440654e-01 -1.15629804e+00 2.03844666e-01 3.71381752e-02
5.75845599e-01 -2.08963692e-01 3.61611068e-01 5.71559608e-01
1.11170721e+00 -1.52085793e+00 7.39675164e-01 1.91023126e-01
7.06200480e-01 -1.00298429e+00 5.51218629e-01 5.77714145e-01
-3.74341518e-01 -6.65511787e-02 -2.96570122e-01 3.66788991e-02
-5.44224679e-02 4.05530706e-02 -6.92903221e-01 7.24819452e-02
2.21406996e-01 -4.53306921e-03 -8.97721291e-01 3.91201317e-01
-4.84999537e-01 7.67488539e-01 -2.19095170e-01 -1.20026143e-02
2.81066984e-01 6.23091161e-01 8.43138516e-01 1.55267656e+00
-3.88136744e-01 1.32937700e-01 5.59928082e-02 7.75507689e-01
-5.22290766e-01 -2.01682374e-01 -1.08433688e+00 -2.72860706e-01
5.90949416e-01 1.01521349e+00 -5.14993608e-01 -1.08277798e-01
-1.82752505e-01 1.08622861e+00 3.28323305e-01 1.53387010e-01
-1.03668523e+00 -6.69982433e-01 7.27045774e-01 2.35060528e-02
-2.35328525e-01 -1.80365384e-01 -2.15405539e-01 -1.10109198e+00
-1.96085215e-01 -1.67489851e+00 3.83934230e-01 -7.17645884e-01
-1.67702556e+00 6.16507053e-01 -4.49216552e-02 -9.19374585e-01
-2.66982645e-01 -8.83878469e-01 -7.55380034e-01 7.95605779e-01
-8.20399761e-01 -1.43799984e+00 3.58392239e-01 9.43707407e-01
3.55443269e-01 -5.31205893e-01 9.90206242e-01 -1.16784863e-01
-5.26725829e-01 1.29015434e+00 -2.64898032e-01 5.06403863e-01
8.31618190e-01 -1.19118571e+00 1.17148197e+00 1.17740667e+00
3.92390430e-01 7.70318270e-01 1.12664354e+00 -5.14581084e-01
-1.21088266e+00 -1.06667387e+00 9.26589072e-01 -1.31020236e+00
9.45449591e-01 -8.73877347e-01 -9.23801064e-01 9.17970955e-01
1.47036120e-01 9.35903415e-02 5.43718755e-01 -2.18496755e-01
-8.37990761e-01 4.92955089e-01 -1.56625736e+00 1.01374114e+00
1.18565524e+00 -1.17980421e+00 -8.90858769e-01 7.60130763e-01
1.01640761e+00 -3.36408734e-01 -6.33445323e-01 3.25507015e-01
4.11533386e-01 -8.06721807e-01 1.26703501e+00 -1.37933540e+00
5.59760451e-01 5.62037015e-03 -1.28177628e-01 -1.08976793e+00
-2.05599800e-01 -1.04567182e+00 2.07126290e-01 1.36884832e+00
3.67270321e-01 -7.82451153e-01 8.11286092e-01 8.57078791e-01
2.66058594e-01 -2.03856885e-01 -7.77242541e-01 -9.04285431e-01
8.03586483e-01 -5.68933308e-01 4.15459931e-01 1.45223486e+00
-5.07706665e-02 2.18406573e-01 -5.30550539e-01 3.94305348e-01
4.84049886e-01 -4.23337519e-01 1.04367185e+00 -9.54768836e-01
-5.65804005e-01 -5.52696347e-01 -3.96730661e-01 -4.69054699e-01
7.42539942e-01 -1.05617678e+00 -5.94993047e-02 -4.98269767e-01
-1.07993465e-02 -8.34860131e-02 2.30042799e-03 6.08059108e-01
-1.64946660e-01 6.07324064e-01 4.74538535e-01 1.93030328e-01
-2.91580647e-01 7.17917979e-02 8.41976166e-01 -4.07488048e-01
3.41440499e-01 3.92884277e-02 -7.04867363e-01 1.07430530e+00
8.38043272e-01 -8.55254233e-01 -2.00620264e-01 -2.05438226e-01
2.69203395e-01 -2.37440035e-01 7.44625092e-01 -1.05481458e+00
2.72546202e-01 -3.63109738e-01 3.15818220e-01 1.91039681e-01
3.45871449e-02 -6.61747277e-01 -1.25449346e-02 5.82057953e-01
-1.00688076e+00 1.40903860e-01 4.00114357e-01 4.13772970e-01
2.52484947e-01 -3.35437983e-01 1.19499683e+00 -2.41751537e-01
-2.41400152e-01 1.84829071e-01 -4.24151242e-01 4.17170197e-01
1.10596383e+00 -3.17855319e-03 -8.96300912e-01 -4.40774173e-01
-7.67856359e-01 5.47930636e-02 7.48987496e-01 4.84805822e-01
2.65624523e-01 -9.46804821e-01 -9.78177249e-01 7.67630339e-02
1.32237747e-01 -6.75924361e-01 7.59957284e-02 -5.53223900e-02
-7.70234644e-01 3.47570390e-01 -4.41203237e-01 1.00099362e-01
-1.42548609e+00 1.10176623e+00 4.52754319e-01 -4.31122094e-01
-4.85813349e-01 1.05097139e+00 2.38584325e-01 -6.08665884e-01
1.90797508e-01 3.72675270e-01 1.07649505e-01 -3.47473949e-01
4.36219066e-01 2.63154596e-01 2.57763034e-03 -6.72693312e-01
-5.63918948e-01 6.82699457e-02 -2.21248865e-01 -3.15323174e-01
7.16949105e-01 3.39058757e-01 1.69564828e-01 2.69383863e-02
1.24075329e+00 3.48013997e-01 -8.45963299e-01 -2.78761089e-01
-3.73730250e-02 -5.18026590e-01 -4.44352001e-01 -1.08591163e+00
-3.55666816e-01 1.08814764e+00 3.49312901e-01 4.80586350e-01
6.34918630e-01 -1.81816086e-01 7.27454126e-01 8.61493766e-01
3.61839712e-01 -6.77798688e-01 1.77952260e-01 4.76345122e-01
1.14657736e+00 -8.62641811e-01 -5.36587834e-03 -3.00605327e-01
-7.16126084e-01 7.66515017e-01 5.57504594e-01 -2.58039534e-01
2.42985994e-01 3.53583068e-01 2.82205075e-01 2.86994249e-01
-7.53841460e-01 3.49262655e-01 2.07579918e-02 9.22595263e-01
5.67062013e-02 -4.85284701e-02 -7.79775996e-03 5.77486098e-01
-6.64921880e-01 -4.77163315e-01 6.42485797e-01 8.75816107e-01
-7.89292902e-02 -1.29087782e+00 -6.10076010e-01 8.21992233e-02
-9.45465267e-01 -2.42249161e-01 -1.05249965e+00 8.94463539e-01
2.84346361e-02 9.77395654e-01 -2.78448164e-01 -5.61353207e-01
3.93991798e-01 3.75148714e-01 4.57892299e-01 -5.61053216e-01
-1.24223709e+00 -7.41754174e-01 3.65109324e-01 -4.92280126e-01
2.84940377e-02 -4.76234913e-01 -6.54940188e-01 -9.84337091e-01
2.00789988e-01 -1.30627349e-01 3.21860880e-01 7.64557779e-01
1.32714987e-01 3.77009474e-02 6.07806921e-01 -7.67993927e-01
-1.00790489e+00 -6.36690974e-01 -3.13333392e-01 9.88590896e-01
2.20313877e-01 -9.40784067e-02 -6.54693186e-01 1.24116763e-01] | [6.009475231170654, 8.071562767028809] |
3889cc04-8e03-40ff-936a-6a84c015f583 | hierarchical-neural-architecture-search-for-1 | 2010.13501 | null | https://arxiv.org/abs/2010.13501v1 | https://arxiv.org/pdf/2010.13501v1.pdf | Hierarchical Neural Architecture Search for Deep Stereo Matching | To reduce the human efforts in neural network design, Neural Architecture Search (NAS) has been applied with remarkable success to various high-level vision tasks such as classification and semantic segmentation. The underlying idea for the NAS algorithm is straightforward, namely, to enable the network the ability to choose among a set of operations (e.g., convolution with different filter sizes), one is able to find an optimal architecture that is better adapted to the problem at hand. However, so far the success of NAS has not been enjoyed by low-level geometric vision tasks such as stereo matching. This is partly due to the fact that state-of-the-art deep stereo matching networks, designed by humans, are already sheer in size. Directly applying the NAS to such massive structures is computationally prohibitive based on the currently available mainstream computing resources. In this paper, we propose the first end-to-end hierarchical NAS framework for deep stereo matching by incorporating task-specific human knowledge into the neural architecture search framework. Specifically, following the gold standard pipeline for deep stereo matching (i.e., feature extraction -- feature volume construction and dense matching), we optimize the architectures of the entire pipeline jointly. Extensive experiments show that our searched network outperforms all state-of-the-art deep stereo matching architectures and is ranked at the top 1 accuracy on KITTI stereo 2012, 2015 and Middlebury benchmarks, as well as the top 1 on SceneFlow dataset with a substantial improvement on the size of the network and the speed of inference. The code is available at https://github.com/XuelianCheng/LEAStereo. | ['ZongYuan Ge', 'Hongdong Li', 'Tom Drummond', 'Xiaojun Chang', 'Yuchao Dai', 'Mehrtash Harandi', 'Yiran Zhong', 'Xuelian Cheng'] | 2020-10-26 | null | http://proceedings.neurips.cc/paper/2020/hash/fc146be0b230d7e0a92e66a6114b840d-Abstract.html | http://proceedings.neurips.cc/paper/2020/file/fc146be0b230d7e0a92e66a6114b840d-Paper.pdf | neurips-2020-12 | ['stereo-depth-estimation'] | ['computer-vision'] | [ 1.09145045e-01 -1.77668810e-01 1.30815744e-01 -4.03661340e-01
-4.96011138e-01 -4.34140563e-01 4.79388386e-01 -1.75537959e-01
-6.76483333e-01 1.50388315e-01 1.91887692e-02 -2.80679584e-01
2.96072569e-02 -7.91201174e-01 -8.38331878e-01 -4.10908073e-01
3.00436616e-01 5.85410774e-01 4.34579074e-01 -2.09154323e-01
3.61744434e-01 5.29314756e-01 -1.84855354e+00 1.67719841e-01
8.16175878e-01 1.37300324e+00 4.37987089e-01 3.33077937e-01
-7.20453560e-02 4.32980984e-01 1.12547306e-02 -6.00461662e-01
6.22618139e-01 -1.81438029e-01 -1.03177524e+00 -1.55378118e-01
1.00800228e+00 -3.84688407e-01 -5.03851950e-01 1.20163131e+00
6.46464884e-01 9.10060853e-02 3.32796127e-01 -9.77753043e-01
-1.45188898e-01 4.65299398e-01 -4.95174557e-01 1.46601528e-01
-1.47977680e-01 2.93813020e-01 1.18858778e+00 -1.11216676e+00
4.70209330e-01 1.22927868e+00 5.99708259e-01 3.70308191e-01
-1.17485511e+00 -6.82049513e-01 3.48774232e-02 3.40544045e-01
-1.39420521e+00 -5.60095310e-01 6.08764052e-01 -4.89824951e-01
1.11088932e+00 1.97262107e-03 7.61763632e-01 6.54535353e-01
-8.68762732e-02 7.95881629e-01 9.40532207e-01 -1.80594489e-01
2.40580305e-01 -3.92822176e-01 1.74134336e-02 9.00163651e-01
9.14935544e-02 2.87198037e-01 -5.07258713e-01 1.20773040e-01
8.05521071e-01 -1.11722261e-01 -1.05442300e-01 -6.11655176e-01
-1.10531557e+00 7.52834857e-01 9.80294883e-01 1.80207163e-01
-3.31890851e-01 2.01434821e-01 5.11018336e-01 1.85163960e-01
2.59007752e-01 5.22680521e-01 -5.27489483e-01 7.42237270e-02
-1.09693992e+00 4.50407565e-01 5.16818583e-01 7.68541157e-01
8.95167351e-01 -1.46128893e-01 -1.45557329e-01 9.03246224e-01
1.12110928e-01 1.57274246e-01 2.96316326e-01 -1.15317702e+00
4.87730563e-01 8.24004948e-01 -2.99349040e-01 -7.31821299e-01
-4.69491869e-01 -6.60243928e-01 -9.51681376e-01 3.84727240e-01
6.69302940e-01 1.79621205e-01 -1.09713566e+00 1.54658997e+00
3.73452991e-01 2.38763973e-01 -2.79345453e-01 1.19865274e+00
9.64326143e-01 2.03025997e-01 -1.02257416e-01 5.31692684e-01
1.49962139e+00 -1.18884218e+00 7.53987283e-02 -6.49374485e-01
4.58012760e-01 -7.69391000e-01 1.05275321e+00 1.62778407e-01
-1.04563177e+00 -6.73250020e-01 -9.27332699e-01 -4.03810143e-01
-2.15424389e-01 -9.80208963e-02 8.45310628e-01 3.66832972e-01
-1.10382152e+00 7.17458963e-01 -8.31756949e-01 -3.75877798e-01
9.31258500e-01 5.18098116e-01 -3.07410270e-01 -2.68516600e-01
-9.41619873e-01 7.00152040e-01 4.18982029e-01 2.37945467e-01
-7.01494932e-01 -9.11831141e-01 -8.22556913e-01 1.42584413e-01
3.51496547e-01 -1.15246987e+00 1.25131249e+00 -9.03786480e-01
-1.33836949e+00 1.23053288e+00 -1.05956092e-01 -6.83372259e-01
7.14250028e-01 -2.76955873e-01 3.22429031e-01 2.54331753e-02
6.63726404e-02 1.06987250e+00 6.44426048e-01 -7.29972899e-01
-7.48608708e-01 -6.29919469e-01 2.07569048e-01 1.97142050e-01
-1.99088529e-01 6.26695752e-02 -9.31454241e-01 -3.53065014e-01
3.60963196e-01 -1.05558109e+00 -4.65433538e-01 2.55048186e-01
-4.47302252e-01 -2.17842877e-01 3.57159883e-01 -4.75683481e-01
8.36877942e-01 -2.08156419e+00 2.32753322e-01 1.76229209e-01
2.21439660e-01 3.68867755e-01 -7.35985860e-02 -3.82695384e-02
8.84974748e-03 -1.66703224e-01 -3.50878060e-01 -6.02338076e-01
1.19684942e-01 -1.04541548e-01 -1.15160488e-01 5.87970734e-01
4.42283526e-02 1.05104196e+00 -6.27014041e-01 -4.98282969e-01
3.83022398e-01 3.58868212e-01 -8.01193714e-01 2.82685697e-01
-4.82738800e-02 4.23137099e-01 -3.40443820e-01 7.41670609e-01
6.52133882e-01 -4.15802449e-01 -1.70395926e-01 -4.62851703e-01
-2.79116064e-01 4.17801738e-01 -1.05070043e+00 2.12822008e+00
-4.65858549e-01 6.78548098e-01 1.47425160e-01 -8.96815121e-01
7.11890936e-01 -3.85873541e-02 3.02765429e-01 -9.74204719e-01
2.29168072e-01 4.96485531e-01 9.95320082e-02 -4.68044244e-02
2.36559421e-01 1.75073832e-01 3.01838722e-02 8.67050514e-02
1.11390222e-02 -1.45271778e-01 1.83151558e-01 -8.24882314e-02
1.10003078e+00 9.08114761e-02 2.07515806e-01 -3.76387089e-01
5.62978983e-01 1.59503728e-01 5.77890694e-01 7.30086923e-01
-1.48560375e-01 8.86892974e-01 2.09989697e-01 -6.92029417e-01
-1.10387433e+00 -8.82906973e-01 -2.63286442e-01 1.05298567e+00
2.49991089e-01 -1.95562765e-01 -9.28604424e-01 -3.41788083e-01
-3.12035643e-02 1.17464371e-01 -5.31012833e-01 1.36492923e-01
-7.60022581e-01 -3.92383814e-01 5.85408270e-01 6.07599795e-01
8.96295071e-01 -1.24349368e+00 -9.72661197e-01 1.29607514e-01
4.22952287e-02 -1.29671896e+00 -5.20601749e-01 2.89460551e-02
-9.23950016e-01 -1.05409276e+00 -7.43782759e-01 -8.07485223e-01
6.39140368e-01 1.72780961e-01 1.30834985e+00 2.43917793e-01
-5.59885144e-01 -1.38032079e-01 8.10191110e-02 -1.13881804e-01
8.38303119e-02 4.66606766e-01 -3.93389851e-01 -9.59759429e-02
3.01496089e-01 -5.76001465e-01 -1.02742922e+00 3.07980269e-01
-8.55147660e-01 4.36542630e-01 6.81670010e-01 8.61215949e-01
6.44638538e-01 -3.22053015e-01 -4.89699431e-02 -7.50612319e-01
1.74049154e-01 -3.28652598e-02 -9.55150843e-01 5.08872885e-03
-5.69398046e-01 1.56972662e-01 4.82888103e-01 4.03936543e-02
-6.91616535e-01 3.76572281e-01 -3.84734273e-01 -5.80898583e-01
-3.20893437e-01 3.31146747e-01 -5.98306730e-02 -3.64469290e-01
4.22597498e-01 8.14713910e-02 -1.54348621e-02 -5.04188240e-01
2.72607505e-01 2.69238740e-01 6.30867422e-01 -4.09983456e-01
7.34079421e-01 5.25752187e-01 2.47384861e-01 -5.50664365e-01
-1.04166019e+00 -6.07698202e-01 -7.93326020e-01 -4.32454422e-02
9.45896566e-01 -9.92740631e-01 -7.73400664e-01 7.89519250e-01
-1.17211783e+00 -4.08959568e-01 2.57424060e-02 1.98764682e-01
-5.05988955e-01 1.49612933e-01 -4.90187645e-01 -2.20477208e-01
-5.70337117e-01 -1.56976902e+00 1.19233489e+00 2.57972032e-01
-2.22367659e-01 -6.94915056e-01 -1.00756094e-01 6.76837981e-01
5.68702102e-01 1.52883604e-01 7.87189782e-01 -2.67385185e-01
-9.04717624e-01 2.70298421e-02 -6.21176004e-01 1.85597375e-01
-2.99292624e-01 -2.18659177e-01 -1.11903632e+00 -2.97897995e-01
-2.82614261e-01 -4.51918453e-01 1.18544269e+00 5.16758382e-01
1.42496240e+00 1.50336802e-01 -1.79235891e-01 1.27971339e+00
1.52776551e+00 -2.40005806e-01 7.40703583e-01 6.58609927e-01
1.02427971e+00 7.16359496e-01 4.56116408e-01 2.11308092e-01
4.68195558e-01 1.00516748e+00 6.78800285e-01 -4.35483694e-01
-2.39156708e-01 -7.21478611e-02 -5.58484718e-02 3.06991845e-01
-3.59377377e-02 2.49118432e-01 -1.07255709e+00 5.72998524e-01
-1.88881969e+00 -7.70514965e-01 5.44298477e-02 2.18953705e+00
6.26970232e-01 2.93122619e-01 3.05447001e-02 -1.13092698e-01
6.04185104e-01 2.28253126e-01 -7.85117149e-01 -2.69577593e-01
-5.81953898e-02 4.04184759e-01 6.59662783e-01 3.48112643e-01
-1.35478675e+00 1.14239228e+00 4.49230146e+00 7.64088809e-01
-1.14297569e+00 -4.94490340e-02 8.18125129e-01 -1.33040816e-01
7.95063227e-02 1.04382992e-01 -8.10393751e-01 3.09400767e-01
4.31881577e-01 1.62628904e-01 6.13940954e-01 8.02519262e-01
1.68699324e-02 5.29652163e-02 -1.20007789e+00 1.33534682e+00
-2.24355251e-01 -1.64496255e+00 -8.93411785e-02 9.57253203e-02
6.86697900e-01 6.67789161e-01 -1.26086920e-02 1.82679906e-01
-5.14983721e-02 -1.22402263e+00 8.74012470e-01 2.35714555e-01
5.38767993e-01 -6.14859283e-01 6.79174960e-01 2.58950561e-01
-1.34333813e+00 -5.69062606e-02 -4.42417443e-01 3.22367884e-02
5.57358637e-02 5.60847223e-01 -3.79083067e-01 3.12997013e-01
9.87760067e-01 6.01814210e-01 -6.57389045e-01 1.36297262e+00
-7.03368485e-02 2.23268569e-01 -3.91426951e-01 2.54753023e-01
5.05011618e-01 -1.79471999e-01 2.61179656e-01 1.11281741e+00
2.62676269e-01 -9.77828875e-02 1.15562528e-01 1.05300224e+00
-3.09734493e-01 -3.98491174e-02 -4.61368591e-01 3.89907837e-01
5.54499984e-01 1.30938029e+00 -7.25184441e-01 -1.05660088e-01
-4.42072004e-01 1.03721833e+00 6.01792514e-01 2.12265681e-02
-6.36300266e-01 -2.30495527e-01 8.42136323e-01 2.35667750e-01
4.47840482e-01 -8.98844898e-02 -6.76752806e-01 -9.02127922e-01
1.16356008e-01 -9.30646598e-01 3.35413903e-01 -4.12933975e-01
-1.16317928e+00 7.23334253e-01 -2.52966017e-01 -9.87143695e-01
-6.30632415e-02 -6.03984237e-01 -6.21919215e-01 1.00076139e+00
-1.63856840e+00 -1.06200874e+00 -6.28566980e-01 6.42384768e-01
6.39720976e-01 -1.70867026e-01 4.58182335e-01 5.70740223e-01
-6.46610141e-01 6.01909518e-01 -3.33611995e-01 3.57939065e-01
5.20185173e-01 -1.08780730e+00 9.43917096e-01 8.80446732e-01
1.36374563e-01 5.34045935e-01 4.31923956e-01 -1.53024778e-01
-1.39653385e+00 -1.04027104e+00 8.67274106e-01 -1.55137271e-01
5.96077025e-01 -4.47326481e-01 -8.20402026e-01 1.98847264e-01
5.22115231e-02 3.41154784e-01 1.31588981e-01 1.18647693e-02
-3.65821600e-01 -2.90541470e-01 -8.30773532e-01 6.30111516e-01
1.44564438e+00 -5.26836395e-01 -3.13039690e-01 1.51537552e-01
4.98891950e-01 -7.98882604e-01 -6.31528199e-01 6.35779321e-01
6.33227170e-01 -1.38616264e+00 1.25023627e+00 -3.30662876e-01
5.82473874e-01 -3.23253453e-01 -1.30942136e-01 -9.73672628e-01
-3.72368068e-01 -3.75247329e-01 2.48406276e-01 9.72306311e-01
4.23635870e-01 -5.59609056e-01 9.79448974e-01 6.86202168e-01
-3.72010648e-01 -1.14040685e+00 -1.05874693e+00 -5.30567288e-01
-4.31476347e-02 -3.20297152e-01 6.63742721e-01 6.01507962e-01
-7.32952952e-01 3.07276040e-01 2.14373246e-02 1.80542842e-01
8.89749646e-01 5.02439082e-01 8.38338971e-01 -1.36105597e+00
-4.77482855e-01 -9.12958562e-01 -4.77704614e-01 -1.21072793e+00
3.16874832e-01 -9.19887900e-01 4.56433594e-02 -1.52541316e+00
1.84955478e-01 -5.64726174e-01 -2.39061072e-01 4.86727327e-01
-2.03330532e-01 4.45819467e-01 3.16678703e-01 3.59926999e-01
-4.57163274e-01 4.21529293e-01 1.24689364e+00 -1.46942288e-01
-8.84658545e-02 -5.65589741e-02 -5.66528976e-01 7.99132049e-01
7.89014637e-01 -3.17068666e-01 -2.15362251e-01 -7.61731565e-01
1.96069047e-01 -1.76616982e-01 7.69371152e-01 -1.18004966e+00
4.54694450e-01 1.82771564e-01 1.61961153e-01 -4.63538677e-01
2.39639238e-01 -6.98091686e-01 5.75026907e-02 5.85613847e-01
-2.78014719e-01 1.10012010e-01 1.60691142e-01 6.32529110e-02
-3.75723004e-01 -6.11590818e-02 1.02530277e+00 -1.58531442e-01
-1.11160171e+00 7.10237086e-01 3.63118112e-01 2.44310707e-01
6.86238825e-01 -4.00535762e-01 -4.25972760e-01 7.82831535e-02
-2.64446795e-01 3.00514668e-01 6.49481058e-01 4.24293786e-01
5.59754312e-01 -1.03240490e+00 -6.65357411e-01 1.22973494e-01
1.05248682e-01 4.50355679e-01 2.85785913e-01 8.97729754e-01
-7.37864137e-01 4.96732622e-01 -3.65326077e-01 -8.29016387e-01
-1.22349596e+00 1.18441358e-01 6.21401846e-01 -2.48900458e-01
-7.50136673e-01 1.08660507e+00 4.96412128e-01 -4.01880950e-01
3.95882607e-01 -2.36438587e-01 -2.57000886e-02 -1.38679415e-01
3.27475756e-01 1.90455213e-01 4.21788007e-01 -5.30159593e-01
-5.20552635e-01 8.54178250e-01 -1.22759931e-01 2.23436773e-01
1.27702630e+00 1.15933903e-01 -2.25461349e-01 -8.50848556e-02
1.30358410e+00 -6.12953842e-01 -1.51466024e+00 -3.63434970e-01
-1.77859124e-02 -5.44570386e-01 4.98362005e-01 -4.50587660e-01
-1.47157443e+00 1.03531837e+00 7.90443838e-01 -1.98573038e-01
1.11392140e+00 1.13214836e-01 1.03349364e+00 3.23128194e-01
5.43800116e-01 -9.69845593e-01 -2.00276583e-01 6.49027765e-01
7.46804237e-01 -1.35973787e+00 -7.74750933e-02 -3.40229750e-01
-3.64137381e-01 9.94270802e-01 8.31135809e-01 -2.74586767e-01
5.91488600e-01 2.19379246e-01 -1.01837581e-02 -4.50882494e-01
-5.04371524e-01 -4.84656960e-01 5.47385097e-01 8.52597952e-02
4.65377361e-01 -4.66728508e-02 -3.36930491e-02 9.66907069e-02
-3.11068922e-01 -1.00554161e-01 -1.03111848e-01 5.83529532e-01
-3.96046668e-01 -9.84584153e-01 -1.18541732e-01 5.09154201e-01
-3.55428010e-01 -3.05659592e-01 -2.88479388e-01 4.70461726e-01
1.79746330e-01 5.69502890e-01 2.50322610e-01 -1.95233926e-01
6.11939847e-01 -2.88456768e-01 4.72852975e-01 -5.19441068e-01
-7.50790238e-01 -1.74417481e-01 -2.25989968e-02 -1.08467603e+00
-3.96683753e-01 -5.70173144e-01 -1.19924986e+00 -3.73642564e-01
4.52610329e-02 -4.61144805e-01 6.10353827e-01 1.00723028e+00
4.06712502e-01 3.68686885e-01 3.62415731e-01 -1.15262866e+00
-4.02569354e-01 -6.52709424e-01 -1.23628579e-01 4.67870295e-01
7.91842043e-02 -6.80093050e-01 -1.45120293e-01 -3.46161842e-01] | [8.842385292053223, -2.246411085128784] |
0adea788-b850-4c52-8c8c-8318928e4dae | semantic-information-extraction-for-improved | null | null | https://aclanthology.org/W15-1523 | https://aclanthology.org/W15-1523.pdf | Semantic Information Extraction for Improved Word Embeddings | null | ['Gerard de Melo', 'Jiaqiang Chen'] | 2015-06-01 | null | null | null | ws-2015-6 | ['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] | [-7.293671607971191, 3.6729776859283447] |
a3d0a85b-7898-4fc1-9e0e-7206974312d8 | anatomy-guided-domain-adaptation-for-3d-in | 2211.12193 | null | https://arxiv.org/abs/2211.12193v2 | https://arxiv.org/pdf/2211.12193v2.pdf | Anatomy-guided domain adaptation for 3D in-bed human pose estimation | 3D human pose estimation is a key component of clinical monitoring systems. The clinical applicability of deep pose estimation models, however, is limited by their poor generalization under domain shifts along with their need for sufficient labeled training data. As a remedy, we present a novel domain adaptation method, adapting a model from a labeled source to a shifted unlabeled target domain. Our method comprises two complementary adaptation strategies based on prior knowledge about human anatomy. First, we guide the learning process in the target domain by constraining predictions to the space of anatomically plausible poses. To this end, we embed the prior knowledge into an anatomical loss function that penalizes asymmetric limb lengths, implausible bone lengths, and implausible joint angles. Second, we propose to filter pseudo labels for self-training according to their anatomical plausibility and incorporate the concept into the Mean Teacher paradigm. We unify both strategies in a point cloud-based framework applicable to unsupervised and source-free domain adaptation. Evaluation is performed for in-bed pose estimation under two adaptation scenarios, using the public SLP dataset and a newly created dataset. Our method consistently outperforms various state-of-the-art domain adaptation methods, surpasses the baseline model by 31%/66%, and reduces the domain gap by 65%/82%. Source code is available at https://github.com/multimodallearning/da-3dhpe-anatomy. | ['Mattias P. Heinrich', 'Philipp Rostalski', 'Carlotta Hennigs', 'Jasper Diesel', 'Lasse Hansen', 'Alexander Bigalke'] | 2022-11-22 | null | null | null | null | ['3d-human-pose-estimation', 'source-free-domain-adaptation'] | ['computer-vision', 'computer-vision'] | [ 2.44782075e-01 6.08184695e-01 -3.64191681e-01 -3.92958999e-01
-1.21054792e+00 -6.60765350e-01 1.94626868e-01 2.37347841e-01
-4.16975558e-01 8.65357101e-01 3.96674037e-01 -1.61525384e-01
-2.51292378e-01 -4.32331204e-01 -7.59362996e-01 -6.65862978e-01
1.12839766e-01 9.55320895e-01 2.50979155e-01 -1.14327036e-01
-1.35418221e-01 5.03148079e-01 -5.89835048e-01 6.02365993e-02
8.44489932e-01 8.78115356e-01 5.11829406e-02 3.45517009e-01
3.50947648e-01 2.31533051e-01 -3.04903984e-01 -5.08685410e-01
3.11159015e-01 -3.76378596e-01 -8.01577508e-01 5.89886568e-02
2.83037901e-01 -3.13548446e-01 -1.10991463e-01 7.15294600e-01
9.56522405e-01 6.41169846e-02 8.13872635e-01 -7.41264105e-01
-1.20753087e-01 2.19341576e-01 -6.27327085e-01 2.74539664e-02
2.67874122e-01 7.60554001e-02 5.37905157e-01 -7.80096889e-01
7.95323610e-01 7.23998725e-01 9.52831686e-01 8.49492908e-01
-1.39421737e+00 -6.33334160e-01 1.05134055e-01 -1.34094700e-01
-1.35944772e+00 -2.57411629e-01 9.57467198e-01 -6.93429708e-01
5.67212224e-01 2.24100947e-02 5.93697786e-01 1.25607932e+00
3.19529295e-01 5.02134144e-01 8.82718086e-01 -3.96192044e-01
3.62018377e-01 1.97619833e-02 -3.70128691e-01 6.22016132e-01
1.50583163e-01 2.25589439e-01 -4.44233268e-01 -3.07883024e-01
8.98017466e-01 -1.80853575e-01 -3.32264066e-01 -9.89481688e-01
-1.15573740e+00 7.84639120e-01 6.42969847e-01 -3.40138674e-02
-5.45519471e-01 -2.33908489e-01 3.50040168e-01 -3.44940335e-01
4.97137487e-01 5.90589404e-01 -6.16966009e-01 -3.86492163e-02
-7.64260709e-01 2.52675265e-01 6.39673412e-01 8.42218995e-01
2.92507529e-01 -3.67787123e-01 -1.34444788e-01 9.22498703e-01
4.59262639e-01 4.01990980e-01 5.76324999e-01 -8.78093123e-01
3.26417476e-01 4.38214123e-01 -3.47867161e-02 -6.30363226e-01
-7.49612987e-01 -8.14534009e-01 -5.69884062e-01 -3.42715643e-02
4.38216954e-01 -3.53542924e-01 -1.05975008e+00 1.83904076e+00
8.74273002e-01 6.69286326e-02 -1.84390515e-01 1.09913516e+00
7.28798866e-01 -3.53878997e-02 3.62140656e-01 -1.83087111e-01
1.25613260e+00 -8.02255332e-01 -2.96091110e-01 -4.40486938e-01
5.38508534e-01 -7.17526853e-01 9.16178703e-01 5.18804848e-01
-1.06526458e+00 -4.04411376e-01 -8.63724470e-01 1.79976389e-01
5.99221252e-02 1.65316448e-01 3.71200889e-01 5.49896955e-01
-5.54622233e-01 6.17608368e-01 -1.25708616e+00 -3.95122319e-01
5.30764878e-01 6.22349560e-01 -4.50433105e-01 3.52565497e-02
-1.05072975e+00 1.10284650e+00 4.88039970e-01 -6.34006411e-02
-7.80582130e-01 -9.61848974e-01 -8.30487490e-01 -4.45406914e-01
3.72411191e-01 -1.00405133e+00 1.43454111e+00 -6.95553482e-01
-1.70238638e+00 9.88764048e-01 6.79551288e-02 -4.33638334e-01
8.48157644e-01 -3.95656645e-01 -3.09153914e-01 4.13103223e-01
1.27118742e-02 7.75085211e-01 6.61017358e-01 -1.29465616e+00
-1.81158960e-01 -6.13400996e-01 -2.45422572e-01 4.74800229e-01
-2.50215501e-01 -4.89790589e-01 -5.12372136e-01 -8.07135582e-01
2.88211554e-01 -1.30098128e+00 -4.80344623e-01 4.04320359e-01
-3.51226985e-01 2.95972675e-01 3.26097131e-01 -8.24827611e-01
1.06092024e+00 -1.98047721e+00 3.01192164e-01 4.11844969e-01
3.75305861e-01 8.00268129e-02 1.14228711e-01 1.76799089e-01
-1.80619419e-01 -3.94338638e-01 -5.73659003e-01 -2.44987011e-01
-2.28555039e-01 2.79211640e-01 4.51825522e-02 6.17943048e-01
1.16233127e-02 6.84678733e-01 -9.29112136e-01 -7.46651113e-01
3.72000039e-01 5.23183525e-01 -9.01207685e-01 1.46433994e-01
3.93285835e-03 1.25154448e+00 -6.23823047e-01 5.62632024e-01
7.23341107e-01 -2.94664919e-01 4.11642402e-01 -3.15251857e-01
3.09933722e-01 2.64366567e-01 -9.10731852e-01 2.33452940e+00
-4.74806458e-01 -1.25940531e-01 -1.05137527e-01 -9.07946467e-01
8.73753190e-01 4.16862249e-01 9.77205694e-01 -3.72674406e-01
2.63292581e-01 4.92610604e-01 1.57145470e-01 -3.98782253e-01
-1.39944449e-01 -7.34063804e-01 -1.18689604e-01 -9.68019217e-02
2.19022691e-01 -3.67717981e-01 -2.80672371e-01 -1.42592117e-01
9.33873057e-01 5.04355967e-01 5.95758319e-01 -1.83976278e-01
4.75631833e-01 5.73773533e-02 6.37661755e-01 2.83044755e-01
-4.23081934e-01 9.79235113e-01 1.94012552e-01 -3.04006100e-01
-7.65258551e-01 -1.50920105e+00 -5.26975393e-01 8.02194536e-01
6.75424784e-02 -1.73082590e-01 -6.53651953e-01 -1.04910851e+00
1.70260012e-01 5.19670188e-01 -7.52741098e-01 -4.71852928e-01
-6.84024096e-01 -6.87481165e-01 4.57375199e-01 7.44172812e-01
1.11832567e-01 -7.28214025e-01 -6.06438816e-01 2.63751984e-01
-3.47022533e-01 -1.10602295e+00 -3.55994225e-01 3.36645782e-01
-1.13498557e+00 -1.01547348e+00 -1.10429931e+00 -5.80023229e-01
8.03747177e-01 -5.56147218e-01 1.02209425e+00 -4.11707520e-01
-1.81402549e-01 3.83003175e-01 -2.97973633e-01 -4.72146153e-01
-4.07525510e-01 3.21720362e-01 1.15310751e-01 -3.20511758e-01
1.73278257e-01 -5.93246579e-01 -1.03950071e+00 3.94414783e-01
-4.74534512e-01 2.37917192e-02 7.94521272e-01 8.92095804e-01
9.35828328e-01 -6.58884168e-01 3.96820575e-01 -9.91377115e-01
3.50659817e-01 -4.57683682e-01 -1.97399676e-01 6.92552179e-02
-6.91585422e-01 5.82732745e-02 3.80972743e-01 -5.11802495e-01
-1.15135550e+00 5.98630786e-01 -4.51016635e-01 -4.40845609e-01
-3.43902856e-01 4.07862931e-01 -4.50487994e-02 -2.06539810e-01
1.09586096e+00 -5.24405390e-02 2.36182272e-01 -5.28232932e-01
2.43295938e-01 3.68317395e-01 7.55643964e-01 -9.06492472e-01
7.87309945e-01 5.45113802e-01 2.10219473e-01 -4.15038168e-01
-8.23024035e-01 -4.84147608e-01 -1.05432463e+00 -6.26586378e-02
7.35636055e-01 -8.36681604e-01 -3.32703531e-01 8.63566156e-03
-9.06097293e-01 -3.56123865e-01 -5.30451775e-01 8.83273244e-01
-7.07096636e-01 4.94337499e-01 -3.62931579e-01 -2.19539508e-01
-4.93824840e-01 -1.22579420e+00 1.27001750e+00 -6.07693195e-02
-6.25430048e-01 -1.08697855e+00 4.98235285e-01 3.18694472e-01
2.04281121e-01 6.38641894e-01 7.04633236e-01 -9.70440328e-01
8.47195908e-02 -2.85556912e-01 1.27758846e-01 2.76813984e-01
2.02332020e-01 -4.68187749e-01 -7.61034727e-01 -4.13997501e-01
-1.31435394e-01 -2.66778946e-01 4.04310435e-01 7.02029049e-01
1.20241296e+00 -6.25938782e-03 -5.62099457e-01 7.44178891e-01
1.17030883e+00 7.02967793e-02 3.13524663e-01 2.85786182e-01
5.53648114e-01 5.86520195e-01 5.94849050e-01 6.61156416e-01
3.79223049e-01 7.91931331e-01 4.71555591e-01 -1.72495812e-01
-1.90038100e-01 -3.95221442e-01 -2.13542610e-01 5.80553174e-01
-1.92128256e-01 1.32960081e-01 -1.29598045e+00 4.92037028e-01
-1.57795453e+00 -3.38665485e-01 2.83343196e-01 2.41221142e+00
1.01068592e+00 2.37518609e-01 2.91207939e-01 -1.81378186e-01
3.95455301e-01 -1.52716562e-01 -7.90971875e-01 5.16512245e-02
5.71996272e-01 4.57438588e-01 5.78763902e-01 5.19181967e-01
-1.20819902e+00 6.58515394e-01 5.61418581e+00 4.77916092e-01
-1.22809935e+00 2.61295915e-01 4.67181146e-01 -1.54517382e-01
-1.77075602e-02 -2.27917776e-01 -5.51161110e-01 3.68544459e-01
7.67991722e-01 4.17878293e-02 -1.67241525e-02 9.58402872e-01
1.62591189e-01 6.36928598e-04 -1.13479197e+00 7.46119738e-01
8.15486014e-02 -9.45154190e-01 -1.87317237e-01 2.52023786e-02
4.76431221e-01 -1.76028982e-02 3.01565658e-02 2.72777885e-01
-2.31404528e-02 -9.35728312e-01 5.29683411e-01 3.83094192e-01
1.00568581e+00 -5.43555379e-01 6.24989569e-01 2.80947477e-01
-9.69250560e-01 1.61401749e-01 -1.50371753e-02 3.29924226e-01
2.93135911e-01 2.99471676e-01 -1.17622602e+00 7.23228872e-01
5.30554891e-01 5.88170886e-01 -4.07537937e-01 1.19029069e+00
-3.55428576e-01 4.03211713e-01 -4.85093713e-01 4.73055005e-01
-1.07086629e-01 1.39879256e-01 6.23018682e-01 1.01633406e+00
1.90621555e-01 1.55567244e-01 2.70710975e-01 5.24706781e-01
1.03540130e-01 1.48360997e-01 -2.87870526e-01 4.69186723e-01
4.22574550e-01 1.15657961e+00 -5.75802505e-01 -5.49644046e-02
-2.22250834e-01 1.04629910e+00 8.59388411e-02 1.16004705e-01
-1.06278849e+00 -8.06355849e-02 2.49069035e-01 4.68312681e-01
-7.54022375e-02 9.19150934e-02 -4.89549428e-01 -9.25356865e-01
5.22388108e-02 -8.53780508e-01 6.32496476e-01 -5.39629281e-01
-1.18791032e+00 6.23691916e-01 3.73349756e-01 -1.63460922e+00
-5.22264957e-01 -6.41589284e-01 -1.55733168e-01 8.23305547e-01
-1.37329614e+00 -1.21283913e+00 -5.28456688e-01 4.88297671e-01
3.28064024e-01 7.26415291e-02 9.54377711e-01 5.36432564e-01
-2.43193254e-01 8.03652227e-01 -1.19363517e-01 4.95771356e-02
1.18930054e+00 -1.25232613e+00 1.00734793e-01 3.51467073e-01
-2.65506953e-01 6.02104485e-01 7.53455043e-01 -6.00236416e-01
-8.36136937e-01 -9.83926475e-01 4.41373676e-01 -6.37078822e-01
4.49697971e-01 -1.04655698e-01 -8.62466455e-01 7.10468829e-01
-3.42367798e-01 4.14940327e-01 8.89882803e-01 8.17416385e-02
-7.60952607e-02 -8.78267586e-02 -1.36477315e+00 3.99386764e-01
1.03206336e+00 -1.86173946e-01 -6.98326766e-01 4.73868996e-01
3.59313697e-01 -1.23820734e+00 -1.29568052e+00 8.94525766e-01
8.30620646e-01 -5.06204307e-01 1.20009351e+00 -5.94649076e-01
3.93777728e-01 -1.15942478e-01 -3.94187346e-02 -1.31593812e+00
-2.22690165e-01 -3.64052653e-01 -1.28996462e-01 6.37223005e-01
4.60630834e-01 -6.61749601e-01 1.21278942e+00 5.57856679e-01
-3.63429666e-01 -1.10630441e+00 -1.13946342e+00 -7.03891993e-01
6.35305345e-01 -2.86316454e-01 2.83340454e-01 8.10310900e-01
1.00553393e-01 2.04557300e-01 -1.26717404e-01 4.22290534e-01
5.81844747e-01 -1.55465126e-01 6.54532313e-01 -1.28687942e+00
-6.70228839e-01 -2.94059545e-01 -2.55178809e-01 -9.93956685e-01
-5.29877990e-02 -8.75803649e-01 8.50300416e-02 -1.54561865e+00
1.49175048e-01 -3.55817467e-01 -2.41021797e-01 5.77742755e-01
-5.22814915e-02 2.20694736e-01 -5.18214181e-02 2.99447775e-01
-2.01643646e-01 5.06854296e-01 1.39274895e+00 1.38870776e-01
-3.31677049e-01 1.87146842e-01 -5.10573149e-01 9.15401578e-01
9.10906494e-01 -5.14633715e-01 -5.20302176e-01 -1.96687460e-01
-1.39139742e-01 1.88550994e-01 4.29284096e-01 -1.08224356e+00
-3.91997024e-02 6.81025442e-03 5.74172497e-01 -3.54325712e-01
3.74896884e-01 -1.02369225e+00 1.19111776e-01 7.46419907e-01
-2.45131254e-01 -2.48412982e-01 4.01305050e-01 5.35703778e-01
-4.57523540e-02 8.22201371e-03 1.11406744e+00 -1.01247437e-01
-3.93073976e-01 3.56337607e-01 1.03818923e-01 3.60592246e-01
1.05534530e+00 -2.78226018e-01 5.67995980e-02 -9.12230164e-02
-1.26458097e+00 8.97887424e-02 3.45914602e-01 2.97909349e-01
5.02040148e-01 -1.16737747e+00 -6.14838839e-01 1.11916095e-01
4.05206025e-01 3.96490633e-01 4.57161576e-01 1.29530632e+00
-6.66176081e-01 3.42589080e-01 -1.49999321e-01 -7.98139572e-01
-9.72974539e-01 4.68608975e-01 8.03773701e-01 -3.95924807e-01
-5.70442498e-01 8.99881899e-01 4.07863259e-01 -7.78498530e-01
2.37876758e-01 -2.70149112e-01 1.76665068e-01 -3.48391950e-01
-8.40500966e-02 1.50766313e-01 3.61340106e-01 -6.45924687e-01
-7.75107265e-01 8.67499709e-01 -6.95262626e-02 1.15232067e-02
1.15590429e+00 2.87572276e-02 5.33130169e-01 3.47063467e-02
1.10571146e+00 1.05767883e-01 -1.50184715e+00 -2.51858413e-01
-1.53562501e-01 -1.76120445e-01 -1.51496172e-01 -1.31360686e+00
-9.00489390e-01 8.54661107e-01 8.62396777e-01 -6.62672698e-01
1.24267244e+00 2.53847361e-01 6.55748367e-01 -1.25261590e-01
2.18367934e-01 -9.11131203e-01 1.36182696e-01 1.04560189e-01
9.88267064e-01 -1.32831764e+00 2.67774343e-01 -6.25244439e-01
-9.62821424e-01 8.71614993e-01 7.35722661e-01 -7.12577477e-02
7.88990855e-01 2.04566926e-01 2.10865363e-01 -2.17294529e-01
-2.83937275e-01 7.45429099e-02 7.35388100e-01 7.80112803e-01
6.72071576e-01 1.60389602e-01 -5.35250068e-01 7.61935830e-01
-2.49491155e-01 1.04046077e-01 -7.80545250e-02 9.40940619e-01
-1.44316375e-01 -1.10442197e+00 -3.90060067e-01 1.97664201e-01
-5.54758608e-01 1.42291546e-01 -1.40690029e-01 1.02613950e+00
2.49634057e-01 3.35210443e-01 -2.87806243e-01 -1.63842261e-01
6.74045920e-01 7.89473057e-02 6.51786268e-01 -8.50279808e-01
-4.02114749e-01 4.22339648e-01 -1.75174490e-01 -4.07964647e-01
-1.12621717e-01 -6.75571263e-01 -1.47339225e+00 3.57625604e-01
-1.85979426e-01 -5.23443297e-02 4.87043858e-01 7.79928505e-01
3.81322861e-01 6.43902063e-01 2.86234885e-01 -9.11057353e-01
-8.13103676e-01 -9.09138262e-01 -2.56453753e-01 4.91667330e-01
2.29510576e-01 -1.00057375e+00 -7.74823725e-02 -5.49260452e-02] | [14.559030532836914, -2.0742273330688477] |
8e281cb5-800c-41fb-94e7-0e2a078db6f9 | symbolic-numeric-computation-of-integrals-in | 2303.10046 | null | https://arxiv.org/abs/2303.10046v1 | https://arxiv.org/pdf/2303.10046v1.pdf | Symbolic-Numeric Computation of Integrals in Successive Galerkin Approximation of Hamilton-Jacobi-Bellman Equation | This paper proposes an efficient symbolic-numeric method to compute the integrals in the successive Galerkin approximation (SGA) of the Hamilton-Jacobi-Bellman (HJB) equation. A solution of the HJB equation is first approximated with a linear combination of the Hermite polynomials. The coefficients of the combination are then computed by iteratively solving a linear equation, which consists of the integrals of the Hermite polynomials multiplied by nonlinear functions. The recursive structure of the Hermite polynomials is inherited by the integrals, and their recurrence relations can be computed by using the symbolic computation of differential operators. By using the recurrence relations, all the integrals can be computed from a part of them that are numerically evaluated. A numerical example is provided to show the efficiency of the proposed method compared to a standard numerical integration method. | ['Tomoyuki Iori'] | 2023-03-17 | null | null | null | null | ['numerical-integration'] | ['miscellaneous'] | [-4.11469907e-01 1.86131243e-02 3.26519579e-01 2.48064220e-01
-4.23024058e-01 -4.14175540e-01 1.87897265e-01 -1.09670654e-01
-2.29217529e-01 1.13971090e+00 -5.82306087e-01 -3.69801044e-01
-1.02589048e-01 -7.89720833e-01 -3.39578986e-01 -9.98986244e-01
-2.22280994e-01 4.18209553e-01 8.38274881e-02 -4.44412142e-01
1.03688836e-01 8.80800784e-01 -1.31785917e+00 -3.55253369e-01
1.12411237e+00 1.39145660e+00 -7.18426406e-01 6.86548769e-01
-1.11061886e-01 1.00049317e+00 -3.21956754e-01 -1.69393979e-02
3.88762444e-01 -6.30160153e-01 -8.55425775e-01 -3.73745561e-02
-4.25647080e-01 -6.09493971e-01 -2.61232294e-02 1.02572334e+00
1.32324949e-01 8.10361147e-01 9.41338897e-01 -1.36622870e+00
-3.05641204e-01 -1.71330109e-01 -3.54689628e-01 -5.24221301e-01
6.27306938e-01 -3.13021839e-01 4.37417805e-01 -1.09226525e+00
3.88217568e-01 1.30025852e+00 1.39600384e+00 -6.94253296e-02
-1.42381310e+00 -1.82972148e-01 -6.16134644e-01 -1.18692584e-01
-2.05991983e+00 1.25438377e-01 5.29038906e-01 -4.43889886e-01
9.97819662e-01 5.74048460e-01 1.19255662e+00 -1.86309934e-01
4.08665925e-01 3.00625056e-01 1.10390770e+00 -6.39818132e-01
3.79431754e-01 2.25767717e-02 4.07400459e-01 9.79984462e-01
2.01492682e-01 8.54152739e-02 3.58064294e-01 -8.99315894e-01
1.20044482e+00 1.59419268e-01 -9.70663875e-02 -1.69796199e-01
-4.77178812e-01 1.19838679e+00 1.57949924e-01 4.03595388e-01
-7.86509633e-01 2.39149347e-01 1.88514337e-01 -1.75781950e-01
5.57796299e-01 1.07536085e-01 1.77560359e-01 1.04668409e-01
-1.07011116e+00 6.71995878e-01 1.39785278e+00 7.18298435e-01
1.11778390e+00 2.74955243e-01 -1.11948527e-01 4.83284086e-01
4.14882660e-01 6.93240106e-01 3.78425941e-02 -1.41543901e+00
-6.39632344e-01 4.13059980e-01 5.93033791e-01 -9.74805713e-01
-2.20594779e-01 1.51087552e-01 -8.10804427e-01 5.74929237e-01
6.85236037e-01 -3.23783875e-01 -3.07026148e-01 8.13931882e-01
6.87892795e-01 2.23910376e-01 3.58368382e-02 6.97728455e-01
3.72015417e-01 1.25931394e+00 -3.42317581e-01 -7.36163080e-01
1.05561233e+00 -8.47211301e-01 -1.03958023e+00 8.53727281e-01
6.85786426e-01 -9.73630607e-01 8.84552598e-02 3.92343432e-01
-1.29145145e+00 -3.39885682e-01 -8.16566467e-01 -2.70349085e-01
-2.11664066e-01 3.26587051e-01 5.58164299e-01 7.55241811e-02
-1.20646560e+00 9.01397705e-01 -7.01833785e-01 2.82279283e-01
-3.18777084e-01 2.82910705e-01 1.17498729e-02 5.00542223e-01
-1.30468738e+00 8.63437474e-01 9.50009525e-02 7.36609459e-01
-3.14936966e-01 -6.06528878e-01 -7.38228679e-01 -2.05801614e-02
-4.51807231e-02 -4.83603001e-01 1.35746992e+00 -8.14030051e-01
-1.79543662e+00 4.05199468e-01 -4.10330147e-01 -2.28155166e-01
5.74121892e-01 2.34966442e-01 -5.30781001e-02 3.31536382e-01
1.10266335e-01 -1.56502262e-01 7.99751937e-01 -1.05121815e+00
-3.45029771e-01 7.24525079e-02 -4.25799824e-02 -1.28635228e-01
3.81670296e-01 -5.36965728e-02 2.39039566e-02 -3.60651553e-01
1.62908554e-01 -9.06763971e-01 -7.09403455e-01 1.10082373e-01
-1.77013904e-01 -3.10116380e-01 7.66253591e-01 -1.10322464e+00
1.41551352e+00 -2.11336875e+00 1.56430617e-01 6.91882908e-01
2.89710462e-02 1.72943860e-01 6.21722817e-01 9.43376482e-01
1.36334389e-01 -2.30796099e-01 -5.32871127e-01 -3.76549214e-02
-2.62708515e-01 3.80042076e-01 -2.96559811e-01 5.47527671e-01
-1.92248315e-01 5.42759061e-01 -1.00500882e+00 -6.44854248e-01
-2.02693194e-02 8.62488270e-01 -3.25771004e-01 -2.00036615e-02
1.26490891e-01 3.77039522e-01 -7.74109364e-01 2.96152860e-01
9.89759505e-01 8.74516070e-02 1.09740749e-01 -2.57097512e-01
-8.66226912e-01 -1.37017250e-01 -1.49040151e+00 7.72942781e-01
-4.83627081e-01 3.66407037e-02 4.91680413e-01 -1.10095656e+00
1.11448658e+00 6.38887703e-01 6.17450714e-01 -2.17320189e-01
1.47224918e-01 8.27195585e-01 -3.80710334e-01 -2.99065650e-01
2.89935499e-01 -4.35838848e-01 1.84012830e-01 1.76942483e-01
-3.44199151e-01 -5.51691949e-01 5.08393288e-01 -5.48477396e-02
5.63810587e-01 1.70057639e-01 7.73326874e-01 -4.85728115e-01
1.24202597e+00 4.20475900e-01 3.62783700e-01 4.67956871e-01
6.24515489e-02 1.15973353e-01 7.91952729e-01 -8.11053753e-01
-1.00345039e+00 -6.59826338e-01 -4.97684300e-01 1.76897928e-01
-6.47190586e-03 -8.38274732e-02 -9.01739419e-01 6.84892610e-02
2.44600952e-01 7.36555517e-01 -4.24119264e-01 1.64793566e-01
-5.71605861e-01 -3.32164377e-01 4.40409720e-01 3.28546137e-01
7.09776282e-01 -5.85430622e-01 -6.03329360e-01 4.24322069e-01
1.04077503e-01 -7.81497121e-01 -7.15639815e-02 -4.75341305e-02
-1.17230964e+00 -8.68170083e-01 -9.53179598e-01 -8.64278078e-01
7.11997271e-01 -5.31707555e-02 4.26530659e-01 2.73100972e-01
-9.03712213e-02 3.65081400e-01 1.29017368e-01 1.56251073e-01
-6.13036096e-01 -5.82829714e-01 -4.62896898e-02 2.03806579e-01
-2.59350657e-01 -6.61728680e-01 -3.54980201e-01 2.39379108e-01
-6.81142211e-01 -2.36145973e-01 7.09068850e-02 8.25251222e-01
1.15727425e+00 3.06949437e-01 2.62202799e-01 -5.22590339e-01
8.87988925e-01 -1.38989747e-01 -1.44031370e+00 5.73278107e-02
-2.98591524e-01 2.00568736e-01 7.69617498e-01 -4.04991299e-01
-1.28026605e+00 3.12717378e-01 3.95798087e-02 -3.47099483e-01
5.50856233e-01 6.39503539e-01 5.11946619e-01 -9.88036931e-01
4.28951174e-01 1.51484773e-01 3.04768056e-01 -4.91109699e-01
3.55308115e-01 5.38063884e-01 6.32196784e-01 -8.47504020e-01
3.15414459e-01 6.15622938e-01 7.90909290e-01 -1.04336107e+00
-4.33475345e-01 -5.49291015e-01 -3.72644067e-01 -3.66596818e-01
7.58200645e-01 -4.12464708e-01 -1.22068405e+00 6.98761463e-01
-1.57192028e+00 -4.82843146e-02 -6.81185484e-01 6.10730231e-01
-6.66239798e-01 4.62486356e-01 -1.12699223e+00 -1.61118388e+00
-2.71276027e-01 -1.28571677e+00 8.19998324e-01 1.45643011e-01
-3.69681627e-01 -1.15821815e+00 3.63202035e-01 -3.75607222e-01
3.59987855e-01 4.58870679e-01 7.58750618e-01 -2.26152822e-01
-3.46966535e-01 -7.64806330e-01 4.24452797e-02 5.29982388e-01
-3.82178158e-01 5.03902853e-01 -4.55133945e-01 2.84334738e-02
7.10981488e-01 2.28531420e-01 1.70603886e-01 4.99989033e-01
3.43221813e-01 -6.26969337e-01 -3.60644281e-01 6.22406542e-01
1.56265581e+00 5.04375160e-01 6.19575322e-01 -1.13848671e-01
3.92773330e-01 5.91844693e-02 4.30726022e-01 5.98975599e-01
-2.07513738e-02 5.88462114e-01 -1.47751734e-01 9.29884911e-02
6.78439260e-01 3.10033038e-02 2.04267412e-01 8.07160437e-01
-8.41373086e-01 8.67258489e-01 -7.30127454e-01 2.72957623e-01
-2.12097955e+00 -8.02612126e-01 -9.59884107e-01 2.23637080e+00
8.44423056e-01 -5.47734678e-01 2.14858964e-01 4.00179327e-01
5.83485544e-01 -4.10634249e-01 -3.89639288e-02 -1.12755430e+00
3.21919799e-01 5.29331267e-01 6.70763791e-01 1.03700578e+00
-7.33034015e-01 4.50640321e-01 7.38387251e+00 9.46817636e-01
-9.18587446e-01 5.78262173e-02 -2.30693677e-03 6.49591565e-01
-1.43953994e-01 5.69957733e-01 -6.98937416e-01 3.60267609e-01
6.36601627e-01 -5.17134130e-01 7.35138059e-01 9.76668477e-01
2.50660956e-01 -6.70468092e-01 -4.71123666e-01 6.43755913e-01
-4.20183241e-01 -9.35175240e-01 -4.12872583e-01 5.76545857e-02
1.19190741e+00 -9.17125225e-01 -3.39692056e-01 3.11348736e-01
1.92780457e-02 -6.98069692e-01 5.75771689e-01 9.75442469e-01
2.90553987e-01 -1.03065097e+00 7.34422028e-01 4.36275572e-01
-1.34068024e+00 2.14304581e-01 -3.01355779e-01 -6.70560896e-01
4.20849591e-01 9.61312532e-01 -5.52736938e-01 4.85699713e-01
1.42726943e-01 1.68595575e-02 3.26903880e-01 1.06137288e+00
-2.54025578e-01 2.33368784e-01 -6.44780457e-01 -2.01620311e-01
5.15403867e-01 -1.43996561e+00 4.62910503e-01 7.04523683e-01
8.22614253e-01 6.20734155e-01 -8.93461034e-02 1.34254277e+00
6.51849627e-01 4.61569667e-01 -4.60950702e-01 -4.87729944e-02
2.32666999e-01 1.23145735e+00 -5.16642809e-01 -7.18431532e-01
-3.58942628e-01 8.31694067e-01 -2.07239222e-02 1.02350509e+00
-7.52510667e-01 -6.06772542e-01 4.22006607e-01 8.73903409e-02
3.97078365e-01 -4.30962175e-01 -3.42482507e-01 -5.81678152e-01
-2.24612113e-02 -5.08142233e-01 1.80442885e-01 -7.72768080e-01
-8.02396059e-01 4.32821214e-01 3.23824525e-01 -1.06854522e+00
-7.91263044e-01 -5.25713205e-01 -4.92544234e-01 1.51966918e+00
-8.66070092e-01 -7.83333004e-01 3.40994388e-01 6.45388663e-01
-5.12719929e-01 3.75777394e-01 1.04114175e+00 1.36429086e-01
-3.06649685e-01 1.60503332e-02 5.97323775e-01 -3.20365608e-01
-1.22439422e-01 -7.24773705e-01 -1.84399769e-01 4.81203586e-01
-9.77435410e-01 7.66456485e-01 7.71379173e-01 -8.48930001e-01
-1.16975105e+00 -2.98486024e-01 1.38690948e+00 4.93691713e-01
7.95341969e-01 3.32488924e-01 -1.23363900e+00 8.56353402e-01
-1.05238318e-01 3.57580423e-01 1.52204275e-01 -4.59983259e-01
1.90749884e-01 -3.52587365e-02 -1.53096437e+00 4.03189778e-01
1.41515404e-01 -4.80057031e-01 -1.77861094e-01 2.70971626e-01
3.24139476e-01 -6.09507501e-01 -1.16504180e+00 2.67461091e-01
6.12706006e-01 -9.43010271e-01 1.06769407e+00 -2.61049986e-01
1.38415739e-01 -7.54848301e-01 1.50197804e-01 -8.99046540e-01
-2.19398335e-01 -1.05630469e+00 -2.79472411e-01 7.58871973e-01
-1.18376732e-01 -1.13412309e+00 -5.88317327e-02 9.31637645e-01
1.59598902e-01 -1.07858622e+00 -1.03410435e+00 -9.53621089e-01
-3.35729495e-02 -9.52530131e-02 4.93434012e-01 7.91296721e-01
4.38001215e-01 -9.47066620e-02 -4.41654056e-01 1.82456933e-02
6.93582535e-01 1.69085592e-01 2.94803381e-01 -1.21967554e+00
-3.78481150e-01 -1.87288791e-01 -4.34654236e-01 -8.23602259e-01
1.25953957e-01 -7.03008890e-01 -6.94244504e-02 -1.62502050e+00
-4.75019187e-01 -3.26983154e-01 3.75547975e-01 2.87405431e-01
2.37309426e-01 -2.29439601e-01 -2.90673412e-02 2.27545440e-01
1.79458242e-02 4.70280439e-01 1.46282113e+00 3.56315464e-01
-3.79138589e-01 1.50036305e-01 1.98007464e-01 1.07544065e+00
3.27630460e-01 -3.32894862e-01 -4.04326320e-02 3.84998530e-01
2.21348390e-01 8.02658081e-01 4.05022144e-01 -1.01159132e+00
2.54851907e-01 -7.14177173e-03 -3.57269011e-02 -7.54461765e-01
3.69980812e-01 -7.65473783e-01 8.85532677e-01 6.36139870e-01
2.38997802e-01 2.04540743e-03 1.20521545e-01 1.75304506e-02
-3.55110526e-01 -9.94011402e-01 6.73308611e-01 9.47724953e-02
-2.34632537e-01 -1.57727525e-01 -8.63405943e-01 -3.06181431e-01
9.30921495e-01 -1.55201685e-02 3.57013226e-01 -5.67952037e-01
-8.34972382e-01 -7.52148405e-02 2.96181738e-01 -1.09000063e+00
3.91358823e-01 -1.94626343e+00 -1.65223405e-01 2.44663000e-01
-8.01058352e-01 2.51375705e-01 2.64164656e-01 1.49342549e+00
-1.26729918e+00 3.64951491e-01 -3.72931138e-02 -3.15857679e-01
-6.89979255e-01 4.65225875e-01 7.09285140e-01 -5.95355272e-01
-2.66030937e-01 4.49233443e-01 -7.62759149e-02 -1.15157954e-01
-1.91982046e-01 -4.41425890e-01 2.23895103e-01 7.22851902e-02
4.39803421e-01 1.32746267e+00 -1.67053387e-01 -8.15745831e-01
-1.93877086e-01 8.09798837e-01 7.80794084e-01 -5.47104359e-01
1.00627637e+00 1.74101830e-01 -9.93116498e-01 4.33883756e-01
1.58429348e+00 1.18481100e-01 -8.24576139e-01 1.48374826e-01
-5.56181550e-01 2.78138127e-02 -8.48641470e-02 -1.16052829e-01
-7.89187014e-01 6.58145726e-01 -1.37718961e-01 5.31532645e-01
9.84068871e-01 -7.12530136e-01 1.09768379e+00 3.52993280e-01
3.98217350e-01 -1.06363559e+00 -8.90249312e-01 8.43364239e-01
9.98459995e-01 -4.92748111e-01 1.83623731e-01 -7.73179889e-01
-1.72514871e-01 1.64209878e+00 -1.98079452e-01 -6.30990326e-01
9.12476897e-01 4.38876450e-01 -1.23202763e-01 1.27560019e-01
5.59648648e-02 -6.82505593e-02 2.83991247e-01 1.16730712e-01
5.57246566e-01 7.63896257e-02 -1.36063242e+00 6.12811923e-01
1.87869892e-01 5.44053018e-01 2.53930718e-01 7.37792850e-01
-2.71791518e-01 -9.34877694e-01 -9.72656548e-01 -1.89459935e-01
-2.16431934e-02 1.60243765e-01 -1.74928203e-01 9.59030151e-01
-1.03376448e-01 5.40193021e-01 -1.35992721e-01 2.78543204e-01
3.61260921e-01 4.82961386e-01 5.65483510e-01 1.85783997e-01
-4.21344191e-01 3.33096236e-01 3.09257451e-02 -4.76423591e-01
-1.19919255e-01 -5.21370351e-01 -2.04127455e+00 -4.69134539e-01
-4.31092903e-02 7.86472201e-01 4.97085214e-01 8.56977880e-01
-1.74166545e-01 3.20445150e-01 5.59019268e-01 -1.13040650e+00
-9.50620532e-01 -4.18491542e-01 -1.05551171e+00 4.21196260e-02
1.51975498e-01 -6.26576424e-01 -6.90821409e-01 -2.82211155e-01] | [6.219139099121094, 3.495626449584961] |
cd3ee58b-de74-4821-b053-3be82e610f79 | a-multi-domain-virtual-network-embedding | 2202.01473 | null | https://arxiv.org/abs/2202.01473v1 | https://arxiv.org/pdf/2202.01473v1.pdf | A multi-domain virtual network embedding algorithm with delay prediction | Virtual network embedding (VNE) is an crucial part of network virtualization (NV), which aims to map the virtual networks (VNs) to a shared substrate network (SN). With the emergence of various delay-sensitive applications, how to improve the delay performance of the system has become a hot topic in academic circles. Based on extensive research, we proposed a multi-domain virtual network embedding algorithm based on delay prediction (DP-VNE). Firstly, the candidate physical nodes are selected by estimating the delay of virtual requests, then particle swarm optimization (PSO) algorithm is used to optimize the mapping process, so as to reduce the delay of the system. The simulation results show that compared with the other three advanced algorithms, the proposed algorithm can significantly reduce the system delay while keeping other indicators unaffected. | ['Xin Li', 'Haipeng Yao', 'Yongjing Ni', 'Xue Pang', 'Peiying Zhang'] | 2022-02-03 | null | null | null | null | ['network-embedding'] | ['methodology'] | [-3.12825561e-01 -3.42607409e-01 -2.51986474e-01 2.79853880e-01
7.21925437e-01 -5.09223878e-01 3.14035147e-01 -1.99952245e-01
-2.81860624e-02 8.95076871e-01 -2.10564688e-01 -4.88406032e-01
-4.40080315e-01 -9.56722736e-01 3.11419010e-01 -7.56528974e-01
-1.32890046e-01 4.52895850e-01 6.61598921e-01 -2.65370995e-01
3.82702798e-01 8.75181913e-01 -1.12568402e+00 -5.82761526e-01
5.53664148e-01 1.01820540e+00 2.35127583e-01 1.65418535e-01
-5.21956742e-01 2.94308037e-01 -9.54716384e-01 -4.03261632e-02
1.73482776e-01 -2.34226272e-01 -5.13887882e-01 -1.96723282e-01
-9.07449126e-01 -1.09701596e-01 -5.70052385e-01 1.00331855e+00
5.62456489e-01 1.95608675e-01 1.73602134e-01 -1.86894917e+00
-4.23942149e-01 4.02944416e-01 -4.52056050e-01 6.51141942e-01
-4.37622480e-02 -8.43162388e-02 7.10189462e-01 -5.11859536e-01
8.49760115e-01 1.17740726e+00 9.10754800e-02 3.74174386e-01
-1.10333395e+00 -7.58679628e-01 1.08783640e-01 5.58305264e-01
-1.40790415e+00 -1.75518453e-01 9.61966336e-01 -1.09870106e-01
5.48984408e-01 4.30959761e-01 8.81646156e-01 6.51100159e-01
3.84161294e-01 1.67731196e-01 5.25094688e-01 -2.26126000e-01
3.04667175e-01 5.75662673e-01 -3.66882801e-01 2.11200535e-01
4.92472380e-01 2.24453673e-01 -4.21782676e-03 -3.32664400e-01
9.31197166e-01 -4.15878631e-02 -3.93560261e-01 -5.99557340e-01
-1.09931099e+00 5.09700716e-01 3.21665615e-01 3.22051108e-01
-6.35488212e-01 -1.43572360e-01 8.60396981e-01 3.79264504e-01
3.09704572e-01 2.22036123e-01 -4.15794373e-01 -2.96245635e-01
-7.70948708e-01 -2.82548964e-01 8.76077950e-01 6.13485336e-01
2.91890293e-01 3.67284179e-01 1.38564929e-01 4.27917302e-01
2.25967497e-01 4.67112392e-01 6.75061122e-02 -7.40283489e-01
6.32495433e-02 7.13169575e-01 5.40505797e-02 -1.26723313e+00
-4.31546986e-01 -4.98798728e-01 -9.20127571e-01 2.75120467e-01
-3.75556529e-01 -3.80755752e-01 -1.95360184e-01 1.42872977e+00
5.52687764e-01 7.01161087e-01 -1.72668591e-01 8.63459706e-01
3.76079261e-01 9.85645175e-01 -9.94064733e-02 -7.93624520e-01
7.63065457e-01 -9.29372191e-01 -9.94508862e-01 3.87191743e-01
2.53515840e-01 -9.47733223e-01 4.75234330e-01 8.57986212e-02
-7.16298282e-01 -5.10977328e-01 -1.21467149e+00 1.01570356e+00
-2.04716340e-01 -2.65779138e-01 5.97580612e-01 7.96960235e-01
-1.20512533e+00 4.57122326e-01 -5.47503233e-01 -6.12668276e-01
1.39965832e-01 5.43083787e-01 -1.13244198e-01 1.21669233e-01
-1.16191924e+00 6.27142251e-01 5.62463462e-01 -1.02861285e-01
-5.54142296e-01 -7.34211028e-01 -2.88197726e-01 2.20726222e-01
6.25909925e-01 -4.88521576e-01 4.60604042e-01 -7.96774149e-01
-1.61251271e+00 3.54004130e-02 2.11323872e-01 5.16280085e-02
2.34493166e-01 4.32975024e-01 -1.31292033e+00 1.98733613e-01
-3.13677520e-01 -2.03071218e-02 5.45694888e-01 -1.21920550e+00
-6.50458872e-01 6.49182722e-02 1.07291363e-01 4.62749749e-02
-5.94767213e-01 3.13700944e-01 -3.55679125e-01 -2.48559907e-01
9.51528549e-02 -7.49039531e-01 -2.81087369e-01 -6.27883449e-02
-4.19600785e-01 -8.24899226e-02 1.36849654e+00 -4.06986386e-01
1.59773791e+00 -2.18630099e+00 3.82247359e-01 7.41392136e-01
5.08193791e-01 6.67764962e-01 -1.11694425e-01 7.17017472e-01
1.09497339e-01 1.92996845e-01 4.60716814e-01 3.40715200e-01
-7.14872628e-02 1.53924748e-01 -1.73978880e-01 3.25124413e-01
-4.13884223e-01 3.27618033e-01 -7.24082947e-01 -5.98916411e-01
5.05363345e-01 3.63069952e-01 -4.16038781e-01 1.65991008e-01
3.75595540e-02 4.46700960e-01 -4.41940576e-01 5.31543434e-01
1.09283376e+00 -4.33143973e-01 7.14613259e-01 -2.46197909e-01
-2.78541654e-01 -4.70195204e-01 -1.47555053e+00 7.70926237e-01
-3.95297080e-01 5.07218897e-01 3.99756283e-01 -8.87594819e-01
1.03902376e+00 4.18633163e-01 8.39361489e-01 -7.05406308e-01
4.63080615e-01 2.77777672e-01 1.19812526e-01 -3.72957617e-01
1.35311142e-01 1.79088920e-01 4.34790730e-01 4.49312627e-01
-4.08926070e-01 6.76672697e-01 -1.93521809e-02 1.53413683e-01
1.20629549e+00 -4.44769651e-01 1.07083693e-01 -1.32577598e-01
9.96489108e-01 -2.13671982e-01 7.88473904e-01 -1.12565607e-01
-6.48652732e-01 -4.97758806e-01 7.23304391e-01 -3.49462092e-01
-1.07100534e+00 -1.19261694e+00 3.80427316e-02 4.19351429e-01
8.57967019e-01 -8.24917257e-02 -3.74948829e-01 -4.11360472e-01
-1.81318130e-02 4.12653476e-01 -1.36251166e-01 -3.98228019e-01
-2.85182834e-01 -3.75834167e-01 1.55874223e-01 -1.84255727e-02
3.16487283e-01 -1.03639889e+00 -1.32620335e-01 6.85142756e-01
2.16905311e-01 -1.14538002e+00 -2.82657981e-01 -2.44301468e-01
-7.14130700e-01 -1.08174694e+00 -2.67066985e-01 -8.67312670e-01
6.36672735e-01 6.01180315e-01 7.03580737e-01 5.05576670e-01
-2.93028235e-01 2.02019528e-01 -4.13197786e-01 1.67607546e-01
-3.02457660e-01 1.96161796e-04 5.14821768e-01 4.28073853e-02
2.04707444e-01 -1.07204008e+00 -6.21909559e-01 5.19302249e-01
-6.75179183e-01 -7.30979815e-02 4.92461413e-01 6.41667485e-01
5.63147485e-01 5.97931087e-01 8.60064447e-01 -6.90581560e-01
8.24034154e-01 -8.00428331e-01 -8.84591341e-01 2.13717431e-01
-8.98369670e-01 -3.00847083e-01 1.08066428e+00 -4.57330853e-01
-6.89236045e-01 -7.79454648e-01 3.19829047e-01 -7.03113735e-01
4.49178934e-01 2.83018619e-01 -5.00577867e-01 -4.97199267e-01
-9.34489965e-02 3.05099070e-01 2.60120064e-01 -2.23291263e-01
6.30678236e-02 6.74079478e-01 -1.03196822e-01 2.06310861e-03
9.30224895e-01 3.40909421e-01 5.26250899e-01 -6.18513286e-01
5.23065403e-02 -7.00461194e-02 -2.44669840e-01 -4.52684999e-01
3.29686314e-01 -3.67272764e-01 -1.26579857e+00 -2.04519019e-01
-1.03268671e+00 1.54399082e-01 1.33911923e-01 6.27960086e-01
-4.71525900e-02 3.77769321e-01 -4.53715593e-01 -7.06290901e-01
-2.43387625e-01 -1.09752607e+00 -8.31319671e-03 5.31208038e-01
2.10297287e-01 -1.24688590e+00 1.65191606e-01 -1.40659988e-01
8.17516446e-01 2.64979333e-01 1.09738946e+00 -5.54269373e-01
-8.43341827e-01 -4.95039560e-02 -7.15644598e-01 3.49297881e-01
2.19877869e-01 6.32171094e-01 -3.59831713e-02 -6.03455126e-01
-1.24470249e-01 6.28838480e-01 -1.45295933e-01 3.16677123e-01
1.11086631e+00 -1.49160087e-01 -6.75536215e-01 8.68534267e-01
1.98208725e+00 7.38137305e-01 6.42235100e-01 5.84842563e-01
5.61811864e-01 2.64316291e-01 8.61828327e-01 1.02579129e+00
1.09682597e-01 6.55053020e-01 1.00885558e+00 -3.51860607e-03
7.09112361e-02 2.73397751e-02 1.46775231e-01 1.27507949e+00
1.97800100e-01 -6.16352081e-01 -5.16863108e-01 3.02155465e-01
-1.54980326e+00 -9.36369896e-01 -2.50894409e-02 2.08886862e+00
-2.28000104e-01 4.65042591e-01 5.56181744e-02 3.03267330e-01
1.21904886e+00 3.34701180e-01 -6.26965106e-01 -6.07215703e-01
-1.25339374e-01 7.64747038e-02 4.89202261e-01 1.97065100e-01
-3.51487994e-01 7.40136147e-01 5.78739166e+00 9.02848959e-01
-1.37030900e+00 8.00558850e-02 -3.99129000e-03 -9.38938037e-02
-3.81263167e-01 3.37094754e-01 -2.76396602e-01 1.04199660e+00
1.12690568e+00 -7.77954638e-01 9.69525874e-01 6.55896127e-01
6.49152219e-01 2.60961831e-01 -4.61428344e-01 1.13502467e+00
-3.83546025e-01 -1.46124506e+00 2.28338659e-01 3.40962082e-01
5.01573324e-01 -3.00768256e-01 -1.17861539e-01 1.18564375e-01
-9.92503986e-02 -5.88759542e-01 3.63238156e-02 4.03249323e-01
8.56563747e-01 -1.30024421e+00 1.10114419e+00 7.82993063e-02
-1.46342003e+00 -2.63641953e-01 -5.23293912e-01 1.49900794e-01
5.32852530e-01 6.21058226e-01 -4.28933591e-01 7.77895689e-01
3.14383239e-01 3.76475155e-01 -1.15461744e-01 1.21175778e+00
1.93317443e-01 3.38123918e-01 -9.57390368e-02 -1.85626253e-01
-8.74300376e-02 -5.92706740e-01 9.39122617e-01 2.44485557e-01
3.52096528e-01 1.08204678e-01 2.48541281e-01 6.98171735e-01
-1.21680826e-01 4.54546392e-01 -1.98779508e-01 -2.69214958e-01
1.41150713e+00 1.41690147e+00 -8.98572624e-01 -1.79322243e-01
-3.42845470e-01 9.49757993e-01 -7.46188313e-02 3.79078150e-01
-1.10370696e+00 -9.54723716e-01 1.12666047e+00 1.09936319e-01
2.30262309e-01 -1.85139477e-01 -9.21303257e-02 -6.08863533e-01
-4.59119119e-02 -4.27252471e-01 -7.07345409e-03 -5.43845534e-01
-8.26387525e-01 8.02598894e-01 -6.13028169e-01 -1.48849189e+00
4.36385840e-01 -1.64034009e-01 -8.48559678e-01 8.49012971e-01
-1.61275518e+00 -4.57337111e-01 -3.36147308e-01 5.70234179e-01
5.87958843e-03 -5.56119382e-01 7.25024223e-01 8.52978945e-01
-1.11510146e+00 4.43979114e-01 6.28978074e-01 -3.83850098e-01
4.36194867e-01 -3.90230089e-01 -5.22599630e-02 8.46002162e-01
-3.76657993e-01 2.52342373e-01 6.93105459e-01 -5.59703648e-01
-1.45537996e+00 -7.07139850e-01 4.38347787e-01 2.76797652e-01
7.00733542e-01 6.41435757e-02 -5.89688003e-01 3.30841124e-01
1.63411438e-01 3.59301597e-01 7.44148612e-01 -3.27722669e-01
3.29572439e-01 -3.79806191e-01 -1.34117699e+00 6.48503065e-01
9.37251210e-01 -2.35100344e-01 2.30045229e-01 1.29940838e-01
1.25551200e+00 -6.29148930e-02 -8.87349546e-01 6.19671866e-02
1.59677073e-01 -8.42146516e-01 8.93727183e-01 -6.23868227e-01
-2.50161231e-01 -5.07723570e-01 -1.85539484e-01 -1.39892173e+00
-5.45930922e-01 -4.62035447e-01 -3.79618049e-01 1.38749552e+00
-1.63439661e-01 -1.10252547e+00 8.67703497e-01 2.16885731e-02
1.09730430e-01 -8.24177623e-01 -9.83153343e-01 -1.06126809e+00
-8.68707716e-01 2.50574619e-01 1.19898331e+00 1.00265968e+00
-9.57231820e-02 2.44218871e-01 -1.88832149e-01 4.89670604e-01
5.99956393e-01 1.27309591e-01 6.53208196e-01 -1.61082494e+00
6.36967346e-02 -5.45035779e-01 -7.43622184e-01 -7.17053592e-01
3.36605877e-01 -7.63011575e-01 -8.58276367e-01 -1.59306204e+00
-2.77279735e-01 -8.11256766e-01 -8.33915174e-01 -2.65098333e-01
2.04152688e-01 -2.82139570e-01 1.88452110e-01 3.38048041e-01
-6.72341108e-01 7.51008689e-01 1.32386291e+00 3.53899181e-01
-3.62482935e-01 6.61148801e-02 -4.86848563e-01 1.69912055e-01
8.99961174e-01 -4.47671086e-01 -8.79867375e-01 2.66047455e-02
-5.61913624e-02 8.22618067e-01 -1.96122080e-02 -9.86065209e-01
1.94052190e-01 -5.57714164e-01 -3.78299430e-02 -5.23579836e-01
2.16572270e-01 -1.45708227e+00 5.86755872e-01 8.50958824e-01
4.05470639e-01 4.06187326e-01 1.44401953e-01 7.51517534e-01
-1.85162842e-01 3.06027621e-01 7.22330987e-01 6.01278484e-01
-1.01658523e+00 7.65356004e-01 -3.13490599e-01 -9.93300304e-02
1.60245490e+00 -4.35775101e-01 -4.25560266e-01 -3.39569449e-02
-6.63967848e-01 3.95682067e-01 4.58292902e-01 3.90762657e-01
7.52772331e-01 -1.51694059e+00 -3.19216430e-01 -3.12817581e-02
-2.84325570e-01 -7.78836787e-01 7.18294740e-01 8.12147558e-01
-9.78555441e-01 5.14936388e-01 -7.36910403e-01 -2.69530833e-01
-1.23956192e+00 8.37014377e-01 5.35861626e-02 -2.95195848e-01
-3.78900260e-01 5.83741605e-01 -3.94373536e-01 -2.20741943e-01
5.28242514e-02 6.76697969e-01 -4.09800380e-01 -1.24822013e-01
3.52360278e-01 9.84104812e-01 -2.01059416e-01 -4.00936216e-01
-7.22332954e-01 1.28902227e-01 -6.82729203e-03 1.24720618e-01
1.36162949e+00 -6.31711006e-01 -7.11667001e-01 -7.71140307e-02
1.16998243e+00 4.07621935e-02 -6.68932021e-01 -3.97510350e-01
-2.13883892e-01 -9.69940901e-01 4.18336034e-01 -3.61602217e-01
-1.51785207e+00 3.57707471e-01 6.97529972e-01 4.88433272e-01
1.19280517e+00 -6.26357794e-01 1.17717886e+00 -2.96229333e-01
7.05740094e-01 -1.02624190e+00 9.25542414e-02 2.64040053e-01
1.36251710e-02 -7.88827837e-01 -7.42269456e-02 -7.78206825e-01
-3.14338207e-01 1.19009638e+00 1.09669423e+00 -4.02827449e-02
9.52712059e-01 3.82541977e-02 -2.04677448e-01 -2.03102365e-01
-8.00012529e-01 2.04974949e-01 -5.56878448e-01 6.03047609e-01
-2.32261837e-01 2.85973907e-01 -5.53430855e-01 4.89777595e-01
2.49005139e-01 -2.02700198e-01 8.54600847e-01 6.75274611e-01
-5.92165172e-01 -1.57521951e+00 -1.53160542e-01 5.12380362e-01
2.50070672e-02 4.87868451e-02 2.71465153e-01 6.45976126e-01
-6.07405156e-02 9.11887586e-01 1.52995810e-01 -9.12877440e-01
2.77913719e-01 -4.03153300e-01 -9.22653750e-02 -2.89304405e-01
-2.48829678e-01 -3.98642242e-01 -1.51188597e-01 -5.70353270e-01
4.61776294e-02 -2.41497874e-01 -1.34901905e+00 -1.27580190e+00
-3.87894243e-01 5.50904810e-01 8.65208685e-01 5.66126347e-01
7.15850949e-01 1.11797893e+00 1.50437129e+00 -2.56779969e-01
-2.42781878e-01 -1.76513851e-01 -1.11219001e+00 -1.50311962e-01
7.39774108e-02 -9.05827582e-01 -6.14107251e-01 -8.90050769e-01] | [5.880164623260498, 1.7115333080291748] |
672e3ee8-c8f7-4e1c-a3a1-82cec2ef474a | action-recognition-by-hierarchical-mid-level | 1508.07654 | null | http://arxiv.org/abs/1508.07654v1 | http://arxiv.org/pdf/1508.07654v1.pdf | Action Recognition by Hierarchical Mid-level Action Elements | Realistic videos of human actions exhibit rich spatiotemporal structures at
multiple levels of granularity: an action can always be decomposed into
multiple finer-grained elements in both space and time. To capture this
intuition, we propose to represent videos by a hierarchy of mid-level action
elements (MAEs), where each MAE corresponds to an action-related spatiotemporal
segment in the video. We introduce an unsupervised method to generate this
representation from videos. Our method is capable of distinguishing
action-related segments from background segments and representing actions at
multiple spatiotemporal resolutions. Given a set of spatiotemporal segments
generated from the training data, we introduce a discriminative clustering
algorithm that automatically discovers MAEs at multiple levels of granularity.
We develop structured models that capture a rich set of spatial, temporal and
hierarchical relations among the segments, where the action label and multiple
levels of MAE labels are jointly inferred. The proposed model achieves
state-of-the-art performance in multiple action recognition benchmarks.
Moreover, we demonstrate the effectiveness of our model in real-world
applications such as action recognition in large-scale untrimmed videos and
action parsing. | ['Silvio Savarese', 'Yuke Zhu', 'Tian Lan', 'Amir Roshan Zamir'] | 2015-08-31 | action-recognition-by-hierarchical-mid-level-1 | http://openaccess.thecvf.com/content_iccv_2015/html/Lan_Action_Recognition_by_ICCV_2015_paper.html | http://openaccess.thecvf.com/content_iccv_2015/papers/Lan_Action_Recognition_by_ICCV_2015_paper.pdf | iccv-2015-12 | ['action-parsing'] | ['natural-language-processing'] | [ 6.09785378e-01 -9.89751220e-02 -6.65273726e-01 -3.75249058e-01
-8.18984270e-01 -5.85014701e-01 6.31894112e-01 -6.77886531e-02
1.81375798e-02 3.54562372e-01 7.29817271e-01 1.21370725e-01
5.42581864e-02 -5.34905493e-01 -8.39162648e-01 -6.17950976e-01
-3.97472292e-01 8.77073184e-02 7.11409628e-01 3.29707950e-01
5.53343147e-02 4.23141003e-01 -1.77556646e+00 1.06632388e+00
4.90520507e-01 1.05959690e+00 6.93932176e-02 7.68814385e-01
1.20760232e-01 1.76103652e+00 -5.01031399e-01 -4.47835140e-02
1.65449828e-01 -8.39995444e-01 -9.63748217e-01 9.66891229e-01
6.27030432e-01 -4.51301962e-01 -5.77882946e-01 9.40610170e-01
-3.83271873e-01 3.85326296e-01 5.69544017e-01 -1.45076203e+00
-3.17542762e-01 5.60463727e-01 -6.23680294e-01 3.42285573e-01
5.01746237e-01 1.85823068e-01 1.11728644e+00 -6.19064331e-01
8.56490374e-01 1.25683117e+00 2.34781578e-01 3.71313304e-01
-1.16156435e+00 -2.54483819e-01 7.35838175e-01 4.68284756e-01
-1.06726801e+00 -4.23127383e-01 6.24067008e-01 -7.61196494e-01
8.23887050e-01 2.67564833e-01 7.44838059e-01 1.24396956e+00
4.37813923e-02 1.31210256e+00 8.92923176e-01 -1.06325597e-01
4.02194381e-01 -5.94660878e-01 8.08510035e-02 7.78210104e-01
-3.12165052e-01 -4.15670574e-01 -6.96791351e-01 -1.23642378e-01
1.11628687e+00 2.97948211e-01 4.83767539e-02 -3.96462977e-01
-1.55837667e+00 4.44673747e-01 -5.88752795e-03 4.77801532e-01
-6.66210115e-01 4.54445481e-01 5.23961127e-01 -2.08129361e-01
7.02944472e-02 -3.91408335e-03 -2.00971052e-01 -4.40743417e-01
-9.78658974e-01 -1.68769602e-02 3.37879628e-01 1.08324111e+00
5.74352026e-01 -1.40708625e-01 -6.08169615e-01 6.24275923e-01
-3.12328655e-02 1.25146247e-02 4.84361470e-01 -1.66417193e+00
5.58208108e-01 8.05264294e-01 4.24357027e-01 -1.09852684e+00
-3.70843895e-02 1.37003496e-01 -7.39551485e-01 -1.49777591e-01
3.38583380e-01 2.89320260e-01 -9.52185154e-01 1.84118557e+00
5.11309862e-01 9.09801900e-01 -8.84093642e-02 7.44760931e-01
4.23677415e-01 8.38052928e-01 5.24172962e-01 -3.86467069e-01
1.41109931e+00 -1.28156710e+00 -6.58252478e-01 -1.66370422e-01
6.78124487e-01 -8.02406594e-02 9.06455398e-01 2.53602266e-01
-1.14223742e+00 -8.27765882e-01 -5.73666036e-01 -5.26905134e-02
-1.77734848e-02 4.39043671e-01 4.29309934e-01 1.30028771e-02
-6.80383921e-01 4.78738129e-01 -1.40544558e+00 -3.25466067e-01
4.44225550e-01 -3.15192156e-02 -5.73047161e-01 8.33193585e-03
-8.88279617e-01 2.44120866e-01 5.49805582e-01 1.68968476e-02
-1.23929107e+00 -1.46946818e-01 -9.65102613e-01 -2.26399098e-02
6.62574828e-01 -3.67453188e-01 1.25180900e+00 -1.35927379e+00
-1.15674603e+00 8.46204042e-01 -5.57177246e-01 -6.82454824e-01
2.97748804e-01 -2.51278162e-01 -6.42670274e-01 7.74635911e-01
3.99660617e-01 5.82751274e-01 9.21851814e-01 -9.92655039e-01
-1.31710029e+00 -3.23745966e-01 3.24732661e-01 2.01768056e-01
-1.13041013e-01 2.80986041e-01 -8.52783918e-01 -9.57125664e-01
1.79085415e-02 -8.90904129e-01 -3.25903684e-01 -1.97207540e-01
-3.11054528e-01 -2.85482049e-01 7.00685918e-01 -6.84316754e-01
1.51863599e+00 -2.40809798e+00 4.28719252e-01 -2.19715491e-01
1.63944244e-01 -2.57736538e-02 -1.94870487e-01 3.97126675e-01
9.49074849e-02 -5.19787110e-02 -2.42509753e-01 -1.95142552e-01
1.21120587e-01 4.54845935e-01 -3.95543993e-01 3.78341854e-01
1.79541916e-01 9.45697486e-01 -1.20355141e+00 -7.75229990e-01
3.14733893e-01 2.21972734e-01 -3.12984765e-01 2.88420528e-01
-4.52582240e-01 7.14794099e-01 -7.93202698e-01 8.16864133e-01
1.07600965e-01 -6.00052178e-01 3.60270590e-01 -2.24165902e-01
1.41735105e-02 1.53886914e-01 -1.28611684e+00 1.68920100e+00
8.88609141e-02 4.36793834e-01 -5.47244102e-02 -1.19477499e+00
2.96522111e-01 2.89452970e-01 1.07972598e+00 -3.95325333e-01
-2.11270139e-01 -1.84629187e-01 -3.19057882e-01 -7.93105960e-01
4.57955003e-01 2.58044541e-01 -5.23694336e-01 4.57948565e-01
3.39252762e-02 6.44818664e-01 6.07562840e-01 2.79175788e-01
1.46270108e+00 4.99725997e-01 4.93251503e-01 1.43327460e-01
5.96885681e-01 8.88839066e-02 1.05453324e+00 8.01002026e-01
-5.95684409e-01 3.02702069e-01 7.50209630e-01 -5.60661256e-01
-7.53108919e-01 -9.33145046e-01 2.58807391e-01 1.59543490e+00
2.38329872e-01 -7.03677475e-01 -9.48048711e-01 -8.49485517e-01
-2.74311990e-01 3.19579244e-01 -8.88681650e-01 2.48894524e-02
-8.95616293e-01 -2.84920901e-01 4.94198024e-01 9.43967938e-01
6.58048630e-01 -1.28480649e+00 -9.92847145e-01 3.44617277e-01
-6.32416248e-01 -1.63432276e+00 -6.72676504e-01 -2.64422387e-01
-8.21107745e-01 -1.18659580e+00 -2.77406842e-01 -7.85436451e-01
5.20361602e-01 3.83228779e-01 1.25300431e+00 -2.78449297e-01
-2.86614150e-01 5.67384720e-01 -6.60991013e-01 4.22223002e-01
-3.64922851e-01 -4.59568053e-01 4.31267321e-02 7.42701888e-01
5.45717657e-01 -4.48150456e-01 -6.91606700e-01 6.37667954e-01
-1.10021150e+00 2.98517615e-01 4.88464624e-01 3.24909747e-01
1.16557908e+00 2.04146981e-01 1.93299055e-01 -6.21222675e-01
-1.14529997e-01 -5.73808968e-01 -3.46319675e-01 4.67124671e-01
4.56763089e-01 -1.86207473e-01 6.11555159e-01 -6.24519646e-01
-9.98492777e-01 5.51837504e-01 4.64109898e-01 -6.22102320e-01
-7.97258973e-01 2.86774427e-01 -3.16394538e-01 5.01298249e-01
2.31439918e-01 4.99454677e-01 -5.43013036e-01 -4.56166506e-01
6.23019516e-01 4.59809452e-01 8.39097679e-01 -7.84364998e-01
2.45386794e-01 8.07083011e-01 -4.59054820e-02 -8.05691838e-01
-1.02487755e+00 -5.31350136e-01 -1.19470274e+00 -5.04536867e-01
1.44504309e+00 -1.11041737e+00 -3.52418363e-01 5.07559776e-01
-8.48078132e-01 -5.87140381e-01 -4.15593624e-01 4.66980904e-01
-9.03530777e-01 6.48226917e-01 -8.57493222e-01 -5.81470311e-01
4.70747113e-01 -1.13229966e+00 1.38621891e+00 6.14397787e-02
-3.44641119e-01 -7.44942784e-01 4.70680036e-02 6.66604817e-01
-4.89655197e-01 7.03837991e-01 6.53885961e-01 -3.95179272e-01
-8.61816108e-01 -3.06454375e-02 8.80751293e-03 2.58398414e-01
3.86029154e-01 1.87173992e-01 -4.09809530e-01 -4.22459207e-02
-1.87842175e-01 -3.34513426e-01 8.72562051e-01 4.74130452e-01
1.47150624e+00 -4.54826713e-01 -4.26260591e-01 3.08515817e-01
9.84066606e-01 4.33743030e-01 5.93535423e-01 5.26559576e-02
9.86334026e-01 5.83071172e-01 9.91840243e-01 5.24553180e-01
3.10512602e-01 8.13514531e-01 3.41791242e-01 2.23814666e-01
3.39021198e-02 -4.29910541e-01 7.40909755e-01 5.54413557e-01
-3.98841769e-01 -8.68205354e-02 -6.48129702e-01 7.39466131e-01
-2.39384890e+00 -1.54752791e+00 -9.43404213e-02 1.91624629e+00
5.59013963e-01 -1.03686722e-02 7.07975507e-01 -1.89465329e-01
8.71088684e-01 5.20607352e-01 -8.37635458e-01 1.26389012e-01
-4.98232339e-03 -3.24346930e-01 2.05111980e-01 1.77215829e-01
-1.56444180e+00 1.00183892e+00 6.27390575e+00 7.25352168e-01
-6.06489003e-01 -4.40053754e-02 6.59320772e-01 -1.92528635e-01
2.43716836e-01 -1.53175250e-01 -5.51157117e-01 6.47918820e-01
8.08528543e-01 1.82680950e-01 9.13851112e-02 9.97809172e-01
4.72877294e-01 -1.15091234e-01 -1.40445244e+00 7.46901572e-01
9.77398902e-02 -1.42649031e+00 1.94191977e-01 1.41047612e-01
9.72099841e-01 -2.36887470e-01 -3.12923253e-01 2.23080724e-01
5.00821471e-01 -9.06460404e-01 9.17099297e-01 5.34465015e-01
6.09861314e-01 -3.56344074e-01 8.45590979e-02 5.00476658e-01
-1.84001756e+00 -3.68286610e-01 -9.13777668e-03 -8.73752311e-03
3.35112095e-01 9.62964725e-03 -1.09983683e-01 3.51854473e-01
6.17046535e-01 1.33638549e+00 -5.31448245e-01 6.43259943e-01
-2.50752747e-01 6.14382982e-01 -2.42435783e-02 5.13210773e-01
5.09197772e-01 -3.74886870e-01 2.54193366e-01 1.42139745e+00
2.23236740e-01 6.08128309e-01 6.76473558e-01 4.59063292e-01
5.73332198e-02 -2.65554845e-01 -4.75240082e-01 -2.94382513e-01
4.71906185e-01 1.02934504e+00 -9.93289173e-01 -7.55176783e-01
-8.18197846e-01 1.20065010e+00 2.55838484e-01 4.42962795e-01
-1.19989836e+00 1.84452981e-01 8.57709050e-01 1.16636537e-01
6.51916444e-01 -2.71393746e-01 3.01935345e-01 -1.52298665e+00
2.85296232e-01 -9.98852193e-01 8.63205254e-01 -7.71834791e-01
-9.80133235e-01 4.57883477e-01 1.80639461e-01 -1.69588208e+00
-4.95492816e-01 -4.79488641e-01 -3.34606528e-01 1.20943725e-01
-8.11865091e-01 -1.16081440e+00 -3.69688004e-01 9.08467114e-01
1.08688188e+00 1.16194785e-01 6.44680262e-01 1.40981808e-01
-7.08977342e-01 -1.08143203e-02 -1.43005863e-01 6.43666565e-01
1.78323120e-01 -1.07276893e+00 2.78319806e-01 1.16068530e+00
7.77257919e-01 4.10491943e-01 3.68826449e-01 -6.47437274e-01
-1.13301253e+00 -1.40310860e+00 6.64510489e-01 -7.22803891e-01
9.53859389e-01 -3.40180278e-01 -8.36631775e-01 1.23750651e+00
-6.40286952e-02 2.96528697e-01 7.96281576e-01 -2.77074873e-01
-5.34396172e-01 8.07506666e-02 -7.19734013e-01 4.84035730e-01
1.49147415e+00 -8.14192057e-01 -8.48626673e-01 3.21338743e-01
4.81555462e-01 -2.22439155e-01 -1.01407647e+00 4.54643995e-01
5.97836316e-01 -1.11319292e+00 1.02746248e+00 -1.12467623e+00
7.09387541e-01 -5.47417998e-01 -3.97609800e-01 -7.23681271e-01
-5.61777472e-01 -4.43988413e-01 -7.94505298e-01 1.02009058e+00
-7.54880533e-02 1.86199516e-01 9.24208283e-01 5.75364828e-01
3.85128222e-02 -6.71795428e-01 -8.86291802e-01 -7.98545361e-01
-4.81808990e-01 -4.52132136e-01 3.49126726e-01 8.48381639e-01
1.30457669e-01 1.10681750e-01 -6.15023732e-01 2.55616158e-01
5.71692705e-01 6.63755238e-01 7.62031555e-01 -9.18271601e-01
-5.16762197e-01 -2.98418462e-01 -6.70938909e-01 -1.55759895e+00
3.55827481e-01 -3.19788247e-01 2.39575639e-01 -1.60783839e+00
4.59135830e-01 7.16346130e-02 -5.32683790e-01 5.92039645e-01
-7.11780488e-02 3.45497668e-01 1.99905604e-01 4.94965643e-01
-1.39232266e+00 2.22088680e-01 1.05711496e+00 -1.61424965e-01
-1.42528400e-01 -5.51806875e-02 -1.98308721e-01 1.11194026e+00
3.79057586e-01 -2.87965953e-01 -5.65245509e-01 -3.58183324e-01
-3.02920103e-01 5.39196968e-01 4.10414517e-01 -1.09161377e+00
1.05074802e-02 -7.57901549e-01 1.80464461e-01 -5.37506223e-01
3.98023993e-01 -8.09342325e-01 4.22491878e-01 1.44751906e-01
-6.61956787e-01 -2.02590168e-01 -3.12521458e-01 9.16256487e-01
-5.90414345e-01 1.86749727e-01 6.25042617e-01 -3.55454654e-01
-1.31162786e+00 5.58842778e-01 -6.96386337e-01 4.20118868e-02
1.44182825e+00 -3.00004184e-01 -9.60617065e-02 -2.85813659e-01
-1.16350496e+00 1.71838522e-01 4.61585045e-01 6.02378547e-01
4.04085368e-01 -1.63850820e+00 -3.79868716e-01 -5.63577451e-02
3.50417703e-01 -3.10007215e-01 4.97324914e-01 9.38585341e-01
-2.33590335e-01 2.67952263e-01 -3.43100071e-01 -7.39432395e-01
-1.42494130e+00 7.86370635e-01 1.17940612e-01 -3.87338936e-01
-9.18244779e-01 6.54579878e-01 5.93021035e-01 3.51409256e-01
4.58542764e-01 -7.23451197e-01 -3.15455914e-01 9.02728587e-02
7.57782996e-01 4.66273874e-01 -6.48202837e-01 -1.40608227e+00
-3.82680058e-01 6.34627223e-01 2.64366359e-01 -1.97355021e-02
1.08204865e+00 -1.98944882e-01 -8.59154090e-02 7.87769139e-01
1.06492066e+00 -2.00996757e-01 -1.99819016e+00 -2.42219299e-01
2.53513664e-01 -6.35942340e-01 -5.33617258e-01 -3.60721052e-01
-8.84222984e-01 5.42752743e-01 1.83610935e-02 3.81679296e-01
1.31221080e+00 4.37497467e-01 8.68038893e-01 2.70746320e-01
5.57630479e-01 -1.29796660e+00 5.24282694e-01 3.32018107e-01
5.25721312e-01 -1.07342255e+00 -2.69417554e-01 -4.83769596e-01
-1.09562206e+00 8.20888877e-01 5.42436719e-01 -1.49637863e-01
2.90441632e-01 1.07916497e-01 -2.58535594e-01 -2.47420371e-01
-7.79148161e-01 -3.94753098e-01 3.41937631e-01 3.85526448e-01
4.36982736e-02 2.31821045e-01 -1.13217361e-01 6.38876975e-01
5.22484362e-01 1.74857110e-01 3.10351342e-01 1.15249801e+00
-6.40475512e-01 -9.29627836e-01 -1.24928094e-01 1.88921630e-01
-5.63010156e-01 3.99877936e-01 -4.57856745e-01 5.31838059e-01
3.80943328e-01 8.92534971e-01 2.89084941e-01 -4.40099359e-01
1.69341728e-01 1.28034845e-01 5.01353920e-01 -5.82425475e-01
-1.17228083e-01 2.90145725e-01 1.79770187e-01 -1.26458824e+00
-1.11215770e+00 -1.06572282e+00 -1.40459514e+00 2.95504540e-01
5.97575784e-01 1.45055056e-01 -1.56851247e-01 1.04063606e+00
5.23566782e-01 3.83361876e-01 5.57176054e-01 -7.72532880e-01
5.77998646e-02 -7.27382243e-01 -7.70489275e-01 1.02388322e+00
3.29335600e-01 -6.67499244e-01 -8.38157386e-02 1.01162195e+00] | [8.379782676696777, 0.6105308532714844] |
37df107c-c6ff-40fd-9be0-532ce3680ff7 | orchard-a-benchmark-for-measuring-systematic | 2111.14034 | null | https://arxiv.org/abs/2111.14034v1 | https://arxiv.org/pdf/2111.14034v1.pdf | ORCHARD: A Benchmark For Measuring Systematic Generalization of Multi-Hierarchical Reasoning | The ability to reason with multiple hierarchical structures is an attractive and desirable property of sequential inductive biases for natural language processing. Do the state-of-the-art Transformers and LSTM architectures implicitly encode for these biases? To answer this, we propose ORCHARD, a diagnostic dataset for systematically evaluating hierarchical reasoning in state-of-the-art neural sequence models. While there have been prior evaluation frameworks such as ListOps or Logical Inference, our work presents a novel and more natural setting where our models learn to reason with multiple explicit hierarchical structures instead of only one, i.e., requiring the ability to do both long-term sequence memorizing, relational reasoning while reasoning with hierarchical structure. Consequently, backed by a set of rigorous experiments, we show that (1) Transformer and LSTM models surprisingly fail in systematic generalization, and (2) with increased references between hierarchies, Transformer performs no better than random. | ['Alvin Chan', 'Bill Tuck Weng Pung'] | 2021-11-28 | null | null | null | null | ['relational-reasoning', 'systematic-generalization'] | ['natural-language-processing', 'reasoning'] | [ 4.55828846e-01 4.73663360e-01 -1.59147874e-01 -3.19138229e-01
-3.49316120e-01 -6.90999269e-01 8.44002068e-01 4.34373975e-01
-3.08451682e-01 7.72429764e-01 5.93080521e-01 -8.69813263e-01
-2.64551908e-01 -1.20825529e+00 -8.74873459e-01 -5.34005225e-01
-2.55246639e-01 7.54691064e-01 3.87299448e-01 -5.00539839e-01
3.87582660e-01 3.59051019e-01 -1.53927541e+00 6.40440822e-01
7.62294233e-01 8.14230263e-01 -2.44295418e-01 6.71474993e-01
-3.12718213e-01 1.87538612e+00 -2.70305812e-01 -7.03577638e-01
-2.77845591e-01 -5.58986545e-01 -1.53081024e+00 -5.66665292e-01
5.75963199e-01 -4.01516855e-01 -2.60537177e-01 9.45492983e-01
3.27845216e-01 5.62903956e-02 5.49975097e-01 -9.06469584e-01
-9.61213946e-01 1.32681501e+00 4.37534861e-02 3.94890785e-01
5.68879128e-01 4.84954506e-01 1.39289546e+00 -4.52750266e-01
6.34175003e-01 1.47574198e+00 9.77006555e-01 6.50508344e-01
-1.52292299e+00 -3.86856794e-01 4.43728626e-01 1.96077630e-01
-8.96744132e-01 -3.40128541e-01 3.64227712e-01 -3.51589233e-01
1.29376280e+00 1.61046073e-01 7.68537998e-01 1.28239107e+00
2.88438678e-01 8.60240877e-01 1.33830357e+00 -2.91044563e-01
2.37814024e-01 -2.53850043e-01 5.43683529e-01 1.08602786e+00
2.59073108e-01 2.46414438e-01 -5.97175360e-01 -9.47695971e-02
4.97480243e-01 -1.58022270e-01 -5.95862679e-02 -3.08402956e-01
-1.38392925e+00 6.58589840e-01 6.89445555e-01 4.99918967e-01
-2.56038159e-01 3.82743597e-01 6.86571896e-01 3.76043528e-01
-2.07445212e-02 8.51913512e-01 -3.27095985e-01 5.47153801e-02
-8.82129967e-01 4.68796849e-01 9.14451957e-01 7.22367406e-01
4.79570985e-01 7.06661344e-02 -5.40201962e-01 2.98983425e-01
1.41984038e-02 3.04339975e-01 4.93787706e-01 -8.46450865e-01
2.53087938e-01 6.21313214e-01 -2.75551826e-01 -7.76944399e-01
-5.28053463e-01 -5.29526114e-01 -8.99953961e-01 -2.60191470e-01
4.52245444e-01 2.52132654e-01 -7.76231349e-01 2.18014169e+00
-1.67227715e-01 -1.03600390e-01 2.30635941e-01 4.94041860e-01
6.98564291e-01 3.01315486e-01 4.52477008e-01 -1.05134018e-01
1.33986795e+00 -6.35911763e-01 -4.75923866e-01 -2.91302711e-01
9.26028192e-01 -8.10701102e-02 1.33806682e+00 4.85197157e-01
-1.64236009e+00 -3.77347797e-01 -1.07534587e+00 -5.31189442e-01
-5.27164280e-01 -5.77896595e-01 1.05989969e+00 4.73875582e-01
-1.52298450e+00 5.13213933e-01 -5.90323865e-01 -5.53802729e-01
3.57674599e-01 1.09494865e-01 -1.69602577e-02 -7.71558806e-02
-1.66914690e+00 1.41718316e+00 6.01393700e-01 6.41970262e-02
-1.28488326e+00 -8.35297465e-01 -8.11222196e-01 2.31323287e-01
3.39680552e-01 -1.12752700e+00 1.46051109e+00 -6.50826514e-01
-1.19595826e+00 1.16240931e+00 -3.09694827e-01 -9.33752120e-01
4.56878096e-01 -2.31000140e-01 1.31080421e-02 3.16141471e-02
-3.70838381e-02 6.19379640e-01 2.95911014e-01 -1.26424992e+00
-2.74398446e-01 -3.73477578e-01 5.04050672e-01 -2.79311575e-02
-2.42764816e-01 -1.05621256e-01 4.12747115e-01 -3.73243839e-01
9.28688496e-02 -5.42769134e-01 -8.95263031e-02 -2.22391546e-01
-5.42687714e-01 -5.18586099e-01 2.66338855e-01 -4.17336464e-01
1.22738206e+00 -1.68419945e+00 3.64133447e-01 4.54122089e-02
4.42010760e-01 -3.27040404e-02 -1.22433431e-01 5.29789209e-01
-2.20783934e-01 5.00667453e-01 -4.91493404e-01 1.03930064e-01
4.12216753e-01 3.29597473e-01 -8.74135792e-01 1.21792234e-01
3.02200675e-01 1.31420231e+00 -1.20989037e+00 -7.20872879e-01
-2.19479367e-01 -1.71435606e-02 -7.17904806e-01 1.43188655e-01
-6.42282665e-01 7.88228214e-02 -1.79552481e-01 4.65649366e-01
3.06733310e-01 -5.78361034e-01 4.47478831e-01 1.67511832e-02
6.67823851e-02 1.05059898e+00 -4.57191676e-01 1.68866229e+00
-4.74423051e-01 3.23540300e-01 -3.17247123e-01 -9.69336271e-01
4.90387708e-01 3.19535464e-01 -1.84221461e-01 -9.98385787e-01
-1.21812873e-01 1.85094163e-01 3.61719072e-01 -4.04193431e-01
3.39548349e-01 -6.60993516e-01 -2.32201949e-01 6.66101754e-01
3.99609767e-02 -1.93850502e-01 4.79216099e-01 5.39604008e-01
1.42621851e+00 7.53305480e-02 3.37295324e-01 -3.71282190e-01
5.54762661e-01 -7.06562996e-02 3.46285939e-01 1.31978190e+00
3.93894166e-02 -3.08322292e-02 7.53061175e-01 -6.49828434e-01
-1.06902337e+00 -1.42521000e+00 -5.77341914e-02 1.47789407e+00
-2.43337750e-01 -2.12004006e-01 -6.14169836e-01 -4.82682765e-01
-4.65579256e-02 1.14105940e+00 -8.48851860e-01 -4.73073840e-01
-7.19568968e-01 -4.98541355e-01 1.27971220e+00 8.27862322e-01
5.87529540e-01 -1.39234352e+00 -8.44982386e-01 1.96716368e-01
-5.60007207e-02 -8.80354404e-01 1.37466371e-01 5.16243160e-01
-1.07060432e+00 -8.89867425e-01 -2.43609086e-01 -6.58377707e-01
4.07186717e-01 -4.19233978e-01 1.89124703e+00 4.76366073e-01
1.11145884e-01 3.46525282e-01 -3.23601440e-02 -1.60708785e-01
-5.19751787e-01 4.99953568e-01 -2.29615778e-01 -7.15484798e-01
3.25876683e-01 -9.01754200e-01 -1.90532357e-01 4.72597480e-02
-1.08354533e+00 2.12131888e-01 7.37246454e-01 1.06314552e+00
3.50122824e-02 -1.27744779e-01 5.08003831e-01 -1.12281513e+00
8.60208631e-01 -3.74322265e-01 -3.95656437e-01 5.43804944e-01
-5.39425552e-01 6.73488617e-01 8.26887250e-01 -1.72417790e-01
-1.15455794e+00 -5.56887925e-01 -2.32903749e-01 3.39633673e-02
-9.01607946e-02 8.68668914e-01 6.11066855e-02 2.86778569e-01
8.50912333e-01 5.72757483e-01 -1.52644783e-01 -1.51537973e-02
5.17920494e-01 -2.02341065e-01 6.63045049e-01 -1.27864516e+00
7.47070491e-01 3.47371846e-01 1.20607100e-01 -3.54395926e-01
-1.39317822e+00 2.10610017e-01 -5.34804821e-01 6.35142028e-02
1.02306628e+00 -5.56553781e-01 -1.28846133e+00 3.20221156e-01
-1.30128312e+00 -9.23643887e-01 -1.90721124e-01 -1.39673963e-01
-7.67740011e-01 1.94768086e-01 -1.13959289e+00 -9.72307384e-01
-3.77654135e-01 -9.60515261e-01 7.26955771e-01 -2.17239216e-01
-7.13330388e-01 -1.14479268e+00 -8.57904106e-02 4.61483672e-02
6.30432189e-01 3.40317227e-02 1.83987176e+00 -6.47540271e-01
-7.60130763e-01 2.63579130e-01 -3.32781225e-01 -4.06804830e-02
-3.35865051e-01 -1.31424084e-01 -9.03748691e-01 -1.08253747e-01
6.36051148e-02 -9.75429535e-01 1.13375950e+00 -1.10793024e-01
1.34967113e+00 -4.94819999e-01 -2.12569341e-01 4.47719514e-01
1.10230803e+00 1.30179837e-01 9.08285081e-01 3.90429765e-01
5.15155911e-01 6.76810265e-01 -4.64005098e-02 5.79213537e-02
6.92544878e-01 7.30912462e-02 1.89619958e-01 2.66234219e-01
-6.63870722e-02 -6.73572302e-01 2.96051025e-01 6.67904139e-01
8.10571611e-02 -8.17565620e-02 -1.39143276e+00 6.76311135e-01
-1.68426561e+00 -1.25125325e+00 1.80581257e-01 1.97470009e+00
1.34309256e+00 4.67857003e-01 -4.11384217e-02 2.68905759e-01
2.48084649e-01 3.38485420e-01 -6.90442204e-01 -7.34019995e-01
-2.78047591e-01 4.71867055e-01 2.84147680e-01 4.59129781e-01
-6.49981499e-01 1.02938974e+00 6.97375536e+00 4.47746634e-01
-9.76447046e-01 -1.19800858e-01 5.66445410e-01 -3.94424349e-02
-8.66048872e-01 9.42812040e-02 -6.03219926e-01 2.44177692e-02
1.18005013e+00 -2.34766118e-02 7.14138269e-01 3.67281944e-01
-4.78975356e-01 -8.50233436e-02 -1.71965992e+00 4.13294584e-01
-5.70059121e-02 -1.48253834e+00 5.21883845e-01 -1.61470324e-01
4.51348305e-01 -2.11037010e-01 1.51611850e-01 8.06770205e-01
1.08273792e+00 -1.46357918e+00 9.45091426e-01 6.56800449e-01
3.71677876e-01 -4.67916667e-01 5.78652859e-01 4.48086470e-01
-7.90397346e-01 -3.74102622e-01 -1.06442630e-01 -2.62132376e-01
-2.90536135e-02 4.35915530e-01 -4.90151167e-01 2.98691094e-01
6.06996775e-01 4.67809796e-01 -8.71539056e-01 4.80696529e-01
-6.60051465e-01 7.17482626e-01 3.45863402e-03 -2.68291980e-01
4.66185689e-01 3.20578575e-01 1.69600576e-01 1.32225776e+00
-2.15487860e-04 1.63578451e-01 -1.11234114e-02 1.03820622e+00
-1.86693162e-01 -4.30517197e-01 -8.76847029e-01 -2.81456023e-01
4.81524587e-01 5.68383813e-01 -4.29252923e-01 -6.85331225e-01
-1.11723188e-02 5.39336741e-01 7.76640534e-01 4.28032041e-01
-6.57139599e-01 -1.40349707e-02 2.87105173e-01 1.94474906e-02
1.12818785e-01 -1.84899002e-01 -7.95286536e-01 -9.88987744e-01
-7.54187852e-02 -1.12518251e+00 7.22030640e-01 -9.84175742e-01
-1.49825048e+00 4.03930992e-01 6.15586564e-02 -3.05037051e-01
-4.62479472e-01 -6.56687081e-01 -4.64469194e-01 7.12713897e-01
-1.32613754e+00 -1.23341918e+00 -4.01474014e-02 4.26672488e-01
1.60545439e-01 9.59966183e-02 1.06354225e+00 -1.07810616e-01
-2.43720680e-01 5.31535566e-01 -5.68432033e-01 3.32703799e-01
3.02634209e-01 -1.46761215e+00 5.20999134e-01 7.41536081e-01
-3.21464078e-03 1.42109394e+00 8.11287463e-01 -5.01445651e-01
-1.53710163e+00 -7.38572299e-01 1.15005207e+00 -7.83963323e-01
9.85080183e-01 -5.38589537e-01 -1.11444390e+00 1.14614534e+00
4.66563374e-01 -2.39030913e-01 4.85023260e-01 6.64056957e-01
-9.51529324e-01 2.36779377e-02 -9.26851153e-01 9.24489558e-01
1.57531393e+00 -1.00225604e+00 -1.18616915e+00 5.21247983e-02
9.47874308e-01 -2.14406937e-01 -8.12160492e-01 4.22850996e-01
6.10865057e-01 -1.12271249e+00 9.36966002e-01 -1.15308940e+00
1.08846533e+00 -1.98667869e-01 -1.65694833e-01 -1.09896910e+00
-4.42151457e-01 -2.85223961e-01 -1.37921929e-01 9.94066238e-01
6.21355355e-01 -8.65290940e-01 5.35268426e-01 4.50095028e-01
-2.68274248e-01 -9.04262245e-01 -6.45828843e-01 -6.26572549e-01
5.62813163e-01 -4.15330619e-01 6.55892730e-01 8.56943846e-01
2.86811501e-01 5.84098458e-01 2.59763841e-02 -1.40035287e-01
5.13159692e-01 1.97790548e-01 3.82766068e-01 -1.23432589e+00
-2.63527274e-01 -6.96664155e-01 -7.16332197e-02 -9.96713340e-01
6.57338560e-01 -1.16668487e+00 2.09269777e-01 -1.83815420e+00
4.03039217e-01 -2.79738873e-01 -4.31220382e-01 7.01729715e-01
-4.66599800e-02 -2.17624620e-01 -5.34074791e-02 -6.14349917e-02
-7.69590557e-01 4.47563469e-01 1.06781983e+00 -3.55547607e-01
4.62793887e-01 -5.97218215e-01 -9.46638286e-01 6.70901179e-01
5.63004017e-01 -2.72687286e-01 -6.26037359e-01 -6.69937491e-01
9.80246723e-01 1.73255026e-01 7.44493783e-01 -9.30117249e-01
4.16724861e-01 4.14615646e-02 4.16321874e-01 -5.88653564e-01
6.84478208e-02 -3.07980776e-01 -3.02720457e-01 7.22101331e-01
-1.22573090e+00 4.51732486e-01 4.04681057e-01 2.89422333e-01
-8.46568123e-02 -6.18620813e-02 6.66992188e-01 -6.81011736e-01
-6.20523989e-01 -2.34719459e-02 -6.07219458e-01 5.18379331e-01
3.15036416e-01 2.68859975e-02 -8.05423081e-01 -3.21731865e-01
-6.61178648e-01 3.22495580e-01 4.24853176e-01 1.21371627e-01
5.60555875e-01 -1.03616619e+00 -5.79488218e-01 -1.19508177e-01
1.54520646e-01 -4.14341837e-02 4.95345294e-02 9.84429598e-01
-3.39068860e-01 7.45424986e-01 -2.23605454e-01 -4.49904501e-01
-6.43011987e-01 9.25589383e-01 4.59482461e-01 -7.76906252e-01
-3.36489737e-01 9.75691319e-01 4.99103516e-01 -8.09597850e-01
8.82291123e-02 -6.90872431e-01 1.25283182e-01 1.26026049e-01
3.12349439e-01 -5.13909124e-02 1.37224615e-01 -6.13319278e-02
-4.07950312e-01 9.03857350e-02 -3.01122338e-01 -2.01395050e-01
1.13516569e+00 2.60252386e-01 -7.12858796e-01 7.67552555e-01
7.64849544e-01 -4.03001696e-01 -5.77386796e-01 -2.86123604e-01
5.90988338e-01 2.13407531e-01 -3.13686132e-01 -1.07796597e+00
-3.28236401e-01 1.17412436e+00 1.18895508e-01 2.91402280e-01
9.58394587e-01 -1.17731892e-01 6.57998741e-01 8.86285543e-01
5.47908723e-01 -5.49171507e-01 1.80737510e-01 1.18009424e+00
7.95236647e-01 -7.16693938e-01 -1.84415683e-01 7.98157752e-02
-3.39224368e-01 7.99163342e-01 6.86480880e-01 5.15137473e-03
6.53700233e-02 3.31255376e-01 -2.97324151e-01 -4.73955512e-01
-1.40138638e+00 -3.08412462e-01 -5.41579686e-02 3.66275519e-01
9.21571374e-01 -1.80401161e-01 -7.16558024e-02 2.56605268e-01
-5.63061714e-01 1.76007673e-01 3.05365950e-01 9.25762773e-01
-2.60933369e-01 -7.05894828e-01 -1.68128058e-01 5.17524123e-01
-3.13355237e-01 -7.45387256e-01 -5.94753623e-01 8.08099389e-01
-1.24625035e-01 6.51937544e-01 7.91393518e-02 -9.39863846e-02
2.40987211e-01 5.01665950e-01 7.80601561e-01 -5.78988194e-01
-9.69583035e-01 -8.04561377e-01 4.36342895e-01 -5.25217414e-01
-1.47967428e-01 -5.22378802e-01 -1.29029155e+00 -4.51046497e-01
4.11044896e-01 1.56589672e-01 -9.53434333e-02 1.21069026e+00
-3.83930914e-02 8.10574889e-01 -1.48644179e-01 -4.51118767e-01
-1.00167072e+00 -8.68850648e-01 -1.79853827e-01 4.49673504e-01
5.10796010e-01 -5.15409589e-01 -2.13253289e-01 -7.35110119e-02] | [9.584216117858887, 7.301270961761475] |
49350c13-5340-4970-aa92-7fe78cd7dcb3 | quantum-mechanics-and-machine-learning | 2103.14536 | null | https://arxiv.org/abs/2103.14536v1 | https://arxiv.org/pdf/2103.14536v1.pdf | Quantum Mechanics and Machine Learning Synergies: Graph Attention Neural Networks to Predict Chemical Reactivity | There is a lack of scalable quantitative measures of reactivity for functional groups in organic chemistry. Measuring reactivity experimentally is costly and time-consuming and does not scale to the astronomical size of chemical space. In previous quantum chemistry studies, we have introduced Methyl Cation Affinities (MCA*) and Methyl Anion Affinities (MAA*), using a solvation model, as quantitative measures of reactivity for organic functional groups over the broadest range. Although MCA* and MAA* offer good estimates of reactivity parameters, their calculation through Density Functional Theory (DFT) simulations is time-consuming. To circumvent this problem, we first use DFT to calculate MCA* and MAA* for more than 2,400 organic molecules thereby establishing a large dataset of chemical reactivity scores. We then design deep learning methods to predict the reactivity of molecular structures and train them using this curated dataset in combination with different representations of molecular structures. Using ten-fold cross-validation, we show that graph attention neural networks applied to informative input fingerprints produce the most accurate estimates of reactivity, achieving over 91% test accuracy for predicting the MCA* plus-minus 3.0 or MAA* plus-minus 3.0, over 50 orders of magnitude. Finally, we demonstrate the application of these reactivity scores to two tasks: (1) chemical reaction prediction; (2) combinatorial generation of reaction mechanisms. The curated dataset of MCA* and MAA* scores is available through the ChemDB chemoinformatics web portal at www.cdb.ics.uci.edu. | ['Pierre Baldi', 'David Van Vranken', 'Aaron Mood', 'Mohammadamin Tavakoli'] | 2021-03-24 | null | null | null | null | ['chemical-reaction-prediction'] | ['medical'] | [ 3.49357754e-01 -8.72464776e-02 -2.03319073e-01 -2.67631054e-01
-9.66878295e-01 -9.27319407e-01 6.33924484e-01 7.65344083e-01
-4.00268286e-01 1.34564209e+00 1.79823115e-01 -7.83980072e-01
-3.49139236e-02 -1.00929260e+00 -8.37835968e-01 -8.58523548e-01
-2.00288042e-01 3.95016402e-01 2.08636560e-02 -1.80561468e-01
6.34978950e-01 1.09034336e+00 -1.23037231e+00 4.15792882e-01
1.01036072e+00 9.96024489e-01 1.22809177e-02 6.43155038e-01
2.11966168e-02 7.65880048e-01 -5.33962905e-01 -3.41080487e-01
4.12011668e-02 -4.65013206e-01 -9.89242196e-01 -8.42124403e-01
4.78637487e-01 1.21290952e-01 -1.39229387e-01 8.35126162e-01
6.55645967e-01 3.57575506e-01 1.14972627e+00 -5.74381471e-01
-9.66768444e-01 6.33687198e-01 -2.26785839e-01 2.90953398e-01
5.58533847e-01 3.13950241e-01 1.32002068e+00 -1.03301525e+00
7.53034294e-01 1.08688378e+00 3.59869629e-01 4.85235125e-01
-1.52202356e+00 -6.42174602e-01 -1.91745892e-01 7.77370334e-02
-1.27547789e+00 -2.50633955e-01 4.77266163e-01 -6.82193995e-01
1.51551235e+00 3.40580881e-01 4.45646226e-01 1.08995426e+00
4.81390446e-01 -5.66997603e-02 1.00132549e+00 -2.19822168e-01
5.46402991e-01 -3.13395672e-02 1.91399828e-01 6.30980134e-01
4.08309311e-01 1.80247530e-01 -4.34227020e-01 -4.76700872e-01
3.43791157e-01 -1.03953239e-02 -2.35180706e-01 -1.15252458e-01
-1.02930415e+00 1.19059920e+00 9.12407696e-01 1.87394679e-01
-4.38279122e-01 1.73792943e-01 1.83619305e-01 2.20749214e-01
1.81478381e-01 1.40337574e+00 -5.47712982e-01 2.97378451e-01
-3.69835526e-01 2.31756553e-01 8.08307946e-01 5.10819614e-01
8.57836664e-01 7.55932108e-02 1.42648488e-01 4.75297987e-01
2.80413568e-01 1.39059454e-01 1.45283759e-01 -6.40761495e-01
9.97913852e-02 3.86996090e-01 4.19394940e-01 -6.49100840e-01
-7.80570507e-01 -1.92460045e-01 -5.25230706e-01 4.42695349e-01
6.37078106e-01 -1.98247880e-01 -9.53524768e-01 1.69523442e+00
3.18095624e-01 -7.35025167e-01 1.95752218e-01 6.68046296e-01
1.03616071e+00 9.19250309e-01 5.73377907e-01 -2.78618276e-01
1.05272698e+00 -6.22859836e-01 -1.40344337e-01 2.47490540e-01
9.46032465e-01 -7.43392825e-01 7.55863845e-01 5.47898948e-01
-9.90236819e-01 -2.41388798e-01 -1.22180688e+00 -1.46095440e-01
-9.84867036e-01 -3.68098617e-01 1.19490051e+00 7.28696585e-01
-7.69461215e-01 1.22934008e+00 -5.09361327e-01 -9.37707126e-02
2.96645969e-01 6.79179490e-01 -5.44505656e-01 1.91434100e-01
-1.37083280e+00 1.27908707e+00 5.16539931e-01 -1.17301181e-01
-8.64579976e-01 -1.01043010e+00 -6.56054616e-01 6.24628030e-02
9.29369032e-02 -3.59360874e-01 1.01074743e+00 -7.71885991e-01
-1.38926077e+00 4.65848416e-01 3.60151082e-02 -2.86121219e-01
1.01910494e-01 1.25975654e-01 -5.13316274e-01 1.05846651e-01
-4.05614562e-02 7.37834156e-01 -1.93546694e-02 -9.64184165e-01
-2.85828952e-02 -1.12460919e-01 6.37675077e-02 -1.35800004e-01
1.19169794e-01 -6.92380220e-02 6.57087922e-01 -2.04860285e-01
-1.64536878e-01 -6.40492320e-01 -3.74587178e-01 -3.06879938e-01
-5.07523119e-01 -1.71854302e-01 3.09628192e-02 -3.67932737e-01
9.30304706e-01 -1.43538713e+00 1.41082168e-01 6.96262360e-01
4.39079762e-01 2.76005983e-01 -2.20364064e-01 9.64281857e-01
-6.74192667e-01 4.30132985e-01 -1.39173821e-01 6.60933137e-01
-2.74705123e-02 -4.91333306e-01 -1.50266886e-01 4.26515192e-01
4.22672004e-01 8.08804989e-01 -9.70223188e-01 8.58722255e-02
1.73401222e-01 6.11637592e-01 -6.09586835e-01 -8.54812488e-02
-4.86129612e-01 3.95660758e-01 -4.48986083e-01 7.70273209e-01
5.55374324e-01 -4.26558554e-01 5.13561487e-01 -4.03131515e-01
-4.74139065e-01 7.27207899e-01 -5.49693167e-01 1.34750962e+00
-1.81310862e-01 1.66054457e-01 -6.50583267e-01 -8.20314527e-01
9.99295235e-01 2.83918142e-01 7.20414698e-01 -8.60974252e-01
1.38941213e-01 4.26118314e-01 5.49375057e-01 -6.99863359e-02
2.81434864e-01 -4.64248300e-01 2.28098631e-02 3.54120821e-01
8.69516730e-02 9.73461419e-02 4.76240009e-01 1.21190690e-01
1.16039586e+00 -6.47054464e-02 3.52059036e-01 -4.82422948e-01
6.17489219e-01 1.94743291e-01 2.43556172e-01 5.95464885e-01
2.91319992e-02 4.05350089e-01 7.17017353e-01 -7.59772956e-01
-1.33155489e+00 -9.66222048e-01 -4.28639710e-01 1.11845279e+00
-1.58833280e-01 -4.83307272e-01 -5.60332000e-01 -4.41489846e-01
2.23712951e-01 6.44620180e-01 -8.37290704e-01 -3.50162715e-01
-1.89750284e-01 -1.14148700e+00 3.56661767e-01 4.85488445e-01
-8.78499672e-02 -1.10492420e+00 1.50490537e-01 4.11822349e-01
3.92078251e-01 -3.48904073e-01 -3.64588171e-01 6.21551096e-01
-3.56118292e-01 -1.29209387e+00 -4.36212629e-01 -2.50601172e-01
4.10657436e-01 9.86831039e-02 1.01735127e+00 -1.25793725e-01
-6.54137909e-01 -2.21079499e-01 -4.94128726e-02 -4.96867180e-01
-6.61440372e-01 8.31884667e-02 -5.19714952e-02 -4.00163352e-01
2.98448950e-01 -7.59643972e-01 -1.00414252e+00 7.14324266e-02
-6.49361551e-01 -3.21419090e-01 8.10565591e-01 5.73916316e-01
6.36074960e-01 -6.40885949e-01 9.59384263e-01 -1.01458132e+00
5.75303912e-01 -6.01925492e-01 -7.22073436e-01 2.19120309e-01
-8.53816569e-01 5.32869220e-01 8.62702131e-01 -3.58765721e-01
-7.12557256e-01 2.60138184e-01 -2.72815645e-01 1.04245014e-01
3.64596979e-03 8.69485736e-01 -2.91982472e-01 -3.29164594e-01
1.16350508e+00 -2.44928047e-01 -3.76342118e-01 -1.97705150e-01
2.72810310e-01 2.52232134e-01 1.13295309e-01 -8.25127006e-01
3.84127736e-01 -6.31469637e-02 5.58277190e-01 -7.28616476e-01
-7.90837705e-01 -1.49951234e-01 -4.86476928e-01 7.24752992e-02
1.01118648e+00 -6.70342386e-01 -1.39744663e+00 -1.64766029e-01
-1.07698810e+00 -1.60428017e-01 2.34201014e-01 5.93345046e-01
-2.71658301e-01 4.99132842e-01 -5.56180298e-01 -5.56136370e-01
-7.55980611e-01 -1.24054646e+00 6.43322349e-01 -6.29863888e-02
-4.60278422e-01 -9.12195623e-01 1.76149234e-01 2.54947037e-01
2.89730012e-01 8.92390192e-01 1.35348725e+00 -7.62698770e-01
-6.18928194e-01 -2.86417216e-01 -1.76967382e-01 -1.78140089e-01
-4.83289687e-03 2.58107573e-01 -9.71868396e-01 -9.01495293e-02
-7.52619207e-01 -4.38175172e-01 1.00196230e+00 3.54001641e-01
1.00508380e+00 -2.04769790e-01 -1.64398089e-01 2.76408195e-01
1.29566848e+00 7.08186626e-01 7.20085204e-01 2.82150060e-01
7.63447762e-01 3.70562911e-01 2.82500029e-01 3.35805476e-01
-7.96387419e-02 4.72985178e-01 4.96025711e-01 1.27234105e-02
2.92041093e-01 -2.72074252e-01 3.49821955e-01 2.63321728e-01
-5.31984925e-01 -3.48359972e-01 -9.05515432e-01 -1.70009747e-01
-1.38011169e+00 -1.20217013e+00 -4.69240844e-01 2.50621605e+00
1.15561414e+00 1.60617232e-01 3.65405619e-01 -9.65735540e-02
5.39499998e-01 -1.28341705e-01 -7.60165691e-01 -7.97644794e-01
-1.76202767e-02 8.52274597e-01 7.02513456e-01 7.11852014e-01
-9.14799511e-01 8.58450472e-01 6.86369419e+00 4.14179176e-01
-1.42800415e+00 -2.80790508e-01 6.65280819e-01 -8.49303007e-02
-4.61286575e-01 1.71229109e-01 -6.02812171e-01 3.67741793e-01
1.47877276e+00 -2.43214279e-01 4.95664597e-01 7.46731102e-01
2.03113928e-02 3.52921262e-02 -1.45075774e+00 5.48907638e-01
-6.40536070e-01 -1.86394656e+00 2.47996628e-01 2.08506539e-01
7.88930058e-01 1.01456605e-01 -6.20902143e-03 3.79957370e-02
2.81604767e-01 -1.40579927e+00 4.94839638e-01 4.75183606e-01
8.62431467e-01 -1.03137994e+00 1.35983393e-01 -2.32265636e-01
-8.84540617e-01 1.50526330e-01 -5.93899369e-01 1.72675908e-01
-4.33590621e-01 6.38035893e-01 -7.23555803e-01 3.27043772e-01
1.06469944e-01 4.18717265e-01 -4.00399566e-01 7.05783069e-01
-7.66268596e-02 1.66095555e-01 -1.36644319e-01 -4.94351298e-01
3.50921243e-01 -4.10020024e-01 -1.20990865e-01 1.08655477e+00
2.12411910e-01 1.74304232e-01 -7.04499260e-02 1.20997822e+00
-3.86764377e-01 1.56837732e-01 -3.83312911e-01 -6.61663949e-01
6.34680569e-01 1.29913986e+00 -5.39353490e-01 -2.80382961e-01
-2.40504563e-01 4.40977693e-01 4.63089138e-01 4.45395619e-01
-8.08510840e-01 -7.28864193e-01 7.40744710e-01 -1.12175636e-01
-7.07639381e-02 -1.40486747e-01 9.50515717e-02 -8.86912048e-01
-4.57558334e-01 -8.03087950e-01 3.05811018e-01 -7.26613402e-01
-1.23298669e+00 2.43038148e-01 -3.21317166e-01 -5.37342012e-01
-5.68698673e-03 -1.14938951e+00 -6.47218645e-01 1.25881028e+00
-1.25351906e+00 -5.61001480e-01 -6.72600940e-02 1.32530242e-01
-1.22176655e-01 5.14349230e-02 1.16453147e+00 3.20532978e-01
-7.16797173e-01 4.62516516e-01 3.56804252e-01 -2.59019256e-01
8.07365179e-01 -1.53695369e+00 4.74664629e-01 2.38938943e-01
-2.72478640e-01 9.40696120e-01 7.64707327e-01 -5.50732553e-01
-1.75027442e+00 -1.03832614e+00 7.00706065e-01 -6.34077072e-01
7.45818317e-01 -4.49589580e-01 -9.80765104e-01 3.85409057e-01
-1.43970430e-01 -1.44995928e-01 1.13813937e+00 9.27160010e-02
-7.40235031e-01 2.38647684e-01 -1.14007592e+00 5.41497409e-01
8.98049533e-01 -6.10472143e-01 -3.98228317e-03 8.98344398e-01
6.55198097e-01 -1.54149517e-01 -1.58142984e+00 1.07645877e-01
6.22527838e-01 -9.63633060e-01 1.22116864e+00 -1.01702774e+00
5.24826109e-01 -1.13289222e-01 -8.63911957e-02 -1.04969835e+00
-6.14835620e-01 -5.27678370e-01 1.33684576e-01 5.17665267e-01
1.09470832e+00 -9.18979168e-01 6.75323069e-01 9.31806684e-01
-3.26053947e-01 -7.32829571e-01 -6.61837339e-01 -4.42515194e-01
5.06794631e-01 1.21173427e-01 6.10030770e-01 1.16246223e+00
3.65619242e-01 6.40950382e-01 1.51119605e-02 1.38221994e-01
2.88370103e-01 2.29969010e-01 2.54548937e-01 -1.34878933e+00
-2.10237369e-01 -6.78285897e-01 -2.57904738e-01 -2.84473926e-01
6.46927506e-02 -1.29454565e+00 -4.16716158e-01 -1.38749683e+00
7.55490214e-02 -1.30906969e-01 -6.47120893e-01 4.56177533e-01
7.09419772e-02 4.87572774e-02 -3.70750219e-01 6.80688117e-03
-2.16663718e-01 3.15240830e-01 1.06939888e+00 -3.03391397e-01
-1.98555112e-01 -4.27024961e-01 -8.84916127e-01 2.06519559e-01
9.03688669e-01 -4.14281130e-01 -2.59838343e-01 4.21317786e-01
5.89640260e-01 8.68795216e-02 2.83232555e-02 -8.81568253e-01
-2.30707869e-01 -4.47501779e-01 9.01099980e-01 -3.57466847e-01
4.69036341e-01 -3.04324329e-01 3.50359887e-01 6.26075387e-01
-6.39851868e-01 -1.33879542e-01 1.13354266e-01 4.31701660e-01
1.49012864e-01 -1.67197764e-01 9.05273139e-01 -1.62479579e-01
-3.00832033e-01 5.51753819e-01 -6.25442147e-01 -4.68745202e-01
8.12403977e-01 -1.39170900e-01 -7.38304913e-01 -4.12999317e-02
-7.77235031e-01 -5.45899905e-02 3.64501983e-01 -5.24825454e-02
4.70686883e-01 -1.15759826e+00 -1.87683329e-01 -1.43864647e-01
2.66070008e-01 -4.93933469e-01 8.89580324e-02 6.69740736e-01
-8.28227222e-01 6.50931358e-01 -2.51301229e-01 3.47213410e-02
-9.69414532e-01 9.22451317e-01 6.99166656e-01 2.85868421e-02
5.28695807e-02 5.20052433e-01 3.41185480e-01 -3.51358235e-01
-1.37882560e-01 -3.40841055e-01 -1.09452493e-01 6.45549670e-02
4.43200648e-01 3.40385377e-01 2.00837210e-01 -3.23345870e-01
-3.77191663e-01 3.56220603e-01 -3.03844631e-01 2.96646297e-01
1.38585854e+00 7.04063892e-01 -1.55276865e-01 -2.13032477e-02
1.38650060e+00 -6.06931411e-02 -8.58824492e-01 2.18240231e-01
2.31606096e-01 -4.46193591e-02 -3.85101400e-02 -1.10872161e+00
-5.22808433e-01 8.72517586e-01 3.48763585e-01 1.32776454e-01
4.68649477e-01 -1.66115299e-01 3.83197159e-01 1.09838724e+00
-5.88843087e-03 -8.37773263e-01 1.09947644e-01 4.19002771e-01
8.72140944e-01 -1.11266947e+00 2.95825392e-01 -2.74286538e-01
-3.58108312e-01 1.34877181e+00 3.71488065e-01 1.55920416e-01
1.63324162e-01 -3.82551581e-01 -2.07763568e-01 -5.11788428e-01
-7.98407018e-01 -3.17775235e-02 3.46519649e-01 3.95635873e-01
1.14943552e+00 1.07329756e-01 -4.50617254e-01 1.37993634e-01
-1.84872746e-01 -4.60581481e-01 5.60771585e-01 9.76132512e-01
-8.45733762e-01 -1.40358126e+00 -1.69819146e-01 3.43080789e-01
-5.92808187e-01 -3.52620065e-01 -1.09644747e+00 6.24010086e-01
-3.19986701e-01 9.39726889e-01 -3.90921533e-01 -2.48788416e-01
2.22705901e-01 2.69645721e-01 7.34574199e-01 -4.93810952e-01
-5.45719862e-01 -2.48294249e-01 4.14028198e-01 -3.61099184e-01
-3.46564054e-01 -2.12156415e-01 -1.61780512e+00 -6.38517976e-01
-4.12329853e-01 5.94571054e-01 7.90252090e-01 5.26923299e-01
5.20692706e-01 3.47317606e-01 7.40166247e-01 -9.82219517e-01
-4.28860784e-01 -8.07901919e-01 -5.28891802e-01 2.68653661e-01
1.01762593e-01 -5.93431652e-01 -2.78406024e-01 -3.70322138e-01] | [5.0212225914001465, 5.62363862991333] |
df8ca513-1b49-4aa0-b9d6-226a884a3254 | learning-end-to-end-goal-oriented-dialog-with | 1808.09996 | null | http://arxiv.org/abs/1808.09996v1 | http://arxiv.org/pdf/1808.09996v1.pdf | Learning End-to-End Goal-Oriented Dialog with Multiple Answers | In a dialog, there can be multiple valid next utterances at any point. The
present end-to-end neural methods for dialog do not take this into account.
They learn with the assumption that at any time there is only one correct next
utterance. In this work, we focus on this problem in the goal-oriented dialog
setting where there are different paths to reach a goal. We propose a new
method, that uses a combination of supervised learning and reinforcement
learning approaches to address this issue. We also propose a new and more
effective testbed, permuted-bAbI dialog tasks, by introducing multiple valid
next utterances to the original-bAbI dialog tasks, which allows evaluation of
goal-oriented dialog systems in a more realistic setting. We show that there is
a significant drop in performance of existing end-to-end neural methods from
81.5% per-dialog accuracy on original-bAbI dialog tasks to 30.3% on
permuted-bAbI dialog tasks. We also show that our proposed method improves the
performance and achieves 47.3% per-dialog accuracy on permuted-bAbI dialog
tasks. | ['Satinder Singh', 'Lazaros Polymenakos', 'Jatin Ganhotra', 'Janarthanan Rajendran'] | 2018-08-24 | learning-end-to-end-goal-oriented-dialog-with-1 | https://aclanthology.org/D18-1418 | https://aclanthology.org/D18-1418.pdf | emnlp-2018-10 | ['goal-oriented-dialog'] | ['natural-language-processing'] | [-2.24738732e-01 4.78985339e-01 3.08430761e-01 -9.92627323e-01
-1.04723454e+00 -8.71161878e-01 5.92231154e-01 -2.04348341e-01
-6.35966420e-01 1.12029433e+00 3.87831300e-01 -5.22897720e-01
1.56294972e-01 -4.87668157e-01 -4.01030958e-01 -3.70444834e-01
1.07795991e-01 1.17933321e+00 3.83286387e-01 -9.43492711e-01
-7.13193268e-02 8.64746794e-02 -7.14141846e-01 3.79776955e-01
4.86677796e-01 5.63384473e-01 1.88969493e-01 1.03724277e+00
-3.24142516e-01 9.17827010e-01 -1.03950465e+00 -4.67800796e-01
6.56618178e-02 -5.59016049e-01 -1.39155722e+00 -8.05128142e-02
3.47396523e-01 -8.45499754e-01 -3.07295531e-01 7.91645110e-01
3.79657030e-01 4.99733567e-01 4.48150843e-01 -1.38878584e+00
-3.38992029e-01 1.01488614e+00 7.34407902e-02 -7.07007647e-02
4.82654601e-01 2.57724732e-01 1.24611807e+00 -6.54876828e-01
3.98025066e-01 1.85245633e+00 3.77484351e-01 1.12156105e+00
-1.13682353e+00 -4.87102151e-01 3.67171079e-01 -1.63123697e-01
-7.05080450e-01 -5.76501071e-01 5.78814626e-01 -3.23759973e-01
1.05768979e+00 7.37884119e-02 5.40528446e-03 1.31588781e+00
1.47450209e-01 7.28258133e-01 1.21401846e+00 -3.72202516e-01
3.16817671e-01 1.29252017e-01 7.47401297e-01 4.15934682e-01
-6.24582350e-01 1.52675703e-01 -4.85331506e-01 -1.88104033e-01
3.24923903e-01 -2.01581940e-01 -1.38492823e-01 9.62675139e-02
-1.05778885e+00 1.34799874e+00 2.50611275e-01 3.01884264e-01
-1.19057767e-01 -1.68591350e-01 3.58202785e-01 7.02195585e-01
1.91984728e-01 5.62090993e-01 -5.55045426e-01 -2.66301990e-01
-3.56142968e-01 7.12657332e-01 1.54282713e+00 1.04713726e+00
4.20669496e-01 1.43829908e-03 -3.22734654e-01 1.05614686e+00
3.47787529e-01 3.77612889e-01 4.32929516e-01 -1.35501266e+00
6.19964957e-01 2.46241793e-01 4.84049082e-01 -3.02816302e-01
-7.17981756e-01 4.05873030e-01 -4.45348680e-01 4.03743595e-01
1.12311995e+00 -6.59759700e-01 -6.20178223e-01 2.25467682e+00
2.61112362e-01 -5.56108057e-01 6.07707739e-01 9.97134268e-01
8.44480515e-01 7.52422035e-01 1.12166606e-01 -1.87247723e-01
1.33928597e+00 -1.22820187e+00 -9.94641781e-01 -2.85585046e-01
5.63229799e-01 -6.84332669e-01 1.28663921e+00 3.56821179e-01
-1.31180108e+00 -4.57111955e-01 -9.10056889e-01 -2.14911476e-02
-4.66744989e-01 -3.10540259e-01 3.58515382e-01 5.92491746e-01
-1.35048735e+00 2.51220942e-01 -4.11012262e-01 -4.15497094e-01
-4.83020484e-01 5.69802225e-01 -1.24765687e-01 2.62722850e-01
-1.52987385e+00 1.14590263e+00 2.98527211e-01 -5.34217097e-02
-1.04434955e+00 -2.22588062e-01 -7.72017956e-01 1.44389600e-01
6.78648293e-01 -2.82703370e-01 2.20539737e+00 -6.09247923e-01
-2.12858319e+00 4.92012113e-01 -2.03606054e-01 -7.05385506e-01
4.76701111e-01 -1.90122619e-01 3.49017195e-02 -1.43780947e-01
-2.51091450e-01 1.09388912e+00 2.91391551e-01 -1.22268546e+00
-7.68135786e-01 -2.93759227e-01 8.38377714e-01 3.00929219e-01
-1.08783424e-01 2.24860627e-02 -3.99067476e-02 -1.50957957e-01
-3.06375355e-01 -1.22958052e+00 -1.62702575e-01 -3.38601321e-01
-3.87259603e-01 -6.99109614e-01 8.66469383e-01 -5.24790227e-01
7.50549912e-01 -1.75628960e+00 1.90822423e-01 -4.69090819e-01
3.06629896e-01 3.46715957e-01 -2.03029901e-01 5.83235204e-01
3.15124452e-01 -2.20859498e-02 -2.07576215e-01 -7.74729788e-01
2.34006360e-01 5.01511574e-01 -4.59886879e-01 -1.93976030e-01
-2.06646156e-02 7.34632969e-01 -7.81333745e-01 -3.11072052e-01
3.32276940e-01 3.01682167e-02 -5.65018535e-01 7.89353907e-01
-6.72990322e-01 7.37995923e-01 -2.61705577e-01 1.08742550e-01
4.93219465e-01 6.94738850e-02 4.30457920e-01 3.12096596e-01
-8.76980871e-02 6.56806350e-01 -6.79357231e-01 1.71914470e+00
-7.44780838e-01 6.82334900e-01 5.00380874e-01 -8.09923589e-01
1.02809286e+00 5.56645811e-01 1.31579533e-01 -4.68962550e-01
2.67076343e-01 -3.84913087e-02 5.26795983e-01 -7.87057877e-02
6.61476076e-01 -4.28546935e-01 -4.91490662e-01 4.82092112e-01
2.75186956e-01 -1.90453306e-01 2.26870209e-01 3.24419826e-01
9.00372922e-01 -2.82827288e-01 1.07274100e-01 -2.98540056e-01
6.25932872e-01 9.52506587e-02 2.70166934e-01 1.13784361e+00
-4.77269977e-01 3.91134620e-01 6.41111195e-01 -2.68490374e-01
-9.22155619e-01 -1.14531410e+00 2.17781201e-01 1.53823507e+00
4.95302901e-02 -1.05849795e-01 -8.77013445e-01 -9.66595650e-01
-2.12020442e-01 1.12321663e+00 -2.33259559e-01 1.02436254e-02
-6.99870706e-01 -4.61748987e-01 4.53806520e-01 4.18959588e-01
7.10927367e-01 -1.14087510e+00 -2.72270858e-01 4.17673588e-01
-5.14866531e-01 -1.24044335e+00 -4.10543114e-01 2.42923856e-01
-5.35807967e-01 -7.70314336e-01 -6.95070922e-01 -7.43392587e-01
2.48519853e-02 1.15286998e-01 1.14356220e+00 -1.94618002e-01
4.47360694e-01 3.88287812e-01 -4.81217772e-01 -3.35455507e-01
-1.19934976e+00 1.90740198e-01 -2.15673773e-03 -3.96022022e-01
4.53470856e-01 -2.63307035e-01 -2.21321449e-01 7.12501347e-01
-4.60226983e-01 -8.50062259e-03 1.77187957e-02 1.30217290e+00
-2.33289167e-01 -3.58099192e-01 1.07627678e+00 -7.92774081e-01
1.12750018e+00 -3.87829274e-01 -7.11029470e-01 2.72882760e-01
-2.24941075e-01 1.47825167e-01 9.40880001e-01 -4.33642775e-01
-1.39798915e+00 -5.46484739e-02 -6.03179038e-01 -3.41879018e-02
-4.44796056e-01 2.36946195e-01 -2.34749749e-01 2.56920278e-01
4.84541833e-01 -2.60326341e-02 2.30272055e-01 -5.08161366e-01
5.11290789e-01 1.09277046e+00 3.25893253e-01 -6.96843922e-01
2.95924604e-01 -4.49395925e-02 -4.26589817e-01 -9.10116374e-01
-7.47898161e-01 -3.28964084e-01 -4.93337035e-01 -2.15239346e-01
9.61865246e-01 -7.72428691e-01 -1.36865258e+00 6.22438431e-01
-1.53896499e+00 -1.06201041e+00 8.88149291e-02 3.06483239e-01
-7.03213692e-01 4.17399049e-01 -8.89341176e-01 -1.16712642e+00
-4.31621850e-01 -1.43531013e+00 9.31239903e-01 2.47287884e-01
-4.93837893e-01 -1.00682425e+00 1.46053704e-02 4.39761251e-01
4.23008800e-01 -4.47863430e-01 8.90524685e-01 -1.55636537e+00
-2.42666289e-01 1.96330443e-01 1.83772519e-02 4.04118299e-01
1.76914379e-01 -5.89076757e-01 -9.08368528e-01 -4.47663516e-01
1.47837847e-01 -8.92093599e-01 3.18734646e-01 1.25531331e-01
3.23361546e-01 -4.95946914e-01 7.66055286e-02 -2.94811279e-01
7.75609374e-01 9.86624539e-01 2.19389811e-01 -5.09755649e-02
2.16791034e-01 9.91281748e-01 8.44450951e-01 3.49995166e-01
6.30128324e-01 1.11559141e+00 3.05771023e-01 1.67855248e-01
-4.48439009e-02 -4.00016569e-02 3.66790593e-01 6.59426033e-01
6.24986470e-01 -7.92886555e-01 -9.40944195e-01 5.84922314e-01
-2.07339501e+00 -6.08797371e-01 5.99682890e-02 2.04370952e+00
1.09465301e+00 2.77814120e-01 5.39204776e-01 -2.66749144e-01
6.00605249e-01 3.34877223e-02 -4.77309495e-01 -6.33579195e-01
2.38625363e-01 -3.07165198e-02 4.59054708e-02 1.36542177e+00
-9.64982986e-01 1.35506618e+00 6.34005785e+00 3.10827315e-01
-8.89896870e-01 2.57847220e-01 6.51675940e-01 1.39242500e-01
1.70641333e-01 -3.11822128e-02 -1.22560227e+00 4.15204436e-01
1.21698713e+00 2.05586441e-02 5.20319819e-01 9.30506349e-01
1.75229102e-01 -1.70578480e-01 -1.33493364e+00 5.18626511e-01
-1.86054528e-01 -9.53018188e-01 -2.74079293e-01 -1.12715572e-01
1.84500933e-01 -2.96880573e-01 -2.17121243e-01 8.53465855e-01
1.03707457e+00 -7.96395242e-01 3.58705908e-01 6.22646045e-03
2.30005816e-01 -6.07141733e-01 7.73400068e-01 6.38719022e-01
-6.75221145e-01 3.57947275e-02 -3.14147741e-01 -2.29502201e-01
4.92380679e-01 -1.42811671e-01 -1.50767994e+00 3.07202309e-01
3.66376281e-01 -1.78507052e-03 -1.76011194e-02 3.60810280e-01
-2.18130574e-01 5.45044065e-01 -3.74184996e-01 -6.12734854e-01
5.18518090e-01 1.13494107e-02 4.97553855e-01 1.17423379e+00
-4.97931764e-02 1.93621814e-01 5.23603618e-01 7.69528568e-01
-2.74741761e-02 -2.20659837e-01 -6.97661161e-01 3.33597243e-01
4.82488036e-01 9.49306250e-01 -2.48667657e-01 -3.13440561e-01
-3.96634340e-01 9.68772531e-01 4.94696915e-01 2.93235213e-01
-8.30041945e-01 -3.47234339e-01 7.32662797e-01 -4.58696008e-01
2.05350406e-02 -4.85715479e-01 6.55185580e-02 -8.83011818e-01
-2.99002051e-01 -1.04428303e+00 3.94121289e-01 -6.41249299e-01
-1.48521614e+00 8.27803612e-01 4.95261401e-01 -5.01836538e-01
-1.01277518e+00 -8.85066211e-01 -7.05045164e-01 9.52339649e-01
-1.20732093e+00 -9.70967174e-01 -1.47111028e-01 6.50323391e-01
1.32501054e+00 -3.42077345e-01 1.03488886e+00 -8.37480798e-02
-2.85416931e-01 6.33772492e-01 7.37997219e-02 3.67051780e-01
9.40602601e-01 -1.54564941e+00 5.30548513e-01 3.84070784e-01
-1.47859588e-01 5.76794267e-01 9.14759159e-01 -4.16193873e-01
-1.04894936e+00 -6.76384985e-01 1.00886297e+00 -4.12207276e-01
6.79056108e-01 -7.15679228e-01 -9.07256126e-01 1.01991677e+00
7.74944782e-01 -8.95080626e-01 4.92998749e-01 4.44814831e-01
-6.38748035e-02 1.66724399e-01 -1.22527218e+00 9.85902250e-01
7.60523915e-01 -2.54155308e-01 -9.92704809e-01 5.96868575e-01
1.16486371e+00 -4.90462661e-01 -6.90391481e-01 2.27039739e-01
2.74994045e-01 -9.60440159e-01 9.23595846e-01 -8.84835720e-01
2.73297042e-01 2.29843631e-01 -3.15412998e-01 -1.47655451e+00
2.93642223e-01 -9.17550266e-01 8.93485770e-02 1.33498573e+00
6.16705179e-01 -7.36976802e-01 8.41488063e-01 1.11886287e+00
-1.44929394e-01 -5.01092434e-01 -1.15775394e+00 -8.23800743e-01
7.37997472e-01 -3.02909222e-02 3.39299977e-01 4.90408599e-01
5.03498256e-01 9.44455147e-01 -8.05164456e-01 -1.54825062e-01
2.71937132e-01 -1.99298561e-02 1.13541675e+00 -9.08223391e-01
-3.92011344e-01 -1.62862018e-01 2.58817971e-01 -1.78913915e+00
5.08713424e-01 -5.19281983e-01 5.42190909e-01 -1.25295401e+00
-2.30040401e-01 -3.48371178e-01 2.51937807e-01 4.42965537e-01
-1.71678334e-01 -4.45909590e-01 5.54860592e-01 2.27343403e-02
-5.36297739e-01 6.81369901e-01 9.77261186e-01 -2.04910323e-01
-4.25595701e-01 2.21362844e-01 -4.59403306e-01 6.63815141e-01
1.08000290e+00 -3.16129714e-01 -5.83381474e-01 -2.73160785e-01
-6.02527022e-01 7.31137812e-01 1.83545604e-01 -8.20066333e-01
3.92509460e-01 -8.49185735e-02 -3.59629840e-01 -5.75575411e-01
1.04786265e+00 -6.84310436e-01 -5.22764146e-01 3.24688941e-01
-8.59113038e-01 1.45080730e-01 2.51043975e-01 5.48732758e-01
-1.96306571e-01 -5.84950805e-01 9.41272676e-01 -3.71996135e-01
-5.71362734e-01 -2.73385346e-01 -7.67710030e-01 1.65853754e-01
9.18315649e-01 3.32066774e-01 -5.04608095e-01 -1.18352163e+00
-1.07674778e+00 6.73737884e-01 -7.20776320e-02 4.58010644e-01
4.38232213e-01 -8.60659838e-01 -6.69899464e-01 -2.59395093e-01
-2.58463085e-01 -2.04614341e-01 1.79578170e-01 3.80116314e-01
-1.71530098e-01 7.20264614e-01 -2.12743431e-01 -5.38015544e-01
-1.53156543e+00 4.09174621e-01 5.80058575e-01 -7.31653810e-01
-3.00535023e-01 7.07590699e-01 5.05263090e-01 -9.74401832e-01
6.32116377e-01 -1.39720678e-01 -2.02594936e-01 -1.66265979e-01
3.67224067e-01 4.82388698e-02 -2.24071369e-01 -4.31962222e-01
-8.80907252e-02 -2.28886697e-02 -5.34790933e-01 -7.48902559e-01
7.72493005e-01 -2.71268308e-01 3.07400525e-01 6.14230394e-01
9.99609947e-01 -2.26452306e-01 -1.16871607e+00 -3.17159951e-01
3.57393846e-02 -2.52380818e-01 -5.42354465e-01 -1.25771880e+00
-4.84799534e-01 1.07654417e+00 4.76149321e-01 5.82672119e-01
7.03768849e-01 -1.14300698e-02 9.44708586e-01 1.06299829e+00
5.40238857e-01 -8.67174387e-01 3.42524707e-01 1.19498324e+00
1.14063001e+00 -1.63133991e+00 -6.75703585e-01 -2.98334956e-01
-9.71057475e-01 1.05478454e+00 1.10173237e+00 1.43641502e-01
4.51990187e-01 1.60739124e-01 4.16453451e-01 1.01628089e-02
-8.99117410e-01 -1.09524734e-01 -3.58948410e-01 5.86747825e-01
4.54665452e-01 -1.01110883e-01 -2.56232113e-01 5.92631221e-01
-3.87087971e-01 -4.43484664e-01 6.08004391e-01 8.83120716e-01
-5.95388770e-01 -1.47822964e+00 -3.64066333e-01 -8.98170993e-02
-2.36758962e-01 -2.11739875e-02 -8.72801125e-01 1.07880926e+00
-7.12562978e-01 1.72259796e+00 -7.34912381e-02 -4.56633568e-01
4.09639299e-01 6.83356643e-01 1.33264929e-01 -6.45534694e-01
-9.49945211e-01 1.19511224e-02 5.45602620e-01 -1.52427897e-01
-2.23594263e-01 -2.88559407e-01 -1.25722837e+00 -4.65733767e-01
-3.08468133e-01 5.44072986e-01 5.36767244e-01 8.76160741e-01
1.02330536e-01 5.12252867e-01 6.61258876e-01 -7.70080090e-01
-1.04240298e+00 -1.39019740e+00 -2.51804054e-01 4.03882325e-01
3.64540875e-01 -6.22324467e-01 -2.72023320e-01 -2.39538953e-01] | [12.847248077392578, 8.012442588806152] |
9296f9fb-8064-4fbc-9352-ee29ce14955b | predicting-beauty-liking-and-aesthetic | 2307.00984 | null | https://arxiv.org/abs/2307.00984v1 | https://arxiv.org/pdf/2307.00984v1.pdf | Predicting beauty, liking, and aesthetic quality: A comparative analysis of image databases for visual aesthetics research | In the fields of Experimental and Computational Aesthetics, numerous image datasets have been created over the last two decades. In the present work, we provide a comparative overview of twelve image datasets that include aesthetic ratings (beauty, liking or aesthetic quality) and investigate the reproducibility of results across different datasets. Specifically, we examine how consistently the ratings can be predicted by using either (A) a set of 20 previously studied statistical image properties, or (B) the layers of a convolutional neural network developed for object recognition. Our findings reveal substantial variation in the predictability of aesthetic ratings across the different datasets. However, consistent similarities were found for datasets containing either photographs or paintings, suggesting different relevant features in the aesthetic evaluation of these two image genres. To our surprise, statistical image properties and the convolutional neural network predict aesthetic ratings with similar accuracy, highlighting a significant overlap in the image information captured by the two methods. Nevertheless, the discrepancies between the datasets call into question the generalizability of previous research findings on single datasets. Our study underscores the importance of considering multiple datasets to improve the validity and generalizability of research results in the fields of experimental and computational aesthetics. | ['Christoph Redies', 'Katja Thoemmes', 'Ralf Bartho'] | 2023-07-03 | null | null | null | null | ['object-recognition'] | ['computer-vision'] | [ 1.89683035e-01 -1.16198629e-01 1.67813063e-01 -5.50338209e-01
-2.81357437e-01 -5.93121648e-01 5.63237786e-01 3.41872305e-01
-3.77253652e-01 1.51585981e-01 3.67930114e-01 2.11445871e-03
-4.46291119e-01 -6.51592791e-01 -3.37149531e-01 -3.82084161e-01
1.25843957e-01 -1.28505245e-01 -3.57141137e-01 -1.82318479e-01
6.08556688e-01 4.94767517e-01 -1.93038750e+00 4.92217958e-01
6.21395111e-01 1.69715166e+00 -1.33571714e-01 3.94957989e-01
-6.42734393e-02 3.26368183e-01 -5.64946234e-01 -8.84995222e-01
4.90069568e-01 -2.09699497e-01 -7.19364703e-01 2.81834006e-01
7.53333151e-01 -2.04829395e-01 -1.36975385e-02 7.84332812e-01
5.53484797e-01 -1.14347674e-01 6.54397786e-01 -1.40043926e+00
-1.29645407e+00 9.52978805e-02 -4.41300839e-01 -1.21127278e-01
2.99142033e-01 6.09351575e-01 1.34612286e+00 -5.31953275e-01
6.59020841e-01 1.19072366e+00 7.68494964e-01 1.17776804e-01
-1.19304812e+00 -6.03417099e-01 -1.22248210e-01 1.71133161e-01
-1.30919838e+00 -4.64584470e-01 6.83963776e-01 -6.77263498e-01
6.38910651e-01 3.15034091e-01 9.54937637e-01 9.66414452e-01
8.80948380e-02 4.07047778e-01 1.67623341e+00 -2.40150958e-01
1.79473862e-01 5.14378965e-01 -3.48930508e-01 3.40169817e-01
3.03589076e-01 1.09458230e-01 -4.92310047e-01 4.47756872e-02
6.91197693e-01 -3.38576823e-01 8.18289742e-02 -1.53953359e-01
-9.49563026e-01 8.53064954e-01 6.92967057e-01 4.09240723e-01
-3.28684777e-01 3.61465327e-02 6.51961267e-01 2.26655632e-01
4.10148501e-01 1.05395710e+00 -3.26429784e-01 -2.19827265e-01
-7.07435668e-01 -2.44503185e-01 5.57947576e-01 6.05474174e-01
5.67422628e-01 1.64952159e-01 -1.64394826e-02 9.51776862e-01
9.82764661e-02 2.78947949e-01 4.17802155e-01 -1.01167178e+00
-1.75712407e-02 6.38956428e-01 -8.57319087e-02 -1.75584257e+00
-4.25983101e-01 -2.47607097e-01 -4.60874021e-01 5.59120595e-01
4.25130337e-01 3.23368520e-01 -4.96403784e-01 1.65137076e+00
-1.24702990e-01 -6.77111566e-01 -2.07794353e-01 1.10881686e+00
1.11157405e+00 2.86837250e-01 5.90456367e-01 2.28335142e-01
1.42671645e+00 -5.44621110e-01 -2.16428772e-01 -3.16602230e-01
2.85977215e-01 -1.20032525e+00 1.54739273e+00 5.82045197e-01
-1.09062099e+00 -8.56513977e-01 -9.49402630e-01 -1.77993253e-01
-7.02046394e-01 4.49729919e-01 8.17174494e-01 7.23864675e-01
-1.22584891e+00 7.53939390e-01 5.28585108e-04 -6.39457464e-01
5.51920593e-01 1.70306891e-01 -5.76964140e-01 2.73020208e-01
-6.02058411e-01 1.10093224e+00 7.61367753e-02 7.44414553e-02
-2.21353084e-01 -6.09915674e-01 -6.64692521e-01 4.56730835e-02
-3.22765291e-01 -3.67802948e-01 7.27054775e-01 -1.86360765e+00
-1.16519868e+00 1.30487525e+00 3.69917899e-01 7.71596208e-02
2.44136289e-01 2.63547502e-03 -5.09252608e-01 4.90499288e-02
7.13293776e-02 9.30931270e-01 4.81457472e-01 -1.47675526e+00
-2.76034623e-01 -2.33778447e-01 1.03295036e-01 6.72465470e-03
-4.90611047e-01 1.90390885e-01 -8.78656060e-02 -7.21869886e-01
-1.55185193e-01 -8.63702238e-01 5.71508482e-02 4.25752550e-01
-1.68109998e-01 -3.37595046e-02 2.30042472e-01 -4.96552944e-01
1.05848682e+00 -2.44378614e+00 -1.43760860e-01 2.41529852e-01
6.86220005e-02 -1.03476621e-01 -5.95102966e-01 5.46916127e-01
-3.30615968e-01 5.15244067e-01 8.09992924e-02 -3.12322706e-01
2.63002336e-01 8.75548832e-03 -3.01115308e-02 4.57105279e-01
3.49894106e-01 1.16394556e+00 -6.17707551e-01 -5.14859378e-01
3.08698595e-01 4.49229151e-01 -2.33963013e-01 2.63618175e-02
1.68238685e-01 2.27740079e-01 -1.05907217e-01 7.54077315e-01
6.48249924e-01 -2.69867003e-01 4.73246500e-02 -6.35389745e-01
-1.98775038e-01 2.11925834e-01 -7.03312814e-01 1.46175492e+00
-6.27088308e-01 1.18409550e+00 -3.15770477e-01 -5.74102759e-01
1.06651640e+00 5.48486896e-02 4.39023495e-01 -1.14243519e+00
4.28501546e-01 2.16440886e-01 1.72134101e-01 -6.17338717e-01
7.02951372e-01 -3.71596277e-01 -7.95160607e-02 5.66908002e-01
8.47475603e-02 -2.93198079e-01 -1.94946229e-02 -2.89062917e-01
5.27502477e-01 1.83096377e-03 2.35725760e-01 -1.39288828e-01
9.25274640e-02 3.17494036e-03 1.21328816e-01 3.93423766e-01
-3.12732846e-01 1.04561114e+00 4.62848514e-01 -5.40442765e-01
-1.32525170e+00 -1.07747245e+00 -4.19084132e-01 9.93275583e-01
5.96558936e-02 -3.73112559e-01 -4.43259388e-01 -2.82263041e-01
7.90581927e-02 5.69275141e-01 -1.12983465e+00 -3.23081285e-01
7.57192448e-02 -6.43033266e-01 3.97958398e-01 3.90320718e-01
2.48820126e-01 -1.41959059e+00 -8.70710492e-01 -2.00838149e-01
1.27975540e-02 -1.07009983e+00 -6.96231276e-02 2.40389109e-02
-7.13878810e-01 -1.05872035e+00 -3.17651600e-01 -4.72163886e-01
6.21932983e-01 3.33550543e-01 1.43676591e+00 3.15571278e-01
-4.72675800e-01 7.57993519e-01 -5.13556421e-01 -3.71834636e-01
-2.99012840e-01 -1.30363971e-01 -2.16947809e-01 7.58122504e-02
3.80995333e-01 -4.85213250e-01 -8.58640730e-01 1.72076881e-01
-1.13777709e+00 -1.13469116e-01 9.33606684e-01 4.84436750e-01
3.25768918e-01 -1.53815625e-02 3.11101496e-01 -2.92163342e-01
7.94617891e-01 -4.99129325e-01 -6.55624941e-02 1.31069332e-01
-8.86549115e-01 -2.00935587e-01 3.89827430e-01 -4.52871621e-01
-7.36118495e-01 -2.99155116e-01 1.82623640e-01 -1.73364639e-01
-5.24796724e-01 4.88249838e-01 1.46045268e-01 -4.56058115e-01
5.72752118e-01 -1.69203371e-01 1.42626449e-01 -3.02221149e-01
2.46464416e-01 6.39884889e-01 3.86953115e-01 -5.50842822e-01
6.57863677e-01 5.14740109e-01 4.05411283e-03 -6.48425281e-01
-6.85539901e-01 -1.17470145e-01 -6.69425011e-01 -6.49502754e-01
8.02869141e-01 -6.53464198e-01 -8.61069381e-01 3.15850437e-01
-9.04842675e-01 -1.51456147e-01 -4.88755316e-01 3.40182483e-01
-4.93039548e-01 4.10400510e-01 -3.42595071e-01 -8.25731337e-01
-3.57887000e-01 -1.30952919e+00 1.16853547e+00 1.83982015e-01
-7.64203846e-01 -9.47756708e-01 -2.29724329e-02 4.70090389e-01
5.11655271e-01 6.08815551e-01 8.76362562e-01 -2.64166743e-01
-2.05938399e-01 -3.97636145e-01 -6.40440226e-01 3.87356192e-01
2.94655204e-01 4.08582062e-01 -1.19767869e+00 9.43233445e-02
-3.16583186e-01 -6.67923629e-01 5.70625305e-01 3.41662079e-01
1.16146541e+00 -3.52213830e-01 3.49449873e-01 4.17367458e-01
1.70784330e+00 2.41805408e-02 9.03046548e-01 8.18024337e-01
2.50338256e-01 1.05868733e+00 4.36518401e-01 5.00515163e-01
2.77908564e-01 6.82824492e-01 4.42278892e-01 -3.27319175e-01
-1.21195942e-01 -7.01961666e-02 2.01872721e-01 4.86662328e-01
-1.09565929e-01 -1.03075523e-03 -6.07026935e-01 5.09671926e-01
-1.36191678e+00 -7.77776897e-01 -6.09043278e-02 2.12856770e+00
6.06796801e-01 -7.84397125e-02 3.44802916e-01 8.02477002e-02
4.60134357e-01 2.07664937e-01 -3.32920194e-01 -1.03636491e+00
-3.23464334e-01 1.54915020e-01 3.17873746e-01 -4.07722853e-02
-8.47220063e-01 6.04553878e-01 7.78832150e+00 6.36799514e-01
-1.27215576e+00 -3.23111445e-01 1.11912036e+00 -3.62156570e-01
-3.31178695e-01 -7.68300071e-02 2.04896048e-01 3.83708805e-01
8.56849134e-01 1.00015946e-01 4.64016795e-01 8.15933883e-01
3.30221951e-01 -3.13507855e-01 -8.76760840e-01 9.23821390e-01
2.15564042e-01 -8.94075453e-01 9.23596695e-02 4.33114506e-02
5.23679793e-01 -3.39528710e-01 7.79777944e-01 -2.06638798e-01
-1.43863708e-02 -1.46057582e+00 1.06511450e+00 6.78938568e-01
9.23524916e-01 -4.73961264e-01 6.39862895e-01 -5.43506145e-01
-8.05060029e-01 -9.98029038e-02 -3.57509375e-01 -4.51145232e-01
-8.79689977e-02 3.69855672e-01 -4.85726386e-01 9.01181847e-02
1.23098671e+00 5.75998127e-01 -1.02057254e+00 1.00019050e+00
1.60600409e-01 2.95257360e-01 -1.04200433e-03 -7.06687644e-02
2.97040343e-01 -3.86225879e-01 4.24282812e-02 1.25306356e+00
2.47632593e-01 -2.46852696e-01 -4.67303455e-01 1.00119841e+00
2.31181666e-01 5.65739453e-01 -5.31085908e-01 -4.25248712e-01
-1.56181650e-02 1.64745367e+00 -1.00752974e+00 8.11891407e-02
-7.03982234e-01 8.79344225e-01 2.75888354e-01 8.62099305e-02
-6.45033240e-01 -4.00427170e-02 6.78163290e-01 1.60404101e-01
6.06866442e-02 -1.43692344e-01 -1.07706904e+00 -6.94026113e-01
1.08126059e-01 -7.10501194e-01 1.46622613e-01 -1.41105497e+00
-1.62401521e+00 5.46050012e-01 -2.44726136e-01 -1.29301667e+00
4.53688443e-01 -8.46803188e-01 -6.36880934e-01 6.57617390e-01
-1.25057161e+00 -1.16209829e+00 -7.13198900e-01 1.65258661e-01
1.19011670e-01 3.29266153e-02 8.28753054e-01 -7.41792917e-02
-2.77215958e-01 5.02224565e-01 4.46314141e-02 -1.79008453e-03
9.45775926e-01 -1.16471112e+00 2.14353174e-01 1.54546052e-01
1.30436376e-01 5.24498224e-01 6.62050605e-01 -2.75425166e-01
-9.01598513e-01 -4.50809747e-01 6.57808900e-01 -2.97519177e-01
6.74179077e-01 -1.39465094e-01 -5.37810624e-01 1.85527354e-01
5.04079342e-01 -4.80057776e-01 1.23187768e+00 1.82319432e-01
-6.51259780e-01 -1.09632321e-01 -1.22973669e+00 6.35939240e-01
8.47157776e-01 -4.23099637e-01 -1.64385691e-01 4.52471748e-02
2.30692044e-01 1.61395282e-01 -1.30190337e+00 1.98011070e-01
1.14170790e+00 -1.47445130e+00 1.03739953e+00 -6.35591149e-01
1.13283885e+00 1.21552810e-01 -2.63682693e-01 -1.20142198e+00
-6.35093451e-01 6.29270002e-02 6.71761453e-01 1.23894894e+00
4.82444048e-01 -3.13169628e-01 5.71838617e-01 9.92879689e-01
8.11028406e-02 -9.03489470e-01 -7.84799993e-01 -4.58490282e-01
2.38005236e-01 -5.77737451e-01 5.85893571e-01 7.43035793e-01
-5.44909388e-02 4.36561480e-02 -2.10231081e-01 -3.56784880e-01
8.38252157e-02 3.10757041e-01 7.34866142e-01 -1.45180035e+00
-3.64601575e-02 -1.07821763e+00 -7.95898199e-01 -1.58323035e-01
3.00498083e-02 -7.26228237e-01 -2.42045701e-01 -1.52572370e+00
4.65212345e-01 -4.26393479e-01 -2.83153415e-01 4.38023537e-01
1.33039458e-02 7.96798170e-01 4.71873522e-01 1.70559287e-01
-4.76646215e-01 4.59143132e-01 1.36359572e+00 -8.56007263e-02
5.66871017e-02 -4.36148196e-01 -1.39327049e+00 6.15276277e-01
9.82942462e-01 -1.50573412e-02 -3.55140060e-01 -5.17493308e-01
6.79507732e-01 -5.67042589e-01 6.20921195e-01 -9.91061866e-01
-2.20553800e-01 -2.85937160e-01 8.12376142e-01 1.71170130e-01
4.81466532e-01 -1.04109275e+00 2.50724882e-01 6.68628886e-02
-5.93002200e-01 3.49411875e-01 5.89883566e-01 2.52716959e-01
-3.87507640e-02 -2.18709707e-01 8.79531443e-01 -1.71139270e-01
-6.70487463e-01 4.34809700e-02 -2.23386273e-01 -1.33588821e-01
8.57843280e-01 -8.21922898e-01 -1.84114262e-01 -6.92799270e-01
-5.13830364e-01 -3.95616382e-01 8.37585330e-01 7.50669777e-01
5.35390019e-01 -1.53824019e+00 -5.75057685e-01 -2.19380617e-01
4.98462886e-01 -7.84052074e-01 2.64187872e-01 8.43643606e-01
-3.70466858e-01 1.50021330e-01 -7.57900059e-01 -2.81641752e-01
-1.05455124e+00 6.41507030e-01 1.95194304e-01 2.14163437e-01
-1.79885790e-01 4.97615099e-01 2.57231921e-01 -2.31384531e-01
-1.89287901e-01 3.51993069e-02 -2.25225940e-01 3.48688960e-01
8.62332359e-02 3.24477553e-01 -1.21113501e-01 -1.18092060e+00
-2.12739766e-01 7.20982611e-01 1.73437417e-01 -1.38904408e-01
1.54536176e+00 -2.40303218e-01 -2.20959708e-01 6.61527395e-01
1.38073826e+00 -7.82449841e-02 -9.72377837e-01 3.40742886e-01
-1.96599320e-01 -7.34624326e-01 2.20118258e-02 -1.02835917e+00
-1.34727240e+00 7.11037099e-01 7.52620041e-01 3.62212360e-01
1.51795232e+00 3.71556021e-02 4.74407822e-01 -1.64835110e-01
1.27899647e-01 -1.30163002e+00 2.69993454e-01 -1.88542098e-01
1.27095735e+00 -1.27036297e+00 1.91944242e-01 -1.09943017e-01
-1.13406682e+00 1.34019363e+00 5.03313601e-01 -1.37894019e-01
4.88103002e-01 -1.80433869e-01 1.66446820e-01 -3.86853069e-01
-3.42875391e-01 -2.23383918e-01 7.44833171e-01 7.23578990e-01
8.10569763e-01 3.73174287e-02 -5.70814073e-01 7.06151485e-01
-7.64924288e-01 -2.08287016e-01 5.61631262e-01 5.29842973e-01
-1.97168320e-01 -8.10178220e-01 -2.26148859e-01 5.58673561e-01
-3.23199004e-01 -1.27851084e-01 -1.03220546e+00 9.42660213e-01
2.56260365e-01 1.09856200e+00 2.18500331e-01 -6.35695875e-01
2.54907519e-01 -1.33394644e-01 4.05049950e-01 -1.90572381e-01
-9.24435139e-01 -3.28421563e-01 5.51780760e-02 -7.65490592e-01
-5.17125666e-01 -8.73954117e-01 -4.74890441e-01 -5.24874628e-01
-2.77999669e-01 -1.90121770e-01 8.57388854e-01 6.29277527e-01
4.67657983e-01 2.18930617e-01 5.67781508e-01 -7.89204657e-01
-2.23412514e-02 -1.00915289e+00 -7.04336047e-01 9.11450684e-01
-7.50162005e-02 -7.77203381e-01 -4.37105715e-01 -8.14887602e-03] | [11.540876388549805, -0.8914440870285034] |
5f7a5e8f-4535-4893-9ffe-9903a4ac350d | it-s-all-in-the-teacher-zero-shot | 2203.17008 | null | https://arxiv.org/abs/2203.17008v2 | https://arxiv.org/pdf/2203.17008v2.pdf | It's All In the Teacher: Zero-Shot Quantization Brought Closer to the Teacher | Model quantization is considered as a promising method to greatly reduce the resource requirements of deep neural networks. To deal with the performance drop induced by quantization errors, a popular method is to use training data to fine-tune quantized networks. In real-world environments, however, such a method is frequently infeasible because training data is unavailable due to security, privacy, or confidentiality concerns. Zero-shot quantization addresses such problems, usually by taking information from the weights of a full-precision teacher network to compensate the performance drop of the quantized networks. In this paper, we first analyze the loss surface of state-of-the-art zero-shot quantization techniques and provide several findings. In contrast to usual knowledge distillation problems, zero-shot quantization often suffers from 1) the difficulty of optimizing multiple loss terms together, and 2) the poor generalization capability due to the use of synthetic samples. Furthermore, we observe that many weights fail to cross the rounding threshold during training the quantized networks even when it is necessary to do so for better performance. Based on the observations, we propose AIT, a simple yet powerful technique for zero-shot quantization, which addresses the aforementioned two problems in the following way: AIT i) uses a KL distance loss only without a cross-entropy loss, and ii) manipulates gradients to guarantee that a certain portion of weights are properly updated after crossing the rounding thresholds. Experiments show that AIT outperforms the performance of many existing methods by a great margin, taking over the overall state-of-the-art position in the field. | ['Jinho Lee', 'Youngsok Kim', 'Noseong Park', 'Joonsang Yu', 'Deokki Hong', 'Hye Yoon Lee', 'Kanghyun Choi'] | 2022-03-31 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Choi_Its_All_in_the_Teacher_Zero-Shot_Quantization_Brought_Closer_to_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Choi_Its_All_in_the_Teacher_Zero-Shot_Quantization_Brought_Closer_to_CVPR_2022_paper.pdf | cvpr-2022-1 | ['data-free-quantization', 'data-free-quantization'] | ['computer-vision', 'methodology'] | [ 1.81123897e-01 -4.81030382e-02 -2.44762987e-01 -2.62921125e-01
-7.40380645e-01 -1.90866098e-01 3.38006496e-01 2.75334269e-01
-9.04340208e-01 1.01256216e+00 -2.97825187e-01 -1.98715404e-01
-1.83495179e-01 -8.72682393e-01 -8.13636363e-01 -9.21033919e-01
1.53862759e-01 1.71085432e-01 2.90805548e-01 -1.52670041e-01
2.04382271e-01 3.20854902e-01 -1.68194675e+00 -1.42372727e-01
1.06097651e+00 1.42355287e+00 1.53716169e-02 5.38576096e-02
-1.39110252e-01 6.50909841e-01 -8.24269652e-01 -7.50994742e-01
4.73747581e-01 -2.64821887e-01 -4.81630862e-01 -2.83715576e-01
5.21392763e-01 -6.16240978e-01 -3.90187740e-01 1.69424534e+00
5.52012265e-01 4.07247901e-01 4.30745304e-01 -1.31606936e+00
-5.84607840e-01 7.30921745e-01 -4.39634472e-01 7.36985952e-02
-2.29910195e-01 2.73372531e-02 9.02583241e-01 -7.68627882e-01
3.26973081e-01 1.14962316e+00 5.39160788e-01 6.58397853e-01
-1.16446388e+00 -1.10329437e+00 -4.25367290e-03 3.08792889e-01
-1.68670380e+00 -6.68688357e-01 5.57376623e-01 -1.79052487e-01
7.03892529e-01 3.50659923e-03 4.22700614e-01 7.25311756e-01
-6.37712628e-02 4.93738443e-01 7.12552905e-01 -4.29006338e-01
7.54996419e-01 3.78971547e-01 1.87490508e-02 6.51740372e-01
3.97945046e-01 2.40483340e-02 -6.42260909e-01 -2.17924789e-01
5.45931756e-01 6.63791597e-02 -5.41934192e-01 -4.57277149e-01
-7.98503041e-01 9.76608157e-01 6.09289229e-01 3.03332150e-01
-1.54421419e-01 3.23738813e-01 5.10711849e-01 4.96496946e-01
3.58279407e-01 3.84554088e-01 -3.28530729e-01 -2.79773980e-01
-1.26758134e+00 1.77995414e-01 5.42219043e-01 7.85521924e-01
8.51266086e-01 3.28845859e-01 -1.54839814e-01 7.39219129e-01
9.05067176e-02 4.34611551e-02 7.93643713e-01 -9.69797432e-01
7.24069655e-01 1.82987168e-01 1.92000121e-02 -8.60358179e-01
2.96919316e-01 -1.80266336e-01 -1.11943400e+00 3.69541198e-01
4.51641798e-01 -1.46978781e-01 -9.71239507e-01 1.89361823e+00
1.67081803e-01 2.06466779e-01 2.28467658e-01 9.10993218e-01
5.51575959e-01 5.74201763e-01 -2.71530021e-02 -3.57046038e-01
1.12114465e+00 -6.68527782e-01 -8.17624986e-01 1.30304135e-02
6.10422671e-01 -2.15093240e-01 1.12886512e+00 3.81259501e-01
-1.15661168e+00 -3.52019012e-01 -1.46496630e+00 9.30875391e-02
-5.32276154e-01 -1.75256819e-01 3.47152770e-01 1.04971409e+00
-1.02015412e+00 1.04530156e+00 -8.17798316e-01 4.92869653e-02
6.73799038e-01 4.82223064e-01 -2.32458949e-01 -6.95872009e-02
-1.56739604e+00 6.77272916e-01 6.43677473e-01 -1.65501982e-01
-4.59247470e-01 -9.06651378e-01 -9.35709596e-01 5.53359330e-01
4.64764774e-01 -2.74731785e-01 1.15697896e+00 -8.34574521e-01
-1.67730713e+00 2.88213164e-01 1.03505746e-01 -7.87499368e-01
7.06747234e-01 -1.70058578e-01 -1.95030961e-02 1.97591260e-01
-1.95180878e-01 6.44953549e-01 9.46806431e-01 -8.63821447e-01
-5.21164060e-01 -3.77168059e-01 -1.76371578e-02 -2.26782635e-02
-1.01440191e+00 -4.09625500e-01 -4.01296645e-01 -6.94520414e-01
-1.38361335e-01 -5.60451508e-01 -1.99459612e-01 3.92268389e-01
-3.21741760e-01 -1.65491238e-01 8.00300717e-01 -2.63441980e-01
1.33564162e+00 -2.49320221e+00 -1.97586387e-01 2.14166433e-01
2.31140330e-01 6.89433932e-01 1.92654088e-01 1.33959400e-02
2.10348275e-02 4.62554693e-01 -3.01319063e-01 -5.23649752e-01
2.25729242e-01 2.85619825e-01 -4.24288988e-01 4.13696140e-01
5.17046973e-02 5.47051251e-01 -8.78363132e-01 -4.41503495e-01
2.44303927e-01 5.98231733e-01 -5.66937447e-01 -5.09800017e-02
-5.54834828e-02 -1.75955996e-01 -2.26861581e-01 3.29107344e-01
5.90432107e-01 -1.72430217e-01 1.43157784e-02 -1.38614684e-01
1.24435484e-01 2.63492495e-01 -1.46884656e+00 1.37833512e+00
-2.16634259e-01 5.40244102e-01 -7.70107331e-03 -1.09321618e+00
7.40142941e-01 4.76977170e-01 2.14241847e-01 -6.64203942e-01
3.46390754e-01 2.51391798e-01 -2.70832121e-01 -6.70877174e-02
4.20848101e-01 -2.01537222e-01 1.53913945e-01 1.51679486e-01
9.50240791e-02 8.60088468e-02 1.60383359e-01 4.76878695e-02
9.05375957e-01 -4.60949600e-01 2.03142464e-01 -1.65815204e-02
2.22311467e-01 -4.98784840e-01 8.48445237e-01 6.89887762e-01
-4.34070081e-01 4.47985709e-01 4.26379651e-01 -1.23004280e-01
-9.82083738e-01 -8.28778863e-01 -1.42006874e-01 8.36919188e-01
3.72257903e-02 -3.22787642e-01 -8.60593557e-01 -3.55172455e-01
1.18143596e-02 8.21721971e-01 -5.41289508e-01 -5.69854259e-01
-5.27235568e-02 -5.45791149e-01 6.93866134e-01 5.13448238e-01
9.58157122e-01 -6.27130508e-01 -7.44008243e-01 8.66385996e-02
8.05693418e-02 -1.01574051e+00 -4.72765088e-01 4.59668785e-01
-8.36164713e-01 -5.62030792e-01 -8.10628891e-01 -5.55513740e-01
5.62454879e-01 2.12168306e-01 6.77409947e-01 1.29106224e-01
-2.89905332e-02 -2.63445795e-01 -2.15284139e-01 -2.66610056e-01
-1.23976521e-01 1.28767222e-01 4.01748747e-01 -5.58791086e-02
3.67095888e-01 -7.11323857e-01 -4.17028934e-01 -1.61012057e-02
-1.10752177e+00 -4.25120741e-01 4.44117934e-01 9.92022157e-01
6.38830364e-01 5.84380329e-01 7.04327166e-01 -6.61257625e-01
6.80208802e-01 -2.25110024e-01 -8.95088136e-01 3.03987235e-01
-9.30480540e-01 3.53633672e-01 8.53377044e-01 -6.78779542e-01
-6.80171192e-01 -6.99198991e-02 5.09168357e-02 -1.13268399e+00
2.78620213e-01 2.41429195e-01 -3.21871907e-01 -4.06334996e-01
5.66387594e-01 1.65487811e-01 6.92126388e-03 -2.12892309e-01
3.89904678e-01 7.52532959e-01 4.45429623e-01 -4.08167511e-01
8.82282019e-01 1.82288587e-01 -1.64894328e-01 -7.92610943e-01
-7.68427253e-01 -2.35793561e-01 -3.43261778e-01 2.21685544e-01
7.05440462e-01 -8.35500896e-01 -8.45514476e-01 2.77475148e-01
-8.23124409e-01 -2.50492811e-01 -5.35585523e-01 5.31774998e-01
-3.79545420e-01 4.15127754e-01 -5.47448933e-01 -9.33951795e-01
-4.95037407e-01 -1.30852914e+00 5.99301577e-01 3.75084996e-01
1.80286035e-01 -7.54718125e-01 -4.07051653e-01 8.94737095e-02
4.21497524e-01 4.99523170e-02 9.21101391e-01 -5.21266937e-01
-4.18891042e-01 -1.34613320e-01 -2.67350912e-01 6.38358414e-01
8.68659765e-02 -5.11148646e-02 -1.05148399e+00 -4.96098518e-01
1.27495781e-01 -6.99718893e-01 9.33734179e-01 1.12448625e-01
1.52468133e+00 -4.89255100e-01 4.92342152e-02 7.93305099e-01
1.48231101e+00 1.31052241e-01 6.37474000e-01 3.46678823e-01
4.18108374e-01 1.12968467e-01 4.97669935e-01 6.27880514e-01
6.83835968e-02 4.33755338e-01 4.95301545e-01 3.41106951e-01
2.16659844e-01 -2.73977816e-01 2.35101342e-01 6.92732334e-01
2.27035627e-01 -1.89081624e-01 -4.83122706e-01 5.22695422e-01
-1.63090312e+00 -8.70287836e-01 4.37199503e-01 2.71692514e+00
1.24236226e+00 5.16822934e-01 -1.57150835e-01 6.04843855e-01
5.86300433e-01 1.69797674e-01 -7.43983388e-01 -4.30533350e-01
2.26131201e-01 1.84068739e-01 8.74813676e-01 1.55688167e-01
-9.94877934e-01 8.30299318e-01 5.98280001e+00 1.30037999e+00
-1.30135870e+00 9.33119282e-02 7.26279855e-01 -2.21401632e-01
-4.45340946e-02 -2.79039174e-01 -1.04522133e+00 6.65751755e-01
1.09789288e+00 -2.84733415e-01 6.14722669e-01 1.01840949e+00
-2.40435138e-01 7.60445073e-02 -1.01019001e+00 1.16261411e+00
4.11589816e-03 -1.35552955e+00 1.56434253e-01 2.17335057e-02
7.32419431e-01 -3.86671238e-02 2.92358607e-01 5.18643498e-01
1.57304168e-01 -9.75913644e-01 7.79992342e-01 1.26549408e-01
1.11250019e+00 -9.81735349e-01 6.95609927e-01 4.63049918e-01
-9.84183371e-01 -1.32586673e-01 -8.13497722e-01 -2.18932945e-02
-2.07164600e-01 7.88805246e-01 -6.45461202e-01 3.21753353e-01
7.32062578e-01 2.83882558e-01 -2.40711436e-01 1.11767423e+00
-2.09628820e-01 5.54575324e-01 -5.28466761e-01 -7.92353898e-02
4.13128257e-01 -1.29057005e-01 2.09581703e-01 7.12937176e-01
4.25908655e-01 2.51313020e-02 -5.24767749e-02 9.01764691e-01
-6.22288883e-01 -7.48923570e-02 -5.32413542e-01 -1.83739349e-01
9.52551365e-01 8.55837464e-01 -3.98163974e-01 -5.43539643e-01
-3.77492160e-01 9.03367519e-01 5.34558117e-01 3.53727460e-01
-7.43729949e-01 -8.35091174e-01 7.47901559e-01 -1.47235870e-01
5.92369914e-01 -5.81168644e-02 -2.67122537e-01 -1.01954627e+00
2.55358458e-01 -5.96047103e-01 1.51414588e-01 -4.33557123e-01
-1.12121737e+00 4.37986910e-01 -9.41563547e-02 -1.22990763e+00
-1.47514373e-01 -3.05270314e-01 -3.77158046e-01 7.85165489e-01
-1.81588662e+00 -3.43543351e-01 -6.37548864e-02 6.40026569e-01
3.52006257e-01 -1.39733166e-01 8.29854310e-01 6.02765203e-01
-6.79040790e-01 1.32145214e+00 5.32995343e-01 1.77343681e-01
6.65356457e-01 -1.04205191e+00 1.74115181e-01 6.30836904e-01
3.57558019e-04 6.19999170e-01 6.02790713e-01 -2.95197487e-01
-1.11279786e+00 -1.17090833e+00 7.56294250e-01 1.38469726e-01
6.43186212e-01 -2.46278018e-01 -1.29636884e+00 3.82675648e-01
-1.84225902e-01 2.75057316e-01 6.36472583e-01 -2.83125937e-01
-4.74563003e-01 -3.51977527e-01 -1.48582792e+00 4.97689128e-01
4.73857760e-01 -4.93721485e-01 -4.90192771e-01 2.00632170e-01
9.30660069e-01 -3.09786826e-01 -1.03642178e+00 2.60694951e-01
3.76339465e-01 -8.19295287e-01 8.13877344e-01 -2.81527102e-01
2.54952103e-01 -8.18241611e-02 -3.19779336e-01 -1.37666559e+00
2.82089543e-02 -7.09436893e-01 -3.32274109e-01 1.25581312e+00
2.04548836e-01 -6.69821858e-01 9.90863979e-01 8.69196415e-01
2.32761458e-01 -9.11713243e-01 -1.38817990e+00 -1.05249739e+00
1.03544250e-01 -1.66834161e-01 7.33612418e-01 1.01633060e+00
5.24048395e-02 4.56113927e-02 -2.73134917e-01 4.28620800e-02
8.83052766e-01 -4.05860305e-01 3.39308232e-01 -1.14617884e+00
-1.76648229e-01 -5.44008851e-01 -6.49863303e-01 -1.02727950e+00
4.83168811e-02 -5.23498058e-01 2.00593382e-01 -1.03106034e+00
-1.36374459e-01 -5.51895440e-01 -4.63123441e-01 5.85966647e-01
-1.13927513e-01 2.65647680e-01 3.30812573e-01 9.60148871e-02
-4.98089254e-01 9.43942130e-01 1.04687893e+00 -2.59499997e-01
-2.46424943e-01 -6.12402745e-02 -6.11239672e-01 5.93415976e-01
8.61906230e-01 -4.83965456e-01 -6.54813945e-01 -3.59239906e-01
7.62912855e-02 3.70520204e-02 1.31803498e-01 -1.27914453e+00
4.58816975e-01 -9.21515450e-02 2.15001002e-01 -2.87060320e-01
5.84513247e-01 -9.02109921e-01 -2.84239948e-01 5.04938185e-01
-3.03240985e-01 -3.61216009e-01 3.52972224e-02 6.71349108e-01
-3.32425684e-01 -5.09203076e-01 1.03891599e+00 -1.78372003e-02
-4.91210401e-01 4.34758127e-01 -1.41090155e-01 8.53017792e-02
9.00359690e-01 -3.22693765e-01 -3.18301558e-01 -1.97244301e-01
-3.44641238e-01 3.85745615e-01 4.42012429e-01 1.04833193e-01
6.17430568e-01 -1.43131387e+00 -1.93910196e-01 4.06066597e-01
-9.73504409e-03 1.73004493e-01 -3.41439657e-02 5.33954978e-01
-2.52437055e-01 2.62563467e-01 -6.14179559e-02 -2.92327404e-01
-1.03184402e+00 6.98919654e-01 3.79884601e-01 -7.63117298e-02
-6.33890688e-01 1.01474404e+00 -9.69646201e-02 -7.15802163e-02
9.76034105e-01 -5.47523618e-01 -1.89855676e-02 5.94692975e-02
8.59116852e-01 2.21884295e-01 2.80541152e-01 -1.85048223e-01
-1.29840046e-01 2.34905705e-01 -2.35293493e-01 -7.89645389e-02
1.11440861e+00 -7.66940927e-03 2.97348082e-01 6.29310727e-01
1.36743939e+00 -5.91884851e-01 -1.39061213e+00 -4.60618049e-01
-2.82161117e-01 -5.71730137e-01 4.06558186e-01 -3.49960357e-01
-1.39619613e+00 1.13894165e+00 7.25234389e-01 2.01198652e-01
1.09486413e+00 -6.17116511e-01 1.07242835e+00 6.68272734e-01
4.10379261e-01 -1.33045864e+00 -4.55251187e-02 3.97349805e-01
2.95130074e-01 -1.24039531e+00 1.34780137e-02 -1.05164289e-01
-4.42847908e-01 9.54954028e-01 5.35293400e-01 4.23399732e-02
7.80005217e-01 5.15558384e-02 -2.28108212e-01 1.62313953e-01
-7.12209761e-01 1.40205294e-01 -5.09560406e-02 3.86610419e-01
-6.56792149e-03 -1.77008703e-01 -1.56583428e-01 4.55107123e-01
-3.12656373e-01 1.83008566e-01 3.62139493e-01 1.01781392e+00
-7.45529652e-01 -8.63031864e-01 -3.74973029e-01 5.42517424e-01
-6.00265205e-01 -1.63605869e-01 -7.36931860e-02 4.47228998e-01
6.15558028e-02 8.00025403e-01 -2.75684595e-02 -4.48618978e-01
2.69104242e-01 9.43000615e-02 1.82161763e-01 -3.59858930e-01
-3.40070993e-01 -3.06737572e-01 -3.90564531e-01 -3.91064256e-01
-1.37874916e-01 -2.34017923e-01 -1.37000382e+00 -4.98383492e-01
-5.79677224e-01 2.21283153e-01 7.37823367e-01 9.20363963e-01
2.24851117e-01 6.13098383e-01 5.26764989e-01 -7.37433732e-01
-1.31313670e+00 -7.76829362e-01 -8.50572705e-01 2.86561400e-01
5.40743291e-01 -8.47851694e-01 -6.27029598e-01 -1.55973658e-01] | [8.744338989257812, 3.031583309173584] |
891bc88f-a636-4cea-9734-1cce8eef092a | optimizing-fairness-tradeoffs-in-machine | 2304.12190 | null | https://arxiv.org/abs/2304.12190v1 | https://arxiv.org/pdf/2304.12190v1.pdf | Optimizing fairness tradeoffs in machine learning with multiobjective meta-models | Improving the fairness of machine learning models is a nuanced task that requires decision makers to reason about multiple, conflicting criteria. The majority of fair machine learning methods transform the error-fairness trade-off into a single objective problem with a parameter controlling the relative importance of error versus fairness. We propose instead to directly optimize the error-fairness tradeoff by using multi-objective optimization. We present a flexible framework for defining the fair machine learning task as a weighted classification problem with multiple cost functions. This framework is agnostic to the underlying prediction model as well as the metrics. We use multiobjective optimization to define the sample weights used in model training for a given machine learner, and adapt the weights to optimize multiple metrics of fairness and accuracy across a set of tasks. To reduce the number of optimized parameters, and to constrain their complexity with respect to population subgroups, we propose a novel meta-model approach that learns to map protected attributes to sample weights, rather than optimizing those weights directly. On a set of real-world problems, this approach outperforms current state-of-the-art methods by finding solution sets with preferable error/fairness trade-offs. | ['William G. La Cava'] | 2023-04-21 | null | null | null | null | ['multiobjective-optimization'] | ['methodology'] | [ 2.91737735e-01 1.35706589e-01 -7.25355387e-01 -9.42060888e-01
-9.35801864e-01 -2.43670315e-01 4.04433221e-01 4.92414683e-01
-1.00405073e+00 1.03568721e+00 1.80822164e-01 -2.98911870e-01
-5.88161767e-01 -6.69726729e-01 -2.10014671e-01 -4.78277504e-01
3.61610919e-01 6.30787611e-01 -4.09248531e-01 2.23086867e-02
5.43122709e-01 1.80622727e-01 -1.38849354e+00 1.48259103e-01
1.37481034e+00 1.06542003e+00 -3.84710819e-01 4.63439107e-01
7.60623291e-02 6.03197813e-01 -7.46667981e-01 -1.14589071e+00
1.28098816e-01 -4.65534031e-01 -8.00542712e-01 -2.72714376e-01
5.27956307e-01 -8.82562399e-02 4.41646367e-01 9.87313032e-01
6.60192609e-01 1.58934623e-01 1.00277865e+00 -1.63965833e+00
-5.87242067e-01 5.28746486e-01 -3.87376070e-01 -9.68925580e-02
-3.61829251e-01 7.67016634e-02 1.42392719e+00 -2.76820004e-01
1.63999960e-01 1.48444676e+00 6.94887519e-01 7.38957584e-01
-1.62415791e+00 -7.21580625e-01 1.70521364e-01 2.24042460e-01
-1.13123858e+00 -7.18346953e-01 3.88162971e-01 -5.05998731e-01
5.51608860e-01 6.57701254e-01 3.25275511e-01 7.25575805e-01
2.24086344e-01 3.03679764e-01 1.16828930e+00 -5.77198982e-01
4.61532593e-01 3.77246678e-01 2.03963190e-01 6.15374804e-01
4.85062569e-01 3.63454103e-01 -3.61983061e-01 -5.36724627e-01
5.82251102e-02 -8.11175033e-02 4.74689901e-02 -5.66256642e-01
-6.99369371e-01 1.34740210e+00 1.50882006e-01 -2.11643100e-01
-1.80642098e-01 2.52280146e-01 4.23578829e-01 3.27529520e-01
8.09792817e-01 9.24502134e-01 -6.14408731e-01 1.22597396e-01
-9.45334554e-01 6.17545962e-01 7.94382870e-01 2.05206960e-01
5.64475477e-01 -3.08457077e-01 -7.28715122e-01 1.12461770e+00
4.80513126e-01 1.19414665e-01 2.65165806e-01 -1.35417593e+00
5.59216678e-01 4.55252051e-01 3.07084322e-01 -8.84956896e-01
-4.15677369e-01 -6.88952744e-01 -5.70505381e-01 6.10566795e-01
6.79782629e-01 -3.26375246e-01 -4.58530515e-01 2.15525842e+00
4.55514222e-01 -3.97998631e-01 -2.46177346e-01 8.56162965e-01
2.80023277e-01 2.73224473e-01 6.83752418e-01 -3.55212390e-01
1.15761495e+00 -8.96057248e-01 -4.63336736e-01 -3.07453364e-01
6.31451607e-01 -5.31315327e-01 1.09164345e+00 2.58399397e-01
-1.15795743e+00 -1.11293040e-01 -9.29895401e-01 -1.68717057e-01
-2.89147586e-01 -5.51079623e-02 4.70864236e-01 1.15496111e+00
-6.90085232e-01 9.05782461e-01 -1.64717123e-01 7.53656700e-02
6.87042654e-01 5.16774118e-01 2.68037282e-02 2.04289123e-01
-1.12812996e+00 1.34001994e+00 3.31158787e-01 -2.91422635e-01
-6.28126264e-01 -1.24548984e+00 -5.51064551e-01 4.37570393e-01
4.63547885e-01 -1.12870407e+00 1.20298243e+00 -1.42900121e+00
-1.48583198e+00 1.09294975e+00 2.31865391e-01 -2.91744739e-01
8.68276536e-01 -1.58524476e-02 -2.10591614e-01 -5.73719084e-01
2.88699605e-02 4.75335062e-01 7.21716344e-01 -1.23103547e+00
-8.53311360e-01 -3.13398421e-01 6.03964664e-02 3.87667745e-01
-5.21589577e-01 3.63571763e-01 1.33884996e-01 -6.70755565e-01
-6.78277850e-01 -7.08065271e-01 -4.53823566e-01 3.70909065e-01
-1.95631206e-01 -1.82568803e-01 9.06161219e-02 -2.94326395e-01
1.47576749e+00 -1.90645742e+00 2.23282203e-01 2.98033416e-01
4.20054346e-01 2.27929682e-01 -3.08810294e-01 -2.23119095e-01
2.34906990e-02 4.11865979e-01 -4.76292342e-01 -6.77694499e-01
3.62516761e-01 3.02135739e-02 1.74909815e-01 5.08888781e-01
7.10521936e-02 6.35801911e-01 -6.99242294e-01 -7.00089157e-01
-1.50138870e-01 2.32843176e-01 -1.01501131e+00 4.21504855e-01
-9.20054018e-02 3.14889513e-02 -4.15225834e-01 2.41586700e-01
5.72009861e-01 -9.27233603e-03 2.29862645e-01 -3.56585442e-05
-6.90276697e-02 3.06109637e-01 -1.25066137e+00 1.08345115e+00
-5.37605226e-01 6.72863051e-02 1.68459967e-01 -1.18441856e+00
6.72210038e-01 -8.74426886e-02 4.03416991e-01 -6.01779342e-01
1.83574364e-01 2.57419437e-01 1.28537580e-01 -2.39590898e-01
4.26893950e-01 -7.61705577e-01 -2.09218308e-01 7.25362301e-01
1.12937577e-01 6.35509416e-02 3.46955098e-02 -3.85494471e-01
5.35611212e-01 -2.10020900e-01 6.79440141e-01 -7.68975973e-01
4.07162607e-01 -3.32107753e-01 9.41753149e-01 7.67176628e-01
-3.26331109e-01 3.99152845e-01 7.36607552e-01 -5.60866237e-01
-9.24365282e-01 -8.62387776e-01 -4.16430920e-01 1.59825921e+00
-2.91004211e-01 -1.36488631e-01 -6.60981417e-01 -8.06801140e-01
4.64764088e-01 9.79359388e-01 -8.58143091e-01 -4.63630646e-01
-2.13380650e-01 -1.35964155e+00 4.68741059e-01 -1.64649859e-02
7.62761757e-02 -6.14016235e-01 -9.61057842e-01 1.23334918e-02
-1.56201914e-01 -4.98638809e-01 -7.75359690e-01 2.19061598e-01
-7.64226317e-01 -9.45655942e-01 -6.56619370e-01 -9.98354331e-02
5.28656363e-01 -3.14581871e-01 1.58656752e+00 3.00286174e-01
-3.29860717e-01 1.75432295e-01 8.08434784e-02 -8.06047261e-01
-3.98254424e-01 1.40044734e-01 -4.76630330e-02 1.96269244e-01
3.28108184e-02 -2.44251892e-01 -5.19139528e-01 3.73991877e-01
-6.60473466e-01 2.21538506e-02 2.92557478e-01 1.06752026e+00
4.50881839e-01 -2.70213842e-01 7.26996481e-01 -1.22604954e+00
9.82472181e-01 -6.17891014e-01 -5.67132473e-01 7.07896471e-01
-1.31162190e+00 4.86897528e-01 5.19648015e-01 -4.27667975e-01
-1.03992724e+00 -4.92543817e-01 2.64446497e-01 -1.56925786e-02
2.75676072e-01 4.38992977e-01 -3.73227805e-01 -6.37385994e-02
7.73285389e-01 -5.93102157e-01 8.35608244e-02 -4.39372689e-01
4.02540475e-01 5.04683912e-01 1.10669918e-01 -9.94544685e-01
3.47068280e-01 -1.83261469e-01 2.95228571e-01 -7.24345446e-02
-1.15100837e+00 1.54901579e-01 -2.46476889e-01 -1.73755631e-01
5.67969561e-01 -5.53178191e-01 -1.05250883e+00 8.54220018e-02
-7.68981218e-01 -5.43443322e-01 -4.93170708e-01 3.83625925e-01
-5.70951521e-01 -1.14360280e-01 7.85717182e-03 -1.00840843e+00
-4.89366114e-01 -1.00169301e+00 7.21258521e-01 1.01426363e-01
-5.01488030e-01 -1.16054916e+00 2.10863322e-01 4.70048398e-01
6.52470291e-01 3.44361931e-01 1.41304958e+00 -6.43700540e-01
3.61975506e-02 1.03238858e-01 -3.17546368e-01 2.79518962e-01
6.68569561e-03 2.55063120e-02 -8.82350206e-01 -1.46665379e-01
-6.26209751e-02 -4.91627097e-01 8.53988469e-01 6.68485522e-01
1.68785965e+00 -9.56215620e-01 -2.75931377e-02 8.46417189e-01
1.40984917e+00 -2.11647451e-02 1.64014250e-01 4.53849047e-01
4.18479353e-01 9.62270141e-01 4.48423177e-01 7.00024486e-01
7.08119571e-01 8.99780691e-01 3.94417107e-01 -9.79753658e-02
3.31667006e-01 -3.87296975e-02 -1.69722717e-02 -7.08643571e-02
-1.58541992e-01 -1.73112005e-01 -8.66665483e-01 1.10026963e-01
-2.08195901e+00 -9.62045968e-01 3.55539829e-01 2.59275818e+00
1.19581509e+00 2.97251809e-02 4.91990328e-01 2.92982534e-02
6.90742910e-01 7.70045146e-02 -6.59034491e-01 -9.78505850e-01
1.45163432e-01 1.14178374e-01 6.14150286e-01 8.34163427e-01
-9.88909423e-01 5.00915885e-01 6.91301203e+00 8.99739623e-01
-9.73341107e-01 1.12453096e-01 1.38098419e+00 -6.62609220e-01
-7.38932848e-01 -1.15431800e-01 -4.88951117e-01 4.40981060e-01
1.03097403e+00 -6.32018805e-01 7.71831751e-01 5.66966653e-01
2.87852883e-01 1.71145108e-02 -1.36711347e+00 7.48763859e-01
-1.18025988e-01 -1.09776413e+00 -3.77947204e-02 6.80509061e-02
7.22799122e-01 -4.50621575e-01 3.45106840e-01 2.78353542e-01
5.63353539e-01 -1.56141329e+00 1.11380780e+00 7.98464954e-01
7.32145607e-01 -9.67515826e-01 5.13629556e-01 3.27885091e-01
-4.98905003e-01 -4.58344221e-01 -2.18399674e-01 -1.10534869e-01
-2.16851264e-01 8.84492457e-01 -2.27211192e-01 2.23495990e-01
5.84666073e-01 2.09867761e-01 -5.45743406e-01 9.80792880e-01
2.60542780e-01 2.80043304e-01 2.14290693e-02 -8.55745003e-02
-4.02756147e-02 -1.18601181e-01 1.21875100e-01 1.16579688e+00
3.01865488e-01 7.40005374e-02 -1.23191983e-01 1.18919337e+00
-3.53400171e-01 4.58945096e-01 -2.97756214e-02 1.70070007e-01
5.65668344e-01 1.26966810e+00 -1.28985018e-01 -4.13416252e-02
-9.38275829e-02 3.96162033e-01 6.85528219e-01 1.39485031e-01
-7.48731852e-01 -2.38154426e-01 1.13700438e+00 -8.45577940e-02
-4.30220366e-01 3.13401461e-01 -8.81441534e-01 -9.60340738e-01
-2.60716707e-01 -1.15567005e+00 8.64370108e-01 -1.60366222e-01
-1.51966965e+00 1.64250895e-01 1.03964023e-01 -9.02671874e-01
-3.58641818e-02 -5.47351599e-01 -6.76197529e-01 1.18917322e+00
-1.65025318e+00 -8.31966639e-01 1.17143705e-01 3.51012856e-01
5.02540451e-03 -2.66735613e-01 9.73438203e-01 3.07452142e-01
-7.34281123e-01 1.13261306e+00 1.86512247e-01 -5.04257560e-01
1.02837074e+00 -1.20296431e+00 -1.36406168e-01 4.00873333e-01
-4.39618319e-01 3.96989882e-01 8.19844842e-01 -1.03245907e-01
-6.39172196e-01 -9.90910232e-01 1.29492688e+00 -5.00087976e-01
2.24048883e-01 -1.08993880e-01 -5.86888850e-01 1.84120312e-01
-7.01678395e-02 5.64434156e-02 1.29090440e+00 5.33023477e-01
-5.45656681e-01 -5.22771060e-01 -1.68322134e+00 6.29795134e-01
8.75373781e-01 -2.60971069e-01 -1.13958366e-01 3.09447125e-02
3.37115258e-01 -1.61337540e-01 -9.12653804e-01 3.15416455e-01
8.53509545e-01 -9.82059181e-01 1.04255223e+00 -1.13844728e+00
7.02928305e-01 1.71286494e-01 -3.72424811e-01 -1.64272392e+00
-5.12118101e-01 -4.80617344e-01 7.04113543e-02 1.20902741e+00
7.24239230e-01 -7.49830127e-01 5.35360098e-01 1.31629062e+00
2.74705172e-01 -9.77739871e-01 -1.07038546e+00 -3.78430992e-01
5.53747475e-01 -1.98835894e-01 9.69365656e-01 1.02403605e+00
-1.37402222e-01 2.18279824e-01 -5.93219459e-01 -2.29815811e-01
1.13389540e+00 8.02661255e-02 4.24758315e-01 -1.56608176e+00
-3.80608559e-01 -1.11984909e+00 6.48936182e-02 3.01133897e-02
5.09388804e-01 -9.74915862e-01 -1.05542354e-01 -9.55180764e-01
5.73608279e-01 -8.16049993e-01 -5.57018995e-01 5.87778091e-01
-6.08018577e-01 -9.01995152e-02 5.62751114e-01 -1.00619629e-01
-4.86204088e-01 6.16998374e-01 9.03778195e-01 -1.36352181e-01
-3.75217274e-02 2.47210085e-01 -1.39058518e+00 5.01477122e-01
8.36755276e-01 -9.21266258e-01 -4.74673033e-01 -4.82419848e-01
2.60158747e-01 9.71996486e-02 2.90347874e-01 -5.76863408e-01
-1.04384646e-01 -1.05759001e+00 4.58773166e-01 5.38858891e-01
-8.40782560e-03 -6.80958450e-01 1.23919316e-01 5.15758276e-01
-1.12255907e+00 1.24572508e-01 -1.96141507e-02 1.95920825e-01
2.01279014e-01 -3.67793262e-01 1.27486062e+00 -3.46883051e-02
-2.46731155e-02 3.61310601e-01 8.44162479e-02 4.44205552e-01
9.35173452e-01 2.13478401e-01 -3.89195621e-01 -1.80314049e-01
-4.31083679e-01 4.51589674e-01 2.86004394e-01 3.81868392e-01
1.81706905e-01 -1.30755997e+00 -1.07583332e+00 -6.95328191e-02
2.07058862e-01 -5.88196516e-01 9.41881463e-02 4.40605193e-01
-1.14267714e-01 2.41550505e-01 -4.61191326e-01 -3.63446549e-02
-1.31694055e+00 2.91224301e-01 8.53979051e-01 -6.94380701e-01
2.19176188e-01 7.50407636e-01 9.35595259e-02 -7.34265983e-01
3.01534534e-01 1.13209978e-01 -2.06174642e-01 2.94778943e-01
3.41439724e-01 7.95709789e-01 -7.15947477e-03 -4.05804068e-01
-3.12524766e-01 4.09744322e-01 2.65212059e-01 -2.23204941e-01
1.28116000e+00 -1.43271014e-01 -7.59603754e-02 1.65675163e-01
1.20769846e+00 -2.82865822e-01 -1.21248794e+00 -1.88092589e-01
1.27719432e-01 -7.31553793e-01 3.11199635e-01 -1.32903588e+00
-1.11370373e+00 6.47114813e-01 7.37781286e-01 -7.34035149e-02
1.17207110e+00 -5.63639104e-01 8.04416537e-02 9.16925892e-02
2.47355830e-02 -1.36975396e+00 -1.54854923e-01 1.58409148e-01
7.92618871e-01 -1.45835221e+00 2.35656857e-01 6.05379343e-02
-8.13980937e-01 9.98074234e-01 5.67468405e-01 2.76176751e-01
6.90305710e-01 1.99781600e-02 1.40973181e-01 2.27322057e-01
-9.47372019e-01 1.60403609e-01 7.22682536e-01 3.55116904e-01
5.91607153e-01 4.59178567e-01 -7.98448741e-01 9.53770638e-01
-2.14171126e-01 1.57771688e-02 1.08717598e-01 4.78956074e-01
-3.36126417e-01 -1.50871217e+00 -2.87305683e-01 9.11676288e-01
-7.96541691e-01 4.82103080e-02 -2.39233226e-01 2.81534195e-01
2.62542367e-01 9.73602295e-01 5.27943894e-02 -2.11454883e-01
4.39886659e-01 6.41079098e-02 3.00425351e-01 -4.38777059e-01
-9.13295329e-01 -5.14406979e-01 3.20366740e-01 -5.31999767e-01
-4.25419480e-01 -6.15269363e-01 -6.23776734e-01 -7.41059005e-01
-6.47847578e-02 2.10466653e-01 5.48728049e-01 1.02199411e+00
1.80257484e-01 4.39329416e-01 7.90134668e-01 -5.72680175e-01
-1.17684579e+00 -3.62587661e-01 -4.65974182e-01 3.99095476e-01
5.40418088e-01 -6.65787756e-01 -3.06989193e-01 -5.58918715e-01] | [8.899652481079102, 5.275594234466553] |
c87b11cd-f028-4b9c-a911-17e0483d2fca | extracting-complex-named-entities-in-legal | 2305.05836 | null | https://arxiv.org/abs/2305.05836v1 | https://arxiv.org/pdf/2305.05836v1.pdf | Extracting Complex Named Entities in Legal Documents via Weakly Supervised Object Detection | Accurate Named Entity Recognition (NER) is crucial for various information retrieval tasks in industry. However, despite significant progress in traditional NER methods, the extraction of Complex Named Entities remains a relatively unexplored area. In this paper, we propose a novel system that combines object detection for Document Layout Analysis (DLA) with weakly supervised learning to address the challenge of extracting discontinuous complex named entities in legal documents. Notably, to the best of our knowledge, this is the first work to apply weak supervision to DLA. Our experimental results show that the model trained solely on pseudo labels outperforms the supervised baseline when gold-standard data is limited, highlighting the effectiveness of our proposed approach in reducing the dependency on annotated data. | ['Abhinav Agrawal', 'Hsiu-Wei Yang'] | 2023-05-10 | null | null | null | null | ['document-layout-analysis', 'weakly-supervised-object-detection', 'named-entity-recognition-ner'] | ['computer-vision', 'computer-vision', 'natural-language-processing'] | [ 8.33696947e-02 2.69711494e-01 -2.80923933e-01 -3.33251208e-01
-1.17329037e+00 -9.05581832e-01 5.77612758e-01 2.63859659e-01
-6.71667337e-01 8.12443793e-01 1.45352513e-01 -6.59560978e-01
-6.46045282e-02 -4.89157289e-01 -6.60965741e-01 -1.56187654e-01
1.88781973e-02 1.49374485e-01 3.84842306e-01 7.66216069e-02
5.70095241e-01 1.09485233e+00 -1.05508935e+00 4.63703632e-01
1.07992077e+00 6.58759773e-01 5.33377528e-02 4.15931761e-01
-5.75469017e-01 9.70744371e-01 -9.71083522e-01 -6.51115954e-01
1.39554530e-01 -1.65914550e-01 -1.10433483e+00 -4.00155485e-02
8.29485357e-01 -1.34868249e-01 -9.75132138e-02 7.55097091e-01
5.52431345e-01 -4.64867577e-02 5.28906763e-01 -7.09616244e-01
-6.18097544e-01 6.81182444e-01 -4.41889673e-01 1.32655099e-01
2.18811244e-01 -2.98407435e-01 1.26962018e+00 -1.10315752e+00
1.17677736e+00 9.21126485e-01 6.91526055e-01 4.34312671e-01
-8.84745181e-01 -4.62640852e-01 2.79905617e-01 4.03938703e-02
-1.45801377e+00 -4.49686944e-01 7.06705451e-01 -1.66716501e-01
1.07913339e+00 -7.06588328e-02 -1.28871605e-01 9.13228989e-01
2.31501609e-02 1.09749258e+00 1.05563939e+00 -7.47290552e-01
2.92173147e-01 2.98074067e-01 5.65344214e-01 7.70638704e-01
5.79342067e-01 -5.61340868e-01 -3.95191014e-01 -6.56229183e-02
2.88675755e-01 -1.95703849e-01 1.14438154e-01 -3.36123049e-01
-8.98682117e-01 4.44613129e-01 -1.75563879e-02 7.37920702e-01
-3.33121330e-01 -2.05645770e-01 4.45731580e-01 3.88611369e-02
3.67157847e-01 8.71262133e-01 -6.62652552e-01 -2.85699695e-01
-1.06028163e+00 -1.13179557e-01 1.17426980e+00 1.14704835e+00
4.74671751e-01 -1.95947841e-01 -2.92723924e-01 6.54068053e-01
2.75442638e-02 4.57081109e-01 4.95215468e-02 -6.01343691e-01
7.84716785e-01 1.03087378e+00 4.07891005e-01 -7.26155698e-01
-4.36591983e-01 -3.89489710e-01 -2.39306569e-01 9.23709720e-02
6.40668869e-01 -3.05918515e-01 -9.71948624e-01 1.02344894e+00
2.12529540e-01 -1.21969707e-01 1.43838510e-01 4.77764875e-01
6.70350075e-01 4.26516026e-01 1.90063611e-01 -2.25983839e-02
1.10166466e+00 -1.01926017e+00 -9.14289653e-01 -5.37302606e-02
1.02420652e+00 -1.15836298e+00 7.17804193e-01 2.67089605e-01
-7.11143434e-01 -3.87162834e-01 -8.80311251e-01 -3.41050416e-01
-7.87689745e-01 7.94562876e-01 7.07141697e-01 9.82807636e-01
-5.53568125e-01 4.09256905e-01 -7.35961676e-01 -6.79241240e-01
7.51759946e-01 2.89204329e-01 -4.90389109e-01 -3.30662757e-01
-9.64761198e-01 7.49601781e-01 3.61234516e-01 2.07583055e-01
-3.86206299e-01 -4.40229744e-01 -8.66987646e-01 -4.05690372e-02
8.93824399e-01 -5.00038117e-02 1.28011382e+00 -2.67018974e-01
-1.22219479e+00 7.09932804e-01 -3.72430772e-01 -4.78236765e-01
2.85349995e-01 -6.77046359e-01 -6.22593343e-01 2.82375384e-02
3.85337085e-01 2.97504246e-01 3.81787568e-01 -1.33884573e+00
-7.66784191e-01 -3.70019704e-01 2.24848270e-01 -2.16907173e-01
-5.65585196e-01 4.95116234e-01 -6.42612159e-01 -6.13339782e-01
-1.96927547e-01 -7.77030945e-01 -2.29426831e-01 -1.62372202e-01
-6.83881700e-01 -5.66289544e-01 9.24662650e-01 -8.12714696e-01
1.54804587e+00 -2.25304484e+00 -3.77480924e-01 2.43463978e-01
-1.18106708e-01 5.15359759e-01 -1.93031996e-01 7.02710927e-01
3.15521471e-02 4.34852123e-01 -3.47037405e-01 -3.15263689e-01
8.10330138e-02 1.65074563e-03 -4.62707311e-01 2.94963926e-01
5.33687651e-01 9.45216000e-01 -8.78552258e-01 -8.90101492e-01
-6.29282296e-02 3.41598660e-01 -6.92838803e-02 -1.28085734e-02
-1.99031569e-02 2.09225342e-01 -7.06561983e-01 1.02305877e+00
6.82752788e-01 -1.08842589e-01 3.04903060e-01 -1.35870188e-01
-5.42004228e-01 3.27558160e-01 -1.44984484e+00 1.73558867e+00
-4.59928006e-01 6.64115489e-01 -6.68811798e-02 -7.39715278e-01
7.61976242e-01 2.62433469e-01 2.56865591e-01 -6.81994677e-01
-8.44943598e-02 5.42349696e-01 -3.19337696e-01 -6.63808465e-01
7.52065301e-01 2.55227029e-01 -4.08012122e-01 1.96322545e-01
5.54863848e-02 2.64433563e-01 6.02511883e-01 4.08391654e-01
1.50511456e+00 3.25989306e-01 4.52566862e-01 2.36884370e-01
4.44012076e-01 3.55188012e-01 5.64678907e-01 1.02167225e+00
-3.63288194e-01 3.64241242e-01 3.99094582e-01 -2.51879275e-01
-1.08583713e+00 -7.37323761e-01 -3.28361422e-01 9.42001224e-01
4.82442323e-03 -5.83364785e-01 -8.22073340e-01 -1.45761919e+00
-1.08741328e-01 8.23958218e-01 -1.86651617e-01 4.60437238e-01
-1.01725495e+00 -3.50817174e-01 1.06142867e+00 6.75150692e-01
3.98375183e-01 -1.31855941e+00 -1.36882916e-01 7.46913031e-02
1.79807052e-01 -1.58034146e+00 -3.12266976e-01 1.09829172e-01
-8.55765224e-01 -1.14533317e+00 -6.23984337e-01 -1.03290093e+00
8.04124296e-01 6.66072816e-02 7.57583380e-01 -1.02353908e-01
-3.90893251e-01 3.43852580e-01 -3.97082657e-01 -4.66190606e-01
-2.54150063e-01 6.41455710e-01 -3.84289265e-01 -1.72800839e-01
4.14326042e-01 -1.53186508e-02 -2.66834825e-01 1.79481402e-01
-9.69903409e-01 -3.60440373e-01 1.20783329e+00 5.23627400e-01
3.92760575e-01 2.90560454e-01 4.46130306e-01 -1.60276639e+00
6.22036994e-01 -1.54273078e-01 -5.23688972e-01 6.65074646e-01
-5.67050874e-01 2.46231139e-01 7.63100684e-01 -6.34441748e-02
-1.71573257e+00 3.15928906e-01 -2.17736855e-01 1.51730299e-01
-8.68159235e-01 5.19119680e-01 -4.18582737e-01 -4.51661833e-02
4.86972660e-01 -2.32636794e-01 -1.02265668e+00 -8.38098764e-01
4.22473788e-01 9.80325639e-01 4.50255245e-01 -4.29264545e-01
8.58477712e-01 5.84860981e-01 4.40360457e-02 -9.99823987e-01
-1.25810397e+00 -9.56242085e-01 -1.15742934e+00 9.67115611e-02
8.82324278e-01 -5.55001318e-01 -1.76781923e-01 2.55450517e-01
-1.41459668e+00 5.48538752e-02 -3.19300056e-01 4.82767791e-01
4.53394689e-02 7.46983171e-01 -6.44281507e-01 -1.01222944e+00
-1.88740045e-01 -6.21157706e-01 1.24272907e+00 2.60062903e-01
-7.17839822e-02 -8.07558894e-01 4.60940748e-02 7.42391050e-01
-2.59020552e-03 1.28084078e-01 6.46114230e-01 -1.11995173e+00
-7.42293358e-01 -6.88811660e-01 -3.50446194e-01 2.28709906e-01
2.98140883e-01 -5.78974746e-02 -9.27198410e-01 6.30590469e-02
-3.85801345e-01 -1.14179775e-01 9.21286583e-01 -1.71485260e-01
7.73313403e-01 -4.95725088e-02 -5.91260374e-01 6.92597926e-02
1.17441094e+00 2.59082586e-01 6.64732158e-01 7.39070058e-01
1.05353487e+00 5.30604959e-01 9.10694063e-01 3.12238544e-01
2.45701402e-01 4.30323422e-01 -1.49151608e-02 -3.47369403e-01
-3.13692629e-01 -2.80518711e-01 2.17072695e-01 5.83373368e-01
-1.44474609e-02 -3.81267935e-01 -1.19480443e+00 7.73807228e-01
-1.65006244e+00 -1.02387834e+00 -3.53243500e-01 1.87168181e+00
8.17850351e-01 4.36954170e-01 -3.37326437e-01 1.35305347e-02
8.33591819e-01 1.56436443e-01 -3.45893145e-01 -2.66957760e-01
-2.35654131e-01 4.59942549e-01 7.93698072e-01 1.29758433e-01
-1.55431712e+00 1.04075241e+00 5.79670238e+00 1.00112092e+00
-7.13273048e-01 7.96349347e-02 2.32856914e-01 3.41881275e-01
6.43914416e-02 3.39794308e-01 -1.35135341e+00 1.55296192e-01
9.50041890e-01 2.91657060e-01 -2.01030806e-01 1.10582399e+00
2.14124992e-01 3.12342942e-02 -8.81569743e-01 5.44540703e-01
3.12504768e-01 -1.17566860e+00 -9.43000317e-02 1.72283500e-01
9.54275012e-01 -1.77722603e-01 -2.73843616e-01 2.99287945e-01
1.91972688e-01 -6.22242808e-01 5.14531016e-01 6.95952415e-01
4.52228695e-01 -7.51726270e-01 9.06468332e-01 2.85954773e-01
-1.09207428e+00 -8.77719298e-02 -3.92827913e-02 3.60440135e-01
4.55299437e-01 7.46522844e-01 -9.99760807e-01 7.10313916e-01
4.50128704e-01 5.39956152e-01 -8.57703507e-01 1.32919407e+00
-5.61082244e-01 9.32146311e-01 -1.43646881e-01 -3.13474936e-03
1.75176620e-01 2.23220289e-01 4.92024213e-01 1.60842919e+00
7.01318681e-02 -1.80171192e-01 2.34037638e-01 4.60581064e-01
-6.33306026e-01 5.08720875e-01 -6.75750434e-01 -5.29948711e-01
4.21932638e-01 1.34019840e+00 -1.10099101e+00 -2.30165586e-01
-6.54169321e-01 1.13930690e+00 5.62323689e-01 3.31813127e-01
-5.55308282e-01 -1.00412428e+00 -1.70649305e-01 1.67294934e-01
7.36717820e-01 -5.62395453e-01 -4.26294148e-01 -1.13986576e+00
3.99632156e-01 -5.84074080e-01 6.42340899e-01 -4.60948467e-01
-1.25978732e+00 4.12530601e-01 -1.18424147e-01 -1.12847638e+00
1.08892255e-01 -7.89114714e-01 -4.47513163e-01 4.68745708e-01
-1.81512833e+00 -1.42101502e+00 2.18739435e-01 1.10357143e-01
5.71871221e-01 -5.15547954e-02 6.73964202e-01 5.28392792e-01
-7.74998307e-01 5.47775686e-01 3.91198039e-01 6.08647287e-01
1.11686206e+00 -1.39051986e+00 5.04568636e-01 1.11378860e+00
4.43017662e-01 1.02966952e+00 2.52284199e-01 -9.63794768e-01
-1.24437594e+00 -1.00191033e+00 1.54935980e+00 -5.97605407e-01
8.12307060e-01 -4.89965618e-01 -8.65467191e-01 5.94540358e-01
2.80520260e-01 -3.69713679e-02 8.76701832e-01 2.72031158e-01
-4.39920843e-01 4.70942259e-02 -9.63523746e-01 4.10594821e-01
9.50593770e-01 -7.11680889e-01 -6.51525378e-01 4.22811061e-01
4.21660423e-01 -3.99170965e-01 -1.03773749e+00 2.04658613e-01
2.66514301e-01 -5.32109022e-01 6.99871957e-01 -5.69585383e-01
7.14322478e-02 -5.78675091e-01 2.46652737e-01 -5.78185260e-01
9.27995965e-02 -3.83149326e-01 -1.14172220e-01 1.82855785e+00
7.43529856e-01 -2.82847226e-01 7.94715703e-01 7.07062483e-01
-2.31805131e-01 -3.82040679e-01 -8.48185718e-01 -9.07786071e-01
-1.55083403e-01 -3.57735276e-01 1.89489499e-01 1.11627805e+00
-8.32858868e-03 4.14921403e-01 -3.17719072e-01 5.10899782e-01
4.47484046e-01 2.49295726e-01 5.22181332e-01 -1.19835460e+00
1.53565193e-02 1.68088134e-02 -3.17234695e-01 -9.33045387e-01
4.54362571e-01 -7.36371279e-01 1.20871738e-02 -1.79597628e+00
2.11391985e-01 -3.14330101e-01 -2.86446154e-01 7.68209875e-01
-5.06081916e-02 1.37124851e-01 1.72995716e-01 2.99144626e-01
-1.19458199e+00 5.99135347e-02 8.55882466e-01 -1.56322688e-01
-2.76874125e-01 -6.04276508e-02 -5.56438446e-01 6.97244227e-01
8.69897246e-01 -7.61374593e-01 1.19891986e-01 -2.96774983e-01
2.81215250e-01 -4.60060507e-01 -4.00221236e-02 -7.65101552e-01
5.16022205e-01 -7.71595836e-02 3.30357015e-01 -8.14374268e-01
-2.83301681e-01 -7.54376709e-01 -5.84743559e-01 -4.79776599e-02
-2.72158831e-01 -2.27440745e-01 2.82544941e-01 5.65749347e-01
-2.96851248e-01 -7.11793244e-01 2.38917843e-01 -2.69201188e-03
-1.11403644e+00 -2.61056036e-01 -3.22680235e-01 1.17746435e-01
9.32615459e-01 -1.79448158e-01 -4.37462389e-01 2.56090254e-01
-4.90760684e-01 -1.64817665e-02 3.83198261e-01 3.79556388e-01
3.46322358e-01 -9.71807241e-01 -4.09535825e-01 -3.77006114e-01
3.12412590e-01 -2.68106051e-02 4.07852717e-02 5.09335756e-01
-4.84741211e-01 9.63338733e-01 2.17110381e-01 -2.98473202e-02
-1.27063048e+00 5.40803075e-01 -2.01095536e-01 -7.49026477e-01
-4.21328634e-01 3.39459717e-01 -2.75849879e-01 -6.31435752e-01
2.37122983e-01 -5.89230508e-02 -5.85951328e-01 7.63897002e-02
3.67271245e-01 6.78395212e-01 3.77529174e-01 -6.77366734e-01
-3.94048452e-01 3.17436904e-01 -5.52423000e-01 1.09999642e-01
1.35971701e+00 1.19025502e-02 -8.64195004e-02 2.82606691e-01
9.88079250e-01 7.69886613e-01 -6.96675181e-01 -5.92668541e-03
9.00660813e-01 -2.63619661e-01 -2.42105201e-02 -9.15476620e-01
-5.49458385e-01 6.27748847e-01 5.21403432e-01 6.15353659e-02
8.86371911e-01 1.68459818e-01 7.97608078e-01 1.14044619e+00
3.06216240e-01 -1.45352793e+00 -1.65950701e-01 5.57761729e-01
3.68726492e-01 -1.27167952e+00 1.20233588e-01 -6.63602233e-01
-5.05774319e-01 1.07043624e+00 4.98095661e-01 2.41311997e-01
2.81625777e-01 3.12670469e-01 1.84824511e-01 -2.41770729e-01
-2.00927198e-01 -4.80974913e-01 4.75254983e-01 3.54670793e-01
7.65150368e-01 -4.53606933e-01 -7.60872662e-01 4.46906239e-01
3.71415466e-01 9.29311290e-02 4.39376235e-01 1.69491518e+00
-2.54867911e-01 -1.65809274e+00 -3.49841058e-01 2.34919235e-01
-9.93341148e-01 -3.22212815e-01 -7.17418373e-01 9.86333311e-01
1.96630388e-01 9.31483388e-01 -3.43725055e-01 4.00160521e-01
8.21232557e-01 2.63553232e-01 2.04267114e-01 -8.31370413e-01
-6.40974164e-01 8.77413452e-02 4.69734609e-01 -3.75614285e-01
-6.16509438e-01 -8.62995505e-01 -1.63627875e+00 4.49350506e-01
-6.14370584e-01 2.29939714e-01 8.26351047e-01 1.01877761e+00
5.54692447e-01 3.25093925e-01 4.27071482e-01 -2.71144390e-01
-3.71291310e-01 -7.32599497e-01 -6.80209100e-01 5.34567595e-01
4.15905565e-02 -3.16433370e-01 -1.42198548e-01 2.89011776e-01] | [9.637215614318848, 9.437087059020996] |
bb8a751b-5e9a-4f96-8ce2-53d9f4b21d17 | generalizing-gaze-estimation-with-rotation | null | null | http://openaccess.thecvf.com//content/CVPR2022/html/Bao_Generalizing_Gaze_Estimation_With_Rotation_Consistency_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Bao_Generalizing_Gaze_Estimation_With_Rotation_Consistency_CVPR_2022_paper.pdf | Generalizing Gaze Estimation With Rotation Consistency | Recent advances of deep learning-based approaches have achieved remarkable performance on appearance-based gaze estimation. However, due to the shortage of target domain data and absence of target labels, generalizing gaze estimation algorithm to unseen environments is still challenging. In this paper, we discover the rotation-consistency property in gaze estimation and introduce the 'sub-label' for unsupervised domain adaptation. Consequently, we propose the Rotation-enhanced Unsupervised Domain Adaptation (RUDA) for gaze estimation. First, we rotate the original images with different angles for training. Then we conduct domain adaptation under the constraint of rotation consistency. The target domain images are assigned with sub-labels, derived from relative rotation angles rather than untouchable real labels. With such sub-labels, we propose a novel distribution loss that facilitates the domain adaptation. We evaluate the RUDA framework on four cross-domain gaze estimation tasks. Experimental results demonstrate that it improves the performance over the baselines with gains ranging from 12.2% to 30.5%. Our framework has the potential to be used in other computer vision tasks with physical constraints. | ['Feng Lu', 'Haofei Wang', 'Yunfei Liu', 'Yiwei Bao'] | 2022-01-01 | null | null | null | cvpr-2022-1 | ['gaze-estimation'] | ['computer-vision'] | [ 2.88085341e-01 -9.74232629e-02 -3.55392247e-01 -6.10039473e-01
-3.99148107e-01 -4.46608126e-01 3.69560927e-01 -6.32516265e-01
-4.11919981e-01 7.01007783e-01 -9.56231728e-02 -4.74063568e-02
1.34395016e-02 1.82261601e-01 -6.49533331e-01 -9.46435332e-01
3.43918115e-01 5.41954860e-02 1.49033010e-01 -4.11020517e-02
4.68246609e-01 1.82186723e-01 -1.80406022e+00 -3.16645443e-01
1.13142145e+00 9.72046018e-01 1.64759889e-01 2.35068783e-01
3.19839209e-01 4.72536683e-01 -5.83354294e-01 -8.54665935e-02
1.94050387e-01 -4.94297624e-01 -6.98316574e-01 1.37587562e-01
8.46977413e-01 -5.83506107e-01 6.97643608e-02 1.08162498e+00
5.40309966e-01 3.29999804e-01 8.84188414e-01 -1.54552042e+00
-8.72784495e-01 -1.90917119e-01 -1.33110595e+00 1.37105241e-01
3.41616720e-01 5.98271377e-02 7.53282249e-01 -8.48108768e-01
7.06561923e-01 1.17858994e+00 2.04110399e-01 1.06356084e+00
-1.14896941e+00 -1.11316288e+00 5.38877606e-01 5.00799596e-01
-1.50775123e+00 -6.85286760e-01 9.50554907e-01 -4.00597513e-01
4.94524509e-01 -8.12230855e-02 9.10131335e-02 1.38241315e+00
-3.19352567e-01 8.27254951e-01 1.42179787e+00 -5.77341974e-01
5.91005459e-02 1.82392985e-01 1.39676780e-01 4.71000761e-01
2.14923412e-01 1.16740182e-01 -8.89009953e-01 2.42443413e-01
6.42609596e-01 -1.62689298e-01 -4.31021154e-01 -7.79324293e-01
-1.13316762e+00 6.24212682e-01 6.48519695e-01 -2.43695736e-01
8.28567054e-03 -1.36695772e-01 -1.86911575e-03 1.58385664e-01
7.20719099e-01 3.52497786e-01 -3.87237102e-01 2.66468246e-02
-7.24625885e-01 8.21500272e-02 3.35151881e-01 1.29201722e+00
7.52480030e-01 -1.75848410e-01 -6.78978115e-02 1.06294036e+00
6.75418615e-01 7.83467591e-01 5.04449248e-01 -9.48274910e-01
2.64838368e-01 3.91902775e-01 3.34819496e-01 -8.81465077e-01
-4.87842321e-01 -1.96064144e-01 -5.56315362e-01 3.80699158e-01
6.95554554e-01 -2.15400875e-01 -1.12449896e+00 2.10969210e+00
6.42701030e-01 2.18830347e-01 -1.29479349e-01 1.32368970e+00
6.81802988e-01 3.05483222e-01 1.34387508e-01 -3.59230936e-01
1.31156886e+00 -1.10581601e+00 -8.77621770e-01 -2.43809611e-01
4.31530625e-01 -7.08160520e-01 1.23168290e+00 4.83568221e-01
-7.25321591e-01 -5.26856422e-01 -9.70400810e-01 -1.73132077e-01
-7.66900107e-02 2.95114219e-01 3.79278928e-01 6.49224758e-01
-1.12396657e+00 -1.00212254e-01 -7.24323869e-01 -5.76134920e-01
5.75140178e-01 5.91783762e-01 -2.81780452e-01 2.09028441e-02
-8.81376028e-01 7.92475820e-01 1.18827380e-01 3.95806581e-02
-6.53090715e-01 -4.36566830e-01 -1.01160145e+00 -3.02605987e-01
4.88342673e-01 -5.32902539e-01 1.38692796e+00 -1.30950475e+00
-1.79574561e+00 1.12081802e+00 -8.28428686e-01 3.25801820e-02
2.42216408e-01 -3.61388683e-01 -4.81250376e-01 1.95895564e-02
2.39886008e-02 1.06872439e+00 1.37866378e+00 -1.36620855e+00
-7.25835443e-01 -5.46230733e-01 -1.88625120e-02 5.19608796e-01
-5.76043487e-01 1.84589520e-01 -5.66268981e-01 -4.08807248e-01
1.15956865e-01 -1.30296803e+00 3.97447109e-01 2.19636619e-01
-2.63790786e-01 -5.74028850e-01 8.98081303e-01 -3.98131371e-01
1.17311549e+00 -2.32346058e+00 2.62013406e-01 2.60138735e-02
5.71581662e-01 1.62145153e-01 -1.35167107e-01 -3.49690437e-01
-1.02704518e-01 -2.28742406e-01 -1.10125780e-01 -5.96947730e-01
3.34043019e-02 -1.33238599e-01 -1.32914364e-01 5.75389683e-01
4.18321222e-01 7.13253975e-01 -8.56752157e-01 -6.14473343e-01
-1.05154347e-02 3.91184032e-01 -4.48456347e-01 2.98870057e-01
-1.18381217e-01 8.29255879e-01 -3.46410334e-01 6.64366305e-01
1.04734838e+00 -5.81734419e-01 1.58448234e-01 -8.30615610e-02
1.20933078e-01 1.90580457e-01 -7.39011049e-01 1.82593131e+00
-1.40542567e-01 9.85883713e-01 -1.35642707e-01 -5.89010298e-01
9.38369453e-01 -7.27556050e-02 1.35027424e-01 -9.13473070e-01
2.09700882e-01 8.02566856e-03 1.35518447e-01 -5.50309241e-01
5.90099871e-01 9.88522917e-02 6.54758811e-02 6.86439037e-01
1.62145153e-01 2.90676713e-01 -1.62905723e-01 -7.14593604e-02
4.63462681e-01 6.87144518e-01 2.87859082e-01 -4.53781448e-02
6.19659722e-01 -3.20799023e-01 5.18580675e-01 2.22555697e-01
-8.33146930e-01 6.49613261e-01 3.40957582e-01 -1.27878889e-01
-8.36100936e-01 -9.15571749e-01 -2.19257474e-01 1.78624499e+00
5.91147482e-01 7.13731945e-02 -8.51260602e-01 -9.31612372e-01
-1.41788915e-01 4.03999120e-01 -8.28694999e-01 -2.17128485e-01
-4.90574747e-01 -5.29074430e-01 2.61104047e-01 4.92277682e-01
4.83458221e-01 -9.26247656e-01 -3.56861413e-01 -5.66623271e-01
-2.73493111e-01 -1.01315093e+00 -7.64866292e-01 -2.25416332e-01
-6.48714066e-01 -8.69246125e-01 -1.09452748e+00 -9.37347293e-01
1.04241514e+00 7.15094566e-01 8.81356895e-01 -1.95135862e-01
3.57996702e-01 2.06153944e-01 -3.98915917e-01 -5.25358737e-01
1.85891807e-01 2.72922426e-01 3.90590847e-01 2.15357572e-01
1.02231228e+00 -3.42262477e-01 -8.02380264e-01 6.39809251e-01
-5.02179027e-01 -1.23695321e-01 5.23424745e-01 8.49907935e-01
5.14768958e-01 -6.37573183e-01 6.18087828e-01 -7.18258321e-01
6.02752626e-01 -4.13596869e-01 -6.66328788e-01 2.21012428e-01
-8.74758661e-01 2.79472768e-01 4.21002172e-02 -7.55760789e-01
-1.45741630e+00 2.41260231e-02 4.80607778e-01 -6.07004344e-01
-4.91068661e-01 3.65995355e-02 -1.89278334e-01 -2.81531215e-01
8.18780780e-01 1.88239086e-02 1.66671932e-01 -4.63555276e-01
2.90223688e-01 9.08363402e-01 3.73649180e-01 -4.60655957e-01
8.27600241e-01 4.26659584e-01 -4.13677879e-02 -6.36199415e-01
-9.61174488e-01 -5.72882712e-01 -6.78827345e-01 -2.46675685e-01
8.39810431e-01 -1.02609313e+00 -8.08921993e-01 7.25968719e-01
-9.78590727e-01 -2.97882140e-01 3.85864705e-01 5.17884493e-01
-4.75731432e-01 4.25340950e-01 4.47067879e-02 -7.50867248e-01
-1.61311522e-01 -1.10630059e+00 1.23166418e+00 7.30569243e-01
-2.08798796e-01 -8.49287212e-01 1.99313417e-01 3.04542184e-01
2.01124892e-01 -5.51772900e-02 3.61079782e-01 -4.48792309e-01
-5.66715181e-01 3.11221719e-01 -5.87754965e-01 1.87669054e-01
2.35076278e-01 -1.66092739e-01 -1.29918325e+00 -4.33445901e-01
-1.83674872e-01 -6.07917428e-01 7.23023951e-01 4.29511815e-01
1.14698255e+00 6.24253675e-02 -5.59612334e-01 9.71566975e-01
9.31977332e-01 6.21088855e-02 4.91610587e-01 4.87345934e-01
8.47924650e-01 7.91354597e-01 1.01598251e+00 2.92135984e-01
5.94328642e-01 8.35806489e-01 3.68029684e-01 1.02494769e-02
-1.92228541e-01 -1.98577702e-01 3.10527503e-01 4.83697027e-01
-3.25343996e-01 -2.35722452e-01 -9.03470397e-01 5.38291335e-01
-1.70663464e+00 -5.36577642e-01 6.35782853e-02 2.25627708e+00
1.01125395e+00 -1.07992277e-01 4.09643143e-01 -2.77699172e-01
8.61626983e-01 9.08802375e-02 -1.09768856e+00 -1.80179626e-01
1.38293952e-01 1.06313840e-01 3.42887580e-01 1.96237713e-01
-1.24401867e+00 1.09615695e+00 5.82637024e+00 5.62779129e-01
-1.40527725e+00 1.95245162e-01 1.81468427e-01 -1.87036410e-01
6.93240389e-02 -2.84732223e-01 -1.07772970e+00 5.97507238e-01
6.51952088e-01 1.31976783e-01 5.15690386e-01 8.46270919e-01
7.11362883e-02 -2.08089843e-01 -1.02905405e+00 1.23917472e+00
5.23195386e-01 -2.86460042e-01 -4.15476263e-01 1.39634788e-01
8.80766869e-01 -1.11211494e-01 6.55301154e-01 2.11815566e-01
-1.24590956e-02 -9.07850921e-01 5.39918005e-01 4.41342980e-01
1.34485137e+00 -5.69931924e-01 3.60510200e-01 1.43381387e-01
-7.82703459e-01 -4.02032547e-02 -3.04598689e-01 -1.02595380e-02
-1.52670711e-01 -1.44680172e-01 -9.32241380e-01 2.21516505e-01
7.82556772e-01 1.08312464e+00 -7.25932896e-01 8.88907552e-01
-5.35881519e-01 5.30231595e-01 -2.94704884e-01 -1.49280414e-01
-1.63583294e-01 -8.40908960e-02 4.18981671e-01 6.87926412e-01
1.01541966e-01 -1.44959614e-01 -4.63901013e-01 8.11251283e-01
-2.79471010e-01 -5.65060973e-02 -4.54329580e-01 3.16522717e-01
7.22937226e-01 1.05524302e+00 -2.29022980e-01 8.56562927e-02
-4.10149664e-01 1.12848043e+00 5.84208727e-01 8.57171953e-01
-9.20685530e-01 -4.48301733e-01 8.74287844e-01 -9.07834098e-02
3.92479569e-01 -1.39075905e-01 -1.99104413e-01 -1.27780068e+00
1.19461134e-01 -9.31799531e-01 1.25586182e-01 -1.08290088e+00
-1.26491177e+00 4.50312644e-01 2.03836069e-01 -1.57905066e+00
-4.36411262e-01 -7.79425681e-01 -1.96349055e-01 1.01437116e+00
-1.98060501e+00 -1.33233643e+00 -7.06293046e-01 7.74122238e-01
2.68360049e-01 -2.34312072e-01 6.39088869e-01 1.21740744e-01
-7.43628860e-01 1.27325964e+00 2.79096514e-01 9.03896093e-02
1.55536485e+00 -1.25868034e+00 1.73593089e-01 6.91721380e-01
-7.29327351e-02 8.77896309e-01 5.45860887e-01 -1.94839776e-01
-8.92638385e-01 -7.88736641e-01 8.03656757e-01 -7.36899972e-01
3.94913346e-01 -4.86834228e-01 -9.63853002e-01 7.59343207e-01
3.90107930e-01 -3.66231985e-02 7.51961231e-01 4.46838021e-01
-7.04581380e-01 -3.05182338e-01 -1.04974449e+00 5.07607281e-01
1.16393936e+00 -5.04024982e-01 -5.46812892e-01 2.33712923e-02
6.03242755e-01 -6.00867748e-01 -5.04974902e-01 3.07126611e-01
7.70501792e-01 -7.45984435e-01 7.89768815e-01 -6.08734667e-01
2.89528847e-01 -4.78997648e-01 3.32639635e-01 -1.15992725e+00
-1.99023396e-01 -6.60420597e-01 -3.53943169e-01 1.12060928e+00
2.34926343e-01 -7.57421792e-01 6.17236853e-01 5.52137554e-01
2.42050514e-01 -2.83918679e-01 -6.47822618e-01 -5.98850548e-01
8.73967111e-02 -7.06498697e-03 5.66044450e-01 1.05418468e+00
-1.03004277e-02 6.04211390e-01 -5.36574244e-01 1.93136960e-01
8.53877306e-01 9.66839045e-02 1.10880256e+00 -1.40446258e+00
5.31297475e-02 -3.65523547e-01 -2.53243744e-01 -1.59430408e+00
4.17617828e-01 -2.98702270e-01 2.46386617e-01 -8.07341158e-01
2.13112503e-01 -3.26368898e-01 -6.10340774e-01 4.53755975e-01
-4.84442770e-01 3.80940080e-01 5.33729419e-02 6.34489715e-01
-9.59141135e-01 4.89104480e-01 1.49389672e+00 2.44511310e-02
-3.36061090e-01 -5.31250276e-02 -8.09828937e-01 5.72129369e-01
7.73048341e-01 -3.96402985e-01 -5.74673653e-01 -6.46831453e-01
1.08510904e-01 -5.34876764e-01 2.24916577e-01 -7.56336629e-01
1.82448700e-01 -1.78152442e-01 3.55597764e-01 -4.92875099e-01
2.55018502e-01 -7.13120997e-01 -5.65466166e-01 -2.30521470e-01
-2.99297243e-01 -3.85985434e-01 2.35225827e-01 7.77936637e-01
-8.15677196e-02 2.42487527e-02 8.62461269e-01 6.10997379e-01
-1.03627694e+00 2.13334292e-01 1.80345520e-01 1.65950567e-01
1.10416448e+00 -4.25179183e-01 -4.38177824e-01 -4.30139840e-01
-6.32875443e-01 2.16330796e-01 7.87801266e-01 5.54254413e-01
5.15932381e-01 -1.24037564e+00 -4.21733230e-01 4.71795321e-01
6.60669386e-01 1.98230043e-01 1.31146282e-01 1.08853269e+00
-7.52830654e-02 4.83763129e-01 -5.24325311e-01 -1.00879538e+00
-1.40657115e+00 7.28574753e-01 2.10255533e-01 2.83777177e-01
-3.95835154e-02 1.08074200e+00 6.92877471e-01 -2.69254863e-01
3.39737207e-01 -2.26723149e-01 -6.33824527e-01 -8.11525136e-02
5.73057413e-01 1.73590168e-01 -2.89031923e-01 -9.43643034e-01
-4.60316628e-01 9.55032885e-01 -4.23077703e-01 9.78199244e-02
1.01425540e+00 -6.43291116e-01 1.43414021e-01 2.20352948e-01
1.13094223e+00 -1.16616227e-01 -1.69199562e+00 -6.03968561e-01
-9.56185460e-02 -6.21439755e-01 -1.09655671e-01 -8.64255071e-01
-9.37723100e-01 1.08898818e+00 9.34736490e-01 -3.08844537e-01
1.49007666e+00 4.02497798e-02 4.60432976e-01 2.92889953e-01
2.76830107e-01 -9.60365474e-01 1.90660775e-01 4.23235983e-01
6.04493558e-01 -1.80779636e+00 -3.98620069e-02 -1.90082431e-01
-8.97136450e-01 6.85246468e-01 9.84224498e-01 6.09619841e-02
5.06461620e-01 -3.52986902e-01 3.72732997e-01 -1.81982238e-02
-4.54445869e-01 -4.35727239e-01 7.71378636e-01 1.10645354e+00
4.02658284e-01 -9.06716958e-02 -1.83742896e-01 3.71225923e-01
8.57009441e-02 1.17265411e-01 2.15798572e-01 6.14402711e-01
-3.71488690e-01 -9.90270138e-01 -4.34160113e-01 8.87547061e-02
-2.08916292e-01 1.82286631e-02 -4.81345713e-01 8.51406395e-01
4.70346920e-02 9.85561728e-01 1.18208714e-01 -3.77395242e-01
8.63438994e-02 1.16794921e-01 6.80116594e-01 -5.43964028e-01
2.78105676e-01 1.76027179e-01 -2.84084857e-01 -4.55856949e-01
-8.73388231e-01 -7.09828258e-01 -9.64575291e-01 -1.75904587e-01
-5.52463531e-01 -2.27085635e-01 5.51964581e-01 8.39736104e-01
6.38561070e-01 1.52399257e-01 6.64051771e-01 -8.39130461e-01
-5.59633851e-01 -1.34707904e+00 -5.57098210e-01 5.85076332e-01
7.64321148e-01 -1.37909627e+00 -4.83917445e-01 4.31393415e-01] | [14.113128662109375, 0.025898484513163567] |
ac2f23e9-d8f9-4283-9a3e-b434544a853c | skip-connections-in-spiking-neural-networks | 2303.13563 | null | https://arxiv.org/abs/2303.13563v1 | https://arxiv.org/pdf/2303.13563v1.pdf | Skip Connections in Spiking Neural Networks: An Analysis of Their Effect on Network Training | Spiking neural networks (SNNs) have gained attention as a promising alternative to traditional artificial neural networks (ANNs) due to their potential for energy efficiency and their ability to model spiking behavior in biological systems. However, the training of SNNs is still a challenging problem, and new techniques are needed to improve their performance. In this paper, we study the impact of skip connections on SNNs and propose a hyperparameter optimization technique that adapts models from ANN to SNN. We demonstrate that optimizing the position, type, and number of skip connections can significantly improve the accuracy and efficiency of SNNs by enabling faster convergence and increasing information flow through the network. Our results show an average +8% accuracy increase on CIFAR-10-DVS and DVS128 Gesture datasets adaptation of multiple state-of-the-art models. | ['Younes Bouhadjar', 'Imane Hamzaoui', 'Amine Ziad Ounnoughene', 'Hadjer Benmeziane'] | 2023-03-23 | null | null | null | null | ['hyperparameter-optimization'] | ['methodology'] | [ 2.54979014e-01 -4.53353882e-01 4.37298603e-03 -1.57469139e-01
1.30142793e-01 -4.44941580e-01 3.87881130e-01 -9.73330587e-02
-9.93887663e-01 9.06988204e-01 -1.83209136e-01 -9.17776525e-02
1.49550125e-01 -5.77076614e-01 -8.32393110e-01 -1.00725496e+00
4.21893373e-02 1.95901424e-01 6.78883731e-01 -1.00232750e-01
3.17867696e-01 8.37723851e-01 -1.33839166e+00 3.39853354e-02
5.64638734e-01 1.06802011e+00 1.92954361e-01 5.32147706e-01
-6.05504736e-02 5.14571846e-01 -7.13722885e-01 1.11744806e-01
1.57443956e-01 -6.66474819e-01 -3.14337134e-01 -7.57631898e-01
-8.83794501e-02 1.90277010e-01 -5.28074741e-01 7.09046364e-01
8.41586590e-01 1.96551532e-01 5.25308788e-01 -1.36376977e+00
-1.98861450e-01 6.74671769e-01 -2.29488894e-01 5.92607558e-01
-5.56937516e-01 4.31753218e-01 3.98032635e-01 -3.36438507e-01
4.53603089e-01 8.99226606e-01 7.92073607e-01 1.13355923e+00
-1.43948936e+00 -1.24285269e+00 2.82111280e-02 3.18567246e-01
-1.31617606e+00 -5.90523958e-01 4.94107276e-01 -5.71807437e-02
1.49215984e+00 -2.54100651e-01 1.28648496e+00 1.34985340e+00
2.46008411e-01 6.60013437e-01 8.79683614e-01 -2.34096169e-01
9.07649994e-01 -4.59996253e-01 1.31526738e-01 3.54409248e-01
6.00620449e-01 6.08352274e-02 -8.88421774e-01 6.54468909e-02
1.18040562e+00 -2.25089714e-01 -9.90964472e-02 -1.15518458e-01
-1.07177782e+00 5.71278214e-01 8.17935944e-01 3.45523536e-01
-3.37013304e-01 8.88110578e-01 2.66299814e-01 -3.28614056e-01
-2.39238083e-01 7.88436890e-01 -3.16801339e-01 -4.19530571e-01
-7.87613451e-01 3.57662678e-01 7.00955212e-01 4.27296609e-01
2.35803857e-01 4.91981089e-01 -5.19427471e-02 8.07649136e-01
1.33901358e-01 5.64862192e-01 6.51770949e-01 -1.16511130e+00
1.21894134e-02 7.56417453e-01 -1.60748929e-01 -5.29129744e-01
-8.37347806e-01 -5.53141057e-01 -1.01705873e+00 4.03887004e-01
6.07192695e-01 -1.85763285e-01 -1.19390404e+00 1.91289270e+00
-2.23766997e-01 4.71166342e-01 1.03898369e-01 7.73991346e-01
5.94753087e-01 6.53577983e-01 3.18366855e-01 1.21478625e-01
9.29123461e-01 -6.81767523e-01 -5.91685355e-01 -5.41010380e-01
4.37046558e-01 -1.94229692e-01 6.85842395e-01 3.40446323e-01
-1.18801165e+00 -2.32758790e-01 -1.21065640e+00 1.47919908e-01
-4.56149995e-01 -1.14446290e-01 7.53338397e-01 5.41574240e-01
-1.09065604e+00 8.11067224e-01 -1.35869610e+00 -4.47393328e-01
8.95467877e-01 1.04924250e+00 1.15263492e-01 3.14751029e-01
-8.48899901e-01 9.04650092e-01 6.35936737e-01 -5.20904586e-02
-4.43480909e-01 -6.60808504e-01 -3.44668239e-01 2.75381237e-01
-1.42973617e-01 -5.93742847e-01 1.12137640e+00 -6.79834306e-01
-1.71953642e+00 4.07490760e-01 -1.58870742e-01 -8.59487116e-01
-7.71119595e-02 3.86414617e-01 -6.12160936e-02 -3.67011577e-02
-6.61754251e-01 1.33281314e+00 2.26850480e-01 -9.32724535e-01
-3.76526117e-01 -2.52703726e-01 -3.25667500e-01 7.75156394e-02
-5.79891741e-01 -2.40061522e-01 -2.77090997e-01 -7.18698263e-01
2.59960324e-01 -1.20584667e+00 -2.60861993e-01 3.60836953e-01
-7.95900896e-02 -9.12968442e-02 7.69979417e-01 -2.32569858e-01
9.24491227e-01 -2.05271244e+00 1.91428661e-01 3.29348505e-01
5.60806505e-02 6.97485983e-01 -2.04144672e-01 -1.66098878e-01
3.25298429e-01 1.17000356e-01 -4.46687967e-01 -1.44134657e-02
-3.83675575e-01 6.29479587e-01 9.60374549e-02 1.26401037e-01
1.39575690e-01 1.15579963e+00 -5.62384427e-01 -1.66459948e-01
5.99405915e-02 8.16084623e-01 -6.14786744e-01 -1.06969371e-01
-1.34658098e-01 6.13720238e-01 -1.77361667e-01 5.03654599e-01
2.40517199e-01 -5.31123757e-01 -3.82203050e-03 -9.86430570e-02
3.03475894e-02 1.92414120e-01 -9.30852592e-01 1.60052979e+00
-2.34682769e-01 8.57405066e-01 -2.69472301e-01 -9.67815042e-01
9.77241278e-01 6.78190738e-02 5.36309421e-01 -1.15591288e+00
4.99475747e-01 3.83039623e-01 6.20894372e-01 -4.44057062e-02
-2.00886995e-01 9.59677249e-02 3.03870648e-01 3.48506987e-01
2.40285099e-01 1.95037141e-01 3.71818900e-01 -1.15179025e-01
1.26361263e+00 -7.00208694e-02 7.60015845e-03 -3.04228634e-01
-1.35504929e-02 -1.07022278e-01 5.56096375e-01 5.94224036e-01
-1.79826438e-01 3.57149959e-01 2.67137945e-01 -4.35543865e-01
-1.08839452e+00 -9.52787876e-01 -1.97696373e-01 7.20625997e-01
2.33023316e-01 1.05706818e-01 -1.01827776e+00 7.50891194e-02
-1.19532794e-01 5.80322802e-01 -6.65378094e-01 -4.55841333e-01
-9.37964261e-01 -1.20290351e+00 1.07477307e+00 1.03475928e+00
7.82333255e-01 -1.42977405e+00 -1.31695580e+00 5.28996468e-01
-9.96222645e-02 -1.29158044e+00 -5.37583791e-02 9.20667470e-01
-1.25154626e+00 -8.18272352e-01 -6.93574071e-01 -6.90256894e-01
7.01293290e-01 -2.39780024e-01 7.02829063e-01 1.66046441e-01
-4.18457717e-01 -3.83535177e-01 -9.20382291e-02 -6.49573326e-01
-2.03093916e-01 3.64030093e-01 9.29501206e-02 -4.07903820e-01
4.34551626e-01 -7.55049407e-01 -8.06366146e-01 3.06497246e-01
-8.54232490e-01 2.95879573e-01 4.92478639e-01 7.86734521e-01
7.81727791e-01 -4.56013530e-01 6.49888456e-01 -4.06609923e-01
4.06243235e-01 -8.90936852e-02 -6.94898009e-01 9.53810886e-02
-7.17042625e-01 6.00096941e-01 7.45257080e-01 -7.77615428e-01
-5.85962594e-01 2.80775428e-01 -2.14804471e-01 -1.57728896e-01
7.18179867e-02 1.89621195e-01 1.83818415e-01 -7.56341398e-01
8.03060293e-01 1.49173617e-01 -8.08352083e-02 -2.15180308e-01
-3.56979460e-01 1.44898042e-01 5.89392006e-01 -2.82957345e-01
2.69067705e-01 4.35463190e-01 3.44673574e-01 -6.74117208e-01
-2.39754155e-01 -1.69921190e-01 -3.66964936e-01 -1.78513229e-01
7.95864284e-01 -4.22104150e-01 -1.03796744e+00 9.53353047e-01
-1.11742699e+00 -8.11556399e-01 -1.64659098e-01 4.73014206e-01
-3.70748818e-01 -3.35697234e-01 -6.79609597e-01 -6.49470091e-01
-5.35364807e-01 -9.61088479e-01 3.72107506e-01 7.93155491e-01
-3.57546866e-01 -8.82660210e-01 8.17950889e-02 -1.75138727e-01
1.03054500e+00 1.30523235e-01 9.64008570e-01 -6.14049792e-01
-4.00603116e-01 -1.40285771e-03 -3.05377126e-01 -1.34117529e-02
-6.04308359e-02 -6.62484095e-02 -1.04687715e+00 -2.54553944e-01
-4.62074786e-01 -3.32391053e-01 1.11609674e+00 9.53070939e-01
1.36649168e+00 -2.18802631e-01 -5.75354397e-01 1.03551602e+00
1.55559182e+00 5.59034884e-01 6.73028588e-01 5.12876451e-01
3.67151678e-01 -5.69345839e-02 -4.47704881e-01 2.96164572e-01
-8.30499679e-02 6.48698628e-01 6.67680860e-01 1.10777885e-01
-3.42243731e-01 1.12282932e-02 4.22407240e-02 6.57713234e-01
-3.58635753e-01 -4.79905248e-01 -1.03442156e+00 3.71840179e-01
-1.78399038e+00 -7.27461755e-01 2.19245344e-01 2.12284517e+00
8.09160769e-01 2.54334122e-01 1.93392739e-01 3.59566569e-01
4.97262448e-01 -2.80651033e-01 -1.27084768e+00 -2.47600675e-01
-5.93563259e-01 5.02789855e-01 7.82123446e-01 4.51539317e-03
-6.46495283e-01 9.15050805e-01 6.89484549e+00 5.82167327e-01
-1.51177347e+00 -2.30330542e-01 5.41534126e-01 -5.89902103e-01
1.18249089e-01 -5.37068009e-01 -1.09217870e+00 6.14433765e-01
1.24802625e+00 1.03344142e-01 9.70908523e-01 3.33182126e-01
1.09394342e-01 -2.09022701e-01 -8.92574251e-01 1.14862013e+00
-1.72005847e-01 -1.72596574e+00 -8.50211903e-02 -7.92066043e-04
7.37422228e-01 4.17182922e-01 -5.72424568e-02 2.70375833e-02
3.43554467e-01 -1.26285672e+00 4.92288589e-01 3.98211509e-01
4.94967312e-01 -7.47464955e-01 8.40128362e-01 2.16509581e-01
-1.02710474e+00 -2.86035180e-01 -3.74848217e-01 -3.83327417e-02
5.51209413e-02 1.59693807e-01 -4.59677339e-01 -4.34853196e-01
1.03108311e+00 5.07425189e-01 -6.42070532e-01 1.61630702e+00
4.71479595e-02 7.65421748e-01 -8.61859858e-01 -7.79947162e-01
1.97483331e-01 1.18439503e-01 1.01824611e-01 1.12156248e+00
4.45260912e-01 2.92513251e-01 -4.08320278e-01 9.37978506e-01
-3.71305615e-01 -2.76617259e-01 -2.39717424e-01 -2.25167990e-01
8.03483665e-01 9.44091976e-01 -1.27262783e+00 -7.29296803e-02
8.85328054e-02 5.98149240e-01 3.13887417e-01 5.19347608e-01
-7.59945452e-01 -3.52378219e-01 5.37024260e-01 5.91538884e-02
2.91285723e-01 -2.92844504e-01 -8.74422133e-01 -6.61443710e-01
-1.23629406e-01 -5.05095541e-01 2.63449475e-02 -6.54391885e-01
-7.90137768e-01 6.57702029e-01 -3.79793525e-01 -6.98844671e-01
-1.52212128e-01 -8.15009475e-01 -5.11840522e-01 5.11130512e-01
-1.35142243e+00 -6.66451871e-01 -4.03162926e-01 5.41531861e-01
2.70063549e-01 -8.24462622e-02 8.98686171e-01 1.48650721e-01
-5.89669287e-01 6.84587181e-01 1.77282035e-01 1.69995487e-01
2.66529739e-01 -7.17034757e-01 6.15441859e-01 6.90357208e-01
1.35085285e-01 4.08888787e-01 5.84053338e-01 -2.72003919e-01
-1.19572544e+00 -9.51236367e-01 4.79930282e-01 1.61025554e-01
3.26633602e-01 -3.97548705e-01 -1.04332590e+00 3.06169868e-01
1.57007370e-02 8.39427486e-02 5.72070479e-01 -3.25912476e-01
-1.97201163e-01 -1.51082754e-01 -1.11112571e+00 8.24926794e-01
1.21819103e+00 1.33688539e-01 -4.47851664e-04 -1.77760974e-01
2.29151458e-01 -3.50418299e-01 -5.32244146e-01 5.45830548e-01
8.45644712e-01 -6.43984854e-01 9.40950453e-01 -4.66802329e-01
9.88011900e-03 -1.00089110e-01 -2.33259927e-02 -1.44560218e+00
-3.15083414e-01 -3.35315794e-01 -3.98526043e-01 6.43885434e-01
5.09370863e-01 -6.51012599e-01 1.21982622e+00 7.06222355e-01
-4.57049869e-02 -9.78674471e-01 -1.20125890e+00 -9.80708838e-01
1.05154164e-01 -9.68806893e-02 2.49458805e-01 3.76450181e-01
5.38795628e-02 1.32932216e-01 7.72864670e-02 -1.36175439e-01
6.06346369e-01 -5.12677491e-01 1.72947183e-01 -1.43455267e+00
-7.94938132e-02 -1.00224710e+00 -5.70032835e-01 -7.96772957e-01
3.11083179e-02 -7.17154503e-01 1.09899037e-01 -1.67802703e+00
3.41828205e-02 -5.64142287e-01 -6.69008136e-01 8.87071252e-01
1.10358626e-01 5.24811029e-01 1.28228113e-01 1.44958094e-01
-4.92648482e-01 4.07618791e-01 8.53661776e-01 -4.64581922e-02
-3.09559524e-01 -7.97681138e-02 -3.39922607e-01 6.61935091e-01
1.31312644e+00 -6.75275981e-01 -2.84351349e-01 -6.64011419e-01
2.26775080e-01 -2.76528865e-01 4.19954300e-01 -1.73604620e+00
7.60346651e-01 -1.22079790e-01 8.10058594e-01 -1.62377462e-01
6.03302002e-01 -6.43476963e-01 9.31041539e-02 9.95957613e-01
-5.28971136e-01 3.07918079e-02 7.52211809e-01 4.46937650e-01
2.31669903e-01 -1.69840440e-01 1.15875041e+00 -4.93206922e-03
-6.21470094e-01 2.21250236e-01 -7.28847504e-01 9.48631093e-02
8.82677972e-01 -5.49612761e-01 -4.66597378e-01 -6.26513362e-02
-4.30557251e-01 1.07683167e-01 3.53681803e-01 2.35604241e-01
5.87519646e-01 -1.08552098e+00 -2.93928057e-01 3.23617429e-01
-1.99979022e-01 -3.49599347e-02 -6.82393881e-03 5.48638403e-01
-5.79001188e-01 4.95352745e-01 -8.53809178e-01 -6.14180684e-01
-1.05576408e+00 9.92624834e-02 6.63682818e-01 -9.91550907e-02
-3.55737150e-01 1.23781300e+00 -3.14267367e-01 -7.10245818e-02
7.19294846e-01 -3.44644129e-01 -2.03766659e-01 -5.31166852e-01
4.00123924e-01 4.68119293e-01 1.15860008e-01 -1.38973415e-01
-5.41798770e-01 5.35868108e-01 1.40908867e-01 -1.88920572e-01
1.53749073e+00 3.03189397e-01 1.62250429e-01 3.57124865e-01
7.70358086e-01 -9.61513519e-01 -1.58249426e+00 -1.37945097e-02
2.72930190e-02 2.38350689e-01 1.95826516e-01 -1.21110809e+00
-1.40348244e+00 1.11033237e+00 9.74035680e-01 -3.25884670e-01
1.10219491e+00 -2.41527855e-01 1.05533373e+00 7.95576811e-01
3.92575681e-01 -1.21124506e+00 1.90927133e-01 5.72635293e-01
4.17840451e-01 -9.83804405e-01 -3.88710260e-01 1.28207922e-01
-3.27328235e-01 1.25815690e+00 9.01858151e-01 -2.96161294e-01
5.09443223e-01 8.75004411e-01 -8.36769268e-02 8.02247673e-02
-7.51157105e-01 2.14576676e-01 -7.85032734e-02 6.20696545e-01
2.50373214e-01 -2.37983450e-01 -2.72952882e-03 4.63778168e-01
-4.65094335e-02 4.60896045e-01 3.37425828e-01 9.83416378e-01
-6.34359896e-01 -9.36633229e-01 -1.31487949e-02 6.68798387e-01
-3.37646395e-01 -1.96089029e-01 -1.89962715e-01 5.08114040e-01
-2.00332671e-01 6.96106493e-01 2.02911422e-01 -3.56261164e-01
2.56643087e-01 2.28493974e-01 7.84108877e-01 -1.04043014e-01
-8.18228483e-01 -9.75572839e-02 -3.38054895e-01 -4.33297247e-01
-3.73149484e-01 -4.06149238e-01 -1.99302173e+00 -4.47925717e-01
-1.71578407e-01 -9.30468291e-02 1.21749532e+00 1.14359915e+00
6.25485182e-01 8.76276076e-01 -8.30815407e-04 -9.74862993e-01
-2.39736572e-01 -8.16224575e-01 -3.08220476e-01 -4.81946580e-02
3.33254784e-02 -7.05497324e-01 -2.44726956e-01 9.27463993e-02] | [8.212352752685547, 2.5348784923553467] |
4145d16f-263f-48e2-8fed-0e7375c038f2 | planning-landmark-based-goal-recognition | 2306.15362 | null | https://arxiv.org/abs/2306.15362v1 | https://arxiv.org/pdf/2306.15362v1.pdf | Planning Landmark Based Goal Recognition Revisited: Does Using Initial State Landmarks Make Sense? | Goal recognition is an important problem in many application domains (e.g., pervasive computing, intrusion detection, computer games, etc.). In many application scenarios, it is important that goal recognition algorithms can recognize goals of an observed agent as fast as possible. However, many early approaches in the area of Plan Recognition As Planning, require quite large amounts of computation time to calculate a solution. Mainly to address this issue, recently, Pereira et al. developed an approach that is based on planning landmarks and is much more computationally efficient than previous approaches. However, the approach, as proposed by Pereira et al., also uses trivial landmarks (i.e., facts that are part of the initial state and goal description are landmarks by definition). In this paper, we show that it does not provide any benefit to use landmarks that are part of the initial state in a planning landmark based goal recognition approach. The empirical results show that omitting initial state landmarks for goal recognition improves goal recognition performance. | ['Heiner Stuckenschmidt', 'Christian Bartelt', 'Lea Cohausz', 'Nils Wilken'] | 2023-06-27 | null | null | null | null | ['intrusion-detection'] | ['miscellaneous'] | [ 2.48308927e-01 2.36047998e-01 -5.48410267e-02 -1.68554246e-01
-4.25008118e-01 -4.41619068e-01 7.17711091e-01 5.16047299e-01
-4.02450114e-01 7.80716836e-01 1.09755574e-02 -2.95459330e-01
-3.29048306e-01 -1.11220884e+00 -2.08280504e-01 -3.88550311e-01
-3.22103888e-01 5.46162188e-01 5.50733626e-01 -4.09768045e-01
7.66777277e-01 5.32432020e-01 -1.47888136e+00 -2.86756098e-01
5.11062443e-01 7.12713718e-01 1.99714199e-01 6.15405977e-01
-1.63875282e-01 9.03756499e-01 -6.10689819e-01 2.13426054e-01
2.99550802e-01 -7.19816804e-01 -1.14016092e+00 3.07187527e-01
-2.04965577e-01 -8.48037601e-02 2.12860778e-02 1.22057414e+00
3.68580781e-03 5.33676982e-01 3.04001957e-01 -1.41271544e+00
4.34760302e-02 3.94585162e-01 -2.78384387e-01 4.86836992e-02
9.79604304e-01 -2.23849028e-01 7.46867955e-01 -3.53682846e-01
5.74383914e-01 7.93005407e-01 4.29120690e-01 4.59620714e-01
-8.71824384e-01 -3.24032009e-01 4.81674403e-01 5.86636245e-01
-1.63645577e+00 -3.83256704e-01 9.91738856e-01 -1.21233292e-01
9.48941112e-01 4.45905805e-01 6.62274182e-01 3.44276875e-01
2.50050217e-01 5.83671451e-01 1.14679968e+00 -7.60159910e-01
7.49509037e-01 -2.71876931e-01 3.23934525e-01 8.65973473e-01
4.33205843e-01 2.46389031e-01 -2.77164787e-01 -2.45664090e-01
8.77915919e-01 8.73867720e-02 -6.77397102e-02 -4.14710104e-01
-9.22191978e-01 8.70881677e-01 -5.22988383e-03 6.70291781e-01
-4.79037791e-01 1.21814303e-01 2.52509564e-01 3.59115541e-01
-7.64260143e-02 5.17738283e-01 5.25686592e-02 -5.52167177e-01
-7.94735193e-01 1.91155389e-01 1.09271562e+00 8.80756438e-01
8.87900770e-01 -7.65698403e-03 3.21175426e-01 1.68238744e-01
4.71608132e-01 1.46075189e-01 4.64943796e-01 -8.28020692e-01
1.82812866e-02 9.75356519e-01 3.18930835e-01 -1.12753761e+00
-9.02595460e-01 3.29530463e-02 -3.92410785e-01 5.57519972e-01
5.31618416e-01 -3.61449569e-02 -8.54046583e-01 1.56210506e+00
5.01723886e-01 3.43389481e-01 2.28446960e-01 7.17394054e-01
5.94876766e-01 3.90487224e-01 1.16087897e-02 -4.70090389e-01
1.22792375e+00 -8.42419207e-01 -6.96478367e-01 -5.33813059e-01
8.31357062e-01 -5.73382795e-01 7.13217139e-01 4.92428422e-01
-7.80803204e-01 -1.39855862e-01 -1.02770460e+00 4.24864382e-01
-4.01114553e-01 -2.81778663e-01 1.06479454e+00 8.18944037e-01
-1.12049842e+00 5.03460050e-01 -1.05175889e+00 -9.64630008e-01
-1.55878231e-01 5.76382995e-01 -4.05440658e-01 -5.07032573e-02
-7.93973923e-01 1.11310828e+00 7.88185537e-01 -2.72521317e-01
-8.46941531e-01 4.71822709e-01 -1.13685048e+00 3.20864692e-02
1.04417253e+00 -6.66350871e-02 1.37057519e+00 -7.03186631e-01
-1.64993227e+00 5.92231750e-01 -2.97400713e-01 -4.88353163e-01
9.79147777e-02 1.76885888e-01 -4.20911610e-01 6.36084229e-02
1.43257063e-02 7.30822086e-02 3.55857849e-01 -1.02623093e+00
-9.75546539e-01 -5.65090716e-01 6.89245880e-01 3.85345906e-01
2.20519260e-01 3.53083134e-01 -3.34235400e-01 9.79025215e-02
5.60286880e-01 -8.87257278e-01 -8.29360843e-01 -4.68224645e-01
-2.35569432e-01 -3.92449081e-01 4.56027716e-01 -4.03651178e-01
1.28405738e+00 -1.93849754e+00 -1.52604371e-01 7.03055918e-01
9.31078419e-02 1.95334688e-01 3.12420838e-02 7.05866277e-01
3.38059291e-02 1.12094417e-01 -2.10950717e-01 -8.36689398e-02
1.21410668e-01 4.42444801e-01 -8.25209916e-03 5.78512847e-01
-2.03882858e-01 4.69110370e-01 -1.05930650e+00 -5.33012450e-01
5.95740438e-01 -1.08303726e-01 -2.93650478e-01 -7.37814754e-02
-1.70377672e-01 3.28210562e-01 -8.42653096e-01 6.08273745e-01
1.96949586e-01 -6.61465004e-02 4.04095918e-01 5.29755056e-01
-4.84186053e-01 4.78463084e-01 -1.59536242e+00 1.58231747e+00
-1.86812609e-01 2.98009664e-01 -1.93691358e-01 -1.25704956e+00
1.08321428e+00 5.76391816e-01 8.01724136e-01 -6.75716221e-01
3.15113366e-01 1.68007523e-01 1.08286716e-01 2.90470794e-02
6.06025577e-01 -1.36600494e-01 -3.31078470e-01 7.10058272e-01
-4.12807882e-01 -2.03703165e-01 6.23917878e-01 -2.31508017e-01
1.34130943e+00 1.81032106e-01 1.27190077e+00 -1.20507360e-01
9.02936935e-01 5.68634152e-01 7.58478701e-01 8.01668108e-01
-5.04129529e-01 3.29585761e-01 3.27415735e-01 -6.43121898e-01
-5.64854324e-01 -5.35197020e-01 3.75704557e-01 7.76379704e-01
5.37687361e-01 -1.02638674e+00 -6.59487069e-01 -6.50983930e-01
-7.16433942e-01 7.38858700e-01 -3.75375509e-01 -2.60690540e-01
-6.06680751e-01 -3.72324616e-01 4.25903022e-01 2.87210792e-01
5.91994882e-01 -1.16956103e+00 -1.23160779e+00 6.08961821e-01
-1.27340123e-01 -9.55210567e-01 -5.77666424e-02 1.76641300e-01
-8.38711023e-01 -1.37268686e+00 -8.55246112e-02 -5.82942843e-01
8.26063931e-01 3.92436832e-01 9.63247597e-01 3.85766566e-01
4.97543775e-02 6.86292946e-01 -7.34229088e-01 -5.12632489e-01
-1.82632625e-01 -1.81125969e-01 1.23644955e-01 -4.16264504e-01
7.06650198e-01 -5.45763612e-01 1.54491559e-01 3.95509213e-01
-6.87092185e-01 2.47856081e-02 2.59234607e-01 5.43528616e-01
6.37954116e-01 5.55669963e-01 2.34616101e-01 -6.78245604e-01
8.78067315e-01 -2.18736589e-01 -8.47386241e-01 3.32320362e-01
-7.12003052e-01 1.39329806e-01 7.81406820e-01 -1.90509126e-01
-6.46084070e-01 2.47782335e-01 -3.13010693e-01 1.66088432e-01
-6.39115989e-01 8.25244367e-01 -1.33386761e-01 -3.72900516e-01
6.27939105e-01 4.27230954e-01 -2.00703964e-01 -2.03732207e-01
-1.04781732e-01 2.60036916e-01 4.83472236e-02 -5.10401964e-01
7.19538033e-01 3.79880279e-01 3.21203589e-01 -7.82206535e-01
-5.06731391e-01 -6.98167086e-01 -4.81085986e-01 -2.30098829e-01
4.62221026e-01 -3.13312203e-01 -8.37988257e-01 -3.42383720e-02
-9.97431636e-01 -2.25020394e-01 -4.47321922e-01 6.05261505e-01
-1.08863747e+00 6.49370968e-01 5.68404421e-02 -1.13903451e+00
-7.56592602e-02 -1.10177696e+00 4.65154707e-01 5.28700531e-01
-4.52704340e-01 -9.76655662e-01 2.51551747e-01 -2.70509213e-01
2.93143421e-01 3.38463277e-01 4.22980219e-01 -1.02544546e+00
-3.77831310e-01 -3.41327190e-01 1.91444322e-01 -2.56505191e-01
4.39681411e-01 -2.90463507e-01 -2.73478210e-01 -5.24713472e-02
3.08875352e-01 1.05398335e-01 1.17611110e-01 2.73607433e-01
4.65319186e-01 -2.46706799e-01 -3.63851786e-01 1.28819555e-01
1.45787680e+00 1.15046430e+00 8.90453279e-01 6.98331535e-01
2.01792747e-01 4.99340773e-01 1.08840656e+00 5.19270837e-01
6.04105771e-01 7.56192982e-01 5.00492752e-01 1.66222051e-01
2.87834913e-01 2.80740578e-02 5.80682158e-01 2.80221552e-01
-4.99586523e-01 2.18388415e-03 -1.20607066e+00 4.66190010e-01
-2.20543861e+00 -1.21053588e+00 7.71602988e-02 2.32319713e+00
2.96571851e-01 1.69589818e-01 2.05261260e-01 4.73120600e-01
5.91054320e-01 5.09119295e-02 -2.60133266e-01 -6.22046769e-01
3.55810881e-01 1.96005642e-01 3.98632377e-01 8.03173423e-01
-1.10550702e+00 1.17166221e+00 6.30912590e+00 3.98433656e-01
-8.74288023e-01 3.21127921e-02 5.17642424e-02 5.31984985e-01
1.62039161e-01 3.41146380e-01 -6.63550198e-01 1.38678644e-02
7.89382994e-01 -6.29972577e-01 5.65673888e-01 1.09606218e+00
1.71880707e-01 -6.25594258e-01 -1.03871965e+00 9.52592552e-01
2.36066744e-01 -8.85310709e-01 -4.84730631e-01 1.49710745e-01
2.61965454e-01 -4.53817427e-01 -4.53811020e-01 4.37436670e-01
5.77559769e-01 -9.93515968e-01 5.32945931e-01 1.27374336e-01
1.47951379e-01 -9.55064416e-01 9.05527294e-01 6.82321072e-01
-1.50241292e+00 5.60619570e-02 -3.84661615e-01 -8.09046984e-01
2.52691835e-01 1.44578993e-01 -9.97353554e-01 8.11378598e-01
1.35224119e-01 3.80285621e-01 -2.99846053e-01 1.37342346e+00
-6.08964086e-01 3.80648136e-01 -7.38813758e-01 -3.01710516e-01
5.07206500e-01 -3.74437988e-01 6.04924202e-01 6.19996011e-01
3.28312635e-01 4.20109808e-01 7.41574585e-01 5.06522715e-01
6.49361074e-01 2.34644562e-01 -9.23104763e-01 -6.54501095e-02
2.13502750e-01 8.51755977e-01 -1.23746991e+00 -2.55773425e-01
-3.62649679e-01 6.08840883e-01 2.34634560e-02 -1.27692902e-02
-7.29018033e-01 -2.46254519e-01 5.00555992e-01 -9.65540931e-02
-2.01331422e-01 -7.22653449e-01 -2.18228817e-01 -8.40835154e-01
-2.60197800e-02 -8.02389026e-01 5.01475930e-01 -2.89446205e-01
-4.97086167e-01 6.26182318e-01 1.72300205e-01 -1.18656564e+00
-7.30454087e-01 -3.36943835e-01 -8.94871175e-01 6.25510871e-01
-1.34126830e+00 -9.76017773e-01 -2.68951535e-01 8.42940331e-01
4.10904646e-01 1.14226237e-01 1.18994951e+00 -6.89865798e-02
-4.12478119e-01 1.90607626e-02 -4.66940641e-01 1.25307545e-01
1.17499672e-01 -1.16626549e+00 3.15812588e-01 1.36825049e+00
4.27936882e-01 6.54773891e-01 9.07332182e-01 -7.90714324e-01
-1.26561201e+00 -6.99685276e-01 1.05391634e+00 -1.03429295e-01
3.92311841e-01 1.70963004e-01 -5.02548158e-01 1.12519324e+00
2.99001858e-02 -3.94988537e-01 6.10377431e-01 3.44400167e-01
3.04221481e-01 2.67194539e-01 -1.22495174e+00 9.99585629e-01
1.11052859e+00 -1.57405034e-01 -9.35411155e-01 4.94471431e-01
1.43183500e-01 -4.45894033e-01 -4.50301915e-01 1.69706658e-01
1.76943615e-01 -1.06685305e+00 6.80911601e-01 -5.15110552e-01
-3.17420393e-01 -8.09845924e-01 -3.32984962e-02 -1.10335791e+00
-4.01133865e-01 -6.27523124e-01 -2.70358268e-02 8.00072312e-01
2.09856868e-01 -8.45255971e-01 9.85248029e-01 9.00918365e-01
-2.55689710e-01 -3.82794619e-01 -1.15987062e+00 -9.98812377e-01
-5.78662157e-01 -7.00409591e-01 7.27997124e-01 6.83481038e-01
6.52116060e-01 2.10081965e-01 -3.20075959e-01 2.22039744e-01
3.65933955e-01 3.10997128e-01 9.12069380e-01 -1.42835617e+00
6.74288645e-02 -3.38398010e-01 -9.37038541e-01 -8.86007190e-01
2.73766145e-02 -5.14283359e-01 2.76682705e-01 -2.13518906e+00
-4.22365874e-01 -6.80541277e-01 -3.30052853e-01 8.99559617e-01
2.62409270e-01 -6.27820492e-02 3.29772353e-01 8.83203838e-03
-9.23354745e-01 1.49019495e-01 9.42767084e-01 9.33111608e-02
-4.74878550e-01 1.78606406e-01 -6.74768567e-01 1.14033711e+00
1.27859092e+00 -5.90523303e-01 -4.04658705e-01 -7.38405809e-02
1.62198409e-01 1.69886991e-01 -6.95884973e-02 -1.40165639e+00
7.17322946e-01 -9.18975115e-01 -3.79938006e-01 -2.90998042e-01
4.74865913e-01 -1.23050332e+00 3.83575946e-01 7.59435654e-01
1.61652654e-01 1.89894408e-01 4.21977751e-02 2.10052341e-01
-4.29021746e-01 -9.09540296e-01 4.09273028e-01 -4.47438806e-01
-1.42716134e+00 -4.92516682e-02 -7.02686965e-01 -3.98767799e-01
1.54596031e+00 -6.43027127e-01 -2.86526848e-02 -5.62634945e-01
-7.22748101e-01 1.39124021e-01 7.49730170e-01 7.06218928e-02
6.60308897e-01 -1.10737431e+00 -2.88193345e-01 -2.59904210e-02
5.40243089e-02 1.72047503e-02 -1.50621846e-01 8.95587087e-01
-6.47349834e-01 6.11064970e-01 -3.50326329e-01 -1.75451264e-01
-1.31175542e+00 5.98063886e-01 7.69553408e-02 -7.81533420e-01
-7.04330325e-01 4.20762479e-01 -1.77968234e-01 -2.29880020e-01
1.14324406e-01 -3.08571488e-01 -6.14486456e-01 -3.22939545e-01
8.92410576e-01 2.41873473e-01 -4.82126884e-03 -6.93400323e-01
-7.50520408e-01 6.60694301e-01 1.65144980e-01 -2.30360612e-01
1.18001831e+00 -1.42475694e-01 1.76890846e-02 3.16162407e-01
2.39677280e-01 4.58988808e-02 -6.87890351e-01 -1.48471117e-01
6.05913162e-01 -5.11368215e-01 -3.95922996e-02 -5.35801768e-01
-6.35193169e-01 5.20283222e-01 3.75892133e-01 5.90597570e-01
1.37044322e+00 -2.19215840e-01 3.46838742e-01 7.03470469e-01
1.42688239e+00 -1.23759103e+00 -3.27239424e-01 1.08856010e+00
5.53999603e-01 -1.17665339e+00 1.18426941e-01 -6.17027879e-01
-5.31532466e-01 1.06422770e+00 5.06619990e-01 -1.98183775e-01
3.29943031e-01 2.72397339e-01 -1.55882552e-01 -2.95541823e-01
-4.61098582e-01 -8.60657036e-01 -4.29659821e-02 8.29207122e-01
-1.09379711e-02 1.16153665e-01 -8.51188421e-01 5.35667658e-01
-1.66358799e-01 1.68664634e-01 9.07073379e-01 1.68719339e+00
-8.91127944e-01 -1.22963369e+00 -7.41813779e-01 1.76322713e-01
-1.46369413e-01 1.09289728e-01 -5.68860114e-01 8.59174907e-01
2.89667808e-02 1.39209640e+00 -1.27630845e-01 -4.73544359e-01
4.81108427e-01 1.33322120e-01 5.43727815e-01 -9.63510752e-01
-5.66836178e-01 -2.08359092e-01 3.50212872e-01 -7.69620597e-01
-5.85165977e-01 -5.49817443e-01 -1.73545516e+00 -3.34105402e-01
-3.39884311e-01 5.96668541e-01 6.39903545e-01 1.22426736e+00
-2.31574923e-02 3.63515228e-01 3.09830904e-01 -5.80904543e-01
-2.12609723e-01 -6.56222343e-01 -7.10723460e-01 8.19077343e-02
-2.88861096e-01 -9.09688115e-01 -1.76943108e-01 -1.35462537e-01] | [3.7677464485168457, 1.316941261291504] |
e45635ce-56b5-4ef3-baa0-2b5f9d92182a | integrating-multiple-sources-knowledge-for | 2305.09893 | null | https://arxiv.org/abs/2305.09893v1 | https://arxiv.org/pdf/2305.09893v1.pdf | Integrating Multiple Sources Knowledge for Class Asymmetry Domain Adaptation Segmentation of Remote Sensing Images | In the existing unsupervised domain adaptation (UDA) methods for remote sensing images (RSIs) semantic segmentation, class symmetry is an widely followed ideal assumption, where the source and target RSIs have exactly the same class space. In practice, however, it is often very difficult to find a source RSI with exactly the same classes as the target RSI. More commonly, there are multiple source RSIs available. To this end, a novel class asymmetry RSIs domain adaptation method with multiple sources is proposed in this paper, which consists of four key components. Firstly, a multi-branch segmentation network is built to learn an expert for each source RSI. Secondly, a novel collaborative learning method with the cross-domain mixing strategy is proposed, to supplement the class information for each source while achieving the domain adaptation of each source-target pair. Thirdly, a pseudo-label generation strategy is proposed to effectively combine strengths of different experts, which can be flexibly applied to two cases where the source class union is equal to or includes the target class set. Fourthly, a multiview-enhanced knowledge integration module is developed for the high-level knowledge routing and transfer from multiple domains to target predictions. | ['Ningbo Huang', 'Ke Li', 'Wenyue Guo', 'Xiong You', 'Anzhu Yu', 'Kuiliang Gao'] | 2023-05-17 | null | null | null | null | ['unsupervised-domain-adaptation', 'pseudo-label'] | ['methodology', 'miscellaneous'] | [ 5.88640034e-01 2.41689733e-03 -3.56999785e-01 -4.53774333e-01
-6.52317762e-01 -4.34904367e-01 4.18254375e-01 1.43014103e-01
-8.20964202e-03 7.37417579e-01 -1.11001402e-01 2.85760243e-03
-5.44383347e-01 -1.13715208e+00 -3.04956347e-01 -8.24154437e-01
3.46690655e-01 6.50021374e-01 4.21256453e-01 -1.17735416e-01
2.35955101e-02 4.40201879e-01 -1.52297306e+00 6.47354349e-02
1.43736267e+00 9.73159492e-01 5.54233670e-01 1.45367011e-01
-5.95733762e-01 5.71742713e-01 -4.82623339e-01 2.52200402e-02
2.95828015e-01 -5.16528428e-01 -7.45109260e-01 5.25635898e-01
-8.54742341e-03 1.11769256e-03 3.16779524e-01 1.06195366e+00
3.69972736e-01 2.81266421e-01 1.04116380e+00 -1.30164695e+00
-2.87295997e-01 5.80391049e-01 -8.08016360e-01 -2.00090066e-01
-2.23032147e-01 -2.92354047e-01 9.43749189e-01 -6.18254662e-01
4.82241660e-01 1.19773722e+00 2.91466534e-01 2.34981045e-01
-9.24265683e-01 -9.27349865e-01 5.11468053e-01 2.54256517e-01
-1.50567603e+00 -1.04471028e-01 9.89122152e-01 -5.80807745e-01
1.35447726e-01 -2.97646094e-02 6.10569119e-01 5.30017316e-01
-2.60292768e-01 6.68435693e-01 1.22152090e+00 -3.33658546e-01
3.94834250e-01 4.80129153e-01 -3.41608711e-02 -1.27796302e-04
3.18838023e-02 -2.14067653e-01 -2.11950481e-01 1.88560449e-02
6.44782841e-01 2.21572161e-01 -1.61294192e-01 -5.31128705e-01
-1.02224815e+00 8.89996707e-01 6.11206293e-01 4.33242321e-01
-4.06579286e-01 -7.57463932e-01 1.37620822e-01 1.10016018e-01
6.14210010e-01 8.05339217e-02 -4.94184256e-01 6.15378618e-01
-8.02808881e-01 2.36249156e-02 5.70163906e-01 9.42026317e-01
1.11146331e+00 -5.91788907e-03 1.27369732e-01 1.21367764e+00
6.48001909e-01 6.25300884e-01 3.02181363e-01 -8.82516503e-01
4.89418864e-01 1.01531529e+00 2.55604535e-01 -9.97402549e-01
-3.43971133e-01 -7.89158404e-01 -1.00048995e+00 2.87081957e-01
1.77776366e-01 -1.43054545e-01 -7.69619167e-01 1.55510020e+00
8.27510238e-01 3.50077689e-01 4.83799487e-01 7.56588399e-01
8.71428967e-01 8.24042857e-01 2.80684412e-01 -3.86007756e-01
1.10336399e+00 -7.39683390e-01 -3.82501781e-01 -5.93972802e-01
1.98472920e-03 -6.34270906e-01 6.33838117e-01 1.11198455e-01
-4.59291577e-01 -6.82136297e-01 -1.05749488e+00 5.34449399e-01
-4.16630000e-01 2.57810533e-01 3.47471386e-01 3.00778747e-01
-2.82377303e-01 2.22060412e-01 -2.87579745e-01 -4.10719037e-01
4.34644252e-01 5.01031019e-02 -1.27603829e-01 -1.86563224e-01
-1.30042148e+00 6.45364344e-01 8.12484622e-01 2.11169459e-02
-7.23484516e-01 -7.02843428e-01 -5.99008858e-01 -7.57292882e-02
6.06869221e-01 -3.59452903e-01 9.62435782e-01 -1.68376005e+00
-1.50126243e+00 7.67150164e-01 -1.29602179e-01 -1.09468987e-02
3.96718323e-01 2.46028617e-01 -7.39573479e-01 -1.59910098e-02
4.34610069e-01 4.62343991e-01 1.00688481e+00 -1.61890984e+00
-1.09438300e+00 -6.18263841e-01 -1.18664280e-01 7.60966420e-01
-2.15051636e-01 -1.34632170e-01 -3.00679028e-01 -5.43545604e-01
5.07101476e-01 -6.15915120e-01 -2.32914031e-01 3.61717567e-02
-1.78005278e-01 -2.34340861e-01 7.56938994e-01 -6.25328302e-01
1.06200421e+00 -2.14215589e+00 2.44018748e-01 6.71289980e-01
-6.02313168e-02 4.95585889e-01 9.53624025e-03 2.65028834e-01
-1.84415460e-01 -1.95322096e-01 -8.31635714e-01 2.23752201e-01
-2.36596629e-01 3.47982764e-01 -1.98809251e-01 1.91473842e-01
1.44144565e-01 1.70027003e-01 -8.60030770e-01 -7.13619113e-01
2.91831344e-01 1.84379533e-01 -1.27776582e-02 2.27518737e-01
-3.20948571e-01 8.67897272e-01 -6.96803808e-01 5.57754636e-01
9.98544812e-01 -3.10486972e-01 3.64814192e-01 -2.16589496e-01
-1.96823299e-01 -3.14177185e-01 -1.80750024e+00 1.36991847e+00
-4.18776840e-01 -7.77479410e-02 3.02257746e-01 -1.37103343e+00
1.39590836e+00 3.26862693e-01 7.14233160e-01 -4.85749632e-01
-1.92376509e-01 4.37067837e-01 -1.48460835e-01 -2.21384883e-01
1.25761367e-02 -3.54746729e-01 1.58012584e-01 3.43119144e-01
-1.29628718e-01 -2.79627919e-01 -3.13345827e-02 -8.41251686e-02
2.85713345e-01 3.31304818e-01 6.05720818e-01 -1.20908007e-01
8.43936622e-01 3.53609294e-01 9.77224946e-01 5.87471247e-01
-6.85322583e-02 5.02899945e-01 6.32487684e-02 -1.35271743e-01
-9.72470999e-01 -1.14809608e+00 -1.94680393e-01 9.67812002e-01
5.05998611e-01 3.30941498e-01 -4.64638114e-01 -6.38369858e-01
-1.12528585e-01 4.20165777e-01 -2.65678197e-01 1.02617078e-01
-3.85532677e-02 -7.05129802e-01 1.09122902e-01 4.14520890e-01
1.04686999e+00 -8.57692122e-01 -2.25406200e-01 3.80408674e-01
-3.90415102e-01 -7.12869704e-01 -5.10990843e-02 -4.69146483e-02
-8.39365900e-01 -1.17590678e+00 -9.05545592e-01 -8.49434197e-01
6.20424867e-01 4.42816079e-01 7.02609003e-01 -2.21141756e-01
2.31206819e-01 8.65279585e-02 -5.38257957e-01 -5.22037804e-01
-4.62202936e-01 5.72394133e-02 -2.89703727e-01 4.63101983e-01
3.76976460e-01 -5.53279400e-01 -4.85666633e-01 6.44792974e-01
-1.04948282e+00 1.96203485e-01 6.24374151e-01 6.54890895e-01
8.07563007e-01 4.86226350e-01 1.02954125e+00 -8.90186131e-01
1.38405219e-01 -7.66537488e-01 -5.61141491e-01 5.49520671e-01
-5.28691292e-01 -3.73302609e-01 5.96185446e-01 -4.10870522e-01
-1.56783688e+00 3.14364135e-01 -1.03194518e-02 -2.12327719e-01
-5.49154043e-01 7.03225851e-01 -7.25609243e-01 1.94129214e-01
6.88928723e-01 1.85939074e-01 1.05848581e-01 -3.89989167e-01
2.76684523e-01 1.11853468e+00 2.28125334e-01 -4.08880562e-01
7.81027555e-01 4.90798444e-01 -2.13232502e-01 -8.53837311e-01
-9.02833819e-01 -7.95526326e-01 -8.14219713e-01 -3.20582122e-01
7.99085557e-01 -1.24919248e+00 3.28222965e-03 8.36056292e-01
-8.34520936e-01 -2.40492418e-01 -2.98663169e-01 6.23516858e-01
-2.84960836e-01 3.06544214e-01 1.50687411e-01 -7.49824405e-01
-1.07748769e-01 -9.86033261e-01 7.48634815e-01 6.10217035e-01
1.59668773e-01 -1.02337909e+00 -1.53172955e-01 3.88861209e-01
2.59207487e-01 1.26599148e-01 7.70802677e-01 -9.58925247e-01
-2.68016517e-01 2.51267225e-01 -4.87539977e-01 6.49075210e-01
4.92250115e-01 3.90238613e-02 -7.94529319e-01 -1.34630697e-02
-1.91331819e-01 -4.34808508e-02 5.07677734e-01 4.05795515e-01
7.44765103e-01 -9.72581878e-02 -5.14029860e-01 2.57287264e-01
1.43297529e+00 5.77927113e-01 3.99019420e-01 3.73146445e-01
7.08088756e-01 1.01695991e+00 9.94861424e-01 5.04030645e-01
5.49173295e-01 5.14996707e-01 1.99944302e-01 -2.23267376e-01
1.06671043e-01 -3.56981456e-02 -1.09692570e-02 6.02877855e-01
4.03756350e-02 -2.04739004e-01 -1.01327634e+00 5.86422026e-01
-1.88818765e+00 -8.37663651e-01 -1.40351970e-02 2.35245562e+00
6.53766751e-01 -2.50288725e-01 -1.97096472e-03 1.00421801e-01
1.14072382e+00 6.49319962e-02 -9.08091366e-01 2.92869806e-01
-1.17135756e-01 -8.02715048e-02 4.12159950e-01 4.51163590e-01
-1.30610418e+00 7.92648613e-01 4.83659506e+00 1.12897873e+00
-9.70910370e-01 8.36671740e-02 3.63564700e-01 5.92833579e-01
-3.84357363e-01 1.71775222e-01 -7.21511066e-01 4.24649507e-01
1.49350226e-01 -9.42647159e-02 2.43610710e-01 8.30425739e-01
2.05516770e-01 -1.68950483e-01 -4.40793723e-01 7.24971056e-01
-1.15698986e-01 -9.10518050e-01 9.96038914e-02 9.26364120e-03
9.78330135e-01 -2.38919288e-01 -3.46173376e-01 9.79661718e-02
5.68688691e-01 -6.12567306e-01 4.17191118e-01 5.82535148e-01
6.50086224e-01 -9.44320202e-01 5.71048379e-01 7.66051173e-01
-1.48296595e+00 -3.38656455e-01 -3.78882051e-01 3.22248220e-01
1.36907473e-02 7.22344637e-01 -5.19903839e-01 1.28196740e+00
6.98135614e-01 1.03287947e+00 -2.47925922e-01 9.40044701e-01
-2.68114775e-01 4.55041766e-01 -3.92757326e-01 4.16501343e-01
9.76020768e-02 -6.02089345e-01 5.93890011e-01 7.62391448e-01
4.43040520e-01 2.80518025e-01 5.94650686e-01 5.98025799e-01
2.62773007e-01 4.58333939e-01 -5.23938596e-01 2.79066741e-01
7.87057102e-01 1.24497795e+00 -7.54422128e-01 -4.51063961e-01
-3.09214205e-01 7.80055642e-01 -1.01497114e-01 5.74039042e-01
-6.66565716e-01 -4.63938534e-01 4.34460908e-01 -1.89628694e-02
2.50249386e-01 8.30356479e-02 -2.18633547e-01 -1.17999494e+00
-2.50538468e-01 -7.71074176e-01 6.72082365e-01 -8.06320429e-01
-1.60793126e+00 2.66053438e-01 2.25921214e-01 -1.72306406e+00
-1.62704829e-02 -2.97993988e-01 -7.34954417e-01 1.11401117e+00
-1.90501857e+00 -1.43237782e+00 -5.91021776e-01 8.65474761e-01
4.51953918e-01 -3.68263423e-01 7.11415291e-01 3.64686638e-01
-4.11186963e-01 9.53915790e-02 4.09265965e-01 3.66893858e-02
6.20602429e-01 -8.83229792e-01 -4.35074598e-01 8.39730859e-01
-2.18585745e-01 2.79796064e-01 3.27373475e-01 -7.29459643e-01
-5.57609200e-01 -1.50169766e+00 5.14359653e-01 3.05010468e-01
4.92465556e-01 2.36118868e-01 -1.18628788e+00 5.37164986e-01
-1.45773083e-01 -2.67124087e-01 7.68523097e-01 -1.35972410e-01
-1.62804842e-01 -5.67315102e-01 -1.29140937e+00 4.51712549e-01
6.88486636e-01 -2.38930494e-01 -4.58746225e-01 3.75496566e-01
4.12234008e-01 -1.77010819e-01 -1.05221105e+00 6.69477463e-01
3.11951339e-01 -7.52202094e-01 9.89144444e-01 -7.42951259e-02
3.22918594e-01 -8.62970769e-01 -1.83564276e-01 -1.50700748e+00
-2.20695078e-01 2.34292120e-01 4.68605101e-01 1.60766602e+00
3.95642817e-01 -9.14023638e-01 4.83327717e-01 1.93002343e-01
-8.08800310e-02 -6.87922612e-02 -9.29855347e-01 -7.10163593e-01
1.66479319e-01 -1.39143348e-01 8.00837338e-01 1.15571010e+00
-4.58829910e-01 3.64332229e-01 -9.64251906e-02 5.53830624e-01
6.76971376e-01 4.93824720e-01 6.61629319e-01 -1.79538357e+00
-2.60569394e-01 -2.91211993e-01 5.32503314e-02 -9.12182629e-01
1.86395094e-01 -9.50408936e-01 1.22136809e-01 -1.93900919e+00
1.26835749e-01 -9.77725446e-01 -1.78281993e-01 4.67314363e-01
-1.31846219e-01 -1.29016787e-01 2.35219747e-02 5.81931233e-01
-1.48128405e-01 5.44727445e-01 1.37139332e+00 -2.38279387e-01
-5.11281192e-01 5.25236845e-01 -7.45950341e-01 7.98127472e-01
8.94427598e-01 -4.09366012e-01 -7.69661844e-01 -1.58279240e-01
-2.66062617e-02 1.26816928e-01 2.82437235e-01 -1.04277694e+00
1.16465762e-01 -6.10342383e-01 2.72618622e-01 -5.81962407e-01
-1.35668246e-02 -1.14943850e+00 4.70077842e-01 5.50938435e-02
-5.21889105e-02 -8.07471395e-01 -5.90973832e-02 6.47137344e-01
-4.13795531e-01 -4.01253879e-01 1.12779319e+00 -3.49730939e-01
-1.02721381e+00 4.85212654e-01 -2.07734630e-01 -8.34897831e-02
1.23405111e+00 -4.22332585e-01 -1.17996214e-02 -3.56120542e-02
-7.72458851e-01 5.15794396e-01 2.88700610e-01 3.92992884e-01
4.13029015e-01 -1.17451620e+00 -9.16930854e-01 1.50195956e-01
4.71312582e-01 6.02247477e-01 6.84361398e-01 5.86680114e-01
-8.70241225e-02 -2.46114075e-01 -2.66456097e-01 -6.89596951e-01
-1.08577025e+00 2.94014484e-01 5.67642033e-01 -2.70833611e-01
-3.25305283e-01 4.30689037e-01 2.53488749e-01 -1.01695156e+00
-1.52353823e-01 3.32887620e-01 -7.94913232e-01 5.81814051e-01
3.70351702e-01 3.72604191e-01 -1.66657940e-01 -9.05456543e-01
-2.46234432e-01 8.96359205e-01 3.13870341e-01 -9.29858163e-02
1.19215202e+00 -4.10357088e-01 -2.61129141e-01 5.50624907e-01
6.69315994e-01 -2.87533313e-01 -1.36860371e+00 -7.85443008e-01
-2.36227646e-01 -2.71741897e-01 -1.86614960e-01 -9.45051908e-01
-1.09158158e+00 9.04381871e-01 7.62057066e-01 -1.19718991e-01
1.39824867e+00 1.06762759e-01 3.88207942e-01 1.42462268e-01
3.04500639e-01 -1.45952940e+00 -2.08294362e-01 3.78250808e-01
5.88500738e-01 -1.39774239e+00 -7.43852481e-02 -4.75742310e-01
-9.67844725e-01 1.02305591e+00 7.14265227e-01 2.29481116e-01
8.35600257e-01 -2.87238508e-01 1.99513674e-01 -8.41262415e-02
-1.70552190e-02 -4.28330928e-01 3.50531012e-01 8.82665157e-01
-5.27153816e-03 3.15660894e-01 -2.67778963e-01 4.82125223e-01
4.51527864e-01 -1.54867806e-02 1.50856674e-01 7.36046135e-01
-7.95728147e-01 -1.10774171e+00 -6.16562724e-01 3.57114375e-01
-6.93494081e-02 1.64371520e-01 -1.75212361e-02 5.88586330e-01
4.97295409e-01 1.17405367e+00 2.72922032e-02 -2.52621561e-01
3.11827958e-01 3.59608559e-03 -4.27562706e-02 -7.87378609e-01
-2.18588680e-01 2.81352401e-01 -3.31284791e-01 -6.52415603e-02
-1.10331011e+00 -5.88017464e-01 -1.40689790e+00 5.34663349e-02
-3.95958811e-01 2.77927190e-01 5.81806660e-01 1.13637877e+00
1.65322229e-01 3.60957235e-01 9.20257330e-01 -5.85583329e-01
-2.06472576e-01 -8.15715730e-01 -9.37605798e-01 2.15640798e-01
-5.99285066e-02 -8.51618588e-01 -3.25403869e-01 1.72643274e-01] | [9.800823211669922, 1.5834035873413086] |
2a2e016c-d48d-4d0c-89c9-5c73721f9bb1 | reducing-the-label-bias-for-timestamp | null | null | http://openaccess.thecvf.com//content/CVPR2023/html/Liu_Reducing_the_Label_Bias_for_Timestamp_Supervised_Temporal_Action_Segmentation_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Liu_Reducing_the_Label_Bias_for_Timestamp_Supervised_Temporal_Action_Segmentation_CVPR_2023_paper.pdf | Reducing the Label Bias for Timestamp Supervised Temporal Action Segmentation | Timestamp supervised temporal action segmentation (TSTAS) is more cost-effective than fully supervised counterparts. However, previous approaches suffer from severe label bias due to over-reliance on sparse timestamp annotations, resulting in unsatisfactory performance. In this paper, we propose the Debiasing-TSTAS (D-TSTAS) framework by exploiting unannotated frames to alleviate this bias from two phases: 1) Initialization. To reduce the dependencies on annotated frames, we propose masked timestamp predictions (MTP) to ensure that initialized model captures more contextual information. 2) Refinement. To overcome the limitation of the expressiveness from sparsely annotated timestamps, we propose a center-oriented timestamp expansion (CTE) approach to progressively expand pseudo-timestamp groups which contain semantic-rich motion representation of action segments. Then, these pseudo-timestamp groups and the model output are used to iteratively generate pseudo-labels for refining the model in a fully supervised setup. We further introduce segmental confidence loss to enable the model to have high confidence predictions within the pseudo-timestamp groups and more accurate action boundaries. Our D-TSTAS outperforms the state-of-the-art TSTAS method as well as achieves competitive results compared with fully supervised approaches on three benchmark datasets. | ['Zihang Shao', 'Chenwei Tan', 'Shenglan Liu', 'Yunheng Li', 'Kaiyuan Liu'] | 2023-01-01 | null | null | null | cvpr-2023-1 | ['action-segmentation'] | ['computer-vision'] | [ 6.15774691e-01 1.57584354e-01 -7.08890259e-01 -5.55006027e-01
-1.08904791e+00 -2.49208376e-01 4.76170301e-01 5.92122190e-02
-5.22807062e-01 6.85395718e-01 3.33434016e-01 7.30093867e-02
1.80316418e-01 -3.38405281e-01 -5.35632253e-01 -6.02112949e-01
-2.89862286e-02 3.51298839e-01 8.20065975e-01 3.91150922e-01
2.38449454e-01 5.95978610e-02 -1.44090772e+00 6.47215188e-01
9.22390342e-01 1.11503279e+00 1.15853503e-01 4.14079189e-01
2.63493303e-02 1.11588717e+00 -4.74960625e-01 4.98752072e-02
2.71145076e-01 -6.74426317e-01 -7.68879175e-01 2.70810485e-01
2.26620927e-01 -4.70739484e-01 -1.08067520e-01 7.91355789e-01
7.93627724e-02 3.52860242e-01 2.51636416e-01 -1.43495989e+00
1.12118438e-01 7.23777175e-01 -7.60438979e-01 1.24397732e-01
3.02313209e-01 7.27271587e-02 1.09451282e+00 -7.99423575e-01
8.04663479e-01 1.22620678e+00 7.03200519e-01 4.65734035e-01
-1.09295094e+00 -7.13341594e-01 6.41384542e-01 4.40929264e-01
-1.17277157e+00 -4.25430596e-01 8.03158402e-01 -2.25005880e-01
7.77483761e-01 2.69404978e-01 7.59642720e-01 1.04743385e+00
-1.51944816e-01 1.18475616e+00 1.14546990e+00 -2.61358082e-01
4.61327553e-01 -3.12785268e-01 -6.46298304e-02 5.20118952e-01
-2.94518977e-01 -2.74724271e-02 -8.39201927e-01 -1.47442207e-01
6.58437550e-01 3.13646607e-02 1.28476426e-01 -1.80625439e-01
-1.53061676e+00 6.02501631e-01 1.40142620e-01 1.80739865e-01
-4.99473751e-01 2.29431883e-01 7.07232773e-01 -1.75504968e-01
5.67797363e-01 2.13391215e-01 -4.20497954e-01 -5.13449490e-01
-1.49443173e+00 2.26643115e-01 2.90288240e-01 8.51874113e-01
7.07975864e-01 -6.64658621e-02 -6.10275328e-01 5.55530787e-01
1.35273829e-01 8.45090374e-02 4.84585434e-01 -1.48471737e+00
5.25480270e-01 6.09551311e-01 1.89536318e-01 -5.92317283e-01
-2.14049041e-01 -1.17836446e-01 -4.43639576e-01 -2.07193583e-01
3.71421963e-01 2.53304560e-02 -1.16940343e+00 1.63144255e+00
7.00101495e-01 8.28193545e-01 -1.08280115e-01 9.64375198e-01
2.88431376e-01 7.26445079e-01 2.79272705e-01 -4.81848985e-01
1.05207205e+00 -1.35475206e+00 -7.21847117e-01 -1.49831817e-01
9.69161689e-01 -4.53151584e-01 1.10408103e+00 3.81571651e-01
-8.48317683e-01 -6.04335248e-01 -8.68725657e-01 2.75581852e-02
1.41280860e-01 2.35922635e-01 3.69807422e-01 2.38566071e-01
-6.60430849e-01 8.03821862e-01 -1.50066948e+00 -1.11372411e-01
5.83778024e-01 2.53535032e-01 -1.82044327e-01 -5.92413060e-02
-1.09273756e+00 4.93001878e-01 6.50056422e-01 2.34672800e-02
-1.07350898e+00 -7.26889253e-01 -1.08334351e+00 -3.38225067e-01
1.03110802e+00 -1.39242709e-01 1.60870099e+00 -1.01526392e+00
-1.49952471e+00 4.20309782e-01 -5.90974927e-01 -8.48579645e-01
7.44028747e-01 -3.74525011e-01 -2.04675496e-01 3.41533214e-01
4.50540364e-01 9.64074492e-01 8.58039677e-01 -1.08146155e+00
-1.17457557e+00 -8.02921206e-02 -1.43497251e-02 2.82050908e-01
-2.02813134e-01 -2.86236871e-02 -8.20279419e-01 -8.47057343e-01
2.86703229e-01 -1.25335431e+00 -5.44778407e-01 -2.74818122e-01
-4.85063404e-01 -2.21413672e-01 1.13529932e+00 -6.15357518e-01
1.62789929e+00 -2.03843331e+00 1.53521076e-01 -3.57309245e-02
-5.27012125e-02 2.91715205e-01 3.34618203e-02 2.06832707e-01
1.42925397e-01 -1.67844132e-01 -4.67426777e-01 -7.92653024e-01
-8.95441994e-02 5.63325644e-01 -4.09065425e-01 2.79174179e-01
7.42108449e-02 6.13949358e-01 -1.19050527e+00 -9.75272357e-01
4.80450869e-01 1.71419203e-01 -5.78635633e-01 2.12510109e-01
-5.75492442e-01 8.61850560e-01 -6.37172043e-01 7.04032660e-01
2.05555275e-01 -2.62521088e-01 1.14353284e-01 -2.29288235e-01
-3.34244430e-01 4.22512472e-01 -1.20709717e+00 1.85731709e+00
-1.66550204e-01 3.21428061e-01 -4.09215420e-01 -7.50536740e-01
5.01498580e-01 3.31754357e-01 9.50123549e-01 -5.05693674e-01
-8.59927684e-02 1.16738960e-01 -4.03448373e-01 -3.79608154e-01
5.99592090e-01 -4.41331416e-02 -1.07706226e-01 4.69841838e-01
-2.44498998e-01 1.60834044e-01 3.61935228e-01 2.86111712e-01
9.40939844e-01 7.52477825e-01 3.19323212e-01 2.28059247e-01
5.95659018e-01 1.79697365e-01 1.23638391e+00 4.81641740e-01
-5.23486257e-01 6.82431042e-01 6.01353824e-01 -4.04542416e-01
-8.59042883e-01 -5.56570590e-01 2.39135891e-01 1.20955920e+00
2.72332609e-01 -8.24717343e-01 -9.72269416e-01 -1.09002066e+00
-2.54245520e-01 8.35038483e-01 -5.90684056e-01 -1.10176682e-01
-9.52662349e-01 -3.01638067e-01 6.70946717e-01 9.31262970e-01
6.26371682e-01 -1.02139318e+00 -1.04471123e+00 3.54548424e-01
-4.93441999e-01 -1.50785458e+00 -7.94982553e-01 4.53133211e-02
-1.01473379e+00 -8.75889122e-01 -6.37748837e-01 -3.56165588e-01
6.05402112e-01 1.05845705e-01 6.40732110e-01 -1.60105437e-01
1.56744808e-01 -6.36813119e-02 -6.13578677e-01 -5.52870668e-02
-4.17247742e-01 6.42890111e-02 5.45634702e-02 2.26836741e-01
2.33987942e-01 -2.19588652e-01 -7.57969141e-01 5.87254882e-01
-8.39000940e-01 4.06924516e-01 3.53123277e-01 5.80026090e-01
1.16221178e+00 -9.64920223e-03 4.99084175e-01 -1.00765574e+00
-5.25575317e-03 -1.99752450e-01 -5.57750523e-01 1.90831438e-01
-6.42603934e-01 1.09407797e-01 5.69751620e-01 -5.85144520e-01
-1.13756323e+00 4.64411944e-01 9.57650542e-02 -7.86039650e-01
-1.22352384e-01 4.53349918e-01 2.10448876e-02 3.96401316e-01
2.34182805e-01 1.05410904e-01 -4.60252613e-01 -5.06264031e-01
2.72091210e-01 4.30690438e-01 7.41718769e-01 -5.48274577e-01
4.50272411e-01 7.09218264e-01 -1.20205127e-01 -4.18259591e-01
-1.20434594e+00 -6.40274763e-01 -8.14436615e-01 -4.19309139e-01
1.04041004e+00 -8.85181487e-01 -4.22576480e-02 4.48610455e-01
-9.21244740e-01 -5.97582698e-01 -4.02337372e-01 6.85494840e-01
-6.26990020e-01 7.04287469e-01 -6.71002150e-01 -9.57922339e-01
-9.95236412e-02 -1.19934857e+00 1.27611578e+00 1.02816924e-01
-4.56094861e-01 -6.07267082e-01 6.83201775e-02 6.30117297e-01
-2.09631652e-01 4.87176776e-01 4.27941203e-01 -7.35091388e-01
-5.26380241e-01 -9.48780701e-02 -2.67304834e-02 2.83580214e-01
2.12893069e-01 -7.47883618e-02 -9.02476668e-01 -2.24699378e-01
-1.69430092e-01 -4.04081106e-01 9.66767013e-01 4.49012905e-01
1.21226573e+00 -2.32868269e-01 -4.26220953e-01 4.18112576e-01
8.85945857e-01 3.29671174e-01 3.80986214e-01 4.08043295e-01
8.25082779e-01 5.23220241e-01 1.55463982e+00 7.25054681e-01
5.59096813e-01 7.40888059e-01 3.63432795e-01 -3.26305744e-03
-8.30683578e-03 -6.48010910e-01 5.13994217e-01 5.94788074e-01
-2.19987020e-01 -1.51024163e-01 -6.90378845e-01 6.23067319e-01
-2.23529458e+00 -8.66522133e-01 -8.60848278e-02 2.24933171e+00
9.87294555e-01 5.69872379e-01 4.77225035e-01 3.50513488e-01
8.12335134e-01 5.35242617e-01 -8.07274818e-01 2.19577588e-02
1.85399428e-01 -1.61554024e-01 5.05224526e-01 3.73832434e-01
-1.30729938e+00 1.32297766e+00 5.44596624e+00 9.55420852e-01
-9.66150105e-01 2.78940946e-01 5.42508900e-01 -2.81506956e-01
6.43413216e-02 3.85805607e-01 -8.30785692e-01 6.15461111e-01
8.72058153e-01 1.63735926e-01 1.03820734e-01 8.86226594e-01
7.15078056e-01 -3.57307464e-01 -1.29901803e+00 8.28179240e-01
-1.07678220e-01 -1.19700181e+00 -1.08386084e-01 -5.55728078e-02
8.41009974e-01 -1.61115155e-01 -1.80362567e-01 3.26098591e-01
8.44541043e-02 -6.18075311e-01 1.13482046e+00 1.37911797e-01
8.38084757e-01 -5.62307954e-01 4.10736918e-01 4.69231606e-01
-1.58319855e+00 -1.15048148e-01 2.88859252e-02 5.19942231e-02
6.86419666e-01 5.51518381e-01 -8.18213224e-01 5.76684892e-01
5.72847545e-01 1.12131321e+00 -3.81546676e-01 8.19958687e-01
-3.83370548e-01 8.43262970e-01 -3.82928610e-01 4.56802815e-01
7.01278687e-01 -8.41892287e-02 4.53437895e-01 1.23199856e+00
1.24208137e-01 2.39904180e-01 6.82981968e-01 4.89529371e-01
1.83799505e-01 -2.18793616e-01 -4.33063246e-02 -3.92192565e-02
7.00136244e-01 1.03704512e+00 -1.20592999e+00 -7.54661024e-01
-3.49570394e-01 1.07178068e+00 -8.04935172e-02 2.13776395e-01
-1.23765385e+00 1.73249140e-01 2.72098571e-01 1.55047864e-01
3.05270731e-01 -2.31877759e-01 -9.25218984e-02 -1.06979156e+00
9.18204263e-02 -9.14743543e-01 8.10180187e-01 -5.49978435e-01
-6.24570310e-01 4.86511201e-01 3.75558436e-01 -1.46367848e+00
-3.95672619e-01 1.94001824e-01 -4.06341046e-01 2.45184064e-01
-1.47163200e+00 -1.33472431e+00 -3.47733945e-01 5.68137646e-01
1.15905142e+00 4.43471223e-01 2.64370233e-01 4.00011152e-01
-7.78350115e-01 4.17995125e-01 -4.51106042e-01 -7.11044818e-02
8.96761477e-01 -1.18214929e+00 4.29816484e-01 1.04894865e+00
1.04369909e-01 1.85028031e-01 5.95659435e-01 -1.04215169e+00
-8.05643022e-01 -1.44101405e+00 8.40614438e-01 -3.08218002e-01
5.14409781e-01 -1.30977362e-01 -9.86731529e-01 9.29366052e-01
-2.81276554e-01 8.33960176e-02 3.69286537e-01 -3.58600438e-01
-2.12126225e-01 -2.17621448e-03 -8.24525237e-01 5.67491412e-01
1.06108451e+00 -4.61117387e-01 -6.53148293e-01 2.45946899e-01
9.64482188e-01 -5.57568312e-01 -7.52766550e-01 5.89971542e-01
4.75434989e-01 -1.07362640e+00 6.92474902e-01 -2.70663440e-01
2.94875681e-01 -5.49352229e-01 -3.41581181e-02 -7.87355542e-01
5.89560308e-02 -9.37336445e-01 -2.81699985e-01 1.28675437e+00
2.34193072e-01 -1.98593900e-01 9.99286294e-01 5.74197829e-01
-3.55593204e-01 -8.82378638e-01 -1.03758001e+00 -8.11316311e-01
-5.03160059e-01 -7.72293031e-01 5.84887862e-01 8.67079675e-01
1.31831139e-01 -3.85094136e-02 -6.84892058e-01 1.62161991e-01
6.16040111e-01 1.39082909e-01 6.03253484e-01 -8.30829561e-01
-1.25172973e-01 -2.17938311e-02 -9.24158022e-02 -1.30380666e+00
1.82049438e-01 -4.39885288e-01 3.49817455e-01 -1.29670525e+00
6.96432963e-02 -5.99575818e-01 -3.70680660e-01 7.05929995e-01
-2.49587372e-01 3.83922398e-01 1.60866305e-01 4.85294431e-01
-1.29741645e+00 5.41724324e-01 1.15983880e+00 1.72314361e-01
-5.88768423e-01 1.33896107e-02 7.43671954e-02 1.02838922e+00
5.94098091e-01 -7.21271038e-01 -6.62188053e-01 -2.53620267e-01
-1.54628798e-01 3.32055926e-01 3.33813846e-01 -1.11473954e+00
3.48378956e-01 -3.85102421e-01 -3.77195934e-03 -1.04690933e+00
3.95389855e-01 -5.83781242e-01 2.41380766e-01 4.55686361e-01
-7.10157931e-01 -1.18196338e-01 -1.71190694e-01 7.91366637e-01
-3.22225630e-01 -1.31517634e-01 8.10302496e-01 -6.86432272e-02
-8.68098021e-01 4.44856495e-01 -3.22205752e-01 4.70333025e-02
1.17070174e+00 -5.21208286e-01 2.15320170e-01 -1.39180452e-01
-7.06649363e-01 4.80337262e-01 5.30842185e-01 3.85662436e-01
4.66465712e-01 -1.25245500e+00 -2.70217031e-01 -6.37386888e-02
8.55890960e-02 3.53822619e-01 2.66218930e-01 1.18579996e+00
-2.61503935e-01 4.45251971e-01 1.76121950e-01 -8.12266707e-01
-1.29903877e+00 5.11293232e-01 -7.55322352e-02 -5.10699332e-01
-1.05323088e+00 6.08462811e-01 2.27888495e-01 1.26898527e-01
5.14466643e-01 -8.43010604e-01 -1.68302611e-01 1.68531090e-01
3.82644951e-01 6.89261138e-01 -1.08700804e-01 -7.68812418e-01
-4.59456801e-01 4.95061427e-01 -1.04844518e-01 -4.37131107e-01
1.03086567e+00 -2.32224181e-01 2.76416630e-01 5.01607716e-01
8.84586394e-01 -2.14010432e-01 -2.00964952e+00 -3.55851889e-01
4.67741191e-01 -4.93318707e-01 -4.70344126e-02 -4.81373817e-01
-9.48354900e-01 4.87858981e-01 2.55579710e-01 -2.02768221e-01
1.15730464e+00 -3.66030075e-02 1.44292068e+00 6.05657883e-02
4.35535491e-01 -1.45566416e+00 3.58714372e-01 3.12176883e-01
3.27395439e-01 -1.06821644e+00 5.79560176e-02 -5.93948603e-01
-1.17700028e+00 7.91785419e-01 6.80617392e-01 9.88835320e-02
1.41299278e-01 1.01512179e-01 2.26066797e-03 6.57178387e-02
-8.27598810e-01 -2.07261115e-01 2.03132808e-01 3.29723030e-01
7.04987124e-02 -1.90501750e-01 -2.43970245e-01 6.35140717e-01
1.52719960e-01 4.42569792e-01 3.02569151e-01 1.32377768e+00
-3.79543781e-01 -1.07874537e+00 -3.40972036e-01 1.70502618e-01
-4.28015530e-01 1.60623521e-01 -2.55925596e-01 6.59949660e-01
2.87184030e-01 9.66469944e-01 -1.01837620e-01 -4.91695523e-01
2.43584350e-01 1.21145099e-01 2.55982846e-01 -7.02841222e-01
-3.69297445e-01 6.28429830e-01 2.80583829e-01 -9.64644372e-01
-8.07671368e-01 -9.74111319e-01 -1.78083050e+00 2.16198772e-01
-4.67346162e-01 1.95440754e-01 1.23403773e-01 1.20970070e+00
3.25946838e-01 4.67108250e-01 5.13060451e-01 -9.99647856e-01
-3.59354645e-01 -8.02420795e-01 -3.06226224e-01 2.98310250e-01
3.28641683e-01 -7.50695467e-01 -2.75672585e-01 4.67125386e-01] | [8.477771759033203, 0.5956428647041321] |
a52d0eae-c31f-472b-8496-dedf0e765ca6 | an-open-world-lottery-ticket-for-out-of | 2210.07071 | null | https://arxiv.org/abs/2210.07071v1 | https://arxiv.org/pdf/2210.07071v1.pdf | An Open-World Lottery Ticket for Out-of-Domain Intent Classification | Most existing methods of Out-of-Domain (OOD) intent classification, which rely on extensive auxiliary OOD corpora or specific training paradigms, are underdeveloped in the underlying principle that the models should have differentiated confidence in In- and Out-of-domain intent. In this work, we demonstrate that calibrated subnetworks can be uncovered by pruning the (poor-calibrated) overparameterized model. Calibrated confidence provided by the subnetwork can better distinguish In- and Out-of-domain. Furthermore, we theoretically bring new insights into why temperature scaling can differentiate In- and Out-of-Domain intent and empirically extend the Lottery Ticket Hypothesis to the open-world setting. Extensive experiments on three real-world datasets demonstrate our approach can establish consistent improvements compared with a suite of competitive baselines. | ['Xipeng Qiu', 'Yuxin Wang', 'Peiju Liu', 'Yunhua Zhou'] | 2022-10-13 | null | null | null | null | ['intent-classification'] | ['natural-language-processing'] | [ 6.57855161e-03 3.85112077e-01 -1.00609159e+00 -6.01597428e-01
-8.08110535e-01 -6.22214496e-01 7.59221494e-01 -1.25157788e-01
-1.68805987e-01 7.14779973e-01 4.50611621e-01 -5.04856169e-01
-8.93073231e-02 -4.80185568e-01 -4.73338097e-01 -1.43677086e-01
-1.44051969e-01 7.65278399e-01 1.87376663e-01 -2.71791160e-01
1.26320407e-01 -1.64551437e-01 -1.21869671e+00 3.64285171e-01
9.37153399e-01 1.08736396e+00 -1.58688292e-01 3.57761741e-01
-2.56895185e-01 7.41954207e-01 -9.85512257e-01 -4.97754693e-01
3.19218189e-01 8.36212337e-02 -8.56897593e-01 -5.61267473e-02
4.82298434e-01 -5.24018645e-01 -3.27563852e-01 8.76109064e-01
4.35135737e-02 -1.29777223e-01 1.11237085e+00 -1.56343150e+00
-6.82917953e-01 6.00215018e-01 -1.69708326e-01 2.92903960e-01
4.33129877e-01 2.08488882e-01 1.56662512e+00 -2.55917251e-01
7.83726811e-01 1.26580644e+00 7.46880829e-01 6.35372281e-01
-1.62554193e+00 -8.49618435e-01 4.84545588e-01 -3.51352215e-01
-9.71306145e-01 -3.96078557e-01 7.93138087e-01 -4.58026379e-01
1.35244548e+00 1.36607200e-01 2.52618283e-01 1.94827235e+00
1.86862186e-01 1.00429785e+00 1.07579541e+00 -2.05714196e-01
4.48058039e-01 3.65777731e-01 7.99082994e-01 5.00872195e-01
7.44078934e-01 3.96333724e-01 -5.91660678e-01 -5.91528714e-01
6.11154258e-01 -2.67509036e-02 -2.17390552e-01 -4.12493199e-01
-9.18335736e-01 1.26253438e+00 2.22254559e-01 1.73219144e-01
1.47022873e-01 -3.65577191e-02 5.98426998e-01 6.25399947e-01
9.67886388e-01 8.52721393e-01 -9.58929062e-01 -2.05357328e-01
-8.36350620e-01 2.57196724e-01 1.20718324e+00 1.17723334e+00
7.67033696e-01 -2.15363249e-01 -4.26353067e-02 8.43373597e-01
2.73846567e-01 -8.69295597e-02 8.02269995e-01 -9.16981936e-01
9.15489197e-01 7.54807413e-01 -1.20301545e-01 -4.55658704e-01
-4.59689945e-01 -3.62414986e-01 -7.13348746e-01 -3.23530436e-02
5.45378327e-01 -1.71075329e-01 -9.73169982e-01 1.89806521e+00
8.67671818e-02 2.04816479e-02 3.24910194e-01 4.62937355e-01
6.82459056e-01 1.55678198e-01 2.02634856e-01 1.89595088e-01
1.51875079e+00 -1.04800856e+00 -2.95989901e-01 -8.51095498e-01
8.96926701e-01 -2.95661569e-01 1.21141231e+00 1.34543553e-01
-4.52035517e-01 -4.18803930e-01 -1.26246309e+00 -1.74918681e-01
-4.53658164e-01 -2.57407099e-01 1.10953152e+00 8.20521414e-01
-7.33100116e-01 6.15436494e-01 -4.73003715e-01 -4.34316188e-01
4.76567864e-01 2.91973621e-01 -2.42909983e-01 1.62881628e-01
-1.52244377e+00 7.29649484e-01 6.56183124e-01 -7.86206305e-01
-9.10975337e-01 -1.06332171e+00 -1.09726131e+00 -5.22624105e-02
4.42461401e-01 -1.02500165e+00 1.67689717e+00 -8.93766642e-01
-1.20365524e+00 1.12099075e+00 1.58327240e-02 -7.37817764e-01
5.26117861e-01 -1.92808434e-01 -3.80338460e-01 -2.08432660e-01
4.95827138e-01 7.01751769e-01 9.29161727e-01 -1.02176821e+00
-5.98110557e-01 -3.46527576e-01 5.21332324e-01 1.08793624e-01
-4.11867589e-01 -2.66765177e-01 -2.60554820e-01 -7.77456284e-01
-1.36718852e-02 -7.87829578e-01 -1.20538473e-01 -3.29968214e-01
-8.21607411e-01 -4.87684876e-01 7.23311365e-01 -2.47816116e-01
1.17664623e+00 -1.95779490e+00 -4.79471475e-01 3.97391766e-02
5.86037219e-01 -3.81915152e-01 9.79444310e-02 1.93955898e-01
-1.72941774e-01 5.58631420e-01 2.09834978e-01 -5.27374566e-01
4.12450075e-01 4.02020216e-01 -6.88209295e-01 3.30769718e-01
2.80811787e-01 5.59028149e-01 -7.27683306e-01 -5.45379400e-01
-8.39126632e-02 -1.11352570e-01 -8.18509161e-01 2.92048693e-01
-6.26096904e-01 1.08159311e-01 -5.95985293e-01 7.94192016e-01
5.22305965e-01 -5.89534402e-01 2.86563694e-01 1.03574470e-01
4.56610501e-01 1.17085671e+00 -7.44955301e-01 1.53200424e+00
-5.95615149e-01 5.74846983e-01 -5.79181351e-02 -9.36479747e-01
6.40212357e-01 2.23651245e-01 1.41211197e-01 -5.62649667e-01
1.65782720e-01 1.77612841e-01 -1.39179990e-01 -1.18175410e-01
6.46263480e-01 -2.57403731e-01 -7.26141930e-01 6.84948564e-01
3.41220617e-01 -2.66340762e-01 8.18809494e-02 1.28501862e-01
1.12249303e+00 -1.51204079e-01 6.92058682e-01 -5.13766706e-01
-2.29088888e-02 1.49771512e-01 4.92422879e-01 1.03795862e+00
-5.66342950e-01 3.89357865e-01 9.64078665e-01 -3.25091243e-01
-9.95902002e-01 -9.57035840e-01 -5.21630764e-01 1.34036219e+00
7.61180222e-02 -3.88101697e-01 -4.66899633e-01 -1.33428025e+00
3.40540618e-01 1.00345182e+00 -7.07091510e-01 -2.00757489e-01
-3.19989920e-01 -6.46240234e-01 9.54026937e-01 5.99639773e-01
4.60998863e-01 -6.19429529e-01 -2.99142301e-01 4.45800610e-02
-3.57224494e-01 -1.32245529e+00 -6.04068518e-01 9.73022282e-01
-1.07975376e+00 -1.05367613e+00 -5.43839693e-01 -8.30475628e-01
3.80427420e-01 1.76416397e-01 1.73042262e+00 -2.39877198e-02
2.53940761e-01 1.17940553e-01 -7.65567794e-02 -4.05673981e-01
-7.16065407e-01 5.95184445e-01 1.49788186e-01 -6.72717333e-01
8.32992196e-01 -5.11671245e-01 -9.24404189e-02 4.67454880e-01
-5.23107827e-01 -1.72882572e-01 4.58252102e-01 9.59533513e-01
-7.05317929e-02 2.14850288e-02 4.75970387e-01 -1.21549511e+00
9.82079566e-01 -8.80020440e-01 -4.36691850e-01 2.96284538e-02
-8.59603345e-01 5.75248837e-01 5.42760968e-01 -5.92813134e-01
-1.15476263e+00 -4.62513596e-01 2.17128377e-02 -4.86854732e-01
-4.83427554e-01 3.27671498e-01 -1.55091032e-01 4.73901242e-01
7.56360769e-01 -1.35444909e-01 -3.56477559e-01 -5.12768805e-01
1.74910694e-01 8.11030984e-01 3.03744793e-01 -9.43699241e-01
7.08069265e-01 4.89395618e-01 -5.96715808e-01 -4.65055585e-01
-1.26855695e+00 -7.30576694e-01 -4.90276843e-01 4.05911475e-01
5.74914217e-01 -1.18614292e+00 -3.99752587e-01 7.95185193e-02
-1.13397920e+00 -5.17148733e-01 -2.47158945e-01 1.78134635e-01
-5.07370114e-01 2.38808438e-01 -6.86627686e-01 -7.27022707e-01
-2.61551529e-01 -1.02805090e+00 1.30714214e+00 1.17469341e-01
-8.55510771e-01 -1.40759480e+00 2.07347453e-01 3.69418591e-01
-4.29765545e-02 -3.88994366e-01 1.09666610e+00 -1.49637258e+00
-3.36963087e-01 -4.90302220e-02 -4.36515696e-02 3.51216286e-01
6.78135902e-02 -4.25865442e-01 -1.34485698e+00 2.86793597e-02
6.13266639e-02 -7.33687997e-01 1.20406890e+00 3.21969390e-01
8.76823545e-01 -3.62184793e-01 -6.70352697e-01 6.92337036e-01
1.06151593e+00 -2.95171794e-02 4.86507088e-01 6.21805847e-01
-1.87534373e-02 4.25690532e-01 6.63371027e-01 3.60634685e-01
2.90749401e-01 6.18594050e-01 5.12067080e-02 1.09793074e-01
6.13700487e-02 -7.22548544e-01 6.13254189e-01 1.18646421e-01
4.60306674e-01 -4.96075481e-01 -6.56589389e-01 6.15390658e-01
-1.56610680e+00 -8.36752057e-01 2.76787311e-01 1.87236834e+00
1.12595725e+00 9.23465371e-01 3.52111518e-01 -3.06006223e-02
5.77565551e-01 4.17014986e-01 -6.70877397e-01 -4.50347245e-01
1.32870264e-02 8.73877406e-02 5.03042996e-01 3.65451068e-01
-1.38369930e+00 8.43744934e-01 7.43196774e+00 9.78062749e-01
-8.33544314e-01 9.60126668e-02 9.74611819e-01 1.59969449e-01
-4.92526114e-01 -3.37775089e-02 -1.35790133e+00 4.91812825e-01
9.52847242e-01 2.41593015e-03 5.74493408e-02 1.46551907e+00
-4.96777892e-01 8.00099690e-03 -1.85936749e+00 6.09471262e-01
-4.09874991e-02 -1.19644773e+00 -2.89551347e-01 2.75475770e-01
7.25359559e-01 1.04711555e-01 1.19639337e-01 1.03656971e+00
1.02880585e+00 -7.66920269e-01 7.45153069e-01 -3.24440002e-01
1.07620704e+00 -3.21262181e-01 6.90900743e-01 6.30188107e-01
-1.01109159e+00 -3.30768116e-02 -4.13249701e-01 -2.35324740e-01
-2.72254854e-01 5.51520050e-01 -1.41027498e+00 1.44946784e-01
7.41095901e-01 6.15731895e-01 -4.49999690e-01 4.51180071e-01
-3.96537691e-01 8.03937554e-01 -4.21117723e-01 -1.73509181e-01
2.54859924e-01 1.38466746e-01 4.82770324e-01 1.06374991e+00
-9.36712325e-02 -3.09939295e-01 2.54128128e-01 1.21849442e+00
-3.47080171e-01 -4.74626988e-01 -8.67246866e-01 -2.42899060e-01
2.97178984e-01 6.85163260e-01 -3.63314539e-01 -5.54986596e-01
-3.08335572e-01 7.61921704e-01 3.25558156e-01 3.44653845e-01
-1.04351652e+00 -2.55242139e-01 1.32630181e+00 -2.68650819e-02
4.46941316e-01 1.93684995e-01 -6.83635116e-01 -1.59253192e+00
-1.07346207e-01 -8.69155109e-01 8.97179484e-01 -3.28383446e-01
-1.88171232e+00 3.45104545e-01 2.19070524e-01 -1.19088101e+00
-5.45472503e-01 -8.24716806e-01 -6.76572919e-01 5.17084360e-01
-1.55613911e+00 -7.94029713e-01 2.78320420e-03 3.35775793e-01
8.84947419e-01 -7.57565945e-02 8.36955428e-01 8.64112526e-02
-5.88939607e-01 9.84090805e-01 1.75092205e-01 4.12638158e-01
8.62860322e-01 -1.27646255e+00 7.94442356e-01 3.57241690e-01
9.81062278e-02 7.61862874e-01 7.26724625e-01 -8.78551126e-01
-7.37080157e-01 -9.90961313e-01 8.62270951e-01 -7.00651169e-01
8.99739385e-01 -7.90248930e-01 -5.75157702e-01 1.22364318e+00
4.81563397e-02 -2.73184270e-01 7.79838562e-01 8.71504068e-01
-1.07123721e+00 2.46433616e-01 -1.33127391e+00 6.10237896e-01
1.29726839e+00 -7.01380730e-01 -1.21813715e+00 4.36465830e-01
1.24372149e+00 -2.56641209e-01 -6.29735053e-01 3.00282896e-01
3.93414795e-01 -1.01877797e+00 1.27457178e+00 -1.07539737e+00
6.13783121e-01 5.25630236e-01 -1.60181254e-01 -1.12425470e+00
-7.05498382e-02 -6.36076450e-01 -2.45736718e-01 1.26598072e+00
5.80709696e-01 -9.17748928e-01 1.02703178e+00 6.77326620e-01
-2.57540435e-01 -5.43614209e-01 -9.43750083e-01 -1.22612691e+00
3.96457940e-01 -8.98309052e-01 5.60302615e-01 8.69622886e-01
4.35627401e-01 6.53668165e-01 -1.85110956e-01 -1.75786309e-03
5.96468449e-01 2.52985507e-01 7.96458304e-01 -1.60358799e+00
-5.34929276e-01 -5.43542206e-01 -1.62497491e-01 -1.60657966e+00
7.18197882e-01 -8.97039711e-01 -4.72205505e-02 -9.92318213e-01
2.97575086e-01 -6.87148690e-01 -8.44809413e-02 5.01105249e-01
-1.30130991e-01 1.46244783e-02 -7.25366473e-02 2.83962399e-01
-8.36747348e-01 4.04827178e-01 6.52086973e-01 -3.80272210e-01
-1.87780708e-01 1.33188814e-01 -1.20675504e+00 1.06680131e+00
7.90720820e-01 -8.55291665e-01 -4.33290780e-01 -4.66399714e-02
-1.33093104e-01 -2.68354658e-02 1.91752329e-01 -7.40206242e-01
-1.27530396e-01 -2.38853768e-01 1.49540037e-01 -5.73311627e-01
4.39113110e-01 -7.61387646e-01 -4.83041137e-01 5.60149193e-01
-6.23661935e-01 -3.67087483e-01 8.95724744e-02 9.76081371e-01
-4.00921656e-03 -3.80537510e-01 6.21391594e-01 -2.13978305e-01
-8.13922942e-01 -8.21579173e-02 -2.97344029e-01 9.03615713e-01
9.09390092e-01 -2.79480100e-01 -8.17734540e-01 -5.00331104e-01
-1.90167978e-01 1.57862276e-01 5.36838889e-01 5.17407358e-01
1.61809072e-01 -1.01875103e+00 -2.34922379e-01 4.77665573e-01
4.38302398e-01 -2.03628659e-01 -6.88553005e-02 4.70804423e-01
-2.30772495e-02 9.16348279e-01 2.94035554e-01 -7.09279537e-01
-9.60588455e-01 6.04707897e-01 3.59668911e-01 -1.17224407e+00
-1.94149300e-01 9.28064406e-01 6.60953224e-01 -9.52234328e-01
4.69448179e-01 -8.21996033e-01 2.56080300e-01 2.39979662e-02
1.52348816e-01 1.21751964e-01 -3.17374617e-02 -9.75172520e-02
-4.42524612e-01 -5.88161696e-04 -3.11956942e-01 4.63465229e-02
1.12541497e+00 7.41761997e-02 4.74068761e-01 4.03028697e-01
1.59130120e+00 -6.73637018e-02 -1.06301081e+00 -1.40780434e-01
2.68696874e-01 -3.38108957e-01 -1.00587577e-01 -7.62271106e-01
-4.84239370e-01 6.86707497e-01 6.74383640e-02 3.42574328e-01
7.39702344e-01 4.37127233e-01 7.87402213e-01 6.37068927e-01
5.36586344e-01 -1.11243653e+00 3.95171732e-01 7.20099449e-01
5.13074219e-01 -1.58093131e+00 -1.46439355e-02 -4.34730291e-01
-7.64615774e-01 9.49198246e-01 7.85625875e-01 -2.07402036e-01
7.97869086e-01 4.22248214e-01 -1.42041460e-01 -1.84646234e-01
-1.22870207e+00 -7.26064071e-02 2.56165475e-01 7.04643488e-01
2.75476635e-01 -5.10005467e-02 2.61690438e-01 1.15512061e+00
-4.83657032e-01 -1.52361780e-01 2.89217740e-01 7.56014407e-01
-2.47097507e-01 -9.25258100e-01 -2.99036622e-01 7.47393787e-01
-4.20012146e-01 -3.17258388e-01 -5.52462518e-01 1.23729599e+00
-1.80526897e-02 9.96701062e-01 2.45226726e-01 -6.15928113e-01
6.71296790e-02 4.01062906e-01 -1.52468018e-03 -8.16735566e-01
-6.50164127e-01 2.98346113e-03 5.69682837e-01 -5.05292356e-01
4.35007270e-03 -4.88344818e-01 -9.71111357e-01 -2.19075620e-01
-5.38693726e-01 2.05636416e-02 2.45462775e-01 8.54195952e-01
5.16281724e-01 2.59742945e-01 2.78443784e-01 -5.51576853e-01
-1.17091930e+00 -1.03633761e+00 -6.27783060e-01 5.74571550e-01
5.62139809e-01 -8.09371591e-01 -1.19193375e+00 -3.46672893e-01] | [11.584813117980957, 7.651508331298828] |
b5d6dda2-c122-483b-aed2-9328301a1621 | synthetic-data-for-english-lexical | null | null | https://aclanthology.org/2020.lrec-1.773 | https://aclanthology.org/2020.lrec-1.773.pdf | Synthetic Data for English Lexical Normalization: How Close Can We Get to Manually Annotated Data? | Social media is a valuable data resource for various natural language processing (NLP) tasks. However, standard NLP tools were often designed with standard texts in mind, and their performance decreases heavily when applied to social media data. One solution to this problem is to adapt the input text to a more standard form, a task also referred to as normalization. Automatic approaches to normalization have shown that they can be used to improve performance on a variety of NLP tasks. However, all of these systems are supervised, thereby being heavily dependent on the availability of training data for the correct language and domain. In this work, we attempt to overcome this dependence by automatically generating training data for lexical normalization. Starting with raw tweets, we attempt two directions, to insert non-standardness (noise) and to automatically normalize in an unsupervised setting. Our best results are achieved by automatically inserting noise. We evaluate our approaches by using an existing lexical normalization system; our best scores are achieved by custom error generation system, which makes use of some manually created datasets. With this system, we score 94.29 accuracy on the test data, compared to 95.22 when it is trained on human-annotated data. Our best system which does not depend on any type of annotation is based on word embeddings and scores 92.04 accuracy. Finally, we perform an experiment in which we asked humans to predict whether a sentence was written by a human or generated by our best model. This experiment showed that in most cases it is hard for a human to detect automatically generated sentences. | ['Rob van der Goot', 'Kelly Dekker'] | 2020-05-01 | null | null | null | lrec-2020-5 | ['lexical-normalization'] | ['natural-language-processing'] | [ 2.61836082e-01 1.63512096e-01 1.24910660e-01 -4.08090353e-01
-5.33506632e-01 -4.85287786e-01 7.27324903e-01 6.74906015e-01
-1.06336534e+00 8.26705337e-01 1.66383535e-01 -2.42151663e-01
3.18227321e-01 -1.12823856e+00 -3.67049485e-01 -3.41477871e-01
4.25645411e-01 5.56875527e-01 4.00255054e-01 -5.57150543e-01
3.32535028e-01 3.43123823e-01 -1.40952897e+00 8.11923370e-02
8.32462013e-01 3.63699794e-01 6.97528198e-02 6.00121677e-01
-6.69936240e-01 5.96361935e-01 -8.67902398e-01 -5.56242168e-01
1.67114899e-01 -4.79234457e-01 -8.51164222e-01 -1.86913371e-01
-7.76101137e-03 2.28430986e-01 2.43498236e-01 1.19921637e+00
5.26418984e-01 1.47706956e-01 6.68365240e-01 -9.20991957e-01
-5.20957053e-01 7.89065063e-01 -1.59763157e-01 1.69580594e-01
6.03193700e-01 -1.82504222e-01 8.32995415e-01 -7.01513827e-01
8.03285837e-01 9.05353963e-01 6.41287625e-01 4.94237304e-01
-1.35898483e+00 -5.60844302e-01 -3.49770129e-01 -1.76219177e-02
-1.49806225e+00 -4.38720256e-01 5.23827672e-01 -5.39537013e-01
1.16946638e+00 2.31451422e-01 2.73446053e-01 1.11783898e+00
-1.94125948e-03 2.39406779e-01 1.08147418e+00 -1.15815985e+00
1.66885644e-01 6.08210027e-01 2.90688902e-01 2.48593628e-01
2.43711442e-01 -3.67387801e-01 -1.79562226e-01 -1.42983556e-01
2.80244231e-01 -3.05409640e-01 -1.24086812e-01 2.11972352e-02
-1.11486697e+00 1.02767980e+00 7.47787505e-02 1.02054083e+00
-3.96207750e-01 -4.53371614e-01 4.67189372e-01 3.86104554e-01
6.36572540e-01 8.32750082e-01 -5.31659544e-01 -5.93567155e-02
-1.01107609e+00 1.91539496e-01 1.11614478e+00 6.85282111e-01
8.44167590e-01 -3.69667798e-01 -1.79346859e-01 1.06260204e+00
1.24129705e-01 4.06994969e-01 1.09199095e+00 -3.46574754e-01
4.98307317e-01 8.27345312e-01 1.18477829e-01 -1.23522806e+00
-6.60595298e-01 -6.88386112e-02 -6.23300135e-01 2.04981714e-02
6.49789751e-01 -3.55677485e-01 -8.29202354e-01 1.71082449e+00
2.33904243e-01 -3.07647318e-01 2.85568029e-01 4.87563491e-01
7.60839224e-01 7.19786048e-01 2.37407014e-01 -3.66045982e-01
1.24944746e+00 -6.51428699e-01 -9.44653511e-01 -1.48727760e-01
1.06953216e+00 -1.23184812e+00 1.35541117e+00 3.99786621e-01
-6.19984090e-01 -4.70480442e-01 -1.01861513e+00 1.16852127e-01
-8.66917193e-01 -1.42035764e-02 2.04833388e-01 9.93883491e-01
-8.44747901e-01 6.63941920e-01 -5.39963186e-01 -9.68460262e-01
1.49269181e-03 3.09345037e-01 -5.35292327e-01 3.25416505e-01
-1.44334459e+00 1.15931439e+00 5.98473251e-01 -4.24375564e-01
5.83122000e-02 -1.94247320e-01 -7.66187191e-01 -1.50336340e-01
3.32939684e-01 -3.86207491e-01 1.19948268e+00 -1.15971029e+00
-1.58570671e+00 1.01523817e+00 -9.95569676e-02 -5.96563756e-01
5.37945330e-01 -7.15589002e-02 -7.80732572e-01 -1.69112176e-01
2.91986436e-01 3.75479698e-01 5.01749814e-01 -9.25147355e-01
-4.78556395e-01 -9.50726047e-02 -5.24365380e-02 -1.71738535e-01
-8.96873057e-01 4.20202434e-01 -2.40643010e-01 -6.57069027e-01
-1.00580275e-01 -9.15394843e-01 -2.70848513e-01 -5.39521813e-01
-3.48722458e-01 -3.63662720e-01 3.50329280e-01 -5.83534658e-01
1.57558215e+00 -1.87993777e+00 -2.17725202e-01 4.56756681e-01
-1.21522829e-01 6.42304420e-01 -6.43463656e-02 6.22981131e-01
-7.72047266e-02 6.36645436e-01 -2.10883573e-01 -4.00519162e-01
9.13546383e-02 3.52044642e-01 -1.93249419e-01 8.42937604e-02
2.28286743e-01 5.43082237e-01 -8.21724832e-01 -7.46513903e-01
8.54198411e-02 3.65332067e-01 -5.80474854e-01 2.57784545e-01
-1.34907349e-03 1.70878902e-01 -1.22727498e-01 -9.18381102e-03
4.31603014e-01 1.50281385e-01 1.28688380e-01 -1.22236144e-02
-3.04402858e-01 3.57139051e-01 -1.29675055e+00 1.37438262e+00
-4.85999852e-01 6.38187945e-01 -3.56115520e-01 -8.33587646e-01
1.11375177e+00 4.18030083e-01 4.00196493e-01 -6.29171491e-01
4.13039386e-01 2.05659181e-01 1.72988828e-02 -7.86975563e-01
7.43868709e-01 -1.85270503e-01 -1.87437713e-01 6.37077391e-01
3.24528009e-01 -2.10007772e-01 8.20392370e-01 1.74867094e-01
1.08558965e+00 -2.11954579e-01 6.40955031e-01 -1.67843431e-01
7.64408410e-01 1.02789074e-01 3.81275386e-01 7.82438636e-01
-1.03256151e-01 6.39230132e-01 4.88282830e-01 -3.14563215e-01
-1.20164084e+00 -5.95509708e-01 -1.76599324e-01 1.06419110e+00
-4.77088243e-01 -8.29939902e-01 -1.05620778e+00 -6.18996263e-01
-5.44986665e-01 9.09210384e-01 -4.52473611e-01 1.19368114e-01
-4.21398729e-01 -1.07278681e+00 6.17761850e-01 1.27669051e-01
9.56639349e-02 -1.25498915e+00 -3.06191981e-01 4.37277675e-01
-2.87319720e-01 -1.26132500e+00 -1.80616483e-01 2.07533166e-01
-4.42356586e-01 -1.10480249e+00 -3.31260890e-01 -6.88635468e-01
6.20641053e-01 -9.99397039e-02 9.91558731e-01 2.15920448e-01
-3.99005227e-02 6.24657124e-02 -7.88081110e-01 -7.08447039e-01
-9.09689844e-01 5.30940473e-01 2.21895248e-01 6.83541447e-02
7.99122870e-01 -3.96682680e-01 4.93916273e-02 2.66534537e-01
-1.36826980e+00 -2.21955508e-01 4.75014329e-01 6.29960060e-01
3.39815527e-01 7.39753619e-02 5.73610485e-01 -1.25445962e+00
1.08011889e+00 -3.13955277e-01 -3.76940072e-01 1.37669414e-01
-6.71181977e-01 1.46657094e-01 9.04386759e-01 -4.14123476e-01
-7.68026948e-01 1.94896355e-01 -7.10855424e-01 2.02566087e-01
-5.20845354e-01 5.69471896e-01 -1.50188789e-01 -3.16113792e-02
1.18847239e+00 -1.49120361e-01 -8.65575895e-02 -4.68581915e-01
1.70659348e-01 1.09084034e+00 1.41500801e-01 -2.93989748e-01
8.80862355e-01 -7.57120326e-02 -4.77045059e-01 -1.17776191e+00
-9.38156128e-01 -5.35534263e-01 -9.31601048e-01 6.79484894e-03
8.64063382e-01 -3.49417418e-01 -2.74414092e-01 2.71364748e-01
-1.28633249e+00 -1.05804592e-01 -3.00668448e-01 6.44418120e-01
-1.22942917e-01 4.65185344e-01 -2.12298438e-01 -6.05944812e-01
-3.16076458e-01 -7.40433991e-01 4.12207216e-01 1.36942565e-01
-6.91228926e-01 -9.45103526e-01 4.47572291e-01 2.66277939e-01
6.04002893e-01 1.36194557e-01 7.33028710e-01 -1.36818504e+00
3.52337539e-01 -4.81805414e-01 -3.50159369e-02 6.83533013e-01
3.11096430e-01 2.99524248e-01 -9.78475273e-01 8.00417438e-02
9.50884447e-02 -1.24031916e-01 4.86445069e-01 -2.14012682e-01
7.47406960e-01 -3.23731780e-01 -5.35040274e-02 9.80857089e-02
1.28757346e+00 -4.06137221e-02 6.89134777e-01 6.43718243e-01
4.72628474e-01 6.78095937e-01 4.82299596e-01 2.19788089e-01
2.95676112e-01 6.78460956e-01 4.05398272e-02 -6.13426380e-02
2.40528747e-01 -2.33245090e-01 3.27672333e-01 1.04586256e+00
-6.00488223e-02 -4.53660101e-01 -1.16835332e+00 3.74937445e-01
-1.70218372e+00 -9.19364870e-01 -4.54141229e-01 2.18729377e+00
1.08712542e+00 4.93065566e-01 9.28245932e-02 5.46747029e-01
7.34042466e-01 -1.67905211e-01 2.98164606e-01 -7.59193003e-01
-1.98655352e-01 4.60544109e-01 4.34216321e-01 5.67115903e-01
-1.21714759e+00 1.03033853e+00 6.02354574e+00 7.19047010e-01
-1.10530245e+00 2.12009698e-01 1.06863081e-01 2.41176143e-01
-1.57814957e-02 -1.04204781e-01 -9.77968693e-01 4.78260309e-01
1.22398889e+00 -4.52154040e-01 1.91160873e-01 5.99066377e-01
3.58245134e-01 -6.36223480e-02 -8.24998140e-01 7.92404294e-01
3.68320733e-01 -8.62099528e-01 3.24603394e-02 -2.68590122e-01
4.78999287e-01 -1.19141951e-01 -5.73595285e-01 5.68026960e-01
1.49354368e-01 -9.04731214e-01 4.02407318e-01 4.11710888e-01
2.03379452e-01 -6.36403620e-01 1.22192359e+00 5.92527747e-01
-7.40657210e-01 3.54153425e-01 -4.48033988e-01 -3.36384922e-01
1.67125911e-01 7.33542264e-01 -1.20510411e+00 2.80377597e-01
4.04313207e-01 2.40285516e-01 -7.97484994e-01 9.30693209e-01
-4.97456998e-01 6.10907853e-01 -5.37976980e-01 -4.17145550e-01
8.94313827e-02 -1.40416548e-02 4.33217496e-01 1.57169402e+00
2.40482703e-01 -1.49404407e-01 1.37290850e-01 5.54778337e-01
-1.28389165e-01 1.03293574e+00 -6.43173099e-01 -1.39049456e-01
2.13009477e-01 1.38090646e+00 -9.21002567e-01 -3.67611945e-01
-4.61479962e-01 8.33172023e-01 4.17017490e-01 -2.96990341e-03
-8.16136301e-01 -8.05964530e-01 1.81804448e-01 3.45321268e-01
-3.25340219e-02 -1.43394321e-01 -1.07824720e-01 -1.25212932e+00
-3.87225486e-02 -7.16157317e-01 3.78205031e-01 -6.57547891e-01
-1.50828528e+00 9.98515785e-01 -8.15336108e-02 -1.05785418e+00
-4.05074686e-01 -7.01595962e-01 -3.78528357e-01 7.38894165e-01
-1.36557615e+00 -7.33986855e-01 -2.14476600e-01 4.97487426e-01
2.39867970e-01 -1.70229971e-01 1.24366379e+00 6.58747435e-01
-4.97843206e-01 7.03636408e-01 -1.07076667e-01 4.07233894e-01
1.28615487e+00 -1.17080414e+00 2.17068925e-01 1.02680099e+00
3.33986104e-01 6.47385180e-01 9.04650211e-01 -5.04288137e-01
-6.60925865e-01 -1.13541925e+00 1.83320117e+00 -4.10020113e-01
9.51767445e-01 -3.64630908e-01 -1.11676908e+00 5.11572182e-01
2.92569548e-01 -2.24523231e-01 1.12151611e+00 1.30420044e-01
-1.47933081e-01 1.37406597e-02 -1.24910927e+00 4.95961636e-01
5.75131536e-01 -3.05419773e-01 -1.06741619e+00 5.78516483e-01
5.87284744e-01 -1.88351914e-01 -8.55327606e-01 1.01477042e-01
1.68488473e-01 -8.05487752e-01 5.24729013e-01 -6.27000272e-01
3.74830097e-01 -4.03010309e-01 -4.51936051e-02 -1.39577758e+00
-9.88338962e-02 -3.67197752e-01 6.36772335e-01 1.77060270e+00
7.87792206e-01 -7.77885020e-01 4.36092734e-01 7.22302258e-01
1.19837239e-01 -1.75611824e-01 -5.80015123e-01 -7.27301717e-01
4.19027992e-02 -6.36705399e-01 5.71414649e-01 1.21531081e+00
1.49363101e-01 6.94560528e-01 -1.88937262e-01 -7.24044070e-02
1.50361136e-01 -4.25229073e-01 1.02975953e+00 -1.33677244e+00
7.10250884e-02 -4.22785223e-01 -5.41841984e-01 -3.40932697e-01
2.89683074e-01 -9.97915745e-01 8.10463578e-02 -1.36056018e+00
-1.03231519e-01 -3.58789802e-01 -8.41725096e-02 5.84347606e-01
-1.80130564e-02 5.74859202e-01 1.76733598e-01 4.03669402e-02
-4.16480511e-01 1.95413411e-01 7.50146866e-01 2.49346364e-02
-5.02787948e-01 -4.18216996e-02 -6.44144237e-01 9.54373240e-01
1.15654767e+00 -8.60441566e-01 3.06528062e-02 -2.57936627e-01
5.45860946e-01 -5.99269927e-01 -3.53974067e-02 -1.12426877e+00
3.47117513e-01 -9.25540105e-02 1.40122443e-01 -1.93691865e-01
-9.97125283e-02 -7.91127205e-01 1.12596020e-01 1.58737630e-01
-2.77644694e-01 8.29936862e-02 1.45634636e-01 5.48723973e-02
-3.29853565e-01 -7.61312842e-01 9.06781912e-01 -8.28442574e-02
-4.77708966e-01 -2.70610839e-01 -6.54211283e-01 1.70626387e-01
1.02039695e+00 -3.63879814e-03 -1.81072816e-01 -1.57268658e-01
-8.48091364e-01 -7.55138919e-02 6.46917105e-01 4.26526129e-01
9.09991190e-02 -1.11459446e+00 -6.44345939e-01 3.36210877e-01
2.63274491e-01 -3.47754091e-01 -2.85847485e-01 6.46779239e-01
-7.03521907e-01 3.46678048e-01 -1.92490354e-01 -3.16224724e-01
-1.22652352e+00 5.94058096e-01 9.17703211e-02 -4.95175570e-01
-2.19537064e-01 4.09133554e-01 -5.87886035e-01 -7.24125803e-01
6.46296218e-02 -4.09798443e-01 -7.67058313e-01 5.17643690e-01
6.23974979e-01 1.19111374e-01 5.74588478e-01 -8.77534509e-01
-2.15816811e-01 4.81038481e-01 -9.70392395e-03 -1.70852661e-01
1.45487642e+00 4.97521386e-02 -3.43947440e-01 5.07863402e-01
1.07004058e+00 4.73914415e-01 -1.75359473e-01 -1.59888476e-01
2.54847914e-01 -3.68503630e-01 -1.38395831e-01 -5.64086020e-01
-6.78237200e-01 6.11396432e-01 4.29148823e-01 7.56437063e-01
9.36602592e-01 -3.38064075e-01 6.51717663e-01 6.63097501e-01
2.30066359e-01 -1.36386764e+00 -2.07180396e-01 9.27223742e-01
5.82495868e-01 -1.43513680e+00 -3.13910507e-02 -3.75306010e-01
-4.82086122e-01 1.33744216e+00 4.50502366e-01 -9.98152569e-02
7.43642569e-01 2.70679891e-01 3.50763530e-01 3.02331746e-02
-3.68630379e-01 -3.43756586e-01 8.21448714e-02 4.19172317e-01
7.52517521e-01 -8.57008621e-02 -1.00144804e+00 8.35432291e-01
-6.16591036e-01 9.88144353e-02 8.42730701e-01 8.44933271e-01
-5.15646636e-01 -1.68386626e+00 -5.36359549e-01 4.79390681e-01
-8.04932117e-01 -1.11651339e-01 -6.62824690e-01 8.21788430e-01
2.66131103e-01 1.19145954e+00 -9.81866494e-02 -4.01781827e-01
6.27911031e-01 3.37710738e-01 1.59230426e-01 -1.06909180e+00
-9.35937703e-01 -2.51018733e-01 2.20459819e-01 -1.69878557e-01
-6.27691507e-01 -6.16609395e-01 -1.21444905e+00 -4.15565729e-01
-4.05368239e-01 4.64293152e-01 8.25672150e-01 1.04632711e+00
2.76252022e-03 2.84051895e-01 3.76157552e-01 -6.97110236e-01
-2.94041246e-01 -1.21041346e+00 -2.66065598e-01 8.26514482e-01
-1.02211237e-01 -3.78116012e-01 -6.29240096e-01 2.81343818e-01] | [10.153371810913086, 9.77626895904541] |
d0b74230-920d-47f0-9ba1-80138d88f9c9 | sheffield-multimt-using-object-posterior | null | null | https://aclanthology.org/W17-4752 | https://aclanthology.org/W17-4752.pdf | Sheffield MultiMT: Using Object Posterior Predictions for Multimodal Machine Translation | null | ['Pranava Swaroop Madhyastha', 'Lucia Specia', 'Josiah Wang'] | 2017-09-01 | null | null | null | ws-2017-9 | ['multimodal-machine-translation'] | ['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] | [-7.346563816070557, 3.7064201831817627] |
e42645c8-b241-4c71-a8c3-c487283a25b7 | multi-task-multi-channel-multi-input-learning | null | null | https://aclanthology.org/D19-6208 | https://aclanthology.org/D19-6208.pdf | Multi-Task, Multi-Channel, Multi-Input Learning for Mental Illness Detection using Social Media Text | We investigate the impact of using emotional patterns identified by the clinical practitioners and computational linguists to enhance the prediction capabilities of a mental illness detection (in our case depression and post-traumatic stress disorder) model built using a deep neural network architecture. Over the years, deep learning methods have been successfully used in natural language processing tasks, including a few in the domain of mental illness and suicide ideation detection. We illustrate the effectiveness of using multi-task learning with a multi-channel convolutional neural network as the shared representation and use additional inputs identified by researchers as indicatives in detecting mental disorders to enhance the model predictability. Given the limited amount of unstructured data available for training, we managed to obtain a task-specific AUC higher than 0.90. In comparison to methods such as multi-class classification, we identified multi-task learning with multi-channel convolution neural network and multiple-inputs to be effective in detecting mental disorders. | ['Prasadith Kirinde Gamaarachchige', 'Diana Inkpen'] | 2019-11-01 | null | null | null | ws-2019-11 | ['3d-object-classification'] | ['computer-vision'] | [ 2.52292007e-01 3.44890594e-01 5.42752966e-02 -6.57136261e-01
-6.38407648e-01 -6.65888339e-02 2.84307182e-01 6.55582070e-01
-6.82492614e-01 5.72809219e-01 3.46531600e-01 -3.18493992e-02
-3.10661167e-01 -6.37614429e-01 -5.13989059e-03 -2.92576700e-01
-3.67670864e-01 5.59656799e-01 -5.47724545e-01 -3.00374061e-01
2.61600882e-01 3.28319639e-01 -1.27949727e+00 1.10912919e+00
6.56987906e-01 9.49609995e-01 -2.56696522e-01 3.00556362e-01
-1.55650884e-01 8.30814898e-01 -4.16326940e-01 -6.32659197e-01
-1.68444589e-01 -9.69159044e-03 -1.07726967e+00 -1.70505106e-01
-3.09708178e-01 -3.92762959e-01 1.57605141e-01 5.97075164e-01
1.02040827e+00 -4.49319258e-02 6.70061052e-01 -1.06926608e+00
-6.36416614e-01 4.52458054e-01 -2.77560294e-01 2.03212142e-01
5.26851535e-01 9.43947434e-02 6.93171978e-01 -8.79591882e-01
5.79730272e-01 1.22002530e+00 1.13891613e+00 6.03585839e-01
-1.07584476e+00 -9.19972420e-01 -3.44894916e-01 2.18407393e-01
-1.16627014e+00 -5.07436872e-01 7.62284994e-01 -7.06665218e-01
1.74404252e+00 -2.60757297e-01 5.57533443e-01 1.31198215e+00
5.81802666e-01 4.61281002e-01 1.27003860e+00 -3.09507281e-01
3.00927758e-02 5.15594840e-01 1.12322263e-01 8.07597995e-01
-1.52814994e-03 -1.86777592e-01 -3.30806077e-01 -7.00583160e-01
2.63325423e-01 7.45857581e-02 2.85248011e-01 3.84812683e-01
-8.38928699e-01 1.16909873e+00 5.18665574e-02 6.87622190e-01
-8.31239820e-01 -2.01031834e-01 1.27491486e+00 4.03035611e-01
9.51531053e-01 6.35543227e-01 -7.82831967e-01 -2.34165609e-01
-1.08963287e+00 -6.73865974e-02 6.07729256e-01 -6.86914194e-03
4.34207946e-01 -2.96970103e-02 -2.28454843e-01 1.23104668e+00
-7.65402615e-02 -9.29083601e-02 9.53397036e-01 -5.67487776e-01
2.11489141e-01 9.30235684e-01 -3.89604807e-01 -1.00358760e+00
-1.00262332e+00 -2.22608298e-01 -1.09036732e+00 1.92706853e-01
-3.80006395e-02 -6.12006664e-01 -7.31241763e-01 1.48005664e+00
-1.64985672e-01 -1.38112530e-01 4.60285366e-01 4.65184808e-01
7.75721550e-01 2.25787178e-01 6.18577361e-01 -1.08491220e-01
1.59784198e+00 -4.27636534e-01 -7.52382100e-01 -3.81317884e-01
1.23961806e+00 -4.13683176e-01 5.91650188e-01 6.60909176e-01
-9.71649528e-01 -2.95248836e-01 -6.78283095e-01 1.19501762e-01
-6.26828074e-01 2.94740260e-01 9.43103492e-01 6.53177083e-01
-1.01411402e+00 6.13691986e-01 -6.02251470e-01 -7.06896067e-01
8.36533487e-01 5.81726491e-01 -7.92566955e-01 7.66525343e-02
-1.44650364e+00 1.43200588e+00 4.85452294e-01 -1.82594240e-01
-7.51167893e-01 -5.03556967e-01 -6.87394679e-01 1.80875361e-01
-3.47067028e-01 -5.96274257e-01 7.15775847e-01 -1.54577291e+00
-1.01443362e+00 1.42023957e+00 -3.85907479e-02 -5.62308073e-01
1.34069100e-01 4.19362076e-02 -6.72677994e-01 2.95610696e-01
6.08003475e-02 6.40441477e-01 4.14772034e-01 -5.71621299e-01
-1.98997781e-01 -5.61131477e-01 -1.64401352e-01 4.50116806e-02
-7.63670206e-01 7.65815794e-01 3.92520875e-01 -4.11232680e-01
-1.80027619e-01 -6.15950465e-01 -3.67561519e-01 -1.56312019e-01
-1.12732418e-01 -5.52296042e-01 5.06279349e-01 -8.74760449e-01
9.54857290e-01 -1.97758961e+00 -3.18836093e-01 1.55561075e-01
3.26001823e-01 3.78895879e-01 5.43792583e-02 5.23228526e-01
-6.52540803e-01 4.50568825e-01 -1.67853311e-01 -5.05243063e-01
-3.02409410e-01 8.03783461e-02 3.92204612e-01 3.69443744e-01
9.07837927e-01 8.52286041e-01 -7.10135162e-01 -4.45787251e-01
5.97940423e-02 6.91618979e-01 -4.88024563e-01 1.51489750e-01
4.16008443e-01 5.39855275e-04 -5.03134787e-01 6.37583673e-01
5.73955357e-01 -1.48199111e-01 1.34176761e-01 1.21745408e-01
1.61336586e-01 2.36454889e-01 -6.54161274e-01 1.45577121e+00
-4.18617070e-01 4.75516528e-01 5.17520793e-02 -1.29007363e+00
1.31989622e+00 9.87257898e-01 8.62609029e-01 -6.76747680e-01
3.12260449e-01 1.72838554e-01 1.76122516e-01 -9.74690616e-01
2.73481086e-02 -9.65947628e-01 1.53198782e-02 4.75160033e-01
3.00793797e-01 3.28654557e-01 -4.26025599e-01 -2.45564580e-02
1.23764122e+00 -2.51800627e-01 4.96239007e-01 -1.98240995e-01
2.95535952e-01 -1.81349277e-01 6.29018128e-01 2.38415316e-01
-4.72453415e-01 4.24782962e-01 7.90601909e-01 -8.08159709e-01
-1.02180767e+00 -1.73189223e-01 -5.27550101e-01 9.85885620e-01
-1.10344005e+00 -3.60576749e-01 -3.29331368e-01 -3.80696625e-01
-1.16025686e-01 5.52595019e-01 -9.87211823e-01 -4.62448299e-01
1.85105242e-02 -1.28783977e+00 1.16231155e+00 5.72454870e-01
3.50864500e-01 -1.40519106e+00 -9.52479780e-01 5.52218139e-01
-1.35828227e-01 -1.00912082e+00 4.03136760e-01 7.83886671e-01
-6.39068842e-01 -8.99274051e-01 -7.09510148e-01 -7.65061915e-01
2.08123296e-01 -6.89874768e-01 1.07317626e+00 4.26699743e-02
-5.82286477e-01 2.68036216e-01 -3.05091918e-01 -5.55630684e-01
-3.58304411e-01 -4.55921032e-02 2.24105999e-01 3.69354971e-02
9.21433628e-01 -5.13172865e-01 -4.39985365e-01 -5.00925124e-01
-8.80986869e-01 -1.36757731e-01 6.71499193e-01 9.18934464e-01
-8.43411088e-02 -1.19911365e-01 1.10043240e+00 -9.79791760e-01
1.26345789e+00 -1.04364133e+00 6.46426797e-01 4.82314490e-02
-7.86907911e-01 -2.04977527e-01 3.55679989e-01 -4.46325392e-01
-8.53546798e-01 1.01101108e-01 -5.16158640e-01 -3.43065023e-01
-6.97689593e-01 9.20531988e-01 4.34618026e-01 8.26583803e-02
7.62534499e-01 -5.68499565e-02 1.82288811e-01 -2.64040738e-01
-3.64668906e-01 9.48879182e-01 -1.93804383e-01 -2.39355356e-01
-2.45625153e-01 1.42431214e-01 9.86407883e-03 -5.65682054e-01
-4.76830482e-01 -5.31060755e-01 -6.86081529e-01 -8.20193738e-02
1.13481414e+00 -8.61056745e-01 -6.23306990e-01 4.07823920e-01
-1.23357642e+00 -3.54458392e-02 2.10270137e-01 5.18632770e-01
-3.07148099e-01 1.49504408e-01 -7.00954795e-01 -1.21184659e+00
-9.56159830e-01 -7.77496815e-01 8.78212392e-01 -1.35867670e-01
-7.01790273e-01 -1.14833546e+00 3.26059282e-01 1.47713393e-01
4.84923750e-01 6.08004510e-01 1.28614557e+00 -1.19112861e+00
8.85327876e-01 -2.71514714e-01 -2.71485865e-01 4.08147097e-01
1.62877291e-01 -9.69313830e-02 -1.23393762e+00 3.89736295e-02
6.84106275e-02 -9.72351730e-01 8.28603208e-01 3.22131217e-01
1.08172333e+00 -1.01934351e-01 -2.29802921e-01 2.16245174e-01
1.40573466e+00 4.41204727e-01 7.89767683e-01 3.93032193e-01
2.29018301e-01 9.33195889e-01 -1.76887587e-02 9.05719042e-01
3.85533482e-01 2.61174440e-01 1.37901127e-01 -4.25427645e-01
4.07593518e-01 4.23242420e-01 2.95066088e-01 1.50571927e-01
-1.19451798e-01 1.67446747e-01 -1.53239620e+00 7.53938258e-01
-1.55536234e+00 -1.02193916e+00 5.36638536e-02 1.61377311e+00
6.58257782e-01 1.00949690e-01 1.28729284e-01 2.91418225e-01
4.84026760e-01 -2.38753989e-01 -3.68513405e-01 -1.31366229e+00
-2.85014421e-01 6.37847960e-01 -4.10030484e-02 6.84375642e-03
-1.02450168e+00 7.13066161e-01 6.80537796e+00 3.56903404e-01
-1.32359695e+00 3.01848590e-01 1.06960618e+00 -1.91499859e-01
3.28036696e-02 -5.34742475e-01 -2.59964198e-01 3.38448465e-01
1.40439582e+00 1.57347560e-01 8.60095397e-02 5.64611793e-01
4.95980948e-01 4.56113704e-02 -8.53095114e-01 8.95594776e-01
1.82641014e-01 -8.57865870e-01 -1.27641603e-01 -4.73657735e-02
3.21284175e-01 2.77292430e-01 -5.44862449e-02 4.87054884e-01
-1.33799249e-02 -1.42533648e+00 -7.55864084e-02 7.33981252e-01
9.55518067e-01 -9.47300792e-01 1.23283648e+00 1.56268984e-01
-6.33867562e-01 -3.09798360e-01 -2.62367070e-01 -4.54159409e-01
-1.66847065e-01 5.26738226e-01 -1.24443233e+00 1.39819950e-01
8.18158209e-01 7.48714507e-01 -3.03073019e-01 6.77334845e-01
3.42821836e-01 6.80213451e-01 -1.53234199e-01 3.36310565e-02
5.36007166e-01 9.60282162e-02 -2.95243207e-02 1.66299868e+00
2.46502087e-01 4.02996629e-01 8.56158063e-02 7.20634520e-01
6.62797913e-02 6.07737541e-01 -8.67099404e-01 -2.79991776e-01
-2.80190200e-01 1.46721530e+00 -4.69561875e-01 -3.71104002e-01
-4.74511355e-01 1.00881970e+00 4.41144377e-01 1.28989890e-02
-4.52581435e-01 -4.20887858e-01 5.45133471e-01 3.65413278e-02
-1.71960816e-01 2.40800217e-01 -5.99882722e-01 -8.89160395e-01
-3.81861567e-01 -7.75885820e-01 6.24274850e-01 -9.44117367e-01
-1.42440045e+00 9.71151531e-01 -3.98181051e-01 -8.05182934e-01
-4.14589345e-01 -7.00993299e-01 -8.00533652e-01 1.16015649e+00
-1.36686158e+00 -1.29097116e+00 1.97990298e-01 7.99997807e-01
2.77128249e-01 -3.33596945e-01 1.62760973e+00 3.71059865e-01
-5.92003226e-01 4.79064047e-01 -1.92926019e-01 3.68864119e-01
9.05793905e-01 -8.87311220e-01 -3.57888758e-01 2.21091643e-01
-7.45575428e-01 4.09313977e-01 2.91637123e-01 -5.96634388e-01
-7.62386680e-01 -9.80112076e-01 1.36374605e+00 -1.33449674e-01
5.53282976e-01 -5.17380163e-02 -1.03090358e+00 5.63048720e-01
4.95868057e-01 -4.53176588e-01 1.40098131e+00 5.58382452e-01
-1.97537169e-01 1.86567217e-01 -1.72589934e+00 -6.83463505e-03
4.27775562e-01 -7.36307800e-01 -7.14845717e-01 4.57295626e-01
-3.59916054e-02 2.17658296e-01 -1.22112596e+00 3.61016542e-01
6.10230803e-01 -8.10836613e-01 6.67631924e-01 -1.21317780e+00
1.13477683e+00 8.16882789e-01 1.36176199e-01 -1.40061736e+00
-6.23613238e-01 -1.19463563e-01 4.19685483e-01 1.03238106e+00
5.53018034e-01 -7.09110737e-01 4.81225699e-01 1.10774899e+00
-5.93134724e-02 -1.19263864e+00 -9.64664280e-01 4.38826680e-02
3.61407876e-01 -4.07883763e-01 1.36884883e-01 1.30603516e+00
6.46121323e-01 4.59501952e-01 -4.41990763e-01 -2.74281442e-01
-8.20023343e-02 -3.34608972e-01 -8.01691115e-02 -1.47891366e+00
6.11914247e-02 -3.94628555e-01 -5.53563416e-01 3.44350398e-01
4.60834384e-01 -1.20247686e+00 -5.18294454e-01 -1.66627765e+00
5.55890560e-01 -3.10242683e-01 -8.60858858e-01 1.20906293e+00
4.38623019e-02 2.20816210e-01 -1.41891047e-01 -1.73758522e-01
-2.52037734e-01 3.48117799e-01 6.67826593e-01 -8.41348320e-02
-1.67627588e-01 -1.64875150e-01 -1.06446838e+00 8.38030338e-01
9.29564416e-01 -6.92681253e-01 -7.30475932e-02 -2.59920508e-01
4.61897820e-01 1.95083335e-01 2.29172081e-01 -7.79564142e-01
-2.08335996e-01 1.97695047e-01 8.05785775e-01 -2.20558614e-01
6.27884269e-01 -6.02106392e-01 -6.11681901e-02 7.00942099e-01
-4.97042865e-01 6.22288436e-02 6.29084766e-01 -1.05035037e-01
-2.53122091e-01 -3.70568961e-01 7.27735043e-01 -4.55495030e-01
-5.59218884e-01 1.46387011e-01 -8.32044065e-01 -3.06703001e-01
8.35568845e-01 -1.17630914e-01 -2.60184892e-02 -3.49672616e-01
-1.16531551e+00 2.88477484e-02 -2.32795343e-01 2.68087089e-01
8.49938989e-01 -1.04173434e+00 -9.92563009e-01 2.76965201e-02
1.70648634e-01 -8.26721430e-01 3.80603105e-01 1.02674043e+00
-2.24955246e-01 5.31607449e-01 -8.10550392e-01 -9.70005617e-02
-1.39135432e+00 4.85785127e-01 5.92382133e-01 -5.97477913e-01
-4.90455627e-01 6.79187536e-01 -2.52988905e-01 -3.98640692e-01
-1.48063183e-01 -1.55267522e-01 -7.44088292e-01 5.06251454e-01
5.26919723e-01 2.04942644e-01 4.82771933e-01 -4.56034601e-01
-6.89326227e-01 1.13437116e-01 -1.35849655e-01 2.51194350e-02
1.96081662e+00 2.14000419e-01 -3.44111174e-01 5.00647604e-01
1.36474109e+00 -5.83503306e-01 -4.25911278e-01 3.78927812e-02
7.84023199e-03 3.10472459e-01 3.90706539e-01 -1.20578706e+00
-9.41673338e-01 1.24303746e+00 9.29248273e-01 1.35083735e-01
1.41204572e+00 -2.25787178e-01 5.21911561e-01 5.24208963e-01
5.89503422e-02 -1.27501357e+00 1.50006600e-02 6.17365301e-01
8.33611250e-01 -1.54126573e+00 -3.22263092e-01 1.83372170e-01
-1.20018387e+00 1.46855175e+00 5.00741839e-01 -2.70153999e-01
9.12834883e-01 3.94488275e-01 -1.59851927e-02 -4.89056915e-01
-1.12243509e+00 -2.05292497e-04 7.70587325e-02 6.35089517e-01
8.48034918e-01 1.56865418e-01 -6.17934585e-01 1.14818537e+00
1.63387954e-01 5.00833631e-01 3.25493157e-01 6.46747708e-01
-2.21890032e-01 -1.08540511e+00 -6.82722405e-02 1.01904440e+00
-1.15279222e+00 -4.15922642e-01 -6.47907078e-01 2.79180855e-01
5.40860415e-01 8.81315351e-01 1.24990135e-01 -5.13401568e-01
9.82925445e-02 8.85766327e-01 -1.77023299e-02 -7.13574290e-01
-1.41590917e+00 -1.43303782e-01 5.26147723e-01 -4.29002672e-01
-6.66910648e-01 -6.99781418e-01 -1.13025177e+00 -2.86620157e-03
7.76002407e-02 -1.34780437e-01 6.15788877e-01 1.20712578e+00
6.54178441e-01 5.88465035e-01 3.86451304e-01 -3.32444727e-01
-1.49814889e-01 -1.43435454e+00 -6.71523988e-01 5.71921587e-01
2.21746013e-01 -6.04387820e-01 -6.43758178e-02 -6.49533942e-02] | [12.885971069335938, 5.978109359741211] |
c5818d4b-8509-4ea2-9bef-3b85722ad70c | estimation-with-low-rank-time-frequency | 1804.09497 | null | http://arxiv.org/abs/1804.09497v2 | http://arxiv.org/pdf/1804.09497v2.pdf | Estimation with Low-Rank Time-Frequency Synthesis Models | Many state-of-the-art signal decomposition techniques rely on a low-rank
factorization of a time-frequency (t-f) transform. In particular, nonnegative
matrix factorization (NMF) of the spectrogram has been considered in many audio
applications. This is an analysis approach in the sense that the factorization
is applied to the squared magnitude of the analysis coefficients returned by
the t-f transform. In this paper we instead propose a synthesis approach, where
low-rankness is imposed to the synthesis coefficients of the data signal over a
given t-f dictionary (such as a Gabor frame). As such we offer a novel modeling
paradigm that bridges t-f synthesis modeling and traditional analysis-based NMF
approaches. The proposed generative model allows in turn to design more
sophisticated multi-layer representations that can efficiently capture diverse
forms of structure. Additionally, the generative modeling allows to exploit t-f
low-rankness for compressive sensing. We present efficient iterative shrinkage
algorithms to perform estimation in the proposed models and illustrate the
capabilities of the new modeling paradigm over audio signal processing
examples. | ['Matthieu Kowalski', 'Cédric Févotte'] | 2018-04-25 | null | null | null | null | ['audio-signal-processing'] | ['audio'] | [ 6.08533084e-01 -1.50038168e-01 5.97100612e-03 9.42604095e-02
-6.40772104e-01 -5.60260117e-01 3.89037937e-01 -1.90849021e-01
-1.07738376e-01 5.03025413e-01 6.11396313e-01 -7.28910742e-03
-5.13684273e-01 -3.74482423e-01 -4.99827743e-01 -9.17442083e-01
-9.67026129e-02 -2.63034344e-01 -3.73606265e-01 -2.61136025e-01
-2.68332679e-02 3.56281042e-01 -1.69775474e+00 4.57447648e-01
6.66433215e-01 8.94628763e-01 3.89634222e-01 7.65440941e-01
2.17059985e-01 4.92255390e-01 -3.63446504e-01 -3.29312170e-03
3.37773383e-01 -4.89412993e-01 -1.95859522e-01 3.05988580e-01
3.54271621e-01 1.30903739e-02 -3.28566253e-01 1.10436833e+00
2.10186198e-01 3.89691830e-01 5.73245704e-01 -9.32736695e-01
-3.91437173e-01 4.24473137e-01 -5.17270207e-01 8.56724828e-02
7.02842593e-01 -6.40850365e-01 1.12481868e+00 -1.45386612e+00
6.38776600e-01 1.04466164e+00 6.79364264e-01 3.88688594e-02
-1.35443187e+00 -2.97381490e-01 1.42502725e-01 1.08841226e-01
-1.24795961e+00 -7.22569048e-01 1.16293991e+00 -5.40414214e-01
7.40227938e-01 5.06784379e-01 6.84581876e-01 6.76554143e-01
1.98149890e-01 8.28034580e-01 9.55565333e-01 -9.14291799e-01
1.73788443e-01 -4.54743922e-01 -2.39044912e-02 4.32502180e-01
2.36713022e-01 -7.46783763e-02 -9.29090023e-01 -3.40216041e-01
7.57262707e-01 8.27577114e-02 -5.30748427e-01 -3.41835171e-01
-1.28332055e+00 7.16296554e-01 -5.77875739e-03 6.53226912e-01
-7.49377370e-01 3.42159927e-01 3.18783253e-01 3.35658073e-01
4.62684005e-01 3.02691042e-01 -6.71061277e-02 4.82256822e-02
-1.32563996e+00 2.79105723e-01 7.58510768e-01 5.37225544e-01
6.25257015e-01 7.14453340e-01 -8.03593099e-02 8.52541506e-01
5.12822866e-01 5.42419314e-01 5.81339300e-01 -9.93107855e-01
4.70274240e-01 -5.19044437e-02 4.78282496e-02 -1.26010609e+00
-2.57366985e-01 -8.79406095e-01 -9.98564124e-01 -1.96956262e-01
2.30551735e-01 -1.66098047e-02 -5.91443419e-01 1.83962905e+00
3.17426801e-01 5.03057778e-01 2.68397361e-01 7.90493667e-01
4.41802979e-01 7.72181928e-01 -5.31440675e-01 -7.70241857e-01
1.21195555e+00 -4.76350605e-01 -1.16025555e+00 -4.79878858e-02
3.12259495e-01 -1.22435474e+00 6.61195040e-01 9.28901315e-01
-1.16432786e+00 -6.06915653e-01 -1.12767088e+00 1.79034382e-01
1.75025910e-01 5.79194188e-01 3.96287978e-01 6.46298110e-01
-9.32051897e-01 5.21552205e-01 -5.89120448e-01 -6.57407194e-02
-2.20003560e-01 1.23590223e-01 -4.75870579e-01 -2.29749247e-01
-1.14356375e+00 5.12288272e-01 -7.10797980e-02 4.64896291e-01
-8.68492484e-01 -6.13256276e-01 -7.03418255e-01 7.89645836e-02
3.27686161e-01 -5.76553047e-01 9.48923647e-01 -8.57295334e-01
-1.29940319e+00 3.94831538e-01 -4.98719007e-01 -4.02076483e-01
1.70326516e-01 -5.96000552e-01 -5.31643212e-01 5.19506097e-01
3.08903195e-02 -2.17705369e-01 1.91080046e+00 -1.18387246e+00
-8.37287009e-02 -4.34221327e-01 -1.92416221e-01 1.56708192e-02
-5.46602488e-01 -2.44784802e-01 7.53366873e-02 -1.43951833e+00
8.78826201e-01 -9.17658269e-01 -3.40916425e-01 -3.72849643e-01
-2.55641818e-01 2.92511910e-01 7.11559534e-01 -8.10086727e-01
1.63008142e+00 -2.53239346e+00 6.12564623e-01 4.18778032e-01
1.21467784e-01 7.04643279e-02 -1.74797356e-01 9.92449701e-01
-4.92329240e-01 -4.81724113e-01 -1.18102781e-01 -5.05687475e-01
-1.49735287e-01 1.31331533e-01 -8.59750509e-01 7.30997741e-01
-4.54899631e-02 3.08016002e-01 -7.55173206e-01 4.42496128e-03
2.52866924e-01 8.23136032e-01 -8.13174844e-01 1.18273437e-01
1.37120232e-01 5.32075226e-01 -3.78302395e-01 3.75373840e-01
4.28148121e-01 8.21361914e-02 3.51240247e-01 -6.95351183e-01
-3.36290628e-01 7.61628300e-02 -1.54351330e+00 1.99592185e+00
-4.88662392e-01 4.23333794e-01 6.06250703e-01 -1.17794752e+00
9.37367082e-01 7.21839547e-01 8.34989846e-01 -8.63619670e-02
-7.05918074e-02 5.87670922e-01 1.47387926e-02 -3.74753833e-01
6.24402523e-01 -4.08660918e-01 1.73642784e-01 2.95693576e-01
1.08934119e-01 -5.55597171e-02 3.90180916e-01 8.33943561e-02
8.35403800e-01 -3.23646776e-02 5.37603498e-01 -3.68068933e-01
9.39501643e-01 -5.82517564e-01 4.25837606e-01 4.90392536e-01
2.81261593e-01 5.74462533e-01 1.46293445e-02 -2.38625944e-01
-8.73250842e-01 -6.61732674e-01 -5.19422293e-02 1.01903903e+00
-6.25665247e-01 -6.85474455e-01 -6.12202168e-01 9.53965634e-02
-5.83228208e-02 2.64001369e-01 -4.67453986e-01 1.35180466e-02
-7.33527780e-01 -4.09148753e-01 2.79564172e-01 2.61869878e-01
1.11573666e-01 -4.05651569e-01 -4.68761235e-01 4.02605385e-01
-3.93852293e-01 -1.12366056e+00 -3.87889057e-01 3.14489245e-01
-1.18722701e+00 -6.55546606e-01 -8.45190883e-01 -6.21089697e-01
6.03387117e-01 6.38687968e-01 7.28417099e-01 -2.37929478e-01
-8.00171867e-02 7.58380830e-01 -7.55847454e-01 -7.20804706e-02
-1.64282039e-01 -2.74172336e-01 4.82122362e-01 8.00044894e-01
-2.18723759e-01 -1.23964024e+00 -3.47533554e-01 1.14543393e-01
-1.11623085e+00 -9.98365507e-02 5.11014581e-01 9.90982413e-01
7.57929265e-01 2.32331336e-01 5.88065147e-01 -5.74623406e-01
8.10156584e-01 -2.82756090e-01 -4.15920615e-01 1.28414661e-01
-1.99673265e-01 1.03973467e-02 8.65604222e-01 -6.71582818e-01
-7.49798119e-01 1.93108201e-01 -1.26650231e-02 -9.48561251e-01
5.48837841e-01 8.94728005e-01 -1.28276348e-01 -1.65155992e-01
6.37337625e-01 5.21211207e-01 -1.51942477e-01 -8.34893346e-01
5.30914247e-01 2.89466441e-01 5.76992154e-01 -6.66664004e-01
9.82463479e-01 5.78812182e-01 3.61074954e-01 -1.37609804e+00
-6.51863933e-01 -6.55358136e-01 -5.18340766e-01 -3.26845050e-01
5.38977385e-01 -9.40389097e-01 -4.24922556e-01 1.06516056e-01
-1.08514154e+00 4.10916746e-01 -5.25057733e-01 1.05032372e+00
-6.75085664e-01 6.75034046e-01 -3.63155246e-01 -1.16831493e+00
-2.40306824e-01 -9.29827988e-01 9.68774080e-01 -3.73607814e-01
-3.16789180e-01 -6.50107145e-01 1.85909167e-01 5.32710738e-02
2.97954530e-01 7.20621571e-02 7.55448997e-01 -1.84653372e-01
-2.76406497e-01 -3.04764867e-01 3.83216500e-01 5.19565463e-01
1.93630114e-01 -2.45274216e-01 -1.04773140e+00 -7.19425082e-01
6.19966447e-01 3.71147692e-02 9.90257740e-01 7.61798978e-01
8.42203796e-01 -3.84134650e-01 3.19839492e-02 6.78515375e-01
1.36779785e+00 2.89764199e-02 3.16419482e-01 -1.01562753e-01
5.77740431e-01 5.31368434e-01 6.69654191e-01 9.07758057e-01
-2.45925233e-01 7.77887583e-01 1.93486482e-01 1.23307839e-01
-9.38043445e-02 -2.87812650e-01 5.90034366e-01 1.45575881e+00
-3.28067750e-01 3.34713496e-02 -4.46950376e-01 4.96433794e-01
-1.85464251e+00 -1.00385106e+00 -1.33046284e-01 2.38221121e+00
4.81355786e-01 -1.82681575e-01 7.01010525e-02 1.00947118e+00
5.36706090e-01 5.13976693e-01 -2.71622092e-01 -2.03481510e-01
2.12331470e-02 3.45603317e-01 1.50615916e-01 6.70653105e-01
-9.63460267e-01 2.69268990e-01 6.45161867e+00 8.99265647e-01
-1.06723356e+00 1.51001096e-01 -1.69194400e-01 -2.33383360e-03
-5.73775351e-01 6.77632540e-02 -4.79471952e-01 5.60605302e-02
6.67537093e-01 -1.65848121e-01 6.58749759e-01 5.76342642e-01
4.22362328e-01 1.23473413e-01 -9.73520756e-01 1.31246364e+00
1.66870579e-01 -1.13153017e+00 1.43379107e-01 1.68927401e-01
5.63343406e-01 -5.96096158e-01 3.29970896e-01 -1.11845858e-01
-4.79611814e-01 -8.05664957e-01 8.64540577e-01 5.87322354e-01
8.53796422e-01 -5.29540360e-01 1.52312651e-01 3.88664305e-01
-1.56871498e+00 -3.09849381e-01 -2.98257679e-01 -1.31292254e-01
4.30541873e-01 1.25972033e+00 -6.77838743e-01 8.44213665e-01
2.77678579e-01 9.28632498e-01 1.62627235e-01 6.89517796e-01
-1.79580688e-01 5.96755564e-01 -4.13708061e-01 4.52151179e-01
8.99105519e-02 -5.98733604e-01 1.09644783e+00 1.10656726e+00
7.86830008e-01 1.31010875e-01 3.45273048e-01 4.77562308e-01
8.69873464e-02 1.91218913e-01 -6.38702631e-01 -3.62453908e-01
1.69609889e-01 1.13758016e+00 -6.08702004e-01 -1.00938566e-01
-4.43020016e-01 6.38274670e-01 -3.60909492e-01 5.48911273e-01
-2.94957787e-01 -5.99126033e-02 6.45628750e-01 2.60988086e-01
5.15142083e-01 -6.33555233e-01 1.52809963e-01 -1.50201118e+00
2.13833541e-01 -1.14272666e+00 9.68115032e-02 -5.80773711e-01
-9.65664923e-01 6.33733511e-01 -1.42860517e-01 -1.74544179e+00
-6.10172927e-01 -3.72936785e-01 -1.66928843e-01 8.66928101e-01
-1.43487215e+00 -1.12707627e+00 1.80439979e-01 9.72301543e-01
5.27730703e-01 -3.36407304e-01 1.03783309e+00 5.42345285e-01
-1.37412906e-01 2.24547043e-01 3.99837047e-01 -2.53081709e-01
3.84327173e-01 -8.90088260e-01 -7.40525648e-02 1.22379208e+00
5.63534915e-01 1.20331824e+00 1.04991853e+00 -6.75103724e-01
-1.79415476e+00 -5.62751055e-01 7.05217123e-01 1.65936127e-01
6.86881542e-01 -1.85989887e-01 -5.96179664e-01 4.99302298e-01
-5.81494793e-02 1.21241607e-01 1.06705105e+00 1.67445123e-01
-4.23603445e-01 -2.84252048e-01 -7.40137637e-01 1.36816531e-01
6.12026334e-01 -9.27309930e-01 -6.78816915e-01 2.00283185e-01
3.67168546e-01 -1.21284932e-01 -8.30533803e-01 3.21762651e-01
8.39019775e-01 -8.55140388e-01 1.11139619e+00 -5.38571954e-01
2.46961296e-01 -7.13550031e-01 -9.02567565e-01 -1.24374413e+00
-4.69299436e-01 -1.16315663e+00 -5.29831231e-01 8.88004005e-01
1.57213360e-02 -4.01447475e-01 5.27695119e-01 -1.11358523e-01
-1.78064018e-01 -6.30363524e-01 -1.15054417e+00 -6.54765964e-01
-4.71313268e-01 -6.96044207e-01 2.21631914e-01 9.81851816e-01
1.43355932e-02 2.65511394e-01 -8.73613715e-01 2.72845954e-01
6.71811104e-01 3.91046762e-01 5.21499276e-01 -1.23154438e+00
-7.82544196e-01 1.00840621e-01 -3.23305249e-01 -1.24994433e+00
4.17364761e-02 -7.78589249e-01 -2.12936908e-01 -1.11781323e+00
-2.68884838e-01 -6.87727854e-02 -2.76513398e-01 -1.02688083e-02
2.57204771e-01 2.35627592e-01 5.70330679e-01 4.53053683e-01
8.13622698e-02 6.22374713e-01 1.32851386e+00 9.28097889e-02
-2.08538398e-01 4.72838245e-02 -4.45721537e-01 6.88974977e-01
1.83123633e-01 -3.66512120e-01 -8.01838636e-01 -3.35242033e-01
6.20787859e-01 4.61430520e-01 4.80738841e-02 -9.85614479e-01
2.56254822e-01 1.74104854e-01 4.04014111e-01 -4.70102489e-01
8.16630244e-01 -9.65507150e-01 5.51649332e-01 3.23877394e-01
-1.75913602e-01 8.31876621e-02 -8.75390545e-02 8.58410537e-01
-6.67190611e-01 -2.40122736e-01 5.45789778e-01 -1.76174060e-01
-2.41900072e-01 -1.42795235e-01 -5.19790709e-01 -2.93480486e-01
5.29804170e-01 -3.15211028e-01 2.59948909e-01 -6.61672056e-01
-1.18066299e+00 -5.64004242e-01 5.74164428e-02 1.91556569e-02
8.61928046e-01 -1.44446921e+00 -7.22311258e-01 6.49868965e-01
-2.78030455e-01 -4.35976446e-01 4.98469263e-01 1.19316280e+00
-1.12134039e-01 4.32194561e-01 -9.73363221e-02 -5.15445650e-01
-1.28036118e+00 4.22047824e-01 -8.69282037e-02 -4.31903034e-01
-5.14786720e-01 7.21309066e-01 2.90967911e-01 1.37563169e-01
-8.20440054e-02 -3.56105030e-01 -3.60947490e-01 5.56041479e-01
7.35780776e-01 4.33609039e-01 3.04525226e-01 -1.09480429e+00
-2.70785421e-01 8.74747336e-01 3.08818221e-01 -6.41890764e-01
1.59150875e+00 -1.73939262e-02 -3.34837675e-01 6.24478817e-01
1.20136845e+00 6.07094169e-01 -7.16113150e-01 -2.92156011e-01
-2.45668083e-01 -7.31322289e-01 3.43364239e-01 -1.73699841e-01
-9.60816145e-01 8.38148892e-01 3.50330740e-01 3.77573192e-01
1.57088637e+00 -6.19302213e-01 6.56570733e-01 4.59364712e-01
5.42730749e-01 -8.63066912e-01 1.75233424e-01 2.83067971e-01
1.21697104e+00 -3.19627672e-01 3.69467109e-01 -6.49691284e-01
-7.89037570e-02 1.38379943e+00 -4.02726829e-01 -3.56757700e-01
7.94904232e-01 1.58126727e-01 -2.50218660e-01 1.86699145e-02
-4.66684878e-01 -2.54837275e-01 5.52742481e-01 4.17194486e-01
6.97595179e-01 6.41211420e-02 -5.43386459e-01 5.45821846e-01
-2.55609781e-01 -1.22975312e-01 4.63615865e-01 8.86363626e-01
-3.97326618e-01 -1.48452199e+00 -9.51594234e-01 2.60753393e-01
-5.05558670e-01 -1.65839925e-01 -1.60638407e-01 2.73691207e-01
-2.32795551e-01 1.07402503e+00 -5.98119199e-01 -5.11879742e-01
2.73979425e-01 1.97864473e-01 5.40227890e-01 -6.47320509e-01
-4.21138883e-01 1.03094375e+00 -2.49086097e-02 -4.79866713e-01
-7.98288941e-01 -8.37548196e-01 -8.18116724e-01 1.41696557e-01
-4.35361087e-01 3.95809770e-01 6.36562109e-01 5.72071910e-01
6.78885803e-02 4.65003699e-01 8.34976614e-01 -1.12200642e+00
-6.31218612e-01 -9.46993291e-01 -1.06723285e+00 2.07582638e-01
8.26062262e-01 -7.74234474e-01 -4.92527992e-01 1.83337435e-01] | [15.40606689453125, 5.6601243019104] |
1447b16d-a8ff-4694-aefd-37fabd3edf54 | self-attention-in-vision-transformers | 2303.01542 | null | https://arxiv.org/abs/2303.01542v1 | https://arxiv.org/pdf/2303.01542v1.pdf | Self-attention in Vision Transformers Performs Perceptual Grouping, Not Attention | Recently, a considerable number of studies in computer vision involves deep neural architectures called vision transformers. Visual processing in these models incorporates computational models that are claimed to implement attention mechanisms. Despite an increasing body of work that attempts to understand the role of attention mechanisms in vision transformers, their effect is largely unknown. Here, we asked if the attention mechanisms in vision transformers exhibit similar effects as those known in human visual attention. To answer this question, we revisited the attention formulation in these models and found that despite the name, computationally, these models perform a special class of relaxation labeling with similarity grouping effects. Additionally, whereas modern experimental findings reveal that human visual attention involves both feed-forward and feedback mechanisms, the purely feed-forward architecture of vision transformers suggests that attention in these models will not have the same effects as those known in humans. To quantify these observations, we evaluated grouping performance in a family of vision transformers. Our results suggest that self-attention modules group figures in the stimuli based on similarity in visual features such as color. Also, in a singleton detection experiment as an instance of saliency detection, we studied if these models exhibit similar effects as those of feed-forward visual salience mechanisms utilized in human visual attention. We found that generally, the transformer-based attention modules assign more salience either to distractors or the ground. Together, our study suggests that the attention mechanisms in vision transformers perform similarity grouping and not attention. | ['John K. Tsotsos', 'Paria Mehrani'] | 2023-03-02 | null | null | null | null | ['saliency-detection'] | ['computer-vision'] | [ 9.61671695e-02 1.11480184e-01 2.15065747e-01 1.15755284e-02
4.90410149e-01 -3.33703369e-01 8.50212395e-01 2.15491518e-01
-5.10453641e-01 7.81495497e-02 3.56897593e-01 -3.11894864e-01
-2.09275693e-01 -8.06317866e-01 -7.11149216e-01 -5.99540174e-01
2.24814475e-01 -7.47937635e-02 6.45686865e-01 -4.69400793e-01
1.02910137e+00 5.11191189e-01 -2.29931068e+00 2.80399829e-01
7.68673360e-01 7.41698623e-01 6.28534675e-01 7.17900157e-01
-1.99446708e-01 8.85361850e-01 -6.39487684e-01 -4.29504693e-01
2.18264908e-01 -6.81445122e-01 -8.61471355e-01 7.07855746e-02
7.85308003e-01 7.90533796e-03 -2.98178792e-01 1.30059206e+00
4.56422955e-01 2.77954102e-01 6.73097432e-01 -1.21631277e+00
-1.49044311e+00 4.30182308e-01 -5.61299145e-01 9.20463622e-01
4.64502275e-01 4.93419945e-01 1.05280840e+00 -1.15572417e+00
2.68031955e-01 1.38944209e+00 4.62517500e-01 4.09229070e-01
-1.11543238e+00 -2.74989694e-01 1.72616586e-01 6.08821630e-01
-9.65075374e-01 -3.32188845e-01 6.45117879e-01 -5.17303944e-01
1.20315480e+00 5.11052847e-01 1.11455107e+00 6.66228056e-01
4.78035480e-01 5.83046913e-01 1.11561322e+00 -6.72329128e-01
1.28414005e-01 2.68001914e-01 2.59947747e-01 4.59373653e-01
4.96571630e-01 3.55136663e-01 -6.92816079e-01 1.45011544e-01
9.77137923e-01 4.02419753e-02 -3.52467984e-01 -1.47119239e-01
-1.22470379e+00 7.46765435e-01 1.03552675e+00 5.52408457e-01
-6.02327406e-01 9.75322276e-02 1.66141003e-01 2.40100008e-02
7.19034299e-02 9.10952806e-01 -6.66773319e-02 4.82377976e-01
-7.44916022e-01 -1.64871007e-01 3.38954329e-01 7.05918193e-01
7.63666272e-01 2.65873015e-01 -4.35550958e-01 5.55586219e-01
4.52456862e-01 3.30622584e-01 7.63791978e-01 -8.51734281e-01
-2.75243614e-02 6.08439326e-01 -1.28792480e-01 -1.31239736e+00
-5.75269043e-01 -5.35543382e-01 -6.12205625e-01 5.55412292e-01
5.91447949e-01 4.91141170e-01 -8.74866366e-01 1.84378231e+00
-1.50386468e-01 -3.25636238e-01 -1.93467960e-01 1.46225739e+00
8.76013815e-01 2.22198859e-01 3.53620917e-01 -8.43710601e-02
1.89257550e+00 -8.23869765e-01 -7.06421196e-01 -3.44460189e-01
-5.71847372e-02 -1.03595400e+00 1.40317690e+00 1.10559754e-01
-1.43804514e+00 -1.01306117e+00 -9.53581452e-01 -2.93897957e-01
-5.05211294e-01 -1.67016238e-01 8.66138279e-01 5.21989405e-01
-1.54329658e+00 4.73784804e-01 -2.53107160e-01 -1.04982984e+00
4.51595336e-01 1.00065649e-01 3.77362780e-02 4.02686715e-01
-1.13730335e+00 1.36047792e+00 4.39136662e-02 2.05625989e-03
-9.70528424e-01 -3.50074351e-01 -7.29964435e-01 4.30272102e-01
7.41785914e-02 -1.10654402e+00 1.17069876e+00 -1.61282468e+00
-9.63996351e-01 1.25708568e+00 -5.00408530e-01 -4.12222981e-01
-2.48034179e-01 -1.89386785e-01 -1.23163231e-01 3.94060254e-01
1.25456184e-01 8.83952677e-01 1.06095600e+00 -1.30251193e+00
-5.09221137e-01 -3.68941873e-01 4.33586299e-01 2.12013707e-01
-2.64870554e-01 2.85013974e-01 6.49867728e-02 -5.64137876e-01
1.12810709e-01 -5.53255320e-01 -3.30908708e-02 1.28428459e-01
-1.27871279e-02 -4.23668772e-01 3.87503862e-01 -1.54127195e-01
1.12357962e+00 -2.09752345e+00 3.82547751e-02 -6.17562421e-02
5.53661346e-01 1.70363680e-01 -1.63349867e-01 4.58562732e-01
-4.53280330e-01 2.46212363e-01 -1.79893792e-01 8.77689570e-02
-2.02432387e-02 1.03180870e-01 -5.04791796e-01 4.04839009e-01
3.35395306e-01 1.09206903e+00 -8.46889555e-01 -4.30312395e-01
2.42406279e-01 2.91282892e-01 -6.26877785e-01 1.58410609e-01
8.70085414e-03 1.11876428e-01 9.83406603e-02 5.56631029e-01
4.83574271e-01 -2.91636288e-01 -2.73893982e-01 -3.73509973e-01
-6.68232560e-01 2.60362118e-01 -4.43258673e-01 1.38287652e+00
1.05092958e-01 1.04345071e+00 -3.12268853e-01 -9.85571563e-01
6.40343845e-01 1.20292000e-01 -1.80930421e-01 -9.33829486e-01
4.00194943e-01 -1.41365588e-01 9.86413717e-01 -6.90770328e-01
8.07350278e-01 -1.86298728e-01 3.69747788e-01 3.89098197e-01
2.10094050e-01 -1.26742616e-01 1.36206612e-01 2.35987499e-01
5.53165317e-01 -2.44999975e-02 4.43542153e-01 -4.93696570e-01
4.76863146e-01 -3.63961868e-02 2.09050849e-01 8.74201834e-01
-4.98077065e-01 7.99318552e-01 3.29568714e-01 -2.67928988e-01
-9.34396267e-01 -1.00038087e+00 4.51442786e-02 1.60332108e+00
3.68395656e-01 -2.37094432e-01 -8.56902540e-01 -1.40730059e-02
-3.19010973e-01 8.33685160e-01 -9.89904523e-01 -5.40891111e-01
-1.68917730e-01 -5.99859536e-01 3.61062735e-01 5.82729578e-01
3.60076189e-01 -1.65155435e+00 -1.46864092e+00 -2.26021677e-01
7.39568546e-02 -5.91483474e-01 -5.43275416e-01 4.86113280e-02
-6.98774695e-01 -1.28453708e+00 -8.44829679e-01 -8.49037826e-01
8.35362256e-01 8.23549867e-01 1.27854371e+00 3.50553870e-01
-5.02167940e-01 8.48295689e-01 -2.02922627e-01 -7.64883697e-01
3.25186029e-02 -2.21170992e-01 -1.26977503e-01 -9.89348367e-02
5.83411694e-01 -3.83617401e-01 -9.32430387e-01 2.06450611e-01
-9.76254940e-01 -9.80765745e-02 7.40501940e-01 5.22797883e-01
1.11589134e-01 -4.05472398e-01 4.00253564e-01 -2.77779043e-01
8.74834776e-01 -4.61974353e-01 -3.73292476e-01 4.49975505e-02
-4.32943970e-01 3.07930727e-02 2.76400894e-01 -3.60264331e-01
-9.31063473e-01 -3.76540184e-01 1.95242703e-01 -5.17606795e-01
-2.79188305e-01 3.53901297e-01 4.83953506e-02 -2.77921200e-01
7.92210996e-01 5.00433147e-01 -2.56629232e-02 -3.09063103e-02
3.47412527e-01 6.26853667e-03 3.87940377e-01 -2.68663049e-01
4.96976912e-01 5.12906730e-01 -6.34460971e-02 -9.48106766e-01
-8.01611066e-01 -3.10886204e-01 -4.57555443e-01 -3.29192579e-01
1.23061693e+00 -4.36628580e-01 -1.06170273e+00 3.64038646e-01
-1.29296792e+00 2.87671071e-02 -6.13156378e-01 5.19410610e-01
-6.75214887e-01 4.47416186e-01 -3.75185519e-01 -9.61423755e-01
-1.96136400e-01 -1.06259406e+00 5.42370975e-01 7.12777317e-01
-4.83102113e-01 -9.12877679e-01 -1.67113125e-01 -1.58041626e-01
7.72444427e-01 -3.50679278e-01 9.63861167e-01 -4.28377539e-01
-6.37084663e-01 5.16252160e-01 -5.92527866e-01 -8.98692906e-02
6.22373335e-02 1.22242175e-01 -1.21180964e+00 1.36288509e-01
4.08417344e-01 -5.81742227e-02 1.26802623e+00 7.94563293e-01
1.03701627e+00 -5.84669001e-02 -8.74321163e-02 4.50372607e-01
1.35826516e+00 3.82198274e-01 9.54101562e-01 2.97244102e-01
4.03528452e-01 1.10352039e+00 3.60567033e-01 1.29546866e-01
4.54706311e-01 3.58913869e-01 8.33391845e-01 -3.60074043e-01
-4.34445977e-01 -3.53556089e-02 6.50461838e-02 5.76204598e-01
-6.28375828e-01 -6.29590750e-02 -7.27581620e-01 8.14304292e-01
-1.62289441e+00 -1.44093740e+00 -4.53770459e-01 2.27726102e+00
6.06673598e-01 1.65646687e-01 5.91163784e-02 8.16368908e-02
8.81012738e-01 1.05695553e-01 -5.05372286e-01 -8.38951051e-01
-4.13478315e-01 2.10088372e-01 2.09914505e-01 1.37944788e-01
-6.69763923e-01 9.33438599e-01 7.52257156e+00 2.64773488e-01
-1.22205365e+00 4.05723862e-02 1.76542208e-01 2.32938137e-02
-3.55251938e-01 1.25452429e-01 -4.35160726e-01 3.31009626e-01
6.70767963e-01 -3.10229123e-01 3.37115705e-01 4.47829545e-01
2.38738477e-01 -6.04878902e-01 -1.03403068e+00 9.37502801e-01
4.77972358e-01 -9.45596874e-01 3.43495607e-01 7.73607790e-02
2.83134282e-01 -2.69831240e-01 4.85600978e-01 1.53303146e-01
-1.00515246e-01 -1.13277125e+00 1.05160487e+00 8.74889493e-01
3.71907741e-01 -3.82696331e-01 3.83924246e-01 1.14950001e-01
-9.88169551e-01 -2.71665931e-01 -7.22747862e-01 -7.05676734e-01
1.22670783e-03 2.18792453e-01 -5.42681515e-01 -5.41295521e-02
7.51964331e-01 4.97338831e-01 -9.05374348e-01 1.48507738e+00
-4.18933392e-01 4.05296981e-01 3.65871161e-01 -1.15746908e-01
1.40685260e-01 5.57023324e-02 6.91795647e-01 1.08651733e+00
6.66889325e-02 6.33345097e-02 -4.13786411e-01 1.26464677e+00
3.54274541e-01 -1.14381418e-01 -7.01927185e-01 1.22447915e-01
2.96047091e-01 1.20278919e+00 -9.60246921e-01 -3.58682096e-01
-5.18944502e-01 9.89815116e-01 2.87730306e-01 3.78430635e-01
-8.81854296e-01 -2.40439907e-01 7.40656137e-01 3.61714810e-01
3.94428968e-01 -1.76690981e-01 -4.82609928e-01 -8.35075378e-01
-4.51370239e-01 -4.45654124e-01 1.15216233e-01 -1.38286984e+00
-1.36008286e+00 5.32070279e-01 -1.96056571e-02 -8.82040501e-01
1.98370963e-01 -7.15968430e-01 -8.06786716e-01 1.17492700e+00
-1.46690762e+00 -9.67777133e-01 -4.25065398e-01 7.86544085e-01
7.43347168e-01 -2.45518275e-02 7.17407465e-01 -1.96695372e-01
-1.59192398e-01 3.39222610e-01 -7.28013515e-01 -3.08470950e-02
6.74137533e-01 -1.36033392e+00 3.17738175e-01 9.17801261e-01
2.53465295e-01 1.27927482e+00 8.39548528e-01 -4.20471460e-01
-1.14654326e+00 -5.23766220e-01 8.20879817e-01 -3.75755727e-01
4.46929187e-01 -7.11435005e-02 -1.13077831e+00 4.27145272e-01
1.06936526e+00 -2.08352968e-01 6.97445333e-01 5.03797345e-02
-3.69625509e-01 3.39266062e-01 -8.83298278e-01 5.80793440e-01
1.15167511e+00 -4.85894531e-01 -1.47043014e+00 -1.34317562e-01
5.78523040e-01 9.65620354e-02 -2.92561084e-01 5.79618104e-02
4.80003715e-01 -1.41104722e+00 1.05029249e+00 -6.79064631e-01
5.59578121e-01 -5.45774400e-01 -1.48430644e-02 -1.43714368e+00
-1.09386432e+00 4.92068939e-02 5.83472103e-02 1.02407289e+00
9.16372389e-02 -9.27719712e-01 1.15361378e-01 3.94502789e-01
-4.08439279e-01 -2.77313352e-01 -5.61499119e-01 -4.54838187e-01
-1.10795699e-01 -1.38685510e-01 1.98869258e-01 9.55299497e-01
-9.79096815e-03 6.32671177e-01 5.98666593e-02 -7.02867135e-02
6.15374327e-01 8.24736357e-02 2.92062193e-01 -1.42498374e+00
-1.78837836e-01 -9.46466208e-01 -4.61866438e-01 -6.96824431e-01
1.46782780e-02 -8.95148575e-01 -1.20066546e-01 -1.78096342e+00
3.52660120e-01 7.35009685e-02 -4.62097704e-01 5.28178871e-01
-3.44923466e-01 2.96468168e-01 5.30332446e-01 3.41780066e-01
-3.72459382e-01 4.07443255e-01 1.49947727e+00 9.79120955e-02
1.07615650e-01 -3.37456435e-01 -1.30833197e+00 9.79953945e-01
8.37234199e-01 9.41551849e-03 -4.50257003e-01 -5.36692798e-01
3.76226991e-01 -4.17425871e-01 8.90121400e-01 -1.01263404e+00
5.15186191e-01 -2.23233610e-01 5.81675828e-01 -5.34509540e-01
2.00160295e-01 -7.75691569e-01 -3.33012551e-01 6.26568317e-01
-2.15082869e-01 1.53199106e-01 5.00104725e-01 1.73715934e-01
-1.79322183e-01 -3.49186331e-01 8.14993918e-01 -5.07205486e-01
-1.10757422e+00 -2.26601839e-01 -8.40519249e-01 6.36997595e-02
7.65965223e-01 -7.41011322e-01 -6.12915277e-01 -1.91243753e-01
-7.69939601e-01 -1.81819811e-01 4.27807420e-01 5.19331872e-01
8.30167711e-01 -1.14912164e+00 -5.87704957e-01 1.93247035e-01
1.23523563e-01 -4.65559512e-01 5.89850470e-02 1.19906926e+00
-1.09524749e-01 4.99041438e-01 -7.58089185e-01 -6.54339254e-01
-1.07740402e+00 1.21117556e+00 4.44269210e-01 6.05772197e-01
-2.13296801e-01 1.00770986e+00 9.72778559e-01 3.32551062e-01
-8.79489556e-02 -4.84896481e-01 -6.88564956e-01 3.45281422e-01
6.56506777e-01 3.13011467e-01 -2.42183134e-01 -7.71591365e-01
-4.23643112e-01 7.45374382e-01 1.99231878e-01 1.35009944e-01
8.73710632e-01 -2.46104017e-01 -3.60709310e-01 7.14875698e-01
4.49622810e-01 7.96318725e-02 -9.45292532e-01 1.11521818e-01
-1.77645251e-01 -5.51457584e-01 -2.59061605e-01 -7.51863599e-01
-9.74010885e-01 1.16161644e+00 5.61824977e-01 7.02700794e-01
1.38418579e+00 2.10671768e-01 5.64688295e-02 -1.04816332e-01
2.19992951e-01 -8.68522584e-01 3.29363912e-01 6.81289554e-01
1.35199404e+00 -1.18420827e+00 -2.82231182e-01 -3.08527142e-01
-6.43369734e-01 8.60921681e-01 1.06349576e+00 -4.32790697e-01
3.31272304e-01 -9.76407826e-02 -2.29131237e-01 -4.90938425e-01
-8.78051877e-01 -8.84341717e-01 5.09875000e-01 8.56358647e-01
8.70382905e-01 -2.28258416e-01 -5.03733099e-01 5.03673136e-01
-4.67470735e-01 -2.36827701e-01 5.96575439e-01 6.34158671e-01
-1.02175283e+00 -3.61353993e-01 -9.07770395e-01 2.28499964e-01
-2.39019677e-01 -5.97360909e-01 -5.86833537e-01 6.84489191e-01
4.85796392e-01 9.39226151e-01 4.68452483e-01 -1.73339352e-01
4.72955495e-01 -1.07438214e-01 8.28500569e-01 -7.67550945e-01
-9.51861799e-01 -1.43157467e-01 -3.85536134e-01 -3.94870758e-01
-7.65635252e-01 -5.37979603e-01 -1.16180789e+00 5.96928671e-02
-2.81368256e-01 9.62678529e-03 3.54908496e-01 7.15048194e-01
1.66659355e-01 1.02355909e+00 3.99578363e-02 -1.00254536e+00
-1.70919716e-01 -1.18491840e+00 -4.60059494e-01 4.43189621e-01
1.74566835e-01 -8.73557568e-01 -4.84392136e-01 1.91899270e-01] | [10.022332191467285, 1.9219814538955688] |
775cbe2f-4505-4484-ad22-fbd3c6c16401 | gridformer-residual-dense-transformer-with | 2305.17863 | null | https://arxiv.org/abs/2305.17863v1 | https://arxiv.org/pdf/2305.17863v1.pdf | GridFormer: Residual Dense Transformer with Grid Structure for Image Restoration in Adverse Weather Conditions | Image restoration in adverse weather conditions is a difficult task in computer vision. In this paper, we propose a novel transformer-based framework called GridFormer which serves as a backbone for image restoration under adverse weather conditions. GridFormer is designed in a grid structure using a residual dense transformer block, and it introduces two core designs. First, it uses an enhanced attention mechanism in the transformer layer. The mechanism includes stages of the sampler and compact self-attention to improve efficiency, and a local enhancement stage to strengthen local information. Second, we introduce a residual dense transformer block (RDTB) as the final GridFormer layer. This design further improves the network's ability to learn effective features from both preceding and current local features. The GridFormer framework achieves state-of-the-art results on five diverse image restoration tasks in adverse weather conditions, including image deraining, dehazing, deraining & dehazing, desnowing, and multi-weather restoration. The source code and pre-trained models will be released. | ['Hongdong Li', 'Wei Liu', 'Tae-Kyun Kim', 'Tong Lu', 'Bjorn Stenger', 'Wenhan Luo', 'Ziqian Shao', 'Kaihao Zhang', 'Tao Wang'] | 2023-05-29 | null | null | null | null | ['image-restoration'] | ['computer-vision'] | [ 3.09416771e-01 -3.29158843e-01 2.27339044e-01 -3.01975101e-01
-6.63600624e-01 -1.17863365e-01 5.82263291e-01 -4.30754244e-01
-2.14282766e-01 5.69715261e-01 8.11738849e-01 -2.67060965e-01
1.99396715e-01 -6.38098776e-01 -6.67962790e-01 -1.30501294e+00
5.53012639e-03 -5.89266181e-01 1.88190684e-01 -3.13918918e-01
3.46144468e-01 6.06697321e-01 -1.46034801e+00 4.39607114e-01
1.36367655e+00 1.04795790e+00 6.01197839e-01 8.97996843e-01
3.17459285e-01 1.00829101e+00 -5.96376836e-01 -1.44500121e-01
4.80201364e-01 -3.27134281e-01 -4.63399440e-01 3.51772308e-01
8.64100516e-01 -8.32235038e-01 -7.52798319e-01 1.10168278e+00
9.95233953e-01 3.85428280e-01 4.76382762e-01 -8.96850705e-01
-1.09665668e+00 3.71530801e-02 -6.72894418e-01 7.31172979e-01
1.26139745e-01 3.65188062e-01 6.97208703e-01 -1.07482278e+00
1.24341220e-01 1.21193004e+00 8.95618737e-01 2.75173575e-01
-8.50295901e-01 -6.86607957e-01 3.55099469e-01 7.86922157e-01
-1.04045308e+00 -5.94607055e-01 5.13056874e-01 -2.17700541e-01
1.11491251e+00 1.68849915e-01 5.46827257e-01 6.25066102e-01
6.09694839e-01 8.81457686e-01 1.37136543e+00 -2.99902260e-01
-6.38942271e-02 -2.77078927e-01 -5.72912246e-02 3.44889194e-01
-3.14902335e-01 5.85596383e-01 -4.96490985e-01 1.66615263e-01
7.50747085e-01 1.73975289e-01 -1.01080763e+00 1.85770616e-01
-7.77464926e-01 6.81515157e-01 1.04645562e+00 1.10208597e-02
-4.83871818e-01 -2.38051545e-02 1.40224248e-01 4.27204728e-01
8.86219025e-01 1.25276431e-01 -4.06818360e-01 4.71799910e-01
-8.93977463e-01 1.50291443e-01 1.45211905e-01 5.34070373e-01
7.07028747e-01 4.53903943e-01 -3.61797988e-01 1.15355170e+00
4.20321196e-01 5.29924154e-01 5.21568596e-01 -7.56713271e-01
3.33242267e-01 5.13383485e-02 4.94557992e-02 -8.35588872e-01
-1.23636588e-01 -8.86134267e-01 -1.33074212e+00 5.73253989e-01
-3.79713058e-01 6.24692999e-02 -1.37084937e+00 1.36455941e+00
3.53372097e-01 8.13407540e-01 7.13110343e-02 1.06718683e+00
9.69895959e-01 1.09723651e+00 -1.01703353e-01 -3.23228180e-01
1.19903409e+00 -1.53684032e+00 -9.83562708e-01 -4.12150383e-01
-1.49071068e-01 -9.84333754e-01 1.03505218e+00 4.41610545e-01
-1.00138640e+00 -9.03162718e-01 -1.14000702e+00 -4.55882877e-01
-2.23965466e-01 1.01551935e-01 3.48316342e-01 4.29630697e-01
-1.56435263e+00 5.41444242e-01 -4.64620948e-01 -9.59225968e-02
4.06336665e-01 -2.31871344e-02 -4.05278355e-01 -6.12979650e-01
-1.08370161e+00 1.11531234e+00 -8.45561922e-02 5.98674536e-01
-1.28076732e+00 -8.01610589e-01 -1.22748959e+00 1.14037313e-01
-1.10470122e-02 -6.98499560e-01 1.08327603e+00 -7.20220268e-01
-1.48520017e+00 5.32590330e-01 -3.44151020e-01 -5.26102424e-01
2.17819661e-01 -5.85884094e-01 -5.11916697e-01 1.42267227e-01
7.60817006e-02 3.99554670e-01 1.36331618e+00 -1.30545223e+00
-7.30934441e-01 -2.70019025e-01 -6.48094043e-02 5.50365388e-01
-1.74490288e-01 -2.97751408e-02 -3.17985952e-01 -1.28098691e+00
-2.64922483e-03 -4.94561493e-01 -3.32051069e-01 9.83854607e-02
-1.83677703e-01 8.89679492e-02 1.05007744e+00 -1.31999588e+00
1.18637669e+00 -2.33088493e+00 -1.00487083e-01 -2.48657301e-01
1.48094296e-01 4.91628349e-01 -4.21702176e-01 2.79149383e-01
-5.07793546e-01 -1.49981558e-01 -4.41319436e-01 -5.09808302e-01
-2.65793025e-01 2.08607122e-01 -7.55627573e-01 8.21006179e-01
1.56917393e-01 6.25592947e-01 -6.14141583e-01 -1.26310736e-01
5.93803167e-01 8.03306282e-01 -4.43438381e-01 5.98106623e-01
3.94990474e-01 2.38108739e-01 7.12860674e-02 6.87081099e-01
1.25147367e+00 -7.29521853e-04 -4.61021483e-01 -6.86674535e-01
-3.39466602e-01 1.65841103e-01 -1.06414342e+00 1.47393751e+00
-5.03995299e-01 5.92783391e-01 3.25393468e-01 -8.67052794e-01
7.19680190e-01 3.50027680e-01 1.68694239e-02 -1.05698705e+00
-9.50637758e-02 -6.52015805e-02 -4.60815400e-01 -5.16888022e-01
6.02101743e-01 -1.63312614e-01 5.66369951e-01 2.17612401e-01
4.93646786e-02 -1.09344929e-01 -4.15379480e-02 2.14559808e-01
8.63200426e-01 1.62974685e-01 2.40768760e-01 -3.20507646e-01
6.88002229e-01 -4.01442468e-01 8.62513006e-01 5.91928244e-01
-2.16517270e-01 9.44392741e-01 -1.77998990e-01 -5.98588049e-01
-8.82027984e-01 -1.12341619e+00 -2.02040166e-01 9.57347870e-01
3.99924129e-01 -1.61500618e-01 -4.44434464e-01 -4.33626980e-01
-3.11367810e-01 3.81237179e-01 -6.67121887e-01 -3.08519155e-01
-4.72454518e-01 -1.10623097e+00 1.22599110e-01 5.01931608e-01
1.15554953e+00 -1.08877993e+00 -2.13211194e-01 8.74603540e-02
-5.17664611e-01 -7.93525934e-01 -9.91403818e-01 1.26172483e-01
-7.66286731e-01 -1.03012156e+00 -9.14950252e-01 -9.27262366e-01
7.44400740e-01 8.70285034e-01 1.08351719e+00 3.17965299e-01
-2.69279003e-01 -4.62815203e-02 -6.11693978e-01 -2.13627711e-01
2.31051832e-01 -4.51482028e-01 -2.49265492e-01 1.84720546e-01
-4.54357207e-01 -7.90002227e-01 -1.08413815e+00 2.63609678e-01
-1.08307064e+00 -4.15380560e-02 6.27202034e-01 1.33544421e+00
4.68705207e-01 3.45877588e-01 2.08312616e-01 -3.38065028e-01
6.02041841e-01 -2.22644344e-01 -4.38052386e-01 1.35904700e-01
-7.10603774e-01 -2.79233634e-01 6.23979867e-01 -2.41696969e-01
-1.43010402e+00 -2.11223394e-01 -4.32124197e-01 -2.05335319e-01
-2.67951209e-02 3.81766260e-01 -3.18641931e-01 -4.00398105e-01
4.72344697e-01 6.01698160e-01 -1.52107850e-01 -9.74799573e-01
2.29534179e-01 7.03734338e-01 9.28721666e-01 -8.29810426e-02
1.04834068e+00 4.42999035e-01 -4.44106102e-01 -6.40656769e-01
-1.02200508e+00 -4.97192562e-01 -1.40611067e-01 -3.16681504e-01
6.64157510e-01 -1.49442101e+00 -5.02357543e-01 1.19102919e+00
-8.57694626e-01 -6.45001411e-01 -1.58118218e-01 2.58473158e-01
-1.43010795e-01 5.79803765e-01 -9.29671288e-01 -5.31464815e-01
-7.91513026e-01 -1.14866173e+00 1.04029095e+00 7.57536948e-01
6.74312890e-01 -7.81045020e-01 -3.69307734e-02 4.22985882e-01
8.43494952e-01 -6.34838417e-02 5.17321825e-01 1.57800987e-01
-5.85853517e-01 1.61633268e-01 -4.62803662e-01 8.76062095e-01
2.37660944e-01 -4.47482586e-01 -1.16930079e+00 -8.14441085e-01
2.42765516e-01 -1.80798471e-01 1.31130397e+00 7.15854764e-01
1.07385683e+00 -3.82242531e-01 -5.88502251e-02 1.21337128e+00
1.53741479e+00 1.78080961e-01 1.30634558e+00 6.02074683e-01
4.87246364e-01 1.96080849e-01 4.72037762e-01 2.45962843e-01
5.94662368e-01 4.05053496e-01 6.51366115e-01 -8.43342423e-01
-7.31397629e-01 7.85271376e-02 5.70448637e-01 6.40115082e-01
1.44583628e-01 -3.96228194e-01 -5.36040008e-01 8.61880541e-01
-1.60550356e+00 -1.04516554e+00 9.68347266e-02 1.87611890e+00
9.81171370e-01 -3.53800595e-01 -4.50686008e-01 1.18168920e-01
6.05240285e-01 5.32967329e-01 -6.14973903e-01 -1.59504727e-01
-3.94951224e-01 1.77248642e-01 4.46156442e-01 8.93159509e-01
-1.31121230e+00 7.91091084e-01 6.85015535e+00 7.97752917e-01
-1.04788470e+00 1.40255526e-01 9.87615824e-01 2.91235149e-01
-1.92745291e-02 -1.00141615e-01 -5.97763181e-01 4.54402894e-01
4.00135875e-01 1.24656305e-01 5.16363382e-01 5.52848876e-01
3.93908203e-01 -3.43062907e-01 -2.59940177e-01 8.85339439e-01
3.81021082e-01 -1.26325786e+00 -1.46664493e-02 -8.17443579e-02
8.03118646e-01 2.49151245e-01 1.44055322e-01 1.44313127e-01
5.08503914e-01 -1.07647395e+00 5.75142801e-01 4.81290191e-01
7.46510208e-01 -7.11678445e-01 9.48644757e-01 9.26157758e-02
-1.38988090e+00 -2.93667793e-01 -3.97337914e-01 -4.46627336e-03
3.50426406e-01 8.92415583e-01 -2.25312963e-01 7.91058242e-01
1.29570687e+00 1.19433379e+00 -4.96640265e-01 1.50993705e+00
-5.41048169e-01 5.34375250e-01 2.55753603e-02 9.54827070e-01
4.75762263e-02 -1.67830944e-01 4.35916454e-01 1.18869472e+00
2.80531436e-01 2.68577486e-01 2.82304794e-01 1.63511902e-01
1.39415875e-01 -4.08201665e-01 -1.72918379e-01 7.41137862e-01
1.27608538e-01 1.28869271e+00 -2.78743148e-01 -3.89099479e-01
-2.95105934e-01 1.18895388e+00 -5.45111261e-02 5.81483901e-01
-7.65836418e-01 -5.72222054e-01 9.71831262e-01 -1.53282180e-01
5.74603915e-01 -1.91714689e-02 -1.28872499e-01 -1.26633286e+00
6.41181394e-02 -1.16716957e+00 4.27260399e-01 -1.40298855e+00
-1.45810819e+00 9.57551956e-01 -3.50885838e-01 -1.20346665e+00
3.25330555e-01 -3.51036966e-01 -7.89827764e-01 1.38510823e+00
-2.46848583e+00 -1.21464550e+00 -7.85389304e-01 1.03736937e+00
7.86866069e-01 -2.04730220e-02 5.16104281e-01 5.92927575e-01
-5.94764411e-01 3.44892144e-01 1.48700282e-01 6.27344996e-02
9.83061433e-01 -1.31230819e+00 4.32228416e-01 1.66934001e+00
-2.94628948e-01 3.82657409e-01 7.25731909e-01 -6.99799597e-01
-1.12895036e+00 -1.22895837e+00 7.34540880e-01 3.81097615e-01
2.76039422e-01 1.00819424e-01 -1.00201285e+00 5.49346030e-01
6.57884240e-01 3.88610005e-01 4.08345193e-01 -3.16129982e-01
-5.59251428e-01 -4.74527001e-01 -1.17947090e+00 4.72875148e-01
8.24609637e-01 -6.18127167e-01 -5.22456706e-01 3.99857014e-01
7.13178515e-01 -6.29756629e-01 -7.52217531e-01 4.63098675e-01
2.11289421e-01 -1.04864669e+00 1.25742602e+00 1.20794430e-01
3.55867594e-01 -8.23487639e-01 -1.99804053e-01 -1.73446727e+00
-7.20481277e-01 -8.21468651e-01 -5.49973510e-02 1.15207791e+00
5.60281277e-02 -8.24344337e-01 2.82625377e-01 1.92853790e-02
-7.40161359e-01 -4.68819141e-01 -8.11252594e-01 -4.95486736e-01
-3.57667744e-01 -1.50063867e-02 5.67746878e-01 8.02176118e-01
-5.28354824e-01 2.23661382e-02 -9.14034426e-01 6.36739075e-01
8.13357353e-01 1.40572339e-02 4.17604387e-01 -6.64859354e-01
-2.14337423e-01 -1.27231227e-02 -3.75623032e-02 -1.43185437e+00
-2.03308403e-01 -4.40265328e-01 4.52197731e-01 -1.95437443e+00
1.92147627e-01 -7.96027333e-02 -3.53886664e-01 6.60982549e-01
-6.09080672e-01 6.48142278e-01 2.18074530e-01 1.70982838e-01
-1.67952284e-01 9.85134542e-01 1.51969349e+00 -2.57521242e-01
8.77772644e-02 -1.90033302e-01 -9.34899092e-01 3.78734291e-01
6.77993834e-01 -2.01926783e-01 -3.67152959e-01 -7.90321112e-01
-3.87824267e-01 -9.89158526e-02 4.89640743e-01 -9.45164263e-01
3.87931287e-01 -4.43730541e-02 6.50135756e-01 -6.13401949e-01
3.49975288e-01 -6.54114842e-01 -5.83928358e-03 4.43493038e-01
-2.12673494e-03 3.19679201e-01 2.49142244e-01 6.20428979e-01
-6.45499587e-01 2.47354984e-01 1.07733154e+00 -3.48281264e-02
-9.37752187e-01 5.53364336e-01 -4.25035983e-01 -3.99808943e-01
6.57255232e-01 -2.11585343e-01 -6.85379267e-01 -4.92616594e-01
-5.48519075e-01 3.83386910e-01 2.58549154e-01 3.53640407e-01
1.00489008e+00 -1.14644325e+00 -9.96442258e-01 4.55215186e-01
-3.47768128e-01 -3.81867290e-02 8.14758897e-01 8.32400858e-01
-5.84913373e-01 -1.38012514e-01 -2.56147385e-01 -2.22271279e-01
-1.36233199e+00 3.69348019e-01 6.60104454e-01 -3.97218585e-01
-1.09109426e+00 9.10840094e-01 4.95329976e-01 -1.70052379e-01
3.26639950e-01 -1.23939000e-01 -3.28266025e-01 -1.38891876e-01
1.34621787e+00 2.92609483e-01 1.77237302e-01 -5.56333780e-01
-1.15903631e-01 5.07941067e-01 -3.04449856e-01 1.86687320e-01
1.64933717e+00 -5.87789237e-01 -3.50910693e-01 -3.59414399e-01
8.57879639e-01 -2.08397552e-01 -1.82716429e+00 -4.17077214e-01
-6.65443778e-01 -8.74118149e-01 6.91377759e-01 -1.18802273e+00
-1.52452552e+00 8.24757814e-01 1.05016530e+00 -9.79095027e-02
2.07280707e+00 -5.83747506e-01 9.48067725e-01 -1.72821790e-01
-1.24470159e-01 -7.31667757e-01 1.83449432e-01 5.84947348e-01
1.35910153e+00 -1.09273899e+00 2.22774595e-01 -3.56027097e-01
-3.77577186e-01 9.35387254e-01 6.67461097e-01 -2.68191814e-01
8.07154238e-01 5.23940444e-01 3.88036400e-01 8.65937769e-02
-6.41673326e-01 -2.47613683e-01 4.13436949e-01 9.45827901e-01
2.11252183e-01 -2.93043196e-01 -1.37913357e-02 1.60086378e-01
9.95306969e-02 -1.88611507e-01 5.44783056e-01 7.87721634e-01
-5.88802755e-01 -8.25010300e-01 -6.33931398e-01 1.83700413e-01
-5.24152815e-01 -7.20357597e-01 2.05022559e-01 2.84741461e-01
2.21229613e-01 1.34618199e+00 -1.04930349e-01 -3.95854145e-01
2.90457934e-01 -5.94181895e-01 1.04969457e-01 -2.75926203e-01
-8.41884077e-01 4.06021774e-01 -1.55658081e-01 -7.54090846e-01
-5.59215665e-01 -3.28150451e-01 -5.07268190e-01 -3.28753471e-01
-4.14537936e-01 2.79211774e-02 3.82075518e-01 8.82783294e-01
4.27101195e-01 7.70083964e-01 8.10964704e-01 -1.28145480e+00
-2.08012506e-01 -1.14952874e+00 -7.58702815e-01 1.71992809e-01
1.00543916e+00 -3.57514679e-01 -4.77229863e-01 3.72554570e-01] | [11.091840744018555, -2.803161144256592] |
c45c4424-fb5a-41e2-9ac6-116c3948987d | amdet-a-tool-for-mitotic-cell-detection-in | 2108.03676 | null | https://arxiv.org/abs/2108.03676v1 | https://arxiv.org/pdf/2108.03676v1.pdf | AMDet: A Tool for Mitotic Cell Detection in Histopathology Slides | Breast Cancer is the most prevalent cancer in the world. The World Health Organization reports that the disease still affects a significant portion of the developing world citing increased mortality rates in the majority of low to middle income countries. The most popular protocol pathologists use for diagnosing breast cancer is the Nottingham grading system which grades the proliferation of tumors based on 3 major criteria, the most important of them being mitotic cell count. The way in which pathologists evaluate mitotic cell count is to subjectively and qualitatively analyze cells present in stained slides of tissue and make a decision on its mitotic state i.e. is it mitotic or not?This process is extremely inefficient and tiring for pathologists and so an efficient, accurate, and fully automated tool to aid with the diagnosis is extremely desirable. Fortunately, creating such a tool is made significantly easier with the AutoML tool available from Microsoft Azure, however to the best of our knowledge the AutoML tool has never been formally evaluated for use in mitotic cell detection in histopathology images. This paper serves as an evaluation of the AutoML tool for this purpose and will provide a first look on how the tool handles this challenging problem. All code is available athttps://github.com/WaltAFWilliams/AMDet | ['Jimmy Hall', 'Walt Williams'] | 2021-08-08 | null | null | null | null | ['cell-detection'] | ['computer-vision'] | [-1.39285982e-01 4.14757580e-02 -4.55744922e-01 6.68864921e-02
-6.36157334e-01 -5.90544343e-01 2.40740418e-01 9.47067857e-01
-5.08256316e-01 7.00962842e-01 -8.77753645e-02 -6.19191587e-01
3.05814356e-01 -7.67494321e-01 8.78594145e-02 -9.29698169e-01
2.90307850e-01 7.07796991e-01 3.73471946e-01 1.17140366e-02
2.91292757e-01 7.39006698e-01 -1.19095671e+00 -1.83382444e-02
5.08150399e-01 5.30475795e-01 1.32327273e-01 9.86675322e-01
-1.58691093e-01 6.90997779e-01 -2.78761238e-01 -3.26513767e-01
-9.60647464e-02 -4.95172352e-01 -8.85277390e-01 -1.65410668e-01
-1.40116677e-01 -2.97232181e-01 2.70547360e-01 7.42806733e-01
2.71478802e-01 -6.22849107e-01 9.04231846e-01 -7.83405006e-01
1.86652958e-01 1.65115938e-01 -6.84360325e-01 4.73624527e-01
1.87972203e-01 -3.39811221e-02 6.09700561e-01 -6.71269536e-01
7.35951722e-01 7.02925503e-01 5.29331267e-01 3.35286468e-01
-1.01958883e+00 -5.96175194e-01 -5.84950864e-01 3.98642898e-01
-1.48601425e+00 -3.42731714e-01 2.82092839e-01 -5.54438889e-01
5.86857855e-01 4.51868743e-01 9.44175839e-01 2.74561077e-01
6.83027744e-01 5.06638646e-01 1.31731045e+00 -5.78351736e-01
3.40101033e-01 3.67016315e-01 6.79948628e-02 8.59623492e-01
6.76777661e-01 -3.09808642e-01 -3.56408544e-02 -9.39539522e-02
5.66578150e-01 1.22933321e-01 -9.77662504e-02 -5.45408465e-02
-9.04776335e-01 6.08475804e-01 1.38178930e-01 8.20825815e-01
-2.05317572e-01 4.90979925e-02 4.92242724e-01 9.81472433e-02
1.94967702e-01 6.01920970e-02 2.00307555e-02 -1.73247442e-01
-1.08224225e+00 -1.07277870e-01 7.10050225e-01 7.99678490e-02
4.43471700e-01 -5.68112493e-01 3.30051124e-01 6.34470940e-01
3.37848961e-01 3.65560591e-01 5.41410029e-01 -7.79674709e-01
-5.60356736e-01 9.34808433e-01 -1.12501234e-01 -9.31875467e-01
-5.88168323e-01 -3.82096916e-01 -9.02876437e-01 3.95254582e-01
8.58912170e-01 2.21871689e-01 -7.92396247e-01 1.08550179e+00
3.78502846e-01 -3.78665090e-01 -1.09842300e-01 6.93520665e-01
7.58188784e-01 3.26961130e-01 2.76641160e-01 -1.22913972e-01
1.69492316e+00 -3.72174829e-01 -6.71986103e-01 -4.74692099e-02
9.53632176e-01 -9.64553297e-01 5.46235204e-01 4.79617178e-01
-9.93236601e-01 1.46235123e-01 -1.13623619e+00 -9.61664021e-02
-6.13969564e-01 2.13403866e-01 7.16838777e-01 8.79699945e-01
-1.20742846e+00 1.14242777e-01 -1.22764206e+00 -1.08683360e+00
4.26171184e-01 3.16897750e-01 -7.29055882e-01 -1.30731553e-01
-6.15005374e-01 1.13786006e+00 2.08152682e-01 -9.71753299e-02
-2.78615862e-01 -4.22973394e-01 -5.22683501e-01 -2.45780587e-01
1.52955860e-01 -4.75873023e-01 1.34616673e+00 -4.75577086e-01
-1.02917480e+00 1.54858017e+00 -5.54575443e-01 -1.85079917e-01
3.74346256e-01 6.79540455e-01 -5.59381209e-02 4.39361572e-01
1.07946202e-01 5.54548204e-01 -4.77487221e-02 -9.28115129e-01
-9.86220837e-01 -3.70132089e-01 -1.90191478e-01 1.08962692e-02
-1.37724012e-01 8.73019081e-03 -4.47077185e-01 -3.27328116e-01
6.29065260e-02 -8.29139113e-01 -1.04808204e-01 4.50625300e-01
-5.97905964e-02 -1.61677033e-01 6.50796950e-01 -8.15984309e-01
1.12863743e+00 -2.06994748e+00 -3.95854473e-01 2.45850801e-01
3.07981223e-01 3.28741044e-01 5.78222096e-01 5.22794724e-01
5.05362786e-02 4.39483166e-01 -6.09158836e-02 1.60515442e-01
-1.79346532e-01 2.41389759e-02 4.60027486e-01 8.49726856e-01
3.58327851e-02 4.45990920e-01 -1.06237936e+00 -8.27064633e-01
3.10277790e-01 7.03409970e-01 -1.34649366e-01 -2.36237481e-01
3.52248907e-01 1.99870914e-01 -1.28768489e-01 1.08148003e+00
4.16351020e-01 -5.15556872e-01 4.48093116e-01 -5.55129647e-02
-1.93969548e-01 -1.42517269e-01 -8.94854963e-01 9.31202710e-01
-5.22954687e-02 6.29221320e-01 2.17989936e-01 -6.88814819e-01
7.62415230e-01 5.01154125e-01 3.31100583e-01 -2.21560389e-01
3.30379605e-01 6.11642003e-01 1.76740393e-01 -3.88919294e-01
1.08718663e-01 -5.26050150e-01 6.32168725e-02 3.71556044e-01
-2.66131461e-01 -2.30201483e-01 6.78572774e-01 2.77203202e-01
1.42468369e+00 -4.46800113e-01 1.21990442e+00 -2.98996150e-01
7.02690184e-01 3.39265287e-01 5.67719042e-01 1.05715521e-01
-5.43149173e-01 3.13873380e-01 7.72472680e-01 -3.82709324e-01
-7.29250848e-01 -8.92953575e-01 -4.25670415e-01 3.14299315e-01
-9.63585526e-02 -1.49995416e-01 -5.81163406e-01 -2.98694730e-01
-1.32144123e-01 3.39167506e-01 -8.42280805e-01 2.15286702e-01
-7.35243130e-03 -6.76587045e-01 1.83820412e-01 2.49347985e-01
4.68372405e-01 -6.56045437e-01 -8.71912599e-01 4.03751731e-02
-4.95133549e-02 -6.35427713e-01 1.34651745e-02 8.20250437e-02
-6.82284296e-01 -1.31030738e+00 -9.10379291e-01 -9.09654021e-01
1.07041526e+00 1.90960571e-01 7.42821693e-01 6.29531384e-01
-8.36913049e-01 9.84751061e-02 -4.34637129e-01 -6.52164876e-01
-7.00844288e-01 -1.91166773e-02 -3.66152257e-01 -3.90190631e-01
7.65195310e-01 -2.18388736e-01 -8.40294719e-01 1.35811418e-01
-9.16939378e-01 2.67915010e-01 8.71807635e-01 4.80539709e-01
8.04642975e-01 1.17207818e-01 3.12156320e-01 -1.02002740e+00
7.29660168e-02 -4.97620106e-01 -3.30857068e-01 -2.75189951e-02
-3.42173576e-01 -4.11492437e-01 4.25641418e-01 2.57310718e-01
-4.15257871e-01 1.64623961e-01 -3.37654948e-01 2.61557430e-01
-3.95840257e-01 6.89515591e-01 3.84310216e-01 -4.42250930e-02
3.96660566e-01 -1.86976880e-01 2.62251407e-01 -1.61077697e-02
-6.03919506e-01 7.34862149e-01 5.36255121e-01 2.27386117e-01
4.98037249e-01 8.35068345e-01 4.05837059e-01 -9.98031139e-01
-4.33141381e-01 -1.01716948e+00 -2.44112015e-01 -5.07425606e-01
8.55446517e-01 -4.41565752e-01 -6.18488729e-01 5.09089649e-01
-7.12480724e-01 -3.05905789e-01 2.00925395e-01 4.56940264e-01
-1.66005000e-01 2.76419461e-01 -6.39304638e-01 -6.81246936e-01
-4.27257657e-01 -1.05142117e+00 6.24960899e-01 5.05924523e-01
-7.06910491e-01 -1.22940147e+00 2.78169006e-01 3.20577592e-01
5.15765369e-01 6.06669307e-01 1.04663146e+00 -2.99684495e-01
-8.53299797e-02 -6.49534583e-01 -1.46214023e-01 -1.73123032e-01
4.10457373e-01 5.15167832e-01 -7.57549405e-01 -1.52915463e-01
-1.91008106e-01 6.02985173e-02 7.14363813e-01 5.32861710e-01
7.38720059e-01 4.92811948e-02 -7.82195926e-01 1.93744406e-01
1.71591449e+00 3.61839622e-01 6.63705111e-01 4.88465965e-01
-3.28427646e-03 8.38550270e-01 7.05515683e-01 2.24537373e-01
3.90462309e-01 1.81613103e-01 4.73090589e-01 -3.54375094e-01
-1.67516619e-01 1.44187048e-01 -1.05077714e-01 5.31615853e-01
-7.16929510e-02 -2.53002435e-01 -1.37207305e+00 8.05707574e-01
-1.26772010e+00 -8.04170609e-01 -1.71713665e-01 2.06255579e+00
7.70701170e-01 8.86228010e-02 6.24209195e-02 6.67815328e-01
6.71735406e-01 -6.17463827e-01 -1.87639952e-01 -4.58875895e-01
2.81264186e-01 9.72569138e-02 4.32901502e-01 4.91553664e-01
-8.24904203e-01 3.65690976e-01 5.84423256e+00 6.27660215e-01
-1.49309576e+00 -5.39621823e-02 1.04929888e+00 1.29595429e-01
1.70173332e-01 -3.90322384e-04 -7.17834771e-01 4.57292587e-01
7.03566790e-01 -4.12229568e-01 -3.37582052e-01 4.46024776e-01
3.87225151e-01 -9.88383591e-01 -8.58689308e-01 7.26537108e-01
-6.79617971e-02 -1.27221537e+00 -4.52742338e-01 3.51301908e-01
1.62559584e-01 -3.22735488e-01 -1.51382238e-01 -5.65099679e-02
6.41910806e-02 -9.52057600e-01 3.00731927e-01 4.38762486e-01
9.18008804e-01 -7.54456043e-01 1.30883038e+00 1.56248555e-01
-9.76411104e-01 3.07234615e-01 7.82441050e-02 1.67076781e-01
2.37896927e-02 7.69311666e-01 -1.43673825e+00 -1.02019506e-02
6.02405608e-01 3.23358774e-01 -7.22177088e-01 1.49312949e+00
-9.05004218e-02 6.45124376e-01 -3.75644296e-01 -2.42655575e-01
2.47040354e-02 1.79748133e-01 3.36613387e-01 1.31020153e+00
5.38998783e-01 1.25048589e-02 -1.35620639e-01 1.25170648e-01
3.31286252e-01 4.14933980e-01 -2.48847902e-01 -2.10961267e-01
5.26597619e-01 1.69171751e+00 -1.62270081e+00 -1.09336808e-01
-2.42995694e-01 5.66351175e-01 -9.20567885e-02 -1.46409631e-01
-4.95581985e-01 -3.55204672e-01 6.23252876e-02 6.90328836e-01
-8.63323361e-02 2.06464499e-01 -2.50441492e-01 -3.81768972e-01
-3.82330269e-01 -6.22099638e-01 5.92343032e-01 -4.87048775e-01
-7.99157143e-01 2.18624890e-01 -1.71729416e-01 -1.13762486e+00
-1.39305219e-01 -7.62938440e-01 -6.93596005e-01 6.64076746e-01
-1.48927474e+00 -1.00460219e+00 -4.87134606e-01 -2.30810549e-02
1.58226207e-01 3.21845949e-01 1.09404731e+00 3.25461403e-02
-6.39728248e-01 3.72790128e-01 -1.71381250e-01 1.25017734e-02
9.37192440e-01 -1.52768981e+00 -3.91845644e-01 4.27917898e-01
-7.97553718e-01 5.55499196e-01 9.08849061e-01 -5.30626357e-01
-1.11361575e+00 -7.52685070e-01 1.19737220e+00 2.73445081e-02
5.95735669e-01 7.43862912e-02 -6.37948871e-01 2.33434260e-01
1.79290503e-01 -1.29850000e-01 1.33474815e+00 -3.80747795e-01
8.88032243e-02 -7.95266181e-02 -1.37116826e+00 6.64820015e-01
-9.74009857e-02 -6.48393854e-02 -1.04499869e-01 3.63994747e-01
-5.26415110e-01 -3.02789241e-01 -9.64581788e-01 2.43094921e-01
7.89143860e-01 -1.22056413e+00 2.91804820e-01 5.35048962e-01
4.20644522e-01 -4.91967142e-01 2.33428597e-01 -1.02811730e+00
-2.98890054e-01 -2.04350680e-01 3.75289857e-01 9.93733346e-01
5.67382276e-01 -7.64764845e-01 1.08026516e+00 2.69128740e-01
8.20000842e-02 -1.10179937e+00 -7.92240620e-01 -2.57780313e-01
6.60744309e-02 2.54208356e-01 1.71629861e-01 8.17581713e-01
5.35251915e-01 -4.56292666e-02 5.28641164e-01 -1.50633588e-01
4.56528634e-01 -3.50607991e-01 6.75006807e-01 -1.25083411e+00
-1.12728104e-01 -6.24877810e-01 -8.84500265e-01 4.96357419e-02
-4.31474179e-01 -8.04080427e-01 -2.08218515e-01 -1.92324078e+00
5.63408077e-01 -2.04344824e-01 2.49101613e-02 6.67005718e-01
-3.09520252e-02 7.37189710e-01 -1.77735016e-01 3.36759090e-01
-2.26296410e-01 -5.59644043e-01 9.49760437e-01 -4.90044728e-02
9.06040370e-02 -3.02268267e-01 -9.76579845e-01 7.39333093e-01
1.25503516e+00 -4.48556513e-01 1.22929271e-02 1.92168042e-01
5.23026228e-01 3.38074565e-02 2.42192268e-01 -1.20182705e+00
3.45390737e-01 -1.26227275e-01 5.18624127e-01 -7.63727427e-01
2.04972133e-01 -7.59835124e-01 5.17597377e-01 9.96973097e-01
1.66002244e-01 -5.34976982e-02 2.68260360e-01 1.07141681e-01
-1.47835523e-01 -3.99133712e-01 1.02720809e+00 -3.15015107e-01
-4.61458057e-01 -7.43024647e-02 -1.10503864e+00 -5.55369794e-01
1.49466157e+00 -6.20399356e-01 -5.34423769e-01 -4.42461520e-01
-4.73220617e-01 2.58223832e-01 1.12570560e+00 -4.06875968e-01
1.65279776e-01 -9.24911201e-01 -8.08674216e-01 -2.10590348e-01
2.61878788e-01 -1.44173115e-01 2.52440006e-01 1.44047987e+00
-1.31732237e+00 5.09329677e-01 -3.15026730e-01 -5.55639923e-01
-1.72071576e+00 1.72604084e-01 3.13751668e-01 -7.76845574e-01
-2.63039589e-01 7.31164098e-01 -4.32296954e-02 2.13286906e-01
-8.14913660e-02 -1.78568333e-01 -5.50964832e-01 1.47656545e-01
8.25831771e-01 5.29627502e-01 3.92822534e-01 -5.99370718e-01
-4.96337712e-01 4.21446532e-01 -5.51167786e-01 -9.79224145e-02
9.34446037e-01 2.09191069e-02 -4.89606172e-01 5.60536921e-01
9.59709764e-01 1.25161096e-01 -4.41307127e-01 4.50315803e-01
-1.98123381e-01 -2.17230454e-01 2.37749308e-01 -8.24433208e-01
-7.16527581e-01 6.12696230e-01 6.15811884e-01 3.22742820e-01
1.19046354e+00 2.92764623e-02 4.19480920e-01 6.41027018e-02
3.58601868e-01 -9.13146973e-01 -2.86684066e-01 1.23448566e-01
6.24591649e-01 -9.88352001e-01 2.47471005e-01 -6.00667536e-01
-2.76902914e-01 1.19761550e+00 9.59441662e-02 -1.42036632e-01
6.94224417e-01 5.19200921e-01 4.40047175e-01 -2.22982183e-01
-6.41289353e-01 -1.22063912e-01 -2.62940407e-01 3.84959072e-01
1.01350272e+00 4.64233160e-02 -1.03003705e+00 1.73456132e-01
-3.52620214e-01 2.79381037e-01 7.96828806e-01 1.31450570e+00
-5.86396575e-01 -1.18297184e+00 -6.07158482e-01 8.30418408e-01
-9.96578276e-01 4.07096118e-01 -6.85946882e-01 8.41163635e-01
7.37536699e-03 9.68801022e-01 3.85296755e-02 8.59783664e-02
-6.74307570e-02 -1.05418123e-01 4.91382420e-01 -5.68363309e-01
-5.93851388e-01 1.45005301e-01 5.13635911e-02 -1.38108328e-01
-5.40377140e-01 -7.38943875e-01 -1.51243711e+00 -5.33906579e-01
-1.91778824e-01 3.22352916e-01 8.77033353e-01 7.22463191e-01
-2.28994831e-01 2.64896870e-01 2.63093054e-01 -5.82820117e-01
1.83992594e-01 -8.06720018e-01 -7.92478204e-01 -1.86468616e-01
2.45362639e-01 -4.85833913e-01 -6.81148469e-01 2.76965201e-01] | [15.071793556213379, -3.117854118347168] |
81ad0731-4cef-4905-954a-d3eb6d006766 | high-resolution-image-editing-via-multi-stage | 2210.12965 | null | https://arxiv.org/abs/2210.12965v1 | https://arxiv.org/pdf/2210.12965v1.pdf | High-Resolution Image Editing via Multi-Stage Blended Diffusion | Diffusion models have shown great results in image generation and in image editing. However, current approaches are limited to low resolutions due to the computational cost of training diffusion models for high-resolution generation. We propose an approach that uses a pre-trained low-resolution diffusion model to edit images in the megapixel range. We first use Blended Diffusion to edit the image at a low resolution, and then upscale it in multiple stages, using a super-resolution model and Blended Diffusion. Using our approach, we achieve higher visual fidelity than by only applying off the shelf super-resolution methods to the output of the diffusion model. We also obtain better global consistency than directly using the diffusion model at a higher resolution. | ['Minjun Li', 'Johannes Ackermann'] | 2022-10-24 | null | null | null | null | ['image-inpainting'] | ['computer-vision'] | [ 4.01998699e-01 7.78530538e-02 2.87820667e-01 -5.26816174e-02
-8.19715917e-01 -3.34531575e-01 7.64285564e-01 -3.31954956e-01
-2.64306664e-01 9.75252509e-01 1.37365073e-01 1.69161439e-01
1.39255688e-01 -1.14303732e+00 -5.21643400e-01 -3.95096600e-01
3.42185557e-01 2.54654825e-01 7.61714041e-01 -2.64045596e-01
2.18602672e-01 5.72786152e-01 -1.36077714e+00 5.08895338e-01
1.19695222e+00 5.23375928e-01 5.20397484e-01 1.09127963e+00
-1.80335358e-01 7.96841919e-01 -6.76002800e-01 -2.09532350e-01
3.36898804e-01 -9.45324779e-01 -7.72691846e-01 1.80070147e-01
6.80916786e-01 -7.07755089e-01 -8.06293488e-02 1.20080948e+00
4.36777860e-01 3.51138599e-02 5.06352007e-01 -3.87710631e-01
-1.22314382e+00 3.56405407e-01 -1.07986856e+00 3.71345997e-01
5.96556962e-01 -6.54442906e-02 3.67454052e-01 -8.09269845e-01
1.21083951e+00 1.28385389e+00 6.94234908e-01 7.43910849e-01
-1.58415222e+00 -4.04943049e-01 -2.24150717e-01 -1.50677800e-01
-1.36991739e+00 -4.78468508e-01 6.01800382e-01 -3.73372674e-01
8.48442078e-01 3.87720875e-02 5.73469639e-01 7.67865479e-01
3.23308587e-01 6.38015661e-03 1.47598481e+00 -5.10186076e-01
2.19869707e-03 6.67068213e-02 -5.48123240e-01 7.75320530e-01
5.69862239e-02 3.75412583e-01 -4.73630458e-01 1.48017108e-01
1.80096042e+00 -3.58080447e-01 -3.18233728e-01 7.98549503e-02
-1.16545403e+00 6.15874588e-01 4.70692426e-01 5.67655981e-01
-6.92420185e-01 1.80812329e-01 -2.16734365e-01 3.17900479e-01
9.08646405e-01 4.43587244e-01 8.10893029e-02 5.75013086e-02
-1.61629844e+00 3.18482108e-02 3.13261032e-01 7.53288150e-01
8.29354167e-01 2.03466341e-01 -2.44167656e-01 1.06365335e+00
1.05328232e-01 2.58338660e-01 3.91620696e-01 -1.53191364e+00
1.22377366e-01 8.91897455e-02 3.50496471e-01 -8.02265763e-01
1.04907341e-01 -5.56632429e-02 -9.63505924e-01 1.07376659e+00
2.86473304e-01 -1.85346976e-02 -1.24121940e+00 1.26677096e+00
4.57125485e-01 8.59248172e-03 4.46413038e-03 9.12737966e-01
4.84681249e-01 1.02076721e+00 -1.75861552e-01 -3.57026905e-01
1.01239336e+00 -1.26662064e+00 -9.33917522e-01 1.36716530e-01
1.84798911e-01 -1.10069537e+00 9.13737416e-01 5.07708609e-01
-1.79533446e+00 -7.59044290e-01 -1.21875906e+00 -5.26022792e-01
-9.36775059e-02 -2.22473033e-02 1.53825924e-01 4.32246029e-01
-1.51029217e+00 9.92620111e-01 -6.39425576e-01 -3.07329327e-01
4.21222806e-01 -1.40781300e-02 -2.70856678e-01 -1.81816220e-01
-1.13816392e+00 1.22930932e+00 1.10538311e-01 -2.86402464e-01
-7.16367126e-01 -8.48923683e-01 -6.44388080e-01 -1.50389761e-01
-1.31678924e-01 -1.01044953e+00 9.27963495e-01 -1.02171290e+00
-1.88010120e+00 7.14318275e-01 -1.86784908e-01 -3.64478379e-01
1.00311863e+00 -5.17410645e-03 -2.79994875e-01 3.80319953e-01
-2.78823469e-02 8.78457725e-01 1.12010610e+00 -1.66494000e+00
-7.65089989e-01 1.57285824e-01 1.47225589e-01 1.95290908e-01
-1.29017308e-01 8.96722525e-02 -5.62864602e-01 -7.69033968e-01
-7.68734701e-03 -4.76093441e-01 -4.36569601e-01 4.82807428e-01
6.63631856e-02 4.61645812e-01 1.00600338e+00 -1.09859073e+00
1.30270219e+00 -1.96711898e+00 1.53419495e-01 -1.05959950e-02
3.53534609e-01 3.35953295e-01 -2.61739165e-01 9.29342881e-02
2.26061508e-01 2.11349860e-01 -4.31414068e-01 -6.11266196e-01
-6.02268875e-01 9.33735445e-02 -1.38061792e-01 6.80372771e-03
3.54104117e-02 6.84020340e-01 -9.33205843e-01 -9.12628174e-01
4.33131784e-01 1.08307016e+00 -4.05929953e-01 4.53408808e-02
-1.41177580e-01 5.25850952e-01 4.03779857e-02 3.03983003e-01
9.28273916e-01 -1.09255329e-01 -1.33402944e-01 -3.21502000e-01
-4.43417370e-01 -1.97937921e-01 -1.26902163e+00 1.81354427e+00
-6.73266470e-01 6.67925239e-01 2.83137590e-01 -8.21854398e-02
9.28211272e-01 1.44121736e-01 3.91239882e-01 -8.02004457e-01
-3.45296234e-01 1.39106482e-01 -3.55406076e-01 -1.61455631e-01
1.00626361e+00 -3.12515110e-01 5.99604487e-01 5.98254502e-01
-1.13622986e-01 -7.30496168e-01 4.36697513e-01 2.54482180e-01
6.78750932e-01 5.11582613e-01 -2.60656625e-02 -2.35152896e-02
3.75956267e-01 1.79264277e-01 1.64437547e-01 6.61868870e-01
2.34158531e-01 1.37442434e+00 9.56612676e-02 -1.76694095e-01
-1.68326497e+00 -1.03106606e+00 -2.64311340e-02 4.79297519e-01
1.78747371e-01 -2.15027198e-01 -1.12825251e+00 -2.57699221e-01
-4.07730520e-01 6.14712715e-01 -6.92187428e-01 9.25182551e-02
-5.82822859e-01 -6.09788120e-01 5.12555420e-01 4.89867747e-01
8.87744486e-01 -7.37410486e-01 -5.49323261e-01 2.29595795e-01
-5.71902767e-02 -1.01712072e+00 -8.05570185e-01 -3.63809705e-01
-1.08795536e+00 -5.16385972e-01 -1.36593521e+00 -6.99338496e-01
9.09273088e-01 2.39551008e-01 9.36384201e-01 2.66716838e-01
-3.57911557e-01 6.88871592e-02 -8.50989446e-02 1.50397018e-01
-9.08099651e-01 -2.26828665e-01 -2.96207935e-01 -1.37338161e-01
-6.66103065e-01 -4.50981349e-01 -7.62929678e-01 1.59717485e-01
-1.15085077e+00 4.68741298e-01 6.04790866e-01 7.01029539e-01
8.59047592e-01 3.23733598e-01 1.76842168e-01 -9.47777212e-01
9.61651981e-01 8.85720775e-02 -7.09530771e-01 1.44183636e-01
-8.88556063e-01 9.42660719e-02 6.10900998e-01 -6.02997839e-01
-1.77494490e+00 -4.18481603e-02 -1.46747991e-01 -5.19694269e-01
2.12662011e-01 1.76869735e-01 3.69193077e-01 -3.26814145e-01
8.87301385e-01 6.00574315e-02 3.59998681e-02 -4.21781391e-01
6.88031495e-01 2.39995390e-01 5.67113757e-01 -2.04157978e-01
8.95539939e-01 8.25523078e-01 -7.33141899e-02 -8.08397889e-01
-5.49436867e-01 2.66331851e-01 -9.86174226e-01 -3.43191147e-01
9.63238120e-01 -7.51197159e-01 2.78174013e-01 5.11511564e-01
-1.23692703e+00 -6.54121637e-01 -6.82638288e-01 3.47346723e-01
-5.02177894e-01 5.57723701e-01 -9.77566779e-01 -5.65373600e-01
-3.02665561e-01 -7.91969895e-01 9.17213738e-01 5.03413856e-01
-1.37884244e-01 -1.18038011e+00 2.97484964e-01 1.23036638e-01
9.74233150e-01 2.19567418e-01 6.32152677e-01 6.11083686e-01
-8.35991621e-01 1.13266625e-01 -5.99491298e-01 3.54884923e-01
2.21437469e-01 4.36696321e-01 -7.33374238e-01 -1.00864107e-02
-9.86739099e-02 -7.91242644e-02 9.59633827e-01 5.74210882e-01
7.85029769e-01 3.26485820e-02 -2.69719720e-01 8.05051148e-01
1.75917053e+00 9.45898220e-02 1.00998998e+00 3.66544724e-01
5.96822083e-01 3.61692518e-01 5.90558529e-01 1.51529714e-01
2.18921304e-01 6.52050197e-01 -1.04860798e-01 -6.68001175e-01
-9.98823166e-01 -3.67304623e-01 2.02746809e-01 4.95554745e-01
-6.98579252e-01 4.05673794e-02 -6.37717426e-01 5.94819307e-01
-1.61966693e+00 -1.23392153e+00 -2.78351039e-01 2.07690763e+00
1.16586173e+00 4.03874591e-02 -1.33991003e-01 -2.20837608e-01
7.82154083e-01 2.93369800e-01 -2.87570477e-01 -6.34101987e-01
-3.03852201e-01 2.37873182e-01 4.70205545e-01 1.14945793e+00
-7.10945964e-01 1.12581801e+00 7.94256353e+00 7.66624212e-01
-1.13032699e+00 3.99838686e-01 5.58310390e-01 -1.59871593e-01
-4.47912961e-01 -5.31207658e-02 -6.08025908e-01 1.89549983e-01
6.86312497e-01 -3.69043916e-01 6.55079246e-01 4.33398217e-01
5.36009192e-01 -4.83109325e-01 -5.96099019e-01 7.94788182e-01
1.74501017e-01 -1.53946567e+00 3.17801833e-01 2.34755591e-01
1.36446345e+00 -4.52033252e-01 1.33952275e-01 -3.45242560e-01
4.74862635e-01 -9.38855469e-01 5.81884384e-01 9.32569683e-01
1.12044573e+00 -6.30986691e-01 2.03559473e-01 2.21734926e-01
-1.12248266e+00 3.84949028e-01 -7.18213379e-01 2.41761893e-01
6.07895613e-01 8.20016861e-01 -3.52477163e-01 4.26102608e-01
6.55762374e-01 3.57672393e-01 -3.60748023e-01 7.92592347e-01
-2.39797488e-01 5.63017800e-02 -2.21650481e-01 6.40768051e-01
-1.77801564e-01 -6.25473559e-01 2.87045479e-01 1.23242724e+00
7.10074306e-01 2.59695202e-01 -3.57260525e-01 1.26584470e+00
-8.18523243e-02 -1.53927639e-01 -5.69878995e-01 1.84373826e-01
2.22270668e-01 1.30021513e+00 -6.95837438e-01 -7.17430413e-01
-3.27279061e-01 1.79774988e+00 2.90295452e-01 4.28420186e-01
-9.35855031e-01 -2.48677328e-01 2.63831526e-01 3.66529614e-01
3.74883205e-01 -4.78514671e-01 -4.45627779e-01 -1.05204690e+00
-2.63695717e-01 -4.88217741e-01 4.05272245e-02 -1.31550598e+00
-8.91746283e-01 9.50638473e-01 -4.39756885e-02 -1.09451210e+00
-4.12505835e-01 -4.92382310e-02 -4.30040717e-01 1.45219731e+00
-1.70540714e+00 -1.19768310e+00 -4.26312000e-01 5.81978858e-01
6.15651488e-01 4.70031619e-01 6.42846644e-01 3.06319475e-01
6.71713799e-03 1.87277377e-01 2.27677360e-01 -2.62419462e-01
8.67656410e-01 -1.30556381e+00 4.02167737e-01 1.02456725e+00
-1.44843414e-01 3.97535175e-01 6.25643373e-01 -8.89528871e-01
-7.66392589e-01 -1.01808619e+00 9.36226726e-01 -3.70175332e-01
2.82630384e-01 7.46315867e-02 -1.15576148e+00 3.59393626e-01
6.65259659e-01 2.67274342e-02 1.36025742e-01 -5.19190073e-01
8.38211253e-02 5.21930866e-02 -1.51075077e+00 6.74624562e-01
1.05388463e+00 -6.10317051e-01 -4.12602097e-01 -5.54079153e-02
5.13047040e-01 -6.78244948e-01 -1.18396342e+00 8.73489380e-02
5.85931599e-01 -1.10098863e+00 1.06663215e+00 6.79356530e-02
9.79163826e-01 -5.96475065e-01 3.66997629e-01 -1.56919503e+00
-6.55690074e-01 -3.60819668e-01 8.16273037e-03 1.33233285e+00
5.56445837e-01 -3.88880134e-01 4.82707322e-01 5.77603638e-01
2.60631800e-01 -3.17432821e-01 -6.38009548e-01 -6.05642736e-01
1.56419259e-02 2.85753697e-01 3.17941993e-01 1.13459969e+00
-3.51505756e-01 1.28079757e-01 -5.88822484e-01 2.14793950e-01
9.28664744e-01 1.10358141e-01 4.54228997e-01 -8.70895922e-01
-1.31951421e-01 -4.92883712e-01 1.00983508e-01 -1.14982510e+00
-4.39821392e-01 -3.68114680e-01 1.03838734e-01 -2.16732454e+00
1.87944815e-01 -3.37708265e-01 1.78193107e-01 -5.62827699e-02
-2.71655828e-01 7.84991622e-01 1.36584044e-01 5.01592100e-01
-5.21807186e-02 2.11972773e-01 1.84388781e+00 4.78309989e-02
-3.36924016e-01 -6.43378913e-01 -4.89594251e-01 7.17399478e-01
6.51767731e-01 -2.85750300e-01 -3.85307789e-01 -8.61037135e-01
-5.77315018e-02 1.19678259e-01 4.53439429e-02 -1.06805038e+00
2.71242529e-01 -2.20457911e-01 8.39229405e-01 -3.74932677e-01
4.67678756e-01 -6.21904910e-01 5.27994812e-01 3.13942552e-01
-4.74445492e-01 9.49598402e-02 1.29950300e-01 3.44094247e-01
-2.51208305e-01 -2.93145835e-01 1.35882163e+00 -3.76621902e-01
-5.99748552e-01 1.86549038e-01 -3.63247931e-01 -4.57554519e-01
1.02113938e+00 -4.80171889e-01 -4.27536130e-01 -5.48024416e-01
-9.45249736e-01 -1.55526012e-01 1.16218638e+00 2.55717695e-01
7.81960547e-01 -1.25961995e+00 -8.07004094e-01 2.14068174e-01
-5.12745798e-01 7.17458650e-02 3.13124537e-01 6.53023064e-01
-9.42178667e-01 -2.14054093e-01 -5.08891106e-01 -4.35557097e-01
-1.33837140e+00 5.24646997e-01 6.30213439e-01 -3.55596185e-01
-7.45827734e-01 6.50332391e-01 5.23378477e-02 1.46961555e-01
-3.08247477e-01 -8.44755545e-02 -1.40397146e-01 -2.08997726e-01
7.53362834e-01 4.65944141e-01 -2.74398923e-01 -6.25555813e-01
1.06594354e-01 1.11676908e+00 -1.82036906e-02 -7.53499091e-01
1.12534821e+00 -5.45149982e-01 -1.29164591e-01 1.09139882e-01
6.69646740e-01 3.96983385e-01 -1.66087818e+00 -9.13644433e-02
-4.92739946e-01 -8.97459209e-01 4.35568780e-01 -8.14727485e-01
-1.17975068e+00 8.31979573e-01 6.69692814e-01 2.84761339e-01
1.26718974e+00 -1.89212114e-01 7.73045480e-01 -3.75677407e-01
2.57145077e-01 -1.41104937e+00 4.94541526e-02 3.30049843e-01
1.04072320e+00 -9.17765200e-01 3.29763710e-01 -5.84476292e-01
-6.89253986e-01 1.17816532e+00 5.60300291e-01 -1.43646911e-01
4.23956633e-01 6.96199477e-01 2.73279041e-01 1.33571967e-01
-7.65165329e-01 -1.82827294e-01 1.14541456e-01 8.36481690e-01
6.32240176e-01 -4.35601234e-01 -5.89986682e-01 -1.73057288e-01
1.15400299e-01 5.00740349e-01 8.05817604e-01 7.54244328e-01
-5.06181538e-01 -1.18307662e+00 -6.25432968e-01 1.73079565e-01
-2.90174276e-01 -2.77378500e-01 -2.67788380e-01 4.50013906e-01
1.41556039e-01 7.91490674e-01 1.00263901e-01 -1.52807027e-01
4.06317174e-01 -2.71033853e-01 8.75904024e-01 -4.51582581e-01
-3.25267315e-01 2.74995297e-01 5.00020869e-02 -7.04333663e-01
-6.54214859e-01 -3.96397054e-01 -1.12952566e+00 -4.44589019e-01
-1.79493591e-01 -1.31645843e-01 6.54198349e-01 4.59835649e-01
4.74506319e-01 6.42889082e-01 3.96485090e-01 -1.18686962e+00
-6.40239045e-02 -9.17237520e-01 -6.22825086e-01 3.13946873e-01
2.82398075e-01 -1.99854806e-01 -2.18318701e-01 6.25390589e-01] | [11.20821762084961, -1.4833232164382935] |
f7e6f8d3-01a8-4999-9b8d-945f4ed6e6f6 | graph-convolution-based-efficient-re-ranking | 2306.08792 | null | https://arxiv.org/abs/2306.08792v1 | https://arxiv.org/pdf/2306.08792v1.pdf | Graph Convolution Based Efficient Re-Ranking for Visual Retrieval | Visual retrieval tasks such as image retrieval and person re-identification (Re-ID) aim at effectively and thoroughly searching images with similar content or the same identity. After obtaining retrieved examples, re-ranking is a widely adopted post-processing step to reorder and improve the initial retrieval results by making use of the contextual information from semantically neighboring samples. Prevailing re-ranking approaches update distance metrics and mostly rely on inefficient crosscheck set comparison operations while computing expanded neighbors based distances. In this work, we present an efficient re-ranking method which refines initial retrieval results by updating features. Specifically, we reformulate re-ranking based on Graph Convolution Networks (GCN) and propose a novel Graph Convolution based Re-ranking (GCR) for visual retrieval tasks via feature propagation. To accelerate computation for large-scale retrieval, a decentralized and synchronous feature propagation algorithm which supports parallel or distributed computing is introduced. In particular, the plain GCR is extended for cross-camera retrieval and an improved feature propagation formulation is presented to leverage affinity relationships across different cameras. It is also extended for video-based retrieval, and Graph Convolution based Re-ranking for Video (GCRV) is proposed by mathematically deriving a novel profile vector generation method for the tracklet. Without bells and whistles, the proposed approaches achieve state-of-the-art performances on seven benchmark datasets from three different tasks, i.e., image retrieval, person Re-ID and video-based person Re-ID. | ['Fan Wang', 'Weihua Chen', 'Chong Liu', 'Hongsong Wang', 'Qi Qian', 'Yuqi Zhang'] | 2023-06-15 | null | null | null | null | ['person-re-identification'] | ['computer-vision'] | [-5.29736653e-02 -9.35457647e-01 -1.26297027e-01 -3.95089626e-01
-8.15480471e-01 -6.44959450e-01 7.75227785e-01 4.78263110e-01
-6.55303180e-01 3.44486594e-01 1.75657406e-01 1.98654577e-01
-6.78329945e-01 -6.99282110e-01 -3.52682859e-01 -5.49356222e-01
-6.20519370e-02 3.23172748e-01 6.03157766e-02 -1.38695374e-01
3.17208529e-01 7.32644796e-01 -1.75498462e+00 9.62085500e-02
6.36811674e-01 1.00054693e+00 -5.28706349e-02 6.55950069e-01
1.83566794e-01 4.63199556e-01 -5.48761427e-01 -8.11200976e-01
2.85574675e-01 -9.51744318e-02 -5.80375791e-01 -5.60597405e-02
7.36074567e-01 -5.49351394e-01 -8.71842980e-01 1.14033246e+00
7.43635118e-01 7.15016484e-01 6.94941401e-01 -1.34194088e+00
-9.79995549e-01 2.20271185e-01 -5.62671125e-01 2.56093651e-01
6.19184732e-01 -3.24732780e-01 9.76221621e-01 -1.00522625e+00
6.23401642e-01 1.16030371e+00 6.63620830e-01 3.94731343e-01
-9.01464403e-01 -6.89348161e-01 1.58860803e-01 6.11696362e-01
-1.96564150e+00 -1.64875701e-01 7.05087066e-01 -2.18337789e-01
7.58558095e-01 6.04807615e-01 8.05590153e-01 5.87828159e-01
-2.62016714e-01 7.42572188e-01 6.27443731e-01 -2.37776801e-01
-2.06286803e-01 -7.23792538e-02 3.80815327e-01 7.01391757e-01
4.34795380e-01 2.18756218e-02 -6.16515934e-01 -4.02947962e-01
5.22633612e-01 5.90208113e-01 -4.13241476e-01 -2.80262917e-01
-1.24758244e+00 6.98536396e-01 7.23817408e-01 2.01805189e-01
-2.24289909e-01 2.17959076e-01 6.73753262e-01 3.85369301e-01
2.90590256e-01 8.75735879e-02 2.31440544e-01 3.16185653e-01
-1.12871838e+00 3.51017267e-01 3.38451117e-01 8.93700063e-01
8.21094573e-01 -4.77941990e-01 -7.46249616e-01 1.13284659e+00
2.17883721e-01 7.07199514e-01 5.70613682e-01 -6.07281148e-01
2.80580878e-01 5.54893494e-01 8.88645872e-02 -1.62242174e+00
-2.39261255e-01 -4.37490761e-01 -1.05177855e+00 -4.09685045e-01
1.28473908e-01 3.80716532e-01 -9.16275561e-01 1.49210227e+00
2.77739108e-01 4.70972300e-01 -6.09106906e-02 1.15216398e+00
1.19041693e+00 6.05496943e-01 -9.07114595e-02 -4.48206440e-02
1.37463140e+00 -1.00130832e+00 -3.86133283e-01 4.15174603e-01
3.14412147e-01 -6.85598493e-01 5.06194949e-01 1.70613006e-01
-8.15354884e-01 -8.48565102e-01 -9.53350902e-01 -5.46424203e-02
-5.63676476e-01 2.04167053e-01 6.23387694e-01 5.83114564e-01
-1.45705533e+00 6.76562905e-01 -1.23918347e-01 -5.27606845e-01
2.38581434e-01 5.63724041e-01 -7.17158198e-01 -4.64352638e-01
-1.27414167e+00 4.28286880e-01 2.69669652e-01 4.51830089e-01
-5.44171512e-01 -6.24372184e-01 -8.06403279e-01 9.71455649e-02
8.77784640e-02 -1.00294638e+00 5.94811440e-01 -8.49439859e-01
-1.20081472e+00 9.53586102e-01 -1.48437873e-01 -3.13466728e-01
3.95158648e-01 -1.64664015e-01 -7.14888692e-01 5.30854106e-01
7.85485879e-02 4.74261016e-01 1.06503057e+00 -1.02308953e+00
-5.35924077e-01 -4.14792597e-01 -7.28184357e-02 4.62973952e-01
-6.44033074e-01 9.29785818e-02 -1.31013465e+00 -7.51458406e-01
-7.95266032e-02 -9.53283191e-01 1.09170668e-01 -1.45500004e-01
-3.73033851e-01 -4.75275695e-01 4.39254820e-01 -7.18204319e-01
1.42604077e+00 -2.01115441e+00 2.85026371e-01 9.55539823e-01
6.27338827e-01 4.92533624e-01 -4.40331131e-01 5.32809973e-01
-7.44361356e-02 -6.03316687e-02 2.62172490e-01 -5.39782941e-01
9.99187455e-02 -3.01804870e-01 -7.89612159e-02 6.78717732e-01
-8.14318508e-02 1.27634203e+00 -9.78042960e-01 -6.88859522e-01
3.79977852e-01 8.04727435e-01 -3.11532766e-01 5.48586659e-02
5.89459062e-01 8.45402554e-02 -3.39961290e-01 7.94643164e-01
8.68752837e-01 -5.09616375e-01 -4.91919294e-02 -6.70919955e-01
9.72323641e-02 -6.01539135e-01 -1.18407369e+00 1.79464972e+00
-1.27284035e-01 5.44246018e-01 -2.76752114e-01 -9.03848946e-01
8.53489697e-01 -1.74934059e-01 4.94781405e-01 -8.93779337e-01
-2.50016581e-02 1.44029960e-01 -5.85337818e-01 -2.37749219e-01
1.04188740e+00 4.72932845e-01 -2.58659525e-03 3.42353731e-01
-3.33256945e-02 5.93950331e-01 2.77758479e-01 6.07615888e-01
9.05870795e-01 -1.32081062e-01 -7.60053694e-02 -1.61644220e-02
1.03239644e+00 -2.04120517e-01 9.51395482e-02 9.49415565e-01
-9.24847275e-02 7.64222026e-01 -1.72447532e-01 -4.19091463e-01
-7.60996282e-01 -9.94753182e-01 1.07050128e-01 1.08803380e+00
8.04093778e-01 -6.84415877e-01 -4.59150612e-01 -5.47460973e-01
2.29895785e-01 -7.65489340e-02 -6.24881089e-01 -1.75574034e-01
-3.99230987e-01 -7.29783654e-01 5.98718703e-01 2.49562666e-01
7.66774237e-01 -6.57610714e-01 1.67536400e-02 -1.18443564e-01
-2.87330955e-01 -8.34486425e-01 -1.07434928e+00 -6.34402871e-01
-4.93469685e-01 -1.15587795e+00 -1.44797349e+00 -8.87823284e-01
8.41138601e-01 8.64974737e-01 9.20194626e-01 5.86158037e-01
-4.84602720e-01 1.07436049e+00 -4.86514270e-01 3.61759722e-01
3.85165453e-01 -5.13396971e-02 1.41925260e-01 4.65480298e-01
4.58404839e-01 -2.29201540e-01 -1.08709323e+00 2.77016968e-01
-9.77044404e-01 -3.23488474e-01 4.61874634e-01 1.07498872e+00
8.06097984e-01 1.21186245e-02 3.05582196e-01 -4.30694640e-01
9.48645532e-01 -2.19337493e-01 -5.63757181e-01 8.71951699e-01
-7.10798860e-01 -1.69878043e-02 3.28087449e-01 -6.15386009e-01
-6.87465787e-01 -2.64432520e-01 3.13769609e-01 -8.97573292e-01
2.50556678e-01 4.71516669e-01 1.70707822e-01 -4.81607139e-01
2.25084782e-01 4.66523528e-01 -1.14592135e-01 -2.32012898e-01
4.92737770e-01 5.81633747e-01 5.79934359e-01 -3.18381667e-01
1.08266068e+00 6.70090675e-01 3.89438085e-02 -6.84163034e-01
-3.51667792e-01 -1.06078780e+00 -4.79537398e-01 -4.79903966e-01
6.58151865e-01 -9.60077107e-01 -1.14044559e+00 7.05674827e-01
-1.08092177e+00 1.85526311e-01 -1.18696764e-02 5.20417750e-01
8.68952647e-02 8.87274563e-01 -4.83759284e-01 -6.20628238e-01
-7.54449308e-01 -9.35834587e-01 1.32946491e+00 5.04433513e-01
2.94604242e-01 -9.17613387e-01 1.18050955e-01 2.08947569e-01
5.26482999e-01 -9.20566171e-02 5.00420928e-01 -6.90495074e-01
-6.23052776e-01 -6.18012309e-01 -7.55429924e-01 3.20547335e-02
2.82663833e-02 -2.97598094e-01 -8.05308342e-01 -6.98328078e-01
-8.61866295e-01 -1.30538605e-02 1.10417223e+00 2.31444016e-01
1.17950356e+00 -5.08592604e-03 -6.89728796e-01 8.49792838e-01
1.40562904e+00 -1.60660684e-01 6.93754911e-01 4.38028336e-01
8.88528526e-01 5.33680499e-01 4.81816888e-01 4.47354943e-01
4.49815452e-01 8.66245031e-01 1.14844721e-02 -1.51914991e-02
-1.80627838e-01 -1.31382883e-01 5.84164225e-02 7.50884712e-01
-2.73826361e-01 -3.82912785e-01 -4.54443604e-01 4.86151516e-01
-2.00761628e+00 -1.15083635e+00 3.10050063e-02 2.59643149e+00
1.75772890e-01 -5.93415439e-01 2.43815780e-01 -4.50442396e-02
1.27880013e+00 1.10572159e-01 -4.58500355e-01 2.22084478e-01
-1.45444453e-01 2.04624146e-01 6.27613068e-01 5.14870167e-01
-1.20490897e+00 8.03681135e-01 4.87745237e+00 1.24512184e+00
-9.23434019e-01 7.26790577e-02 4.75952864e-01 -7.09473416e-02
-2.86438614e-01 -1.71887904e-01 -9.20467138e-01 3.23861063e-01
3.11195105e-01 -3.42895061e-01 7.13165104e-01 5.06371379e-01
-3.43459547e-01 7.91580752e-02 -9.27052259e-01 1.95623672e+00
7.06352234e-01 -1.34452641e+00 5.79790413e-01 -1.47499382e-01
4.83129293e-01 -2.73884147e-01 2.27499366e-01 1.97934106e-01
-7.62875676e-02 -8.61533642e-01 4.12373275e-01 9.19298112e-01
9.83633459e-01 -8.91002953e-01 7.81055510e-01 -4.64802295e-01
-1.66772401e+00 -4.02747421e-03 -4.02891219e-01 5.47317505e-01
1.09969310e-01 4.48383659e-01 -2.31439814e-01 1.09247947e+00
9.32592928e-01 9.08215284e-01 -9.93296683e-01 1.41323352e+00
2.86531270e-01 -7.46022463e-02 -2.00103655e-01 -1.32735491e-01
-1.02643678e-02 -2.20232174e-01 5.63738048e-01 1.38119745e+00
3.84803295e-01 4.71648462e-02 4.61855792e-02 3.42864931e-01
-2.73977727e-01 3.50845635e-01 -3.06445688e-01 2.32197836e-01
4.66478229e-01 1.48859107e+00 -7.29388952e-01 -3.98746729e-01
-2.73522168e-01 1.52363658e+00 1.97701707e-01 7.48293221e-01
-8.26243162e-01 -8.34062338e-01 5.24800062e-01 -9.33069214e-02
2.58131862e-01 -5.42348102e-02 7.78007030e-01 -1.31857860e+00
1.61185756e-01 -5.02122343e-01 7.29104757e-01 -7.25864828e-01
-1.69267488e+00 7.41319954e-01 1.61148414e-01 -1.31112361e+00
-1.86508253e-01 -4.82577413e-01 -2.37212405e-01 8.10019851e-01
-1.90848553e+00 -1.38727331e+00 -7.81551421e-01 1.17181456e+00
6.56006485e-02 -3.95495892e-01 4.56834674e-01 8.03516388e-01
-4.24981713e-01 1.23584139e+00 1.97549924e-01 3.72818708e-01
9.33221161e-01 -8.26529682e-01 3.82686704e-02 8.24766636e-01
2.79923320e-01 9.91636753e-01 1.19493837e-02 -7.43260860e-01
-1.62259829e+00 -1.36269462e+00 8.02373230e-01 -5.19929789e-02
4.77123260e-01 -1.21481478e-01 -6.88586593e-01 2.09773049e-01
-1.70029506e-01 2.80979842e-01 6.48665607e-01 -2.76724230e-02
-6.61175728e-01 -4.80861425e-01 -9.03280139e-01 5.40123343e-01
1.27850544e+00 -8.73107016e-01 -1.37844577e-01 3.64321977e-01
4.78565305e-01 -2.75042802e-01 -7.92078495e-01 2.86090642e-01
8.79727662e-01 -7.89231837e-01 1.47289920e+00 -4.40304995e-01
-3.26843113e-01 -5.59313893e-01 -2.25863084e-01 -7.87671506e-01
-5.30427873e-01 -7.82603443e-01 1.21423174e-02 1.37468410e+00
-8.69100317e-02 -7.63859689e-01 5.90204477e-01 4.39725101e-01
2.52041936e-01 -3.51395249e-01 -7.68466294e-01 -7.60865331e-01
-5.75872958e-01 -2.00338602e-01 7.23927557e-01 7.67074823e-01
-2.79951960e-01 -7.33194128e-02 -5.00036597e-01 2.91709989e-01
8.46360683e-01 3.17600012e-01 7.66562283e-01 -1.24096167e+00
-3.08482230e-01 -4.43632811e-01 -9.02104795e-01 -1.08825731e+00
1.99019670e-01 -1.31692600e+00 -3.15971106e-01 -1.57837224e+00
7.56916046e-01 -4.34698254e-01 -7.18721807e-01 1.12123787e-01
-2.49797195e-01 8.00804079e-01 4.91629720e-01 6.25909507e-01
-1.25614572e+00 4.00721759e-01 8.03634822e-01 -4.59854335e-01
-1.93438172e-01 -1.10856079e-01 -4.73272383e-01 1.41488418e-01
3.64831954e-01 -1.70366868e-01 -3.68263662e-01 -4.68446702e-01
4.10808891e-01 -1.82415470e-01 9.69077051e-01 -1.04170048e+00
7.64535129e-01 4.34590071e-01 6.43147588e-01 -7.19149888e-01
4.67660636e-01 -6.28218114e-01 3.17134023e-01 1.87651023e-01
-4.08196330e-01 4.48839247e-01 1.16062001e-03 9.61244285e-01
-3.92663538e-01 -1.86788440e-01 2.80472666e-01 7.48094469e-02
-9.11376774e-01 7.73830533e-01 7.62292789e-03 -3.00354302e-01
8.37151766e-01 -4.17749643e-01 -3.45061839e-01 -5.72950423e-01
-6.58604622e-01 3.69662315e-01 3.26738566e-01 4.82717872e-01
1.16833127e+00 -1.72048211e+00 -7.22791612e-01 -1.22192306e-02
4.49955583e-01 -6.25713885e-01 6.16072953e-01 6.49778903e-01
-3.89525473e-01 4.63822991e-01 1.46952316e-01 -5.87353528e-01
-1.69986963e+00 6.20767951e-01 2.80437142e-01 -4.12116647e-01
-5.36504507e-01 9.31319654e-01 9.49807912e-02 -1.72815472e-01
2.35338435e-01 1.95832312e-01 -3.88355255e-01 2.69369155e-01
8.28601182e-01 5.18610895e-01 8.04782361e-02 -8.88966501e-01
-5.35373449e-01 9.83782113e-01 -2.85346657e-01 -1.48162870e-02
1.00351119e+00 -2.96323895e-01 -3.09915900e-01 -1.25993416e-01
1.49677694e+00 -1.88343041e-02 -6.81558728e-01 -3.39444429e-01
-2.10584700e-01 -6.75287008e-01 3.15978490e-02 -4.50771093e-01
-1.26484895e+00 4.55357105e-01 1.04381597e+00 -1.80709049e-01
1.30608881e+00 -7.41373748e-02 9.11858439e-01 5.52017391e-01
4.33489919e-01 -1.02677095e+00 3.03954571e-01 1.80604368e-01
6.50131941e-01 -1.06701565e+00 2.59105384e-01 -2.34301418e-01
-3.65138859e-01 1.08018696e+00 2.29284033e-01 -1.56827420e-01
6.71427727e-01 -4.77633566e-01 -3.98076236e-01 -3.91055882e-01
-3.21502686e-02 -4.81666028e-01 7.90123045e-01 6.41281068e-01
-3.62252071e-02 -5.25435880e-02 -2.13069528e-01 3.54123235e-01
3.69592458e-01 -7.20717832e-02 -2.51588523e-01 5.40727556e-01
2.28535552e-02 -9.82526422e-01 -4.65185165e-01 6.43411636e-01
-1.98818967e-01 -4.07709181e-01 -2.52407640e-01 7.11264193e-01
-9.23133567e-02 9.96559501e-01 1.25467211e-01 -7.86558807e-01
1.15823552e-01 -3.34821582e-01 6.05076671e-01 -1.58048466e-01
-6.98490560e-01 -6.00396916e-02 -2.35603809e-01 -6.77120864e-01
-7.03168750e-01 -5.14586508e-01 -8.63446116e-01 -4.57171470e-01
-5.98050117e-01 3.63111049e-01 3.94485205e-01 6.05867922e-01
6.30521595e-01 1.83331609e-01 6.42342329e-01 -9.11499023e-01
-1.14846483e-01 -4.68028843e-01 -5.67758799e-01 7.49304056e-01
1.97767407e-01 -6.92960858e-01 -2.34684691e-01 -1.57255992e-01] | [14.708818435668945, 0.9314708709716797] |
3a449a0c-8ef8-4c0e-873e-8b7c89ad0b38 | learning-from-good-trajectories-in-offline | 2211.15612 | null | https://arxiv.org/abs/2211.15612v2 | https://arxiv.org/pdf/2211.15612v2.pdf | Learning from Good Trajectories in Offline Multi-Agent Reinforcement Learning | Offline multi-agent reinforcement learning (MARL) aims to learn effective multi-agent policies from pre-collected datasets, which is an important step toward the deployment of multi-agent systems in real-world applications. However, in practice, each individual behavior policy that generates multi-agent joint trajectories usually has a different level of how well it performs. e.g., an agent is a random policy while other agents are medium policies. In the cooperative game with global reward, one agent learned by existing offline MARL often inherits this random policy, jeopardizing the performance of the entire team. In this paper, we investigate offline MARL with explicit consideration on the diversity of agent-wise trajectories and propose a novel framework called Shared Individual Trajectories (SIT) to address this problem. Specifically, an attention-based reward decomposition network assigns the credit to each agent through a differentiable key-value memory mechanism in an offline manner. These decomposed credits are then used to reconstruct the joint offline datasets into prioritized experience replay with individual trajectories, thereafter agents can share their good trajectories and conservatively train their policies with a graph attention network (GAT) based critic. We evaluate our method in both discrete control (i.e., StarCraft II and multi-agent particle environment) and continuous control (i.e, multi-agent mujoco). The results indicate that our method achieves significantly better results in complex and mixed offline multi-agent datasets, especially when the difference of data quality between individual trajectories is large. | ['Baoxiang Wang', 'Furui Liu', 'Kun Kuang', 'Qi Tian'] | 2022-11-28 | null | null | null | null | ['starcraft-ii', 'continuous-control', 'starcraft'] | ['playing-games', 'playing-games', 'playing-games'] | [-5.52854896e-01 -7.02109635e-02 -1.76747024e-01 4.35902268e-01
-6.68813586e-01 -5.49787700e-01 5.95179021e-01 2.71598101e-01
-5.51206529e-01 1.13458896e+00 -1.73158050e-02 -2.11184751e-02
-3.67714792e-01 -7.65303135e-01 -8.26932967e-01 -1.05023515e+00
-3.68762463e-01 9.69785810e-01 1.22833669e-01 -4.30982560e-01
1.21027678e-01 2.51129210e-01 -1.20619547e+00 -2.57433057e-01
1.10791564e+00 5.63616276e-01 4.62016970e-01 8.32322001e-01
3.45644951e-01 1.34212768e+00 -8.99005413e-01 -1.25451103e-01
2.31780171e-01 -4.56494868e-01 -4.39061135e-01 1.64498851e-01
-4.84639078e-01 -5.62256634e-01 -4.22652304e-01 1.09373987e+00
4.85482126e-01 6.61393940e-01 5.15859187e-01 -1.63216329e+00
-5.99043489e-01 8.24997187e-01 -5.39950252e-01 1.40431240e-01
2.33137067e-02 8.10643494e-01 8.10856402e-01 -2.79865324e-01
3.40327024e-01 1.22476029e+00 2.04086155e-01 4.83548552e-01
-7.73181796e-01 -3.80768210e-01 5.53350687e-01 2.37471402e-01
-7.95330405e-01 2.73219659e-03 5.55969477e-01 -1.11518912e-01
7.87757456e-01 1.47942044e-02 7.91791022e-01 1.01204622e+00
4.98490304e-01 8.14664304e-01 9.73401129e-01 9.53019857e-02
5.26890337e-01 -1.08318239e-01 -3.58820111e-01 7.27757215e-01
2.64145762e-01 3.97768378e-01 -1.26563281e-01 -2.40909606e-01
8.64077091e-01 1.80270508e-01 1.44489765e-01 -7.31565580e-02
-1.41854680e+00 6.82236731e-01 3.96905780e-01 -1.05608076e-01
-8.46199512e-01 3.96769702e-01 2.95060486e-01 5.75351775e-01
1.58078164e-01 5.76637328e-01 -3.77430499e-01 -2.95005888e-01
-2.31060922e-01 5.75450420e-01 6.69939578e-01 7.83251643e-01
6.39930665e-01 5.51067770e-01 -2.53021002e-01 5.76859593e-01
1.76749587e-01 5.76079369e-01 5.94480932e-01 -1.26641548e+00
4.98267412e-01 4.33869779e-01 6.53824687e-01 -9.72334802e-01
-3.61971229e-01 -2.40069687e-01 -8.18284750e-01 5.70320725e-01
3.79068911e-01 -7.34249592e-01 -6.17348552e-01 1.76299560e+00
4.64519799e-01 5.09848416e-01 4.75892752e-01 1.13636661e+00
2.52552032e-01 8.79352152e-01 2.82348003e-02 -5.21753192e-01
8.65309238e-01 -1.40947306e+00 -5.36485374e-01 -3.48487645e-02
2.60764360e-01 -1.94523782e-01 7.74118543e-01 4.07065123e-01
-1.27666330e+00 -3.55411112e-01 -8.11949492e-01 7.31044054e-01
-7.63055608e-02 -7.31977597e-02 3.00726742e-01 6.63870126e-02
-8.69309723e-01 7.84942150e-01 -1.05304623e+00 2.55838223e-02
8.32167268e-02 4.41848248e-01 5.01943678e-02 1.90067783e-01
-1.04896188e+00 8.79208446e-01 3.58808339e-01 -7.92114995e-03
-1.72480273e+00 -4.27292168e-01 -4.56174582e-01 -1.52717205e-02
9.79999840e-01 -5.54639995e-01 1.32978749e+00 -1.04401934e+00
-1.95556319e+00 -1.66160181e-01 4.43316996e-01 -4.42421019e-01
6.26632571e-01 2.65206653e-03 -3.03191215e-01 1.46722808e-01
8.64468440e-02 2.78539360e-01 9.60250497e-01 -1.33068871e+00
-9.04466569e-01 -4.25812192e-02 4.88389879e-01 7.89921224e-01
-9.33787972e-02 -2.46253639e-01 -9.06046033e-02 -4.82324392e-01
-7.09617853e-01 -1.17700851e+00 -6.77717805e-01 -6.39929414e-01
-1.37117893e-01 -3.84165883e-01 5.08365691e-01 -3.99116307e-01
9.19160306e-01 -1.81364405e+00 7.61768460e-01 1.15139022e-01
2.20805094e-01 2.61141866e-01 -3.38917196e-01 7.79783070e-01
5.60240090e-01 -1.00521296e-01 4.91673015e-02 -3.56200963e-01
1.08885556e-01 3.99049878e-01 -1.91165313e-01 4.36129242e-01
-8.59512165e-02 7.69038200e-01 -1.37115812e+00 -3.34392577e-01
1.12895116e-01 -2.79552862e-02 -4.89586502e-01 5.60658336e-01
-6.39390349e-01 7.59346366e-01 -9.08988297e-01 4.90972489e-01
3.91998231e-01 -2.01141343e-01 3.96512717e-01 2.57781327e-01
-2.59978890e-01 -4.71787006e-01 -1.13463962e+00 1.41449583e+00
-3.68575633e-01 3.41473222e-02 2.75327802e-01 -7.83327043e-01
7.15446770e-01 3.39816749e-01 7.76255012e-01 -6.95108771e-01
1.70385644e-01 1.49211824e-01 1.47922590e-01 -4.54938471e-01
9.01698053e-01 2.08407313e-01 -7.95771107e-02 6.66650593e-01
-1.39892131e-01 1.60264358e-01 2.73646444e-01 2.05924764e-01
1.18109965e+00 2.65586772e-03 1.28623143e-01 7.53649250e-02
2.37888545e-01 2.67050445e-01 7.60153890e-01 9.69191492e-01
-4.80763018e-01 1.27478868e-01 4.88477886e-01 -3.24345797e-01
-1.10561097e+00 -7.86408365e-01 9.30505753e-01 1.04592872e+00
6.16939485e-01 -9.47517231e-02 -4.77081507e-01 -8.47367167e-01
2.05311775e-01 6.09460652e-01 -4.26208705e-01 -2.11789474e-01
-7.69089937e-01 -8.69422913e-01 3.22347403e-01 1.70710668e-01
5.07497132e-01 -1.63801253e+00 -8.55323970e-01 6.81054175e-01
-9.86235589e-03 -7.48421073e-01 -7.00744867e-01 -2.04595938e-01
-5.69599867e-01 -1.28404617e+00 -6.97690129e-01 -4.78492618e-01
6.35432541e-01 2.75548190e-01 7.93239653e-01 2.09736764e-01
2.08494559e-01 6.68272436e-01 -5.31888425e-01 -1.70656279e-01
-5.47815800e-01 -9.68444422e-02 3.05865288e-01 1.41553029e-01
-1.98822856e-01 -4.86187339e-01 -6.06191158e-01 3.09258878e-01
-7.68911242e-01 -1.28359348e-01 5.97427249e-01 1.06357455e+00
5.54275572e-01 3.88130546e-01 9.57406163e-01 -4.29465383e-01
1.11510694e+00 -1.01958513e+00 -8.40179920e-01 2.92561650e-01
-5.64030528e-01 6.75765751e-03 1.36866295e+00 -8.62217605e-01
-9.02441323e-01 -4.24714208e-01 3.23502898e-01 -7.14545548e-01
7.82543048e-02 5.59483469e-01 1.12947628e-01 1.48389354e-01
2.01814011e-01 4.37309921e-01 4.03126687e-01 -1.06243566e-01
2.51195669e-01 3.15467685e-01 1.81390405e-01 -9.60836947e-01
5.95640004e-01 1.78132564e-01 -7.76096210e-02 -4.35572356e-01
-3.19969088e-01 9.31540728e-02 1.83150262e-01 -6.39612854e-01
6.71308160e-01 -7.38668919e-01 -1.36785793e+00 9.22404349e-01
-9.11534190e-01 -9.38340545e-01 -2.72100896e-01 5.78930676e-01
-8.18406999e-01 1.24267638e-01 -8.00756037e-01 -9.94462192e-01
-1.31729275e-01 -1.42896795e+00 4.63244379e-01 6.18486345e-01
5.59962153e-01 -1.06525588e+00 5.53596854e-01 -4.95045967e-02
3.29023004e-01 3.07881385e-01 5.37408531e-01 -5.86084783e-01
-7.40739763e-01 2.36420214e-01 2.44264960e-01 7.33559728e-02
1.03665125e-02 -7.29536265e-02 -2.40111515e-01 -8.07478726e-01
-1.94807723e-01 -6.63285613e-01 3.16915244e-01 3.82129371e-01
7.34939992e-01 -7.59394169e-01 -1.75178707e-01 1.46243155e-01
1.54434907e+00 5.97869694e-01 2.92749941e-01 4.72324699e-01
7.27647185e-01 3.47224653e-01 9.72197592e-01 8.52705061e-01
8.74316573e-01 4.13803905e-01 9.11432922e-01 3.44097137e-01
2.06184268e-01 -3.01096052e-01 9.10952449e-01 9.79749084e-01
-3.93834978e-01 -4.59136605e-01 -5.37134290e-01 5.24270952e-01
-2.50546479e+00 -1.11038840e+00 3.01345170e-01 2.14630175e+00
6.29328251e-01 -2.23891437e-02 5.43102205e-01 -5.55924714e-01
7.39237070e-01 1.28445268e-01 -1.12041485e+00 -2.35652223e-01
-8.31258371e-02 -4.12238985e-01 5.25212467e-01 4.17919159e-01
-7.38282740e-01 9.56032574e-01 4.97097397e+00 8.62071812e-01
-8.31499696e-01 2.22073779e-01 5.34044862e-01 -2.64756352e-01
-4.51351665e-02 -1.04974695e-01 -4.59092677e-01 8.81559968e-01
9.72367048e-01 -3.72832239e-01 1.22788632e+00 7.26814985e-01
5.83868027e-01 -2.09598035e-01 -7.36577928e-01 7.28029490e-01
-2.34051794e-01 -1.34527957e+00 -1.70068294e-01 1.88267902e-01
9.42372322e-01 2.36589581e-01 1.72693968e-01 6.37196839e-01
1.17102253e+00 -8.51652741e-01 9.06618357e-01 7.69988239e-01
3.16746056e-01 -1.09414184e+00 5.79117656e-01 6.95392489e-01
-1.20577300e+00 -5.01080036e-01 -3.50656301e-01 -1.60027713e-01
1.05466083e-01 -7.71947503e-02 -5.35482466e-01 9.10383880e-01
4.34823185e-01 6.79645240e-01 -1.23015173e-01 8.71202707e-01
8.63203034e-02 3.84681076e-01 -1.06103964e-01 -3.76401931e-01
6.22700632e-01 -6.27246857e-01 1.00263500e+00 3.97101313e-01
3.43615830e-01 2.50692487e-01 8.75522435e-01 6.11120105e-01
1.43958822e-01 -2.04306677e-01 -5.05784214e-01 -1.51183292e-01
7.17268050e-01 1.32517600e+00 -5.59778988e-01 -3.63466293e-01
-2.68341452e-01 7.39799798e-01 7.34890401e-01 5.53329587e-01
-1.09529412e+00 -1.46329120e-01 8.61468852e-01 -4.76330996e-01
2.29212239e-01 -4.91007537e-01 3.86715293e-01 -1.18043077e+00
-2.00152650e-01 -1.10772824e+00 3.05236161e-01 -4.86692846e-01
-1.41947341e+00 5.46182394e-01 -2.96250224e-01 -1.42769921e+00
-5.23048460e-01 -9.90898013e-02 -9.13790226e-01 4.18241978e-01
-1.51418316e+00 -9.26038623e-01 -8.67017955e-02 6.95085764e-01
5.95617473e-01 -6.85913205e-01 5.38690567e-01 1.51094809e-01
-9.93827343e-01 3.09039861e-01 5.65615237e-01 -9.24942121e-02
4.49275821e-01 -1.32151580e+00 -1.13756629e-02 6.18863106e-01
-1.73638538e-01 2.42756307e-02 6.84536040e-01 -8.10708702e-01
-1.90507019e+00 -1.33550537e+00 -2.45023653e-01 -6.29589632e-02
8.77631903e-01 2.88356692e-01 -6.25633657e-01 6.60535157e-01
4.71412927e-01 -1.10118777e-01 6.72432128e-03 -4.48489279e-01
4.64251459e-01 -1.08170509e-01 -1.03423154e+00 8.88430893e-01
7.11026311e-01 1.44221514e-01 -2.12740734e-01 3.44879240e-01
8.78657937e-01 -5.16463876e-01 -7.71034241e-01 -6.18206747e-02
8.51964653e-02 -8.20260882e-01 6.75870717e-01 -8.65445316e-01
2.97759026e-01 -4.22889352e-01 9.06475559e-02 -2.15226388e+00
-2.70297468e-01 -9.54903424e-01 -5.05266011e-01 9.00712609e-01
5.96248545e-02 -8.73404384e-01 5.92002273e-01 4.00833756e-01
-3.99874151e-01 -7.69955516e-01 -8.07714343e-01 -9.12773252e-01
5.86629026e-02 8.37529004e-02 8.50841165e-01 6.92115605e-01
-3.55436048e-03 -1.51533252e-02 -8.16597462e-01 4.19069767e-01
9.32247818e-01 1.21588722e-01 9.17003036e-01 -5.71743369e-01
-6.13894761e-01 -4.45096791e-01 1.76887348e-01 -9.03870642e-01
3.89160365e-01 -6.98959470e-01 4.93883379e-02 -1.36268747e+00
4.01980206e-02 -8.75537395e-01 -4.01793242e-01 2.93610424e-01
-2.84976214e-01 -3.55053157e-01 6.06528640e-01 4.71463948e-01
-1.13047564e+00 9.79420483e-01 1.63059115e+00 -2.34690398e-01
-5.09989381e-01 5.85025847e-02 -3.90252829e-01 4.67484444e-01
9.99516487e-01 -5.08653045e-01 -5.35938084e-01 -5.02649486e-01
4.87859622e-02 7.42152214e-01 4.41184282e-01 -9.60738778e-01
5.23951828e-01 -8.81496608e-01 -1.51495412e-01 -2.74610698e-01
3.43802303e-01 -7.32754052e-01 2.57269591e-01 6.25658870e-01
-1.86591506e-01 5.86152613e-01 -1.54526561e-01 1.00584114e+00
-8.88221785e-02 -1.34346023e-01 5.53139627e-01 -5.01736641e-01
-6.60615325e-01 6.70504808e-01 -4.82141495e-01 3.42157662e-01
1.39476669e+00 2.92536169e-01 -5.85977197e-01 -5.38971722e-01
-7.13810384e-01 1.07589889e+00 4.22434926e-01 2.30491608e-01
6.67502224e-01 -1.36797214e+00 -8.01789522e-01 -1.34571686e-01
-3.95651996e-01 -2.37557795e-02 5.03155828e-01 8.86825919e-01
-1.96427926e-01 -1.74831256e-01 -4.34242040e-01 -2.36754879e-01
-6.68599427e-01 5.52709103e-01 4.97907013e-01 -6.78250730e-01
-6.55021369e-01 1.87433138e-01 -2.74958342e-01 -3.19067359e-01
7.16212988e-02 -2.55647283e-02 -3.79660249e-01 2.68905796e-02
4.21912938e-01 8.44244599e-01 -4.15844679e-01 -4.59485263e-01
-4.62692231e-03 1.11932665e-01 -7.95109943e-02 -3.34863216e-01
1.45241177e+00 -5.37990071e-02 -3.65947932e-02 4.17936265e-01
5.81060231e-01 -2.00124845e-01 -1.91005647e+00 -4.85182181e-02
-3.63453209e-01 -3.38625580e-01 -2.38135964e-01 -7.62624562e-01
-1.26192653e+00 1.13105886e-01 6.34345114e-02 4.86174852e-01
8.20880949e-01 -2.86104470e-01 8.84212434e-01 3.74507904e-01
7.87739098e-01 -1.43002760e+00 5.51485240e-01 6.92043543e-01
8.29753399e-01 -1.27651834e+00 -2.33298063e-01 5.14691174e-01
-1.37869298e+00 9.60859656e-01 9.61388230e-01 -6.59321725e-01
3.24026346e-01 1.40236700e-02 -1.57814354e-01 -6.77049831e-02
-1.16436398e+00 -2.23743722e-01 -3.04289281e-01 5.59561074e-01
-6.11373246e-01 4.61501926e-01 -2.79985443e-02 6.49188399e-01
2.59539843e-01 -1.86240017e-01 8.98216486e-01 8.41042697e-01
-4.28566337e-01 -9.08141673e-01 -3.42742532e-01 3.86358231e-01
-4.83018495e-02 4.01602864e-01 8.96306187e-02 7.50994980e-01
-1.46314412e-01 1.10373092e+00 7.43935704e-02 -3.71251732e-01
2.49706179e-01 -5.48884630e-01 3.59732121e-01 -3.64288867e-01
-9.87099826e-01 7.05918893e-02 6.81285327e-03 -5.34507394e-01
-3.46073836e-01 -6.58306241e-01 -1.61019611e+00 -5.67351282e-01
2.03532400e-03 3.44084024e-01 2.87906051e-01 8.76480758e-01
5.75026274e-01 8.08354557e-01 9.98246789e-01 -9.79209900e-01
-1.05496287e+00 -7.98142910e-01 -5.75068712e-01 3.20256859e-01
5.39915442e-01 -9.12334144e-01 -2.99853027e-01 -5.01303077e-01] | [3.75110125541687, 2.028992176055908] |
c53db476-f3e2-4719-a0de-8e1a590e5ac0 | analyzing-codebert-s-performance-on-natural | null | null | https://openreview.net/forum?id=canvJzQNs09 | https://openreview.net/pdf?id=canvJzQNs09 | Analyzing CodeBERT's Performance on Natural Language Code Search | Large language models such as CodeBERT perform very well on tasks such as natural language code search. We show that this is most likely due to the high token overlap and similarity between the queries and the code in datasets obtained from large codebases, rather than any deeper understanding of the syntax or semantics of the query or code. | ['Anonymous'] | 2022-01-16 | null | null | null | acl-arr-january-2022-1 | ['code-search', 'code-search'] | ['computer-code', 'computer-vision'] | [-5.84860623e-01 6.63633496e-02 -7.72700250e-01 -3.54858845e-01
-6.64317012e-01 -6.93155587e-01 5.51451445e-01 7.78999329e-01
-3.22088331e-01 1.60247520e-01 5.60444772e-01 -6.66981816e-01
1.18064567e-01 -7.22396672e-01 -7.31261075e-01 2.28473186e-01
-3.85427624e-01 1.88580707e-01 5.33463955e-01 -2.70819634e-01
3.57160985e-01 -1.95110157e-01 -1.38401794e+00 5.91295242e-01
1.03999138e+00 2.91167647e-01 4.97639090e-01 6.69252455e-01
-7.36739457e-01 1.24147701e+00 -4.88243222e-01 -7.13254452e-01
-2.04069600e-01 -3.09611410e-01 -1.09288538e+00 -5.41087806e-01
2.41349056e-01 -8.14327672e-02 -2.49513656e-01 1.28309488e+00
-1.65933147e-01 -1.55663714e-01 2.21972793e-01 -1.14761209e+00
-5.12901604e-01 8.25644493e-01 -5.92343330e-01 1.17736436e-01
6.99329078e-01 -1.48102090e-01 1.41337967e+00 -7.52387702e-01
1.03206599e+00 1.20422614e+00 8.09459925e-01 3.40801090e-01
-1.17960358e+00 -2.51214296e-01 -3.28761369e-01 -4.39657718e-02
-1.62963963e+00 -4.16505605e-01 1.75001025e-01 -1.04207957e+00
1.59511495e+00 -5.37726283e-02 2.92571694e-01 4.58509088e-01
5.60414791e-01 5.40489197e-01 2.32098028e-01 -7.09029198e-01
-1.07310861e-01 4.53687727e-01 3.40332896e-01 1.13774288e+00
3.61680508e-01 -8.84554833e-02 -1.65919617e-01 -9.67622459e-01
1.85872808e-01 -7.63261765e-02 -1.58385292e-01 -5.23480117e-01
-1.14862275e+00 1.10559404e+00 4.27952379e-01 5.72728634e-01
-1.81772217e-01 6.02670789e-01 9.44990754e-01 5.33900678e-01
-7.83187300e-02 7.54444838e-01 -7.26357043e-01 -5.88692307e-01
-6.47190452e-01 1.78919360e-01 1.24830306e+00 1.11062491e+00
1.04405332e+00 -2.01333344e-01 3.70014876e-01 7.83749163e-01
5.67518651e-01 4.31750029e-01 7.29108274e-01 -8.49001646e-01
5.80389500e-01 9.14614677e-01 -1.00451652e-02 -9.55877483e-01
7.69086555e-02 -7.81504065e-02 5.23867488e-01 4.27891798e-02
3.40014279e-01 1.81565240e-01 -2.72018969e-01 1.69792497e+00
-2.48554349e-01 -4.32480514e-01 4.10375930e-02 3.38775873e-01
7.65860379e-01 3.57019037e-01 2.81121790e-01 2.04718143e-01
1.34012735e+00 -6.20899856e-01 -2.50952303e-01 -7.21144378e-01
1.40234733e+00 -9.93839562e-01 1.04447556e+00 -3.68945122e-01
-8.00680578e-01 -1.64395273e-01 -8.14450145e-01 -2.39393324e-01
-3.05635959e-01 1.07735470e-01 8.36409628e-01 5.70125043e-01
-8.53387296e-01 1.33122802e-02 -8.69551599e-01 -6.91035271e-01
5.19208983e-02 -1.69920996e-02 -2.41965368e-01 -2.52444208e-01
-8.90684128e-01 7.28321254e-01 5.06747127e-01 -6.67608380e-01
-6.70449555e-01 -3.99973810e-01 -1.14912593e+00 1.87984332e-01
3.11667621e-01 -2.07733825e-01 1.44947898e+00 -1.08868694e+00
-4.93374616e-01 1.19397283e+00 -5.57542980e-01 -4.29450959e-01
-1.72936648e-01 -2.38478169e-01 -2.92437643e-01 -3.02191406e-01
4.48370904e-01 7.53027424e-02 2.89126396e-01 -1.09769082e+00
-5.01727819e-01 -2.46104732e-01 5.01314282e-01 -3.61788630e-01
3.19114588e-02 7.03410685e-01 -4.27690446e-01 -3.28775078e-01
-1.78323120e-01 -9.04903948e-01 1.03052348e-01 -4.07622103e-03
5.19451536e-02 -5.13165951e-01 2.89731324e-01 -7.06201613e-01
1.50136983e+00 -2.62285185e+00 -1.62551522e-01 7.98330382e-02
3.34646672e-01 9.74111483e-02 -2.27225780e-01 1.03560138e+00
-2.00943023e-01 4.58035469e-01 8.11651275e-02 4.23838139e-01
9.60175693e-02 3.00154716e-01 -7.31375992e-01 5.84513545e-01
-7.11726025e-02 1.19297445e+00 -1.10821509e+00 -6.73436046e-01
-4.59873885e-01 -1.94643527e-01 -7.50113845e-01 -1.71171874e-02
-7.07530677e-01 -6.45332098e-01 -6.89354897e-01 5.02759993e-01
3.12426547e-03 -4.11272883e-01 -3.42373922e-02 4.19095784e-01
-1.93332076e-01 1.03201163e+00 -5.21387458e-01 2.06457567e+00
-8.13300431e-01 8.61583173e-01 1.39841422e-01 -6.12941325e-01
4.47197527e-01 5.17429054e-01 2.16550082e-01 -7.29075015e-01
-4.80036110e-01 4.61215049e-01 1.89645961e-01 -9.53899622e-01
3.57326090e-01 -9.13949534e-02 -3.18805099e-01 9.67877507e-01
-7.90436715e-02 -1.58974394e-01 3.26886803e-01 6.82173789e-01
1.29184365e+00 1.16059691e-01 4.90809649e-01 -5.21803558e-01
3.30471039e-01 6.18400753e-01 3.66748422e-01 8.39617729e-01
1.56888708e-01 1.49886357e-02 8.88240397e-01 -2.92524904e-01
-9.29385483e-01 -8.78090143e-01 -2.26823866e-01 1.43972063e+00
2.72651240e-02 -1.47962749e+00 -4.26054507e-01 -5.58565140e-01
1.95632547e-01 5.36022961e-01 -4.53985453e-01 -2.86250889e-01
-5.91990650e-01 -1.97245151e-01 9.28545058e-01 5.05030870e-01
-5.88483289e-02 -7.56352186e-01 -6.49142683e-01 2.52447188e-01
-1.16132677e-01 -8.65963161e-01 -6.57481730e-01 4.67778258e-02
-7.48012006e-01 -1.33055758e+00 3.04614902e-01 -8.74194145e-01
4.87257421e-01 1.87505290e-01 1.71322060e+00 8.71831179e-01
-3.90910625e-01 3.87751549e-01 -4.18109953e-01 -1.23436540e-01
-1.02106655e+00 1.80790827e-01 -6.59296393e-01 -9.62177336e-01
7.32318163e-01 -2.51772523e-01 1.05027407e-02 2.54671872e-01
-9.35833931e-01 -5.41988075e-01 6.20376170e-01 7.30311036e-01
-4.29369926e-01 -5.90244681e-02 2.05027223e-01 -9.31439459e-01
8.05268705e-01 -1.02912629e+00 -8.68554533e-01 5.15442908e-01
-6.28147066e-01 5.63474834e-01 4.64843094e-01 -2.30589136e-01
-7.70041347e-01 -7.97880441e-02 -2.38089919e-01 1.22465320e-01
-7.28058442e-02 1.07159662e+00 3.71535569e-01 3.49600837e-02
9.95729566e-01 2.82872468e-01 -1.26180992e-01 -4.75101203e-01
1.92089543e-01 6.20172799e-01 2.80019432e-01 -1.06846356e+00
7.48769641e-01 1.68012798e-01 -6.27677798e-01 -7.76919782e-01
-7.87749067e-02 -6.01588070e-01 -1.70614704e-01 4.19198453e-01
8.09931517e-01 -1.05072272e+00 -5.35401106e-01 -8.93824250e-02
-1.45673120e+00 -3.11948359e-01 -8.03296939e-02 4.65126067e-01
-2.58266181e-01 3.55564058e-01 -9.31173384e-01 -5.57724237e-01
-3.04223131e-02 -1.19625723e+00 9.96539414e-01 -2.89389729e-01
-8.04071665e-01 -1.22368193e+00 4.22811598e-01 -6.05303347e-02
6.72283828e-01 -2.01021925e-01 2.00323343e+00 -7.39705384e-01
-7.76893497e-01 -5.43959439e-01 -2.96958715e-01 -1.65635794e-01
1.26326948e-01 3.45179260e-01 -3.70212168e-01 -3.39363545e-01
-1.31374121e-01 -5.09839058e-01 3.21334392e-01 -3.07254225e-01
4.38623965e-01 -3.69208455e-01 -5.89866340e-01 2.67068744e-01
1.95654929e+00 1.35698915e-01 4.05893326e-01 3.21789056e-01
4.24071312e-01 4.88520950e-01 1.48886755e-01 2.22793713e-01
3.93757015e-01 6.67004585e-01 1.22492708e-01 3.79426092e-01
-1.19060390e-01 -5.41209817e-01 5.65570176e-01 9.43780959e-01
7.40167499e-01 2.74756879e-01 -1.61943614e+00 9.46025193e-01
-1.79302084e+00 -7.84091771e-01 -4.46151167e-01 2.04636431e+00
8.61063600e-01 -2.14939415e-02 -1.34445190e-01 -5.72512031e-01
3.51469368e-01 6.46570325e-02 -2.54416466e-01 -6.61754310e-01
3.63930464e-01 -6.11949265e-02 4.64268863e-01 5.47475815e-01
-2.95052648e-01 6.17431343e-01 7.74127197e+00 6.66199148e-01
-1.10135925e+00 2.13514969e-01 -2.23960444e-01 1.87243298e-01
-6.26188338e-01 4.75750983e-01 -4.01133865e-01 4.61587816e-01
9.87760723e-01 -9.07483935e-01 6.71514630e-01 1.16853857e+00
-1.50680915e-01 -2.99767107e-01 -1.35138965e+00 6.07053638e-01
-9.05314460e-02 -1.15649438e+00 -1.82308108e-01 4.60261814e-02
6.76074922e-01 8.74130666e-01 -6.18704259e-01 5.28277159e-01
8.62686396e-01 -9.82733071e-01 9.22495246e-01 9.35466290e-02
5.24642587e-01 -9.05596018e-02 6.69567645e-01 5.77178955e-01
-1.29532063e+00 -1.90232635e-01 -4.24319535e-01 -6.19868301e-02
-4.36832100e-01 3.71274203e-01 -7.38579154e-01 9.98301432e-02
3.33310455e-01 5.07362008e-01 -9.13802505e-01 1.15911150e+00
-3.06710362e-01 5.24523735e-01 -9.21984017e-02 -2.43985981e-01
2.19088182e-01 2.32975796e-01 4.11808282e-01 1.54114795e+00
6.15046695e-02 -3.25978160e-01 4.56435412e-01 1.23999524e+00
-1.53712481e-01 3.62327546e-01 -8.35080028e-01 -6.11629009e-01
4.02276397e-01 8.43407333e-01 -4.16773826e-01 -4.80435342e-01
-1.38163030e+00 4.53029364e-01 4.74919796e-01 4.01998848e-01
-6.63073897e-01 -7.20041156e-01 7.60698557e-01 1.54970109e-01
2.51459777e-01 -3.80839795e-01 1.13415562e-01 -1.39860785e+00
4.91672218e-01 -1.08046591e+00 3.87910783e-01 -9.33259666e-01
-8.94332230e-01 6.30200922e-01 -8.99599642e-02 -6.38238072e-01
-5.52343726e-01 -3.53812277e-01 -6.82637811e-01 9.48918998e-01
-1.09080410e+00 -4.78832185e-01 2.58297920e-01 4.39077020e-01
2.60999829e-01 -1.06685556e-01 9.13728654e-01 4.37338799e-01
3.04322951e-02 2.32765973e-01 1.11336820e-01 6.16933048e-01
5.91687441e-01 -1.12247598e+00 7.86443949e-01 8.74444544e-01
3.36969107e-01 1.50764215e+00 8.15616667e-01 -5.34289956e-01
-1.63987911e+00 -5.72550058e-01 1.08920300e+00 -8.83393228e-01
1.24329901e+00 -5.41212142e-01 -1.20447576e+00 8.51889133e-01
4.61256802e-01 5.30719059e-03 5.80367386e-01 3.28322500e-01
-9.56400156e-01 2.00859338e-01 -8.69942486e-01 1.02846764e-01
7.34215736e-01 -1.36933494e+00 -8.20028305e-01 3.55860293e-01
8.61063957e-01 -1.44338384e-01 -8.58817697e-01 -5.36945872e-02
4.74730939e-01 -6.33736968e-01 5.94416857e-01 -1.07024968e+00
7.71603346e-01 -2.02823147e-01 -4.36373502e-01 -9.84350622e-01
1.15327053e-02 -3.47029060e-01 2.47809485e-01 1.09711504e+00
7.40658641e-01 -4.92121905e-01 5.31406581e-01 9.35255229e-01
1.61477253e-01 -4.27978575e-01 -5.70079982e-01 -8.34376037e-01
2.04745159e-01 -4.99788314e-01 5.15651286e-01 1.14618564e+00
7.98767447e-01 3.88899326e-01 4.55645293e-01 -2.09763199e-01
1.92430586e-01 6.60281837e-01 6.62724495e-01 -1.12353611e+00
-5.07022440e-01 -3.89436066e-01 -5.19680440e-01 -1.04007423e+00
6.89867735e-01 -1.13896060e+00 2.73304701e-01 -1.35733938e+00
5.64645410e-01 -6.15825295e-01 3.01719278e-01 5.86162865e-01
3.07819247e-02 -3.25383514e-01 -2.16915011e-01 4.77659136e-01
-7.63947785e-01 2.68883616e-01 4.16703820e-01 -2.15105782e-03
1.05804615e-01 -2.71576136e-01 -6.74222946e-01 6.08418941e-01
4.03930187e-01 -1.06360698e+00 -3.95050019e-01 -1.04020059e+00
1.13280749e+00 5.06173313e-01 1.38488099e-01 -5.90412021e-01
3.68432134e-01 -2.00024620e-01 -6.03539169e-01 3.20558608e-01
-3.24067801e-01 -8.73021364e-01 -3.74445058e-02 8.92213643e-01
-6.56353652e-01 5.22998869e-01 4.79166508e-01 4.98472363e-01
-4.84171391e-01 -9.57960188e-01 5.00863194e-01 -4.91889209e-01
-9.99764979e-01 -2.40318954e-01 -7.27191210e-01 6.16177559e-01
7.56530583e-01 9.61425975e-02 -7.46583521e-01 -2.63388276e-01
-1.18012369e-01 1.17558509e-01 1.13318360e+00 8.82208645e-01
-4.06653285e-02 -1.08736038e+00 -4.13771898e-01 1.51449904e-01
8.22386086e-01 -6.80729091e-01 -7.89654493e-01 6.70155287e-01
-6.35279536e-01 7.48361707e-01 2.44624525e-01 -3.14408481e-01
-1.09809399e+00 6.92967772e-01 4.25118446e-01 3.38933319e-02
-2.73156166e-01 4.32896674e-01 1.60358399e-01 -5.63761175e-01
-5.13323806e-02 -4.30631787e-01 5.34061372e-01 -5.98609447e-01
6.09113216e-01 -7.85364956e-02 1.02753993e-02 -6.75852060e-01
-7.68560827e-01 4.37801003e-01 -1.19601591e-02 2.32290521e-01
1.21598542e+00 1.14040747e-01 -9.22739685e-01 4.63167518e-01
1.52560544e+00 6.21307015e-01 -4.30316091e-01 -4.59161043e-01
6.25553608e-01 -7.20103025e-01 -4.14565295e-01 -4.70206529e-01
-7.57002115e-01 5.50391138e-01 1.71045586e-01 3.98171097e-01
2.11226985e-01 6.46977365e-01 6.64044499e-01 8.80191565e-01
9.08224285e-01 -6.51629448e-01 -1.68913141e-01 6.70086741e-01
4.83262181e-01 -1.07017946e+00 -4.58456539e-02 -1.96449757e-01
-5.37117384e-02 1.05219936e+00 4.29832399e-01 -6.84567392e-02
6.03395581e-01 5.34104526e-01 1.14198796e-01 -5.55202305e-01
-1.18007171e+00 -3.59177515e-02 -1.32543325e-01 3.29715788e-01
1.09606540e+00 -3.45240861e-01 -5.36843598e-01 2.11497068e-01
-6.96767541e-03 1.68426856e-01 6.02736533e-01 1.22996342e+00
-6.84575856e-01 -1.41689396e+00 -7.48198554e-02 5.08730173e-01
-7.72914886e-01 -4.18880224e-01 -5.09838760e-01 6.42578721e-01
-4.25346866e-02 9.73767757e-01 -1.52931809e-01 -2.24628747e-02
8.86737332e-02 3.12749118e-01 2.61837661e-01 -1.32491219e+00
-7.43337393e-01 -6.15700960e-01 2.07803100e-01 -7.34315217e-01
7.55119920e-02 -4.76868540e-01 -1.58086061e+00 -3.54275614e-01
-5.60165644e-01 8.27118516e-01 7.93875039e-01 8.99788380e-01
4.21684831e-01 -1.51302025e-01 -1.03808921e-02 3.53982627e-01
-4.66361731e-01 -5.51582813e-01 -5.45161307e-01 7.56807625e-01
2.59544581e-01 -2.31684446e-01 -3.36307913e-01 2.30945960e-01] | [7.545039176940918, 8.066950798034668] |
6022525e-15ce-4e15-ad69-bfcb28de325c | deepsleep-2-0-automated-sleep-arousal | null | null | https://www.mdpi.com/2673-2688/3/1/10 | https://www.mdpi.com/2673-2688/3/1/10/pdf | DeepSleep 2.0: Automated Sleep Arousal Segmentation via Deep Learning | DeepSleep 2.0 is a compact version of DeepSleep, a state-of-the-art, U-Net-inspired, fully convolutional deep neural network, which achieved the highest unofficial score in the 2018 PhysioNet Computing Challenge. The proposed network architecture has a compact encoder/decoder structure containing only 740,551 trainable parameters. The input to the network is a full-length multichannel polysomnographic recording signal. The network has been designed and optimized to efficiently predict nonapnea sleep arousals on held-out test data at a 5 ms resolution level, while not compromising the prediction accuracy. When compared to DeepSleep, the obtained experimental results in terms of gross area under the precision-recall curve (AUPRC) and gross area under the receiver operating characteristic curve (AUROC) suggest a lightweight architecture, which can achieve similar prediction performance at a lower computational cost, is realizable. | ['Robert Fonod'] | 2022-03-01 | null | null | null | ai-2022-3 | ['sleep-quality-prediction', 'sleep-micro-event-detection', 'sleep-arousal-detection'] | ['medical', 'medical', 'medical'] | [-2.11590491e-02 1.51389912e-01 -2.74890624e-02 -5.06634414e-01
-5.73738277e-01 -6.88731000e-02 -2.56157704e-02 4.22918856e-01
-7.85276651e-01 1.00491893e+00 2.17720658e-01 -8.27913359e-02
-1.91943765e-01 -3.12927634e-01 -6.25769317e-01 -4.21637505e-01
-4.70292002e-01 -4.74795327e-02 -3.16892527e-02 2.88324207e-02
-3.02869063e-02 4.17523645e-02 -1.49552310e+00 2.12687895e-01
6.98931336e-01 1.69929361e+00 -1.54992700e-01 7.03023076e-01
5.87200224e-01 6.19023919e-01 -3.83179456e-01 -5.43440342e-01
-1.30667146e-02 -2.51443624e-01 -5.12759924e-01 -8.59431624e-01
2.77918309e-01 -1.81566626e-01 -6.65148377e-01 7.88450181e-01
1.14910364e+00 -2.83559319e-02 1.77969903e-01 -8.09376061e-01
-2.85501301e-01 4.27480519e-01 1.66592553e-01 8.83224666e-01
2.07357630e-01 -1.09571621e-01 1.01635504e+00 -6.47026360e-01
2.15837225e-01 3.74322563e-01 1.02367413e+00 5.11508286e-01
-1.06365705e+00 -9.57087517e-01 -7.29417562e-01 3.97505999e-01
-1.56633949e+00 -3.77937168e-01 3.50956053e-01 -8.35258663e-02
1.03349125e+00 3.78337473e-01 8.35238099e-01 1.37604392e+00
1.04861391e+00 3.25957865e-01 9.68046784e-01 1.62428066e-01
3.85693341e-01 6.41417503e-02 1.63211048e-01 4.91626859e-01
1.90567568e-01 1.05150342e-01 -9.28987145e-01 -1.36218667e-01
2.80666798e-01 8.05040821e-02 -3.36225986e-01 2.82529145e-01
-7.80846834e-01 4.83088076e-01 5.12744427e-01 3.33607852e-01
-6.20149851e-01 -5.69806732e-02 7.92386413e-01 2.79606253e-01
3.13478470e-01 4.30795103e-01 -6.04660809e-01 -8.98128390e-01
-1.03441596e+00 4.50377576e-02 8.06691170e-01 7.21664369e-01
2.01435521e-01 -4.14257459e-02 -3.32826912e-01 6.12000406e-01
-8.25401321e-02 3.21348250e-01 9.10246074e-01 -6.50206804e-01
2.69293547e-01 5.70348561e-01 -5.57371452e-02 -7.86810279e-01
-1.00456762e+00 -7.82107234e-01 -9.26285446e-01 -4.49599445e-01
-1.98052436e-01 -4.45479989e-01 -2.73114324e-01 1.50222874e+00
-3.07334363e-01 3.61931384e-01 1.09657466e-01 9.58033621e-01
1.00010765e+00 3.57095778e-01 8.00011903e-02 -3.38392198e-01
1.40701485e+00 -6.97287023e-01 -8.38239312e-01 -2.52577215e-01
3.47755879e-01 -2.72082567e-01 7.07107008e-01 6.33535147e-01
-1.08207190e+00 -6.15451694e-01 -1.50795877e+00 -8.26118439e-02
-2.61046231e-01 3.72459680e-01 5.17758667e-01 7.57007241e-01
-1.01558638e+00 1.11500287e+00 -1.04320097e+00 -2.52705395e-01
5.47776818e-01 7.61572897e-01 -5.20547152e-01 2.58703142e-01
-1.40896654e+00 8.07132661e-01 5.98122358e-01 2.42579728e-01
-7.54194856e-01 -6.71902120e-01 -5.69959760e-01 5.70528626e-01
-4.64317761e-02 -3.76830041e-01 1.02665019e+00 -5.43897748e-01
-1.47195184e+00 5.78229070e-01 1.56052351e-01 -1.13481653e+00
8.69533867e-02 -5.11794388e-01 -9.66886401e-01 2.12846041e-01
-3.01428974e-01 4.14545894e-01 3.49008232e-01 -1.07144721e-01
-3.53699237e-01 -5.56170225e-01 -3.91490966e-01 1.23205990e-01
-6.76094294e-01 -4.76457216e-02 -3.66031617e-01 -3.91925305e-01
-2.99562663e-01 -8.94056797e-01 -8.91485885e-02 -2.19637766e-01
4.16489737e-03 -2.88269445e-02 4.85453218e-01 -5.67588866e-01
1.59696555e+00 -2.24250817e+00 -8.35110992e-02 -1.05829136e-02
3.51179689e-01 6.67728066e-01 5.93754277e-02 5.34322143e-01
-2.33537436e-01 -3.20882708e-01 -1.20809853e-01 -2.64748633e-01
-1.55512184e-01 4.69264686e-02 2.78548837e-01 5.43041646e-01
-2.58334607e-01 8.13342690e-01 -8.33313048e-01 -2.64618155e-02
3.06336194e-01 4.73129243e-01 -5.76780438e-01 4.82653379e-01
3.93710464e-01 1.96281195e-01 -2.10559025e-01 2.69750714e-01
4.73815441e-01 -1.65555328e-01 4.92486842e-02 -7.72344619e-02
-1.35558575e-01 4.59024221e-01 -5.97231269e-01 2.08394289e+00
-4.10554618e-01 8.91918600e-01 -1.91533819e-01 -7.19679892e-01
1.26122129e+00 4.43136066e-01 7.37248123e-01 -1.05084801e+00
5.01327693e-01 1.42902374e-01 1.83575436e-01 -6.03864908e-01
3.17775518e-01 -1.25877157e-01 -3.61843556e-02 1.04733378e-01
5.05016565e-01 6.73211813e-01 -2.10410659e-03 -7.57422596e-02
1.34722519e+00 -2.44440198e-01 5.42439580e-01 -5.80642462e-01
5.30060351e-01 -7.91095436e-01 5.93391716e-01 6.98412657e-01
-4.19654608e-01 6.31445289e-01 5.21240652e-01 -7.02053845e-01
-8.06817293e-01 -9.92356420e-01 -5.89059234e-01 6.95012033e-01
-1.71903819e-02 -1.16407990e+00 -8.49001229e-01 -1.46192655e-01
-3.46332431e-01 4.50788707e-01 -6.33915484e-01 -6.39740050e-01
-8.35149437e-02 -9.05345976e-01 9.04134691e-01 7.35762119e-01
5.73156953e-01 -1.07003546e+00 -1.10641420e+00 5.07677972e-01
-3.24822706e-03 -1.18502474e+00 -1.79548636e-01 5.25362074e-01
-7.13417768e-01 -1.10668373e+00 -4.59992439e-01 -3.88819188e-01
-8.42432305e-02 -6.07320964e-01 9.91779864e-01 -4.07624066e-01
-3.94433886e-01 -4.24144380e-02 -2.97499061e-01 -4.97080684e-01
1.99150249e-01 1.85439467e-01 4.19155777e-01 1.37004167e-01
9.56002951e-01 -9.38954949e-01 -1.15273809e+00 -3.82687040e-02
-4.81949180e-01 -2.36217931e-01 6.93200827e-01 7.60382652e-01
6.65290833e-01 -2.68887877e-01 8.04217994e-01 -4.99718368e-01
7.21267760e-01 -4.94894594e-01 -2.48930037e-01 -7.65895024e-02
-8.51706445e-01 -2.83516318e-01 9.71812129e-01 2.67133806e-02
-4.02990907e-01 -1.17941275e-01 -7.57823169e-01 -4.77117240e-01
-7.09340200e-02 3.89937758e-01 2.66305029e-01 1.00746319e-01
7.66701758e-01 5.07183492e-01 -1.08355604e-01 -5.01916289e-01
-1.63312033e-01 1.09054530e+00 7.54881859e-01 -3.34164426e-02
-2.82404155e-01 1.01479128e-01 -5.16214184e-02 -9.58487272e-01
-8.15112472e-01 -6.51894629e-01 -4.74391848e-01 -2.91937236e-02
1.05880105e+00 -1.15446353e+00 -1.19800770e+00 2.84721136e-01
-6.81641519e-01 -7.97262117e-02 7.29228929e-02 7.08481252e-01
-5.73680460e-01 -4.93903123e-02 -5.49691021e-01 -6.09751463e-01
-1.25294161e+00 -8.81592989e-01 7.75032282e-01 2.83090800e-01
-3.86011332e-01 -5.77632606e-01 2.44769186e-01 3.97017300e-02
4.79140103e-01 3.10714126e-01 4.34652179e-01 -1.07930696e+00
8.90117809e-02 -5.62326193e-01 -2.32603014e-01 7.12783813e-01
-8.48363042e-02 -5.44307709e-01 -1.11539996e+00 -4.03996170e-01
1.61168665e-01 -4.28825051e-01 5.60636997e-01 3.64816189e-01
1.76549947e+00 -3.10260504e-01 3.59789375e-03 8.08929980e-01
1.36031914e+00 3.52058023e-01 9.73833919e-01 1.03327639e-01
1.87988117e-01 -2.72117645e-01 1.72238886e-01 9.45186555e-01
2.60626674e-01 4.65231359e-01 4.08541113e-01 2.47423723e-01
3.59889150e-01 -6.27669692e-02 1.01268180e-01 1.02269125e+00
-3.71753834e-02 -1.42805889e-01 -8.80108654e-01 3.66587698e-01
-1.64086616e+00 -7.55794704e-01 6.56886250e-02 2.28226185e+00
5.06272316e-01 3.37458789e-01 7.28169456e-02 1.67096600e-01
3.21535945e-01 1.23436764e-01 -5.03221571e-01 -8.27710629e-01
2.44150341e-01 7.88099229e-01 4.78016347e-01 -1.14210628e-01
-9.34575498e-01 3.72904003e-01 6.20154762e+00 6.04768395e-01
-1.34257591e+00 1.94035545e-01 5.07643878e-01 -5.35160124e-01
5.57252288e-01 -6.67037964e-01 -5.44148386e-01 9.89407241e-01
2.16580987e+00 -2.27685124e-01 4.75807905e-01 9.88284528e-01
2.24562809e-01 -7.39165815e-03 -1.14776647e+00 1.38224590e+00
2.06493974e-01 -1.36199987e+00 -5.69719791e-01 -1.66782856e-01
2.93130010e-01 6.26280546e-01 -6.65005436e-03 5.52778006e-01
-7.84697413e-01 -1.18582010e+00 2.60377109e-01 6.13543749e-01
1.16567576e+00 -1.09403884e+00 1.21482348e+00 1.72233820e-01
-8.84948075e-01 -3.26334774e-01 -4.30153757e-01 -2.49281108e-01
-3.00352782e-01 2.90568799e-01 -5.18104017e-01 4.58715916e-01
1.08297908e+00 8.36125076e-01 -5.79459488e-01 1.11355376e+00
1.90692112e-01 6.89816356e-01 -1.79135785e-01 -3.02288294e-01
2.82101125e-01 2.15295613e-01 2.35766560e-01 1.26965284e+00
3.69362593e-01 2.56847441e-01 -3.24190855e-01 5.92386365e-01
-3.85911524e-01 1.90863013e-01 -3.52780372e-01 -1.05560273e-01
3.26106995e-01 1.28369248e+00 -2.21876830e-01 -1.53850049e-01
-4.09086406e-01 9.28711116e-01 2.71176964e-01 -1.03540748e-01
-9.00131226e-01 -8.52042496e-01 5.71447134e-01 -7.15719908e-02
1.67378023e-01 2.00183675e-01 -5.21615148e-01 -1.04952276e+00
6.87975250e-03 -7.56318092e-01 4.12035078e-01 -6.82320952e-01
-7.77405262e-01 9.77413833e-01 -4.17562664e-01 -1.25570440e+00
-1.26381975e-03 -6.34184539e-01 -5.88445425e-01 7.25993693e-01
-1.02258110e+00 -4.40259546e-01 -3.87673795e-01 6.26847982e-01
2.45409727e-01 -2.27942601e-01 1.24404991e+00 7.08794057e-01
-6.61251605e-01 8.54383230e-01 3.73873353e-01 -3.73985432e-02
4.64933574e-01 -1.10534883e+00 4.71250117e-02 6.11339569e-01
-2.90868700e-01 7.64039338e-01 5.87702811e-01 -2.18621701e-01
-1.27592504e+00 -1.17189336e+00 9.40047741e-01 -3.31921339e-01
6.24284208e-01 -5.44133782e-01 -6.04224205e-01 6.02571607e-01
1.50057495e-01 7.64387846e-02 1.16247129e+00 2.69272745e-01
2.16186464e-01 -5.27527571e-01 -1.08930016e+00 3.62122893e-01
5.08045137e-01 -5.41909933e-01 -5.68719745e-01 8.04107711e-02
4.22381461e-01 -6.42096043e-01 -1.42034376e+00 4.21846658e-01
1.00239420e+00 -1.19018459e+00 5.91992021e-01 -5.62534451e-01
6.13400102e-01 4.09323394e-01 -7.21644685e-02 -1.14390051e+00
-2.69807398e-01 -9.10414100e-01 -5.04949093e-01 6.50429726e-01
3.86475325e-01 -3.92964780e-01 8.33693266e-01 6.10396564e-01
-5.82037151e-01 -1.42654657e+00 -1.28806639e+00 -6.16160035e-01
-3.29680860e-01 -4.37796891e-01 4.03063476e-01 3.62776220e-01
2.98086017e-01 5.05277872e-01 -7.69133449e-01 -6.11653067e-02
1.79929405e-01 -3.00713122e-01 2.49938905e-01 -1.23642993e+00
-4.20441926e-02 2.01221481e-01 -8.29001188e-01 -4.92543787e-01
-1.50503084e-01 -7.39551723e-01 -2.02905953e-01 -1.14680612e+00
1.24224178e-01 2.92498227e-02 -1.19320869e+00 3.95327538e-01
9.07769948e-02 4.65733409e-01 -2.02036038e-01 -2.79511631e-01
-7.56791472e-01 7.66809225e-01 7.84190536e-01 3.03426415e-01
-3.39694619e-01 7.95553923e-02 -5.56468546e-01 5.08742630e-01
8.71861279e-01 -4.89387751e-01 -4.47483331e-01 -1.42503366e-01
1.67705342e-02 3.67561817e-01 -1.20836206e-01 -1.69119263e+00
2.39839658e-01 6.24267876e-01 6.31138265e-01 -4.52132076e-01
7.14525044e-01 -7.92552352e-01 9.81251150e-02 6.16224289e-01
-4.95033175e-01 2.20088646e-01 3.99283409e-01 4.89273459e-01
-8.10399652e-02 1.24207154e-01 8.58394444e-01 1.41923472e-01
-5.62488496e-01 5.80480695e-01 -3.28136146e-01 2.00223662e-02
9.13797617e-01 1.12529518e-02 -1.73563540e-01 -7.62027577e-02
-9.22989368e-01 4.93321568e-02 -1.92211077e-01 5.36903203e-01
6.61038518e-01 -1.15965497e+00 -4.25346226e-01 4.33718413e-01
3.96092296e-01 -6.88388944e-01 7.64605880e-01 9.69079375e-01
-5.54413736e-01 8.75733674e-01 -7.39853084e-01 -3.15756053e-01
-1.01747024e+00 1.93251580e-01 5.04328251e-01 -3.42190079e-02
-9.27303970e-01 9.07475889e-01 -3.97721767e-01 -6.15137182e-02
3.01622927e-01 -1.69329375e-01 -2.04831198e-01 -1.17701694e-01
9.47944582e-01 5.50498307e-01 6.53149486e-01 -2.93616652e-01
-5.67406893e-01 -1.80606484e-01 -1.57607317e-01 3.11749399e-01
1.53431368e+00 1.25202328e-01 -1.19204642e-02 4.15656894e-01
1.42355704e+00 -5.32928765e-01 -9.87797678e-01 1.21019363e-01
-2.03746721e-01 -1.31000325e-01 4.00756299e-01 -9.87521470e-01
-7.96725392e-01 9.41229045e-01 1.32716036e+00 -1.58032984e-01
1.22354853e+00 -4.43137646e-01 1.22639716e+00 5.19362390e-01
1.58843234e-01 -1.14900208e+00 -1.37557119e-01 3.95530939e-01
5.09533107e-01 -1.01549375e+00 -5.98690584e-02 4.00937438e-01
-6.32024705e-01 1.20736933e+00 8.63837779e-01 -3.46060097e-01
7.97340631e-01 8.29895437e-02 -2.44147316e-01 -3.56350064e-01
-1.03113151e+00 2.29818836e-01 3.06267947e-01 2.20972911e-01
5.78557670e-01 1.58028960e-01 -7.69032598e-01 1.36878955e+00
-5.66555560e-01 4.21527594e-01 3.58291626e-01 4.52485144e-01
-2.73013920e-01 -5.44158041e-01 5.15320957e-01 9.62553859e-01
-1.06064832e+00 -3.13877821e-01 3.65636870e-02 3.70349616e-01
1.30582064e-01 7.77565300e-01 2.48333365e-01 -7.96054184e-01
3.76897275e-01 1.32977560e-01 1.60300136e-01 -4.50060487e-01
-7.71465003e-01 -4.18581486e-01 1.00668229e-01 -8.55373740e-01
-2.13219263e-02 -1.29697502e-01 -9.79885757e-01 -2.67593294e-01
1.26050845e-01 3.94392312e-01 7.09428310e-01 9.74527359e-01
8.38380635e-01 6.64862335e-01 4.36387181e-01 -7.07502723e-01
-1.87218890e-01 -1.26658440e+00 -8.17179501e-01 2.62864102e-02
5.42290688e-01 -3.48252118e-01 1.07144140e-01 -4.86551613e-01] | [13.484424591064453, 3.5222930908203125] |
966c03d8-7795-4109-8e8b-13e5dc8727ba | the-task-2-of-cips-sighan-2012-named-entity | null | null | https://aclanthology.org/W12-6321 | https://aclanthology.org/W12-6321.pdf | The Task 2 of CIPS-SIGHAN 2012 Named Entity Recognition and Disambiguation in Chinese Bakeoff | null | ['Houfeng Wang', 'Zhengyan He', 'Sujian Li'] | 2012-12-01 | the-task-2-of-cips-sighan-2012-named-entity-1 | https://aclanthology.org/W12-6321 | https://aclanthology.org/W12-6321.pdf | ws-2012-12 | ['chinese-named-entity-recognition'] | ['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] | [-7.406741619110107, 3.6685519218444824] |
609bf0b8-4eb0-41aa-bc0a-b3b2f897ccb4 | multistage-pruning-of-cnn-based-ecg | 2109.00516 | null | https://arxiv.org/abs/2109.00516v1 | https://arxiv.org/pdf/2109.00516v1.pdf | Multistage Pruning of CNN Based ECG Classifiers for Edge Devices | Using smart wearable devices to monitor patients electrocardiogram (ECG) for real-time detection of arrhythmias can significantly improve healthcare outcomes. Convolutional neural network (CNN) based deep learning has been used successfully to detect anomalous beats in ECG. However, the computational complexity of existing CNN models prohibits them from being implemented in low-powered edge devices. Usually, such models are complex with lots of model parameters which results in large number of computations, memory, and power usage in edge devices. Network pruning techniques can reduce model complexity at the expense of performance in CNN models. This paper presents a novel multistage pruning technique that reduces CNN model complexity with negligible loss in performance compared to existing pruning techniques. An existing CNN model for ECG classification is used as a baseline reference. At 60% sparsity, the proposed technique achieves 97.7% accuracy and an F1 score of 93.59% for ECG classification tasks. This is an improvement of 3.3% and 9% for accuracy and F1 Score respectively, compared to traditional pruning with fine-tuning approach. Compared to the baseline model, we also achieve a 60.4% decrease in run-time complexity. | ['Deepu John', 'Barry Cardiff', 'Rajesh Panicker', 'Xiaolin Li'] | 2021-08-31 | null | null | null | null | ['ecg-classification'] | ['medical'] | [ 2.26035655e-01 5.77856004e-02 -2.53731050e-02 -3.70235652e-01
-3.60662669e-01 -6.79437891e-02 -5.35126567e-01 6.51024938e-01
-4.64920640e-01 6.47462189e-01 -5.22450097e-02 -5.30433238e-01
-1.44775495e-01 -7.80197144e-01 -2.35840529e-01 -3.39560747e-01
-8.56993198e-02 -8.36241469e-02 1.69084035e-02 2.88454980e-01
1.89194512e-02 3.66232902e-01 -1.04572022e+00 5.76348126e-01
6.69865429e-01 1.34252560e+00 -2.87905633e-01 9.67475295e-01
3.72737914e-01 6.13082111e-01 -8.66130054e-01 -2.75945008e-01
3.19929451e-01 -3.98782521e-01 -4.36830789e-01 -3.78854722e-01
2.57966280e-01 -2.76316375e-01 -3.08284432e-01 6.64072394e-01
1.07345021e+00 -3.08610678e-01 1.97380260e-01 -7.16803432e-01
1.13349110e-01 5.71057916e-01 -6.04475558e-01 8.38723540e-01
-1.33942738e-01 -1.53612971e-01 3.83468658e-01 -4.51634765e-01
7.63860047e-02 4.20224607e-01 1.61851799e+00 4.42938358e-01
-6.77191377e-01 -9.26238894e-01 -6.02247059e-01 1.08336695e-01
-1.56855690e+00 -3.09300870e-01 6.38328671e-01 4.59335595e-02
1.42802405e+00 5.03015339e-01 1.10415304e+00 5.36856949e-01
5.07496417e-01 3.22529882e-01 6.14347637e-01 -4.13135797e-01
1.67772859e-01 -2.38621026e-01 2.86163390e-01 7.61816204e-01
8.42359066e-01 2.90326420e-02 -6.01326644e-01 -3.62233073e-01
6.97382867e-01 4.15543199e-01 -3.23660187e-02 4.59711015e-01
-7.19988406e-01 4.90854323e-01 3.89250070e-01 2.53827155e-01
-3.84331971e-01 3.06888163e-01 8.92737925e-01 2.11543754e-01
1.43961176e-01 3.64744216e-01 -9.31359589e-01 -4.99471068e-01
-1.08782339e+00 3.05327177e-01 6.93342090e-01 8.13329339e-01
-7.45902583e-03 4.67076331e-01 -2.31608033e-01 6.91555798e-01
-7.66247213e-02 1.24879345e-01 5.67086935e-01 -5.61667860e-01
5.39159060e-01 9.76060033e-01 -3.25136125e-01 -1.03345823e+00
-8.00058842e-01 -1.11055887e+00 -1.53638721e+00 -1.49218842e-01
1.11013064e-02 -5.78651547e-01 -9.78379846e-01 1.12247837e+00
1.27296492e-01 3.92794430e-01 3.36196758e-02 4.58068252e-01
1.02307761e+00 2.79781073e-01 2.02508822e-01 -3.48222196e-01
1.41764247e+00 -7.39737391e-01 -5.66333890e-01 7.22284168e-02
1.00149012e+00 -6.93573594e-01 6.81752563e-01 6.27478957e-01
-1.19871938e+00 -4.41030711e-01 -1.35953009e+00 1.02848142e-01
1.32948652e-01 4.76809770e-01 7.60692179e-01 1.04045367e+00
-7.19444513e-01 9.14713323e-01 -9.51697290e-01 -2.16925949e-01
1.30495739e+00 8.87361586e-01 9.40169320e-02 2.00610593e-01
-8.61624956e-01 5.97335398e-01 3.16291094e-01 1.64130539e-01
-6.23298883e-01 -9.71243143e-01 -4.77822214e-01 2.74384379e-01
-1.98336467e-02 -8.04056764e-01 1.02952945e+00 -6.86502814e-01
-1.16700196e+00 5.89504063e-01 -7.31966272e-02 -1.04538596e+00
4.12258565e-01 -4.78987873e-01 -4.08261329e-01 2.99458373e-02
-3.24555963e-01 3.61134470e-01 4.10245806e-01 -1.11609995e-01
-7.09108353e-01 -4.05346453e-01 -4.39931303e-02 2.19049484e-01
-4.55776930e-01 1.95274185e-02 -1.43958122e-01 -8.63476396e-01
2.43629277e-01 -7.51468778e-01 -4.33445722e-01 1.26556799e-01
-1.52034402e-01 1.47100031e-01 8.28667760e-01 -6.63563490e-01
1.56313419e+00 -1.86309338e+00 -6.23415470e-01 4.10636030e-02
5.91280878e-01 7.89509237e-01 5.26063323e-01 -7.90215433e-02
-4.47192267e-02 2.51624376e-01 -1.69363156e-01 -1.79502349e-02
-9.27874625e-01 2.44713411e-01 3.72141480e-01 3.77166450e-01
1.96514912e-02 1.05407465e+00 -2.20341519e-01 -3.89219254e-01
1.41981214e-01 6.76337004e-01 -7.02802241e-01 -2.52762735e-01
5.83134234e-01 3.58640879e-01 -3.16177785e-01 8.29060733e-01
5.27588189e-01 -1.00944869e-01 2.17495412e-01 -4.37949210e-01
2.71386236e-01 2.91042536e-01 -1.05452931e+00 1.71134472e+00
-2.94589788e-01 5.00557363e-01 -3.70220244e-01 -1.31637859e+00
9.84508693e-01 7.04555511e-01 5.09841144e-01 -5.54697812e-01
6.08240962e-01 1.03695653e-01 4.56298262e-01 -5.32238364e-01
-2.01062292e-01 -1.36422306e-01 1.55862361e-01 1.45526424e-01
-3.18031199e-02 5.68520904e-01 -1.81726083e-01 -2.02926606e-01
1.44117999e+00 -2.96892136e-01 5.37117064e-01 -5.36165833e-01
3.49055797e-01 -1.88719094e-01 1.00616765e+00 7.51616657e-01
-2.86296129e-01 5.81754684e-01 2.55993217e-01 -1.40462494e+00
-1.01494026e+00 -4.61500287e-01 -2.33961493e-01 3.08246344e-01
-3.08127791e-01 -6.87438071e-01 -7.71030664e-01 -4.19211686e-01
-2.65428841e-01 -8.30416456e-02 -5.30310988e-01 -3.48414361e-01
-9.20710742e-01 -1.18262517e+00 1.26639020e+00 1.10813308e+00
9.00689900e-01 -9.77832913e-01 -1.58616984e+00 3.66671592e-01
1.44968033e-01 -1.06220281e+00 5.05709983e-02 3.23249042e-01
-1.98050106e+00 -1.11287403e+00 -4.43834633e-01 -8.17616701e-01
5.89580715e-01 -3.11643004e-01 1.30047214e+00 6.17095470e-01
-8.96451771e-01 -5.31517148e-01 -2.61190623e-01 -1.16156399e+00
1.79363102e-01 3.51099730e-01 1.44320317e-02 -5.03529906e-01
4.60000098e-01 -8.14087152e-01 -1.14995229e+00 -2.67328948e-01
-4.34110582e-01 1.57064628e-02 6.67244852e-01 8.90120625e-01
5.57445467e-01 1.89407304e-01 8.17750692e-01 -1.17432296e+00
6.95427477e-01 -9.77059379e-02 -3.13321143e-01 -5.97522520e-02
-1.07052362e+00 -4.01019365e-01 6.00013614e-01 -3.72027487e-01
-4.16174769e-01 3.27132523e-01 -1.34595409e-01 -2.80643255e-01
-4.19228189e-02 5.46977818e-01 2.69057065e-01 -1.18632153e-01
7.93385267e-01 -2.51254383e-02 -2.65273660e-01 -4.88235593e-01
-5.65858185e-01 6.75320029e-01 2.73934692e-01 -1.00724608e-01
1.29860058e-01 2.06964791e-01 3.77078623e-01 -5.14704764e-01
-7.14672506e-01 -2.41809428e-01 -4.95658159e-01 1.01355791e-01
5.61999619e-01 -9.78574753e-01 -8.02187324e-01 2.31314003e-01
-9.63309050e-01 9.95660797e-02 -2.70341575e-01 4.21005160e-01
1.18943036e-01 2.06435338e-01 -6.48019373e-01 -8.27622533e-01
-1.44379795e+00 -7.77832568e-01 6.31736994e-01 1.84427753e-01
-5.40661216e-01 -6.46838903e-01 -4.77673501e-01 1.69850662e-01
6.70624852e-01 7.09878743e-01 9.27893400e-01 -5.83323240e-01
-6.67765811e-02 -7.21105039e-01 -6.32756427e-02 3.64930719e-01
5.84705211e-02 -5.07256746e-01 -8.86259019e-01 -2.78982371e-01
4.28750992e-01 1.71373393e-02 6.34897053e-01 7.78063238e-01
1.54712808e+00 -3.42428178e-01 -4.58587110e-01 7.79696882e-01
1.64320421e+00 6.09259665e-01 7.64150441e-01 1.53573100e-02
6.68777287e-01 -2.13087052e-01 1.08821377e-01 6.37583792e-01
-1.71653524e-01 1.95236579e-01 1.70254409e-01 -3.55349243e-01
-7.98611492e-02 3.44321430e-01 -4.27851439e-01 1.06698513e+00
-3.80849838e-01 1.46778762e-01 -1.28135335e+00 5.50346017e-01
-1.71293461e+00 -5.62727571e-01 -3.45760643e-01 2.06796050e+00
7.30351031e-01 4.33883876e-01 3.31882946e-02 9.79101181e-01
4.24548894e-01 -4.39485043e-01 -6.44224763e-01 -7.59803951e-01
1.03788950e-01 1.07020557e+00 6.36122108e-01 4.69012819e-02
-1.04972553e+00 3.71861428e-01 5.99689722e+00 3.54186624e-01
-1.18106234e+00 3.85607302e-01 1.11048305e+00 -5.67650497e-01
7.25714087e-01 -2.31317058e-01 -7.23043442e-01 2.87267804e-01
1.25988603e+00 3.60612199e-02 -1.31646782e-01 7.43866205e-01
2.38611713e-01 4.89509255e-02 -8.75507116e-01 1.51255119e+00
1.36534609e-02 -1.59048927e+00 -2.91960109e-02 -1.28753006e-01
5.53647578e-01 -8.90706331e-02 -3.61704499e-01 1.18831061e-01
-6.11722887e-01 -1.33281446e+00 1.36541396e-01 2.20508039e-01
9.75068033e-01 -1.26518512e+00 1.40639579e+00 3.16962481e-01
-1.19693756e+00 -3.38860691e-01 -4.71972466e-01 -6.69849813e-01
-3.02489456e-02 8.37118506e-01 -7.83784389e-01 2.87561089e-01
1.26223838e+00 6.37790203e-01 -4.11482066e-01 1.29360199e+00
4.58994627e-01 9.88728404e-01 -5.15189707e-01 -4.06612270e-02
-1.16612941e-01 4.05009806e-01 2.64693052e-01 1.24930668e+00
3.06093425e-01 4.35715139e-01 7.74109513e-02 1.58589229e-01
-2.73628861e-01 1.77984998e-01 -4.66192573e-01 3.70391130e-01
3.58566672e-01 9.71314013e-01 -8.26348245e-01 -6.98523700e-01
-1.22685902e-01 5.16797185e-01 -8.41079056e-02 -1.69117242e-01
-9.48440611e-01 -8.31295133e-01 2.19955340e-01 4.26039070e-01
7.12251812e-02 1.72983915e-01 -1.20980656e+00 -7.56164551e-01
2.97573090e-01 -9.11824942e-01 3.33565325e-01 -1.88132018e-01
-5.53326070e-01 8.23707938e-01 -4.17161167e-01 -1.19021785e+00
2.40327805e-01 -2.53542989e-01 -6.18494213e-01 7.71195233e-01
-1.00151932e+00 -9.35088992e-01 -6.78313792e-01 4.55469072e-01
6.63395107e-01 -1.82167649e-01 1.22522819e+00 7.50429332e-01
-6.04894459e-01 1.01906466e+00 -4.00821716e-01 3.55495900e-01
9.56203714e-02 -8.16956162e-01 3.14894736e-01 7.68984854e-01
-5.80431931e-02 7.64492929e-01 2.12918758e-01 -5.81735671e-01
-1.08475256e+00 -1.35068035e+00 9.28564787e-01 -1.52471542e-01
-2.49248013e-01 -1.33643046e-01 -7.32550979e-01 2.95140624e-01
3.86859151e-03 3.42155725e-01 1.02368677e+00 2.12669477e-01
1.11474857e-01 -4.97622520e-01 -1.34689879e+00 3.40275079e-01
9.36065912e-01 -6.48493245e-02 -2.71682143e-01 1.07784525e-01
2.18392700e-01 -6.85169220e-01 -1.05288601e+00 1.05730140e+00
9.96034086e-01 -8.25636744e-01 8.58075440e-01 -6.58152759e-01
2.65964776e-01 -9.65294614e-02 2.92472392e-01 -3.96751851e-01
-2.53670394e-01 -6.82836354e-01 -5.53132117e-01 6.64469302e-01
4.77504343e-01 -3.85295570e-01 1.27417731e+00 4.37265307e-01
-1.79756224e-01 -1.67477942e+00 -8.99801075e-01 -4.65514213e-01
-1.96166739e-01 -5.54688811e-01 4.31816041e-01 7.57214963e-01
-3.02916408e-01 5.48227191e-01 -3.51975232e-01 -2.27961391e-02
4.32454348e-01 -2.94516861e-01 2.42176846e-01 -1.55406272e+00
-1.53711379e-01 -8.03286135e-02 -8.16070795e-01 -4.51279223e-01
-8.43351483e-01 -7.41432846e-01 -4.91230279e-01 -1.39109945e+00
2.78965086e-01 -5.63461483e-01 -7.42862940e-01 6.84022248e-01
2.70635765e-02 7.65037000e-01 -1.84510425e-01 1.02799669e-01
-3.37583601e-01 -1.47936150e-01 8.47486019e-01 3.88252176e-02
-5.46996653e-01 1.66664809e-01 -8.01965594e-01 9.65863466e-01
1.44095016e+00 -1.04650521e+00 -5.03290176e-01 -6.08255506e-01
2.91681409e-01 9.08449963e-02 1.29624620e-01 -1.66770649e+00
2.51983315e-01 5.50173342e-01 1.01671147e+00 -6.08338833e-01
8.15808252e-02 -5.32362103e-01 2.34785169e-01 1.21869075e+00
-1.19645104e-01 5.15981197e-01 6.22155309e-01 5.13464093e-01
-3.72633450e-02 -1.19762160e-01 7.21949816e-01 -1.76628679e-01
1.10821545e-01 3.92172873e-01 -3.56104821e-01 -1.10059790e-01
9.51291859e-01 -7.56433725e-01 3.29009622e-01 1.97256744e-01
-9.51696277e-01 -2.32716098e-01 -2.69239396e-01 1.07844353e-01
8.01364660e-01 -1.06252289e+00 -6.12525403e-01 1.22461431e-01
-1.75960079e-01 3.66034150e-01 2.66230941e-01 1.01383793e+00
-1.23833549e+00 5.58216631e-01 -3.53166640e-01 -6.62450850e-01
-1.82895923e+00 -5.82733713e-02 5.78260899e-01 -4.30243582e-01
-1.18949544e+00 9.33606386e-01 -4.55714256e-01 3.03117394e-01
4.34458822e-01 -8.15709949e-01 -2.03071848e-01 -5.22699416e-01
6.26352787e-01 7.15830266e-01 6.22840047e-01 3.28376666e-02
-5.17470837e-01 6.37348771e-01 -2.10941974e-02 5.78951716e-01
1.37108696e+00 4.42931682e-01 5.11226393e-02 8.97255912e-02
9.79721963e-01 -3.75922024e-01 -6.33917570e-01 1.95148438e-01
-1.86416693e-02 -1.00563481e-01 3.20713401e-01 -1.09442866e+00
-1.42795718e+00 1.13670969e+00 1.32493556e+00 -2.75202364e-01
1.58363318e+00 -6.25599563e-01 1.27751708e+00 6.05901301e-01
1.56649262e-01 -8.68878067e-01 -2.07599744e-01 1.57167345e-01
4.18756038e-01 -1.02802098e+00 4.35903609e-01 -3.26090276e-01
-4.27827090e-01 1.18573487e+00 4.76385236e-01 -5.59386432e-01
1.12641919e+00 6.27857745e-01 8.14135093e-03 -3.44420046e-01
-6.11690283e-01 5.42543054e-01 1.55952886e-01 5.48738182e-01
6.17488742e-01 8.29087719e-02 -6.26595259e-01 8.22721541e-01
-1.98915303e-01 6.47675097e-01 1.72815964e-01 1.22036541e+00
-2.71648616e-01 -9.53687727e-01 2.27920279e-01 1.14655459e+00
-1.43576550e+00 -4.30056602e-01 -6.25361921e-03 4.23470676e-01
6.15692317e-01 8.61517191e-01 -1.14638329e-01 -5.11541009e-01
3.55085582e-01 1.01239793e-01 4.40694153e-01 -6.48722053e-01
-1.30892169e+00 3.85023132e-02 6.89608306e-02 -3.39027852e-01
-3.39171231e-01 -2.00178921e-01 -1.13117325e+00 -1.20347090e-01
-4.67494786e-01 -4.64433320e-02 6.15666628e-01 7.59022951e-01
8.64296854e-01 1.03196216e+00 -1.46701783e-01 -1.20790310e-01
-2.98879504e-01 -1.22052109e+00 -3.01471889e-01 1.52099133e-02
1.15242794e-01 -2.65456468e-01 2.97296166e-01 2.75622815e-01] | [14.068614959716797, 3.275242328643799] |
ccba85b4-1caf-43cf-b59b-957746c14da5 | reknow-enhanced-knowledge-for-joint-entity | 2206.05123 | null | https://arxiv.org/abs/2206.05123v3 | https://arxiv.org/pdf/2206.05123v3.pdf | REKnow: Enhanced Knowledge for Joint Entity and Relation Extraction | Relation extraction is an important but challenging task that aims to extract all hidden relational facts from the text. With the development of deep language models, relation extraction methods have achieved good performance on various benchmarks. However, we observe two shortcomings of previous methods: first, there is no unified framework that works well under various relation extraction settings; second, effectively utilizing external knowledge as background information is absent. In this work, we propose a knowledge-enhanced generative model to mitigate these two issues. Our generative model is a unified framework to sequentially generate relational triplets under various relation extraction settings and explicitly utilizes relevant knowledge from Knowledge Graph (KG) to resolve ambiguities. Our model achieves superior performance on multiple benchmarks and settings, including WebNLG, NYT10, and TACRED. | ['Bing Xiang', 'Zhiguo Wang', 'Patrick Ng', 'Sheng Zhang'] | 2022-06-10 | null | null | null | null | ['joint-entity-and-relation-extraction'] | ['natural-language-processing'] | [-1.02407724e-01 5.88088870e-01 -6.80648863e-01 -1.91953331e-01
-8.77485335e-01 -3.56582314e-01 6.16691649e-01 1.00454144e-01
6.89531043e-02 1.11794293e+00 2.44646460e-01 -4.27179903e-01
-4.30455267e-01 -1.20518804e+00 -5.09157479e-01 -1.09688580e-01
1.76933452e-01 8.89576972e-01 3.34497362e-01 -4.70867723e-01
-3.16636115e-01 1.94726691e-01 -9.87444699e-01 1.74916908e-01
9.78361070e-01 8.08141947e-01 -3.60122889e-01 2.32466251e-01
-3.54744315e-01 1.34322762e+00 -5.22547245e-01 -1.14911127e+00
7.68992677e-02 -1.87020376e-01 -1.39530408e+00 -2.08925113e-01
3.74201685e-02 -1.71449706e-01 -8.47636938e-01 9.56583917e-01
2.71414369e-01 -6.17142506e-02 5.47453821e-01 -1.32610548e+00
-8.88405859e-01 1.33625865e+00 -5.75599611e-01 1.97149307e-01
4.91894931e-01 -3.36663842e-01 1.36593795e+00 -8.57575178e-01
9.05164182e-01 1.21649349e+00 5.65810740e-01 1.65699303e-01
-1.03007948e+00 -6.31938100e-01 2.05552384e-01 5.33292055e-01
-1.74914157e+00 -4.20845538e-01 5.91129303e-01 -1.10920273e-01
1.52083480e+00 2.02535018e-01 4.74113673e-01 1.12257028e+00
-9.51324925e-02 7.83920348e-01 8.06935787e-01 -3.62976283e-01
-2.87210196e-01 -1.32187337e-01 4.25024271e-01 6.87205791e-01
7.33619750e-01 -2.74959534e-01 -7.78836846e-01 -1.61510319e-01
6.98119521e-01 -3.03436488e-01 -1.43594027e-01 -5.44822253e-02
-1.01419663e+00 6.68073952e-01 2.72270352e-01 2.20159501e-01
-2.22839579e-01 -7.10293576e-02 2.57789671e-01 2.39039958e-01
5.40656388e-01 5.01247585e-01 -7.16371715e-01 -3.62173729e-02
-4.88184541e-01 1.43483937e-01 1.18730283e+00 1.57063222e+00
8.27275515e-01 -3.01159590e-01 -2.78570235e-01 7.44801879e-01
1.80514544e-01 2.32179314e-01 5.99007010e-02 -6.74901724e-01
1.10298860e+00 1.00626063e+00 -1.46048546e-01 -1.28982008e+00
-3.13489914e-01 -5.90059757e-01 -8.78319323e-01 -7.09319353e-01
1.22338280e-01 -3.29223216e-01 -8.71624351e-01 1.50984788e+00
3.97313356e-01 7.53815919e-02 6.57028735e-01 5.31436384e-01
1.49361634e+00 3.34793091e-01 7.72158131e-02 -3.25424641e-01
1.41788459e+00 -1.01900172e+00 -1.22938728e+00 -4.89452928e-01
5.70920467e-01 -6.22632742e-01 5.58825016e-01 8.45836774e-02
-8.90653670e-01 -1.70140684e-01 -1.01385808e+00 -3.89960617e-01
-6.05117321e-01 4.46673222e-02 1.36500025e+00 3.44005078e-01
-7.84529388e-01 2.63763249e-01 -8.09615612e-01 -3.96126926e-01
5.12274444e-01 3.87582958e-01 -4.16185975e-01 -1.07996136e-01
-1.64728570e+00 1.08886445e+00 9.81234670e-01 3.12320143e-01
-4.32320565e-01 -4.64797080e-01 -1.19652104e+00 1.65727377e-01
1.32438779e+00 -1.16127169e+00 1.08061528e+00 1.62980258e-02
-1.34157562e+00 6.44078910e-01 -3.05795729e-01 -5.53287625e-01
2.36057848e-01 -6.37362897e-01 -6.38792396e-01 -1.03223614e-01
2.90204108e-01 1.48393869e-01 1.34124488e-01 -1.20752370e+00
-4.40089107e-01 -3.00007671e-01 2.98259407e-01 2.72079647e-01
-1.57854721e-01 1.44729391e-01 -1.11165774e+00 -5.85075259e-01
4.03918684e-01 -6.61446273e-01 -1.60179317e-01 -9.13473010e-01
-1.11694121e+00 -6.25220835e-01 7.67599702e-01 -5.35473168e-01
1.55271459e+00 -1.65327597e+00 1.42963037e-01 1.75233901e-01
5.92860460e-01 3.41625154e-01 2.20542878e-01 6.76729381e-01
-2.66314037e-02 3.40641141e-01 2.33497005e-02 -1.25912711e-01
1.00166023e-01 5.47539294e-01 -4.89579916e-01 -4.12882507e-01
5.49171865e-01 1.49768853e+00 -9.88631785e-01 -9.99096870e-01
-1.18661746e-01 3.41400653e-01 -1.31164119e-01 1.43696785e-01
-3.26576710e-01 -2.42560287e-03 -7.82292783e-01 8.14880788e-01
4.92160231e-01 -7.03336656e-01 7.36224651e-01 -4.07647938e-01
4.88608688e-01 7.04287767e-01 -1.13515818e+00 1.50964046e+00
-5.34678623e-02 2.73969084e-01 -4.64006484e-01 -7.67327964e-01
8.71155798e-01 4.12641376e-01 3.68810952e-01 -2.79571176e-01
5.95923625e-02 4.89411503e-02 -6.55511441e-03 -6.38976216e-01
6.28914416e-01 8.27101916e-02 -3.86049226e-02 2.88316011e-01
3.00408751e-01 -7.32779503e-02 4.57052708e-01 7.00092793e-01
1.28506923e+00 2.32940033e-01 6.48484945e-01 1.66712403e-01
1.89787477e-01 2.30505113e-02 9.43641424e-01 6.24881089e-01
3.46369207e-01 3.13065350e-01 9.38962579e-01 -3.84002328e-01
-2.66521364e-01 -9.04459476e-01 1.23552509e-01 6.32202923e-01
1.39444172e-01 -1.21041310e+00 -1.64208576e-01 -1.02795660e+00
-4.10171710e-02 4.83756870e-01 -5.07339895e-01 -2.19931945e-01
-5.63022614e-01 -1.19864321e+00 7.59118378e-01 8.41050208e-01
6.42040431e-01 -8.01819682e-01 2.38928229e-01 2.00844631e-01
-6.67204916e-01 -1.89748609e+00 6.57841936e-02 2.21370578e-01
-5.92460990e-01 -1.36199486e+00 1.98904723e-01 -6.22880578e-01
4.48740125e-01 7.66443834e-02 1.60095751e+00 -2.86957435e-02
4.55326475e-02 9.45107080e-03 -4.11501944e-01 -2.61142313e-01
-1.86489284e-01 6.73323274e-01 -2.71803170e-01 -3.65586817e-01
6.57990634e-01 -5.73487222e-01 -4.90083285e-02 7.65320212e-02
-6.76133454e-01 3.58738035e-01 6.42758250e-01 6.00967109e-01
6.15102172e-01 5.15208662e-01 5.62314630e-01 -1.40124106e+00
7.82251656e-01 -4.95814651e-01 -3.10350507e-01 7.24199831e-01
-8.21556270e-01 2.40490064e-01 3.05403501e-01 6.34200647e-02
-1.25282955e+00 -3.13973844e-01 8.19106847e-02 -4.73981164e-02
-6.17168797e-03 1.08292067e+00 -7.04274297e-01 3.65515321e-01
2.27720246e-01 -6.46806136e-02 -5.98754108e-01 -4.24523622e-01
5.07070601e-01 3.70162159e-01 7.49059856e-01 -1.09412217e+00
1.03593946e+00 1.45781443e-01 2.29171336e-01 -3.85638416e-01
-1.48632395e+00 -1.50421441e-01 -8.00207198e-01 4.90713954e-01
6.28147960e-01 -1.00181258e+00 -6.61515951e-01 2.21204922e-01
-1.22649837e+00 -1.25641882e-01 -2.19869703e-01 4.26803499e-01
-9.15977582e-02 4.43994880e-01 -9.85992312e-01 -5.43855846e-01
-7.13164032e-01 -8.02776337e-01 8.71522546e-01 3.29999983e-01
-1.98384285e-01 -9.84354913e-01 -1.06385052e-01 6.39227033e-01
4.59151895e-04 4.54845786e-01 1.03683698e+00 -6.63305044e-01
-9.32162166e-01 -1.42236844e-01 -5.11423886e-01 -8.46412107e-02
6.11181319e-01 1.98026329e-01 -6.61530793e-01 2.41352648e-01
-5.80980480e-01 -5.72211683e-01 7.83058286e-01 -2.35092267e-01
9.65142906e-01 -4.41444486e-01 -7.66438365e-01 5.92117310e-01
1.12619901e+00 1.31262377e-01 7.45817780e-01 2.27858365e-01
1.10868311e+00 4.29804146e-01 5.94143450e-01 -6.83259517e-02
1.20389271e+00 6.88020349e-01 -1.81456916e-02 -3.31874341e-02
-1.83801681e-01 -4.81360644e-01 -1.33135125e-01 9.75894868e-01
-4.75200981e-01 -3.43318164e-01 -1.09571850e+00 5.26774466e-01
-2.24831581e+00 -8.26781213e-01 -2.75090456e-01 1.58045363e+00
1.46867299e+00 3.65848571e-01 -2.03422368e-01 2.02737097e-02
4.54284072e-01 -5.84078766e-03 -3.94687593e-01 1.96676075e-01
-5.11028290e-01 5.18111885e-01 1.72783285e-01 4.25677001e-01
-1.19972146e+00 1.47782242e+00 6.51379013e+00 7.25019813e-01
-5.38564205e-01 -2.63830740e-02 3.52239937e-01 1.62654012e-01
-3.38237435e-01 3.93644065e-01 -1.19185793e+00 -7.35818893e-02
5.25101423e-01 -3.89165878e-01 1.29815623e-01 5.15765607e-01
-5.04559100e-01 1.64515853e-01 -1.19870937e+00 8.47487330e-01
-3.54457684e-02 -1.42911172e+00 2.95312345e-01 -1.45505983e-02
6.57807171e-01 -1.25100002e-01 -2.00676069e-01 6.18735552e-01
1.01602912e+00 -1.07849145e+00 1.06312662e-01 5.20714283e-01
6.09884620e-01 -8.38447988e-01 9.57617402e-01 6.35351688e-02
-1.34256184e+00 3.68652940e-01 -1.28544033e-01 -7.93983415e-02
1.69298828e-01 8.74276459e-01 -1.02210796e+00 1.47990799e+00
6.49623036e-01 8.62100124e-01 -8.08125675e-01 4.17915195e-01
-8.37854087e-01 5.82529008e-01 -3.05631071e-01 3.46437067e-01
-8.57520625e-02 -3.75365950e-02 3.05196017e-01 1.14534700e+00
-7.12606534e-02 2.99071133e-01 3.20832372e-01 8.49307537e-01
-3.90005559e-01 2.81931236e-02 -6.34246051e-01 -1.77320644e-01
6.29682899e-01 1.38641715e+00 -6.51794851e-01 -4.14705485e-01
-5.34151793e-01 4.86673117e-01 8.21466446e-01 5.16894460e-01
-6.92830563e-01 -4.49176818e-01 4.84897643e-01 -2.04088524e-01
1.86906904e-01 -2.26851672e-01 -1.69796765e-01 -1.57455361e+00
2.05628440e-01 -9.48608220e-01 8.22363734e-01 -7.32260108e-01
-1.32174063e+00 7.28611648e-01 2.41929278e-01 -5.95536053e-01
-3.21996033e-01 -3.03844005e-01 -1.73997909e-01 6.71717703e-01
-1.49551928e+00 -1.47849560e+00 -2.42850795e-01 5.27620971e-01
5.18632643e-02 -1.12846956e-01 8.95233452e-01 4.36214924e-01
-9.97089386e-01 6.93968475e-01 -7.11198866e-01 7.31636703e-01
6.47870898e-01 -1.46482229e+00 5.44765472e-01 9.71827328e-01
4.88935560e-01 1.12102318e+00 4.20259297e-01 -1.03364420e+00
-1.47651327e+00 -1.19360447e+00 1.15572298e+00 -7.93851852e-01
9.02125597e-01 -3.10678333e-01 -9.55117285e-01 1.33083367e+00
3.59981477e-01 1.32634878e-01 8.32605600e-01 7.82939672e-01
-5.84139705e-01 -1.45533353e-01 -5.98529875e-01 6.91613913e-01
1.47885764e+00 -5.39151132e-01 -7.62798727e-01 3.81907314e-01
1.08249784e+00 -8.24162662e-01 -1.41781962e+00 9.22758460e-01
2.17162713e-01 -4.44046080e-01 9.23935354e-01 -8.26207817e-01
5.75052023e-01 -3.61764669e-01 -9.87152755e-02 -1.05775142e+00
-2.02575892e-01 -1.04513299e+00 -9.94520009e-01 1.70214486e+00
7.87260413e-01 -7.75240898e-01 7.27675378e-01 8.18339348e-01
7.37671852e-02 -9.54027712e-01 -5.30171692e-01 -6.29402936e-01
-2.24514052e-01 -3.37441772e-01 1.08201015e+00 1.20351470e+00
1.45238563e-01 1.21578860e+00 -3.25708389e-01 2.50709593e-01
4.71442044e-01 5.39911449e-01 9.97744381e-01 -1.22079396e+00
-3.18486512e-01 -1.65168956e-01 -2.41124138e-01 -9.49869454e-01
3.25764626e-01 -9.89659369e-01 -4.76088405e-01 -2.04591107e+00
5.03088355e-01 -5.47286272e-01 -2.39749491e-01 1.02230430e+00
-6.60102129e-01 -9.91966948e-02 -1.19527385e-01 1.06553353e-01
-8.81184042e-01 5.58376968e-01 1.35207188e+00 -1.09303482e-01
-6.43931702e-02 -1.18246928e-01 -1.29466188e+00 6.25594139e-01
6.49966717e-01 -3.84373218e-01 -7.33085096e-01 -6.06837988e-01
8.63536537e-01 9.11859721e-02 2.39613414e-01 -5.46607196e-01
4.30601448e-01 -2.68320531e-01 2.31637418e-01 -9.44362342e-01
2.35675588e-01 -5.45984864e-01 2.13672608e-01 -6.52553290e-02
-5.89706413e-02 -9.43955313e-03 5.22945710e-02 5.14255345e-01
-3.96727622e-01 1.77473620e-01 -1.21588171e-01 5.28288558e-02
-6.52185678e-01 4.75014806e-01 2.90570170e-01 4.77672517e-01
7.18068004e-01 2.94807494e-01 -8.85123789e-01 -2.98681259e-01
-7.13256955e-01 4.75781679e-01 -7.67472154e-03 5.76835334e-01
5.47309458e-01 -1.41990995e+00 -7.54132271e-01 -1.44282907e-01
2.46127516e-01 6.77168906e-01 -1.35592058e-01 7.61806488e-01
-1.81887820e-01 6.32518828e-01 3.95195842e-01 -8.56826603e-02
-1.07427430e+00 6.92142606e-01 2.63701648e-01 -9.26412880e-01
-7.41680682e-01 6.91452503e-01 -4.40754965e-02 -3.89799863e-01
1.77144576e-02 -4.86223966e-01 -3.81316304e-01 -1.33873582e-01
1.91909194e-01 8.82575288e-02 3.01107258e-01 -3.99829924e-01
-4.03275430e-01 3.23594481e-01 -3.87477517e-01 3.04217845e-01
1.39824355e+00 -3.26355062e-02 -5.05502641e-01 1.51282385e-01
6.46412313e-01 1.84057057e-01 -6.35871768e-01 -7.00662851e-01
4.40739304e-01 -1.16229743e-01 -2.09289253e-01 -8.49404037e-01
-1.23624361e+00 2.30532438e-01 -6.28834486e-01 3.28889161e-01
9.63474870e-01 4.08488393e-01 9.70344186e-01 7.05720961e-01
4.56371486e-01 -8.83697331e-01 -2.68840373e-01 7.63845503e-01
6.29644752e-01 -1.24625516e+00 4.39442098e-01 -1.41048050e+00
-6.80179179e-01 7.33165324e-01 1.01166248e+00 4.94005024e-01
6.56519234e-01 6.61805570e-01 -3.87995355e-02 -6.91283166e-01
-1.07963037e+00 -7.29831100e-01 4.52288955e-01 5.93006253e-01
5.23440778e-01 1.59950063e-01 -2.52716452e-01 1.04626775e+00
-4.52383310e-01 -6.17416389e-03 1.50393769e-01 9.09796715e-01
1.37627229e-01 -1.38018382e+00 1.18763588e-01 4.91272271e-01
-6.83450520e-01 -3.79289508e-01 -8.69796872e-01 1.00019467e+00
3.70838530e-02 1.15160656e+00 -5.51183641e-01 -5.37830353e-01
3.62551719e-01 1.82723925e-02 6.21801913e-01 -9.66373980e-01
-3.06364834e-01 -1.10219009e-01 6.81130826e-01 -3.27752471e-01
-5.34153163e-01 -2.83836305e-01 -1.25652218e+00 -2.44272485e-01
-5.68628550e-01 4.68443245e-01 -1.81523845e-01 1.23556626e+00
4.99171883e-01 7.86258757e-01 5.44187836e-02 2.49123573e-01
-3.79452407e-02 -9.93912160e-01 -3.63974810e-01 1.98902473e-01
-2.01728940e-01 -8.20275068e-01 1.56138510e-01 -4.31577861e-03] | [9.205423355102539, 8.502617835998535] |
e1875e98-dcaf-428a-97cc-0c07c6c7413c | object-discovery-via-contrastive-learning-for | 2208.07576 | null | https://arxiv.org/abs/2208.07576v2 | https://arxiv.org/pdf/2208.07576v2.pdf | Object Discovery via Contrastive Learning for Weakly Supervised Object Detection | Weakly Supervised Object Detection (WSOD) is a task that detects objects in an image using a model trained only on image-level annotations. Current state-of-the-art models benefit from self-supervised instance-level supervision, but since weak supervision does not include count or location information, the most common ``argmax'' labeling method often ignores many instances of objects. To alleviate this issue, we propose a novel multiple instance labeling method called object discovery. We further introduce a new contrastive loss under weak supervision where no instance-level information is available for sampling, called weakly supervised contrastive loss (WSCL). WSCL aims to construct a credible similarity threshold for object discovery by leveraging consistent features for embedding vectors in the same class. As a result, we achieve new state-of-the-art results on MS-COCO 2014 and 2017 as well as PASCAL VOC 2012, and competitive results on PASCAL VOC 2007. | ['Daijin Kim', 'Junhyug Noh', 'Danica J. Sutherland', 'Wonho Bae', 'Jinhwan Seo'] | 2022-08-16 | null | null | null | null | ['weakly-supervised-object-detection'] | ['computer-vision'] | [ 2.67206550e-01 -1.25207767e-01 -5.47558963e-01 -6.29478693e-01
-1.00246656e+00 -3.17743748e-01 6.83703005e-01 2.39416078e-01
-7.12981105e-01 6.97844386e-01 -2.55356371e-01 3.64882320e-01
2.05333814e-01 -5.44642270e-01 -1.19175279e+00 -6.90127194e-01
-9.03422311e-02 3.62032950e-01 5.74099541e-01 3.24335605e-01
-1.11307353e-02 2.89557636e-01 -1.87024784e+00 6.55101418e-01
6.68090999e-01 1.25643611e+00 4.71213788e-01 3.05310518e-01
-3.86133790e-02 8.81300092e-01 -5.64453661e-01 -4.13029432e-01
3.10132772e-01 -1.75562039e-01 -8.05151343e-01 4.09752548e-01
1.04331481e+00 -2.38499701e-01 -3.60625014e-02 1.27456450e+00
3.68059367e-01 -1.80678815e-01 7.29354799e-01 -1.37509239e+00
-5.68214953e-01 4.30625290e-01 -8.30089450e-01 3.19948912e-01
6.96209669e-02 3.33273172e-01 1.31255198e+00 -1.29490471e+00
6.27245545e-01 1.09010935e+00 5.76915145e-01 4.28306937e-01
-1.36545885e+00 -7.32313871e-01 4.74136889e-01 4.28072095e-01
-1.45396161e+00 -2.57620782e-01 7.70921588e-01 -3.19500774e-01
6.28331006e-01 2.28000343e-01 4.17970628e-01 1.02483916e+00
-4.08249676e-01 1.36664236e+00 1.08656502e+00 -4.25976396e-01
1.84443817e-01 4.67523009e-01 1.63954794e-01 7.66014218e-01
5.24177670e-01 -1.25991672e-01 -5.48057497e-01 5.18407569e-06
3.37487727e-01 1.94312423e-01 -1.84933003e-02 -6.86535358e-01
-1.26011944e+00 5.99613547e-01 7.20780611e-01 -1.34142647e-02
-1.96518570e-01 1.39869556e-01 4.66212332e-01 -1.60743415e-01
6.03869200e-01 4.05869246e-01 -5.19515753e-01 3.08964044e-01
-9.93517458e-01 3.02975267e-01 4.60359335e-01 1.14164090e+00
9.50912058e-01 -2.67587602e-01 -4.94325310e-01 8.77239048e-01
1.51330724e-01 4.67207283e-01 2.01092005e-01 -8.92164171e-01
3.62096846e-01 7.40878224e-01 2.50687420e-01 -4.53427911e-01
-1.00825787e-01 -1.06841969e+00 -3.89544517e-01 2.30363056e-01
2.84287244e-01 2.74654865e-01 -1.09903944e+00 1.68893814e+00
5.23469985e-01 3.87271672e-01 -1.28518328e-01 8.66279006e-01
9.00673270e-01 3.77618164e-01 2.50768721e-01 1.57914925e-02
1.27369297e+00 -1.27661335e+00 -5.53621054e-01 -3.43303829e-01
5.86657703e-01 -4.92670238e-01 1.15610027e+00 2.86024094e-01
-9.26894307e-01 -5.69072723e-01 -1.05288863e+00 3.07759158e-02
-3.34257931e-01 5.51703215e-01 6.66027904e-01 2.20073149e-01
-5.42535245e-01 2.97292978e-01 -6.66850388e-01 -5.32658398e-02
1.11476767e+00 1.03062950e-01 -4.09215361e-01 -4.63621259e-01
-7.71255553e-01 7.79364467e-01 3.73702645e-01 3.35973091e-02
-1.40046322e+00 -9.32701409e-01 -9.26111281e-01 -4.61739041e-02
8.33247364e-01 -4.00597036e-01 1.17625737e+00 -1.01870167e+00
-8.86326551e-01 1.32701349e+00 -2.41516292e-01 -7.19587684e-01
6.55832469e-01 -4.41136450e-01 -1.11083798e-01 2.24064976e-01
6.28527582e-01 8.31321537e-01 8.76033425e-01 -1.68763494e+00
-9.80029166e-01 -3.04068476e-01 6.11533457e-03 7.57790804e-02
-3.89614940e-01 7.79119208e-02 -6.19233131e-01 -5.18660784e-01
-9.80489142e-03 -7.07795501e-01 -1.96002752e-01 4.85566974e-01
-6.09111905e-01 -8.02218795e-01 9.18653429e-01 -1.53581470e-01
7.99394786e-01 -2.16043925e+00 -2.42522120e-01 -2.59550750e-01
1.56075537e-01 3.86549771e-01 -2.77852267e-01 -1.73160061e-01
9.84787643e-02 -1.89236060e-01 -3.75267118e-01 -7.91368484e-01
-1.81936601e-03 1.08659469e-01 -2.05882400e-01 5.94257593e-01
5.88120162e-01 8.22046697e-01 -1.27402043e+00 -6.96854293e-01
2.27706984e-01 1.55138850e-01 -4.69206750e-01 3.75722408e-01
-4.88616586e-01 1.20889097e-01 -2.57249385e-01 9.03128266e-01
7.67859340e-01 -5.96884370e-01 -2.26809084e-01 -4.80152905e-01
-1.72339063e-02 1.74161419e-01 -1.10944641e+00 1.76442492e+00
-3.21697444e-01 3.94709110e-01 -6.62737712e-02 -1.22995055e+00
5.24590135e-01 -8.83748308e-02 4.35038865e-01 -5.02019823e-01
-1.97446927e-01 1.96921438e-01 -2.41431355e-01 -4.43856269e-01
2.31388673e-01 1.22465841e-01 1.52786359e-01 4.58260216e-02
3.67566079e-01 2.40561757e-02 3.39265168e-01 3.31879646e-01
1.11022115e+00 2.77506053e-01 1.67507902e-01 -1.92629144e-01
3.59400898e-01 6.05132878e-02 8.11351001e-01 9.87922549e-01
-4.35338169e-01 7.41927385e-01 2.66574085e-01 -2.90291965e-01
-9.04331148e-01 -1.19797218e+00 -5.87338507e-01 1.27213717e+00
3.18404883e-01 -3.45850289e-01 -6.16441607e-01 -1.21086311e+00
1.51476815e-01 5.57726324e-01 -7.99286306e-01 -2.92973500e-02
-3.12605947e-01 -8.38246346e-01 3.14509988e-01 5.97224295e-01
5.71542561e-01 -1.00742388e+00 -2.79157937e-01 5.32915071e-02
-1.67079002e-01 -1.61893809e+00 -5.24948359e-01 4.05176908e-01
-5.24464846e-01 -1.20397830e+00 -7.43788183e-01 -9.18997586e-01
9.69669104e-01 4.10717815e-01 1.33045042e+00 -5.99802546e-02
-6.77824974e-01 1.91228792e-01 -3.37179571e-01 -7.35580206e-01
1.54869527e-01 -8.60482454e-02 -2.73469165e-02 2.34679699e-01
4.88545537e-01 8.02327991e-02 -7.31907248e-01 3.85712326e-01
-7.12888420e-01 -9.40091461e-02 7.84993708e-01 8.75160754e-01
1.02562368e+00 -6.18760735e-02 7.21050084e-01 -1.15350246e+00
-3.39087695e-01 -3.90591651e-01 -6.15403295e-01 2.49319613e-01
-5.02281308e-01 -1.45789132e-01 5.08230031e-01 -4.45276111e-01
-1.00061250e+00 4.06909376e-01 1.04090743e-01 -7.42777646e-01
-2.50818431e-01 7.92338327e-02 -3.23705375e-01 -8.27687047e-03
6.21176481e-01 4.17553663e-01 -3.61632466e-01 -4.90606070e-01
2.92092979e-01 5.55463016e-01 6.74653828e-01 -5.76210558e-01
7.09762037e-01 8.50120366e-01 -1.78724930e-01 -6.25354648e-01
-1.75348139e+00 -9.96962845e-01 -5.49521148e-01 -1.01155400e-01
8.26834977e-01 -1.30846274e+00 -5.71090996e-01 4.23368722e-01
-1.09253478e+00 -2.76169807e-01 -6.25994027e-01 5.42910278e-01
-4.48493928e-01 1.13282099e-01 -3.00156057e-01 -8.00216675e-01
-6.62086457e-02 -1.08080554e+00 1.33931732e+00 -7.40805343e-02
8.92007649e-02 -4.83913690e-01 -1.94833785e-01 6.05395496e-01
7.75550976e-02 2.75437653e-01 3.05972964e-01 -5.85572064e-01
-8.35951746e-01 -5.83022237e-01 -5.26907742e-01 8.72545779e-01
-4.35603261e-02 -1.78090766e-01 -1.13773525e+00 -3.52086514e-01
-2.41119087e-01 -8.16749036e-01 1.28522313e+00 2.86010236e-01
1.67359686e+00 -3.03360075e-01 -3.99944633e-01 5.66995263e-01
1.44705737e+00 -4.54948783e-01 4.01626348e-01 2.24300399e-01
9.62084472e-01 6.47332788e-01 9.74592686e-01 2.97470063e-01
3.81014943e-01 7.65868127e-01 7.67575920e-01 -2.07154065e-01
-4.24226761e-01 -2.73495436e-01 2.28558734e-01 2.28584230e-01
2.81280428e-02 -3.90640125e-02 -5.58032572e-01 9.44145858e-01
-1.80665123e+00 -9.18203950e-01 -1.33338377e-01 2.38106227e+00
1.15747857e+00 4.14150208e-01 1.18137889e-01 7.58157000e-02
9.27484512e-01 1.66307852e-01 -6.92635655e-01 4.63446826e-01
-2.16408163e-01 3.16953994e-02 6.57486200e-01 1.04736462e-01
-1.45207524e+00 8.98555875e-01 5.69117165e+00 1.15599453e+00
-8.81210148e-01 4.04756099e-01 7.39418685e-01 -3.59254897e-01
2.71316953e-02 -1.42796353e-01 -1.26255071e+00 6.16281569e-01
3.92412275e-01 2.22847402e-01 -1.32335365e-01 1.30789328e+00
-9.13850497e-03 -1.74641907e-01 -1.38897252e+00 1.01006269e+00
2.99626827e-01 -1.39844036e+00 7.58524612e-02 -1.48395672e-01
1.02205336e+00 3.12280536e-01 9.45506096e-02 3.25571418e-01
1.35590360e-01 -8.26370299e-01 9.00043190e-01 2.07636371e-01
8.23458016e-01 -4.53706175e-01 7.55957842e-01 3.71582538e-01
-1.17446184e+00 1.10855922e-02 -4.84387130e-01 2.97944129e-01
-1.13599062e-01 9.16660666e-01 -7.28410125e-01 2.65565336e-01
9.61856782e-01 1.01281428e+00 -7.43864238e-01 1.37748885e+00
-4.47134256e-01 8.31796825e-01 -3.42414588e-01 4.91769984e-02
2.55087435e-01 3.94432157e-01 5.02964377e-01 1.38810003e+00
-1.82165802e-01 -2.51001835e-01 4.71261680e-01 1.02768719e+00
-5.37261665e-01 -1.11604847e-01 -2.59950757e-01 3.79294127e-01
4.07944590e-01 1.36698842e+00 -6.79863572e-01 -5.61204791e-01
-4.74546701e-01 1.01767802e+00 5.44990838e-01 2.68635899e-01
-8.47379029e-01 -2.19146535e-01 3.87900174e-01 2.91815847e-01
5.11670947e-01 1.71756014e-01 -2.92783052e-01 -1.15993977e+00
3.73839170e-01 -5.46644151e-01 5.16704977e-01 -6.62037671e-01
-1.58024907e+00 2.46326074e-01 -1.30526721e-02 -1.40373814e+00
2.11476371e-01 -6.14958525e-01 -5.78166306e-01 5.14335692e-01
-1.97532022e+00 -1.31171346e+00 -3.69059175e-01 3.93834829e-01
7.36422658e-01 3.01318336e-02 5.54499567e-01 4.59867895e-01
-5.00282943e-01 8.22216332e-01 -8.84667784e-02 2.51648098e-01
9.20476377e-01 -1.57037961e+00 -7.82822743e-02 7.19965160e-01
3.01420033e-01 3.02496195e-01 6.38671219e-01 -3.82760048e-01
-1.02856398e+00 -1.66596723e+00 7.24087119e-01 -6.59555972e-01
3.47611845e-01 -5.30738413e-01 -9.01287198e-01 5.46088338e-01
-9.52327996e-02 1.06501532e+00 5.03887415e-01 8.53983834e-02
-6.82001293e-01 -4.05285835e-01 -1.21418571e+00 1.84171379e-01
1.13229883e+00 -4.52350050e-01 -3.79427880e-01 9.20233130e-01
8.69803011e-01 -2.25121021e-01 -6.02347672e-01 7.09005058e-01
3.04951489e-01 -6.95644677e-01 1.11886990e+00 -6.21435106e-01
4.34460104e-01 -5.89563668e-01 -1.76483780e-01 -8.64827335e-01
-1.27828792e-01 -1.99094608e-01 -4.01065618e-01 1.17377090e+00
4.82392490e-01 -1.91075996e-01 1.03490269e+00 1.68917477e-01
-2.73497194e-01 -9.74295735e-01 -7.99044251e-01 -1.22260904e+00
-2.89865822e-01 -5.11497855e-01 3.43203545e-01 9.65301216e-01
-3.97781104e-01 1.85982421e-01 -3.04477632e-01 2.95564771e-01
1.21177256e+00 8.97821859e-02 8.13952625e-01 -1.05644488e+00
-3.56865287e-01 -2.03248754e-01 -6.88196898e-01 -1.00640702e+00
2.47758701e-01 -8.69769037e-01 4.77656543e-01 -1.37619936e+00
8.20563376e-01 -5.34324586e-01 -6.35123610e-01 6.03664637e-01
-4.86990124e-01 7.40527272e-01 4.97106798e-02 3.40630531e-01
-1.41356015e+00 7.02030182e-01 1.14749575e+00 -2.47275382e-01
2.68734634e-01 1.43263415e-01 -4.72193241e-01 8.66297245e-01
4.09088999e-01 -7.62229860e-01 -1.41436040e-01 -2.95449108e-01
-1.29046589e-01 -5.43624163e-01 6.51836038e-01 -9.86033440e-01
1.49833679e-01 -1.04403161e-01 4.37209040e-01 -8.96838665e-01
2.69681185e-01 -6.68449223e-01 -5.47779262e-01 3.22019607e-01
-5.72145879e-01 -5.70426226e-01 -4.79645357e-02 9.03337777e-01
-3.44889164e-01 -2.83928573e-01 1.04807758e+00 -1.20453641e-01
-9.03428793e-01 6.16345406e-01 1.88328519e-01 3.18439007e-01
1.25603032e+00 -6.25117049e-02 -3.07886600e-01 1.85581371e-01
-6.06786847e-01 4.84825730e-01 3.01660776e-01 3.31941426e-01
5.17917633e-01 -1.42208862e+00 -9.57827687e-01 -8.96667391e-02
8.63481104e-01 5.64696312e-01 8.26932043e-02 6.29010141e-01
-1.27089933e-01 2.42343694e-01 3.13890517e-01 -1.02750778e+00
-1.23795962e+00 4.90900844e-01 1.10244542e-01 -2.06074029e-01
-4.71125454e-01 1.21997654e+00 4.48911309e-01 -4.36507910e-01
6.60594165e-01 -4.32158895e-02 -2.45972760e-02 -1.63019568e-01
7.05637634e-01 1.73550576e-01 9.29078236e-02 -4.66705471e-01
-5.22830367e-01 2.21085221e-01 -3.55504304e-01 2.75985181e-01
1.29993761e+00 1.66877747e-01 2.10851599e-02 5.75005651e-01
1.37574279e+00 -2.38073200e-01 -1.69466496e+00 -6.73259556e-01
7.95318410e-02 -6.90831721e-01 2.91262381e-02 -8.56570959e-01
-1.00253189e+00 7.69177079e-01 7.14631200e-01 -1.16296194e-01
7.89385974e-01 4.30320114e-01 6.19158685e-01 6.15324140e-01
3.08083057e-01 -1.23400342e+00 4.51783717e-01 2.52512872e-01
6.80711627e-01 -1.83419788e+00 8.68649036e-02 -6.46394610e-01
-5.32937467e-01 5.33592284e-01 1.01371455e+00 -2.37128228e-01
5.11321247e-01 1.78038985e-01 -2.37156600e-01 -8.23638588e-02
-6.93235874e-01 -5.65092206e-01 4.72670346e-01 4.65739489e-01
2.89884806e-01 2.80370787e-02 -1.21027201e-01 5.56090176e-01
4.06413496e-01 -9.41685867e-03 2.33548328e-01 8.96302521e-01
-5.96814573e-01 -6.77887619e-01 -1.79059535e-01 8.12218785e-01
-6.25556946e-01 -1.26794010e-01 -8.32917690e-02 5.19438922e-01
4.69463170e-01 7.43686616e-01 1.28319517e-01 -6.72171861e-02
2.40211621e-01 -3.43069315e-01 6.39129996e-01 -1.09706509e+00
-1.75512269e-01 -2.41762832e-01 -3.65724750e-02 -6.50653481e-01
-6.61837220e-01 -6.44382596e-01 -1.15980494e+00 2.88153976e-01
-7.29739785e-01 6.79940656e-02 6.41490281e-01 9.28774953e-01
1.65198267e-01 4.69654351e-01 6.94099605e-01 -8.96605611e-01
-7.40092337e-01 -9.43484128e-01 -6.19581163e-01 7.36328304e-01
4.55458432e-01 -7.84532905e-01 -4.99290347e-01 1.63485318e-01] | [9.295049667358398, 1.273382544517517] |
5edac015-012b-4ba9-a3ee-47fe38236f77 | binary-classification-of-proteins-by-a | 2111.01975 | null | https://arxiv.org/abs/2111.01975v1 | https://arxiv.org/pdf/2111.01975v1.pdf | Binary classification of proteins by a Machine Learning approach | In this work we present a system based on a Deep Learning approach, by using a Convolutional Neural Network, capable of classifying protein chains of amino acids based on the protein description contained in the Protein Data Bank. Each protein is fully described in its chemical-physical-geometric properties in a file in XML format. The aim of the work is to design a prototypical Deep Learning machinery for the collection and management of a huge amount of data and to validate it through its application to the classification of a sequences of amino acids. We envisage applying the described approach to more general classification problems in biomolecules, related to structural properties and similarities. | ['Osvaldo Gervasi', 'Noelia Faginas-Lago', 'Andrea Lombardi', 'Marco Simonetti', 'Damiano Perri'] | 2021-11-03 | null | null | null | null | ['classification'] | ['methodology'] | [ 3.37357596e-02 -2.93125473e-02 2.74599083e-02 -6.67525828e-01
-8.40182304e-02 -4.14923638e-01 4.63019222e-01 7.96197474e-01
-6.09624565e-01 1.04771245e+00 -4.02848274e-02 -6.26610816e-01
-3.19614649e-01 -8.30480039e-01 -9.12395179e-01 -7.75166690e-01
-3.17720175e-01 8.08651268e-01 1.70593970e-02 -3.22696209e-01
1.95267111e-01 1.17012227e+00 -1.63021135e+00 8.02858949e-01
3.68260980e-01 1.06086254e+00 4.44777310e-01 5.68542123e-01
-4.92716640e-01 7.71238625e-01 -4.58657950e-01 -1.54891163e-01
-1.04255110e-01 -1.29646391e-01 -1.12198758e+00 -2.17200935e-01
4.66953039e-01 -2.66467594e-02 2.32430488e-01 7.45503366e-01
2.49530524e-01 -1.90866411e-01 8.12407136e-01 -5.14150083e-01
-4.38903302e-01 4.19750243e-01 5.92991412e-01 -2.50201039e-02
5.69769204e-01 -5.25708834e-04 7.77982235e-01 -7.39412844e-01
9.57526326e-01 9.67451453e-01 7.73085713e-01 4.17640567e-01
-1.25012410e+00 -1.38422102e-01 -4.43317384e-01 4.56592649e-01
-9.94623125e-01 -4.22326848e-02 1.45206124e-01 -7.65552580e-01
1.67087722e+00 -3.59457247e-02 6.71064138e-01 1.02463830e+00
5.64907432e-01 3.16422492e-01 7.40607440e-01 -4.97969270e-01
5.58016419e-01 -1.86347976e-01 5.00063360e-01 5.99282682e-01
2.44292110e-01 2.89967712e-02 -2.13200286e-01 -2.17554495e-01
2.63238847e-01 1.20866299e-01 1.93813205e-01 -7.28011131e-01
-1.06614888e+00 8.13960791e-01 5.31570792e-01 7.04404712e-01
-6.14420772e-01 -1.79170445e-02 8.98688138e-01 4.54548627e-01
2.16637373e-01 5.47562957e-01 -1.01726627e+00 -6.31612912e-02
-6.17988825e-01 5.84550202e-01 1.34668291e+00 7.52843082e-01
9.25019622e-01 -4.34287786e-01 3.98814559e-01 2.88135439e-01
3.28658819e-01 -5.21897823e-02 6.41931832e-01 -5.42295218e-01
-5.90439029e-02 7.27662444e-01 7.82916322e-03 -4.46569413e-01
-9.24408913e-01 1.99198369e-02 -6.98021889e-01 6.07123792e-01
4.86792415e-01 1.12935258e-02 -7.10577250e-01 1.37101209e+00
2.42361113e-01 -3.12948465e-01 5.06725788e-01 8.73483956e-01
9.98831093e-01 4.50661451e-01 4.77901071e-01 -4.41177329e-03
1.38401544e+00 -4.05729532e-01 -4.95454431e-01 7.18369484e-01
1.04117346e+00 -6.78904176e-01 4.44107860e-01 8.09276044e-01
-7.54488826e-01 -9.04927254e-01 -1.30081391e+00 -2.77044207e-01
-1.11051714e+00 2.16686219e-01 9.01093960e-01 3.15272242e-01
-7.42832422e-01 1.29802215e+00 -6.92944884e-01 -5.26916385e-01
2.71917611e-01 5.96623123e-01 -9.41042185e-01 3.38288903e-01
-1.24937737e+00 1.11836779e+00 1.21551394e+00 -1.52553752e-01
-4.17870581e-01 -5.20609677e-01 -7.66343236e-01 3.84244114e-01
-8.70329067e-02 -5.73379517e-01 1.13915610e+00 -1.12440073e+00
-1.22583032e+00 1.09704518e+00 3.01437080e-01 -8.95720363e-01
2.15513930e-01 -6.84140101e-02 -4.59337413e-01 2.84243710e-02
-3.93005997e-01 5.10827839e-01 2.03642562e-01 -6.98065042e-01
-5.35498500e-01 -4.83338058e-01 1.46136269e-01 -4.33588117e-01
-6.07843027e-02 5.21584824e-02 3.00988466e-01 -2.87937552e-01
-2.32414365e-01 -5.83131790e-01 -3.49732131e-01 -2.09260479e-01
-1.80406749e-01 -5.75568497e-01 5.17822504e-01 -4.82186437e-01
5.81324041e-01 -1.63054919e+00 3.69044930e-01 2.27655470e-01
3.57902944e-01 7.19906867e-01 1.47828862e-01 9.84721303e-01
-6.33149981e-01 -2.07956627e-01 -1.39611006e-01 3.05617809e-01
-7.94380158e-02 3.61314386e-01 -1.40239969e-01 5.59807241e-01
2.04815015e-01 7.08982229e-01 -4.48216379e-01 -2.61693150e-01
4.43686604e-01 4.98702049e-01 -2.09597334e-01 3.80295485e-01
-7.40478933e-01 7.29268640e-02 -5.95773816e-01 4.38374102e-01
8.16615283e-01 -3.40670139e-01 6.01243973e-01 -4.62120444e-01
-5.10227978e-01 3.07758719e-01 -5.52995324e-01 1.70242071e+00
-3.11303772e-02 4.52469766e-01 -2.61837065e-01 -1.48503077e+00
1.19268548e+00 3.71501207e-01 8.87909412e-01 -2.63167799e-01
2.48994455e-01 2.58785337e-01 2.65910059e-01 -6.97976530e-01
1.03723794e-01 1.68602280e-02 3.23564827e-01 5.55614471e-01
5.17856658e-01 5.88724792e-01 5.84189892e-01 -1.58512503e-01
1.07504022e+00 6.59648776e-01 5.70561111e-01 -5.41720867e-01
1.04507363e+00 2.55106837e-01 -8.17953721e-02 2.86676913e-01
2.62654871e-01 1.73916787e-01 5.25225699e-01 -1.52982581e+00
-1.75748765e+00 -5.76422334e-01 -5.69094300e-01 1.05818164e+00
-6.01351142e-01 -4.81532842e-01 -9.95416820e-01 -4.72604334e-01
1.78404927e-01 2.79821157e-02 -6.00725651e-01 -3.34736542e-03
-4.75768238e-01 -7.30937183e-01 2.35335022e-01 2.44848520e-01
9.36393142e-02 -1.63347566e+00 -9.14161086e-01 6.40991747e-01
5.97394943e-01 -7.74613976e-01 5.59016228e-01 9.04372931e-01
-6.62368774e-01 -1.28651345e+00 -6.15018487e-01 -8.76894295e-01
1.49542719e-01 -5.28200150e-01 1.12938344e+00 1.35454893e-01
-6.25925362e-01 -3.77719730e-01 -5.30790269e-01 -5.70277274e-01
-9.47348356e-01 5.13937891e-01 7.04484358e-02 -1.22858360e-01
8.98365140e-01 -5.86078942e-01 -4.40795124e-01 -1.21856354e-01
-1.23073697e+00 -5.58253862e-02 6.98050082e-01 6.00744426e-01
4.93706733e-01 -3.84749293e-01 3.60883772e-01 -1.15932512e+00
4.40861315e-01 -4.89039391e-01 -6.75929129e-01 2.58005053e-01
-6.91207170e-01 5.88180542e-01 1.01480412e+00 -1.82448223e-01
-3.29261303e-01 7.55300283e-01 -8.20023715e-01 1.24932073e-01
-8.82658958e-01 6.67224050e-01 -3.74699444e-01 -2.57684797e-01
6.11831188e-01 2.58801728e-01 2.29396552e-01 -9.70027745e-01
3.24490219e-01 7.17935383e-01 5.15402675e-01 -4.20265496e-01
2.27552354e-02 1.39135793e-01 4.58390117e-01 -6.09000862e-01
-2.81993359e-01 -5.16301095e-01 -1.13983929e+00 1.41328350e-01
9.37699080e-01 -4.80196744e-01 -1.28168488e+00 3.91881406e-01
-1.34281361e+00 1.44242076e-02 3.32459174e-02 5.73494256e-01
-1.06116903e+00 6.90116346e-01 -5.21692455e-01 -1.02961451e-01
-5.70687711e-01 -1.08035731e+00 9.74100173e-01 -2.53925294e-01
-1.86288372e-01 -9.82748151e-01 4.52674031e-01 -1.71709955e-02
1.46210954e-01 1.46418571e-01 1.24798155e+00 -1.40590441e+00
-3.46688867e-01 -1.14769883e-01 -9.31542218e-02 4.48942572e-01
2.79453527e-02 1.11135490e-01 -7.32399821e-01 -1.63232967e-01
-2.59378284e-01 -5.42234540e-01 9.09718096e-01 6.64016083e-02
1.34457707e+00 -7.14948680e-03 -2.51103312e-01 7.42547452e-01
1.73756754e+00 3.71124893e-01 5.57143390e-01 6.72724962e-01
4.14755821e-01 6.74895287e-01 4.01691347e-01 4.69137698e-01
-4.03576523e-01 8.07347715e-01 6.51993036e-01 -2.24390700e-01
2.65564740e-01 3.12425017e-01 -9.95143950e-02 1.41949296e-01
-2.85565138e-01 -3.20048928e-01 -1.05377793e+00 -6.25830069e-02
-1.89938664e+00 -8.82132828e-01 -3.40072632e-01 1.83949029e+00
7.58707464e-01 -1.39504420e-02 1.39695957e-01 1.15320661e-04
5.14642656e-01 -3.84992152e-01 -4.57495332e-01 -8.36448371e-01
-9.37995315e-02 5.38503468e-01 5.92072785e-01 1.88599408e-01
-1.16545856e+00 6.41855478e-01 6.96651411e+00 4.02803302e-01
-1.21030235e+00 -4.23692703e-01 2.57295996e-01 4.29745227e-01
3.10307533e-01 -2.75753945e-01 -8.96055222e-01 4.73963588e-01
1.74842000e+00 6.32680207e-02 1.36653066e-01 9.70598221e-01
5.61259910e-02 1.34881780e-01 -1.34401643e+00 6.08307481e-01
-1.44357577e-01 -2.00451469e+00 3.49024326e-01 1.31654248e-01
-1.14146613e-01 2.19137490e-01 -2.59903878e-01 -4.46565524e-02
1.06038600e-01 -1.28805876e+00 4.08668488e-01 7.95003593e-01
5.55650413e-01 -7.63503909e-01 9.91768777e-01 2.06738353e-01
-6.80203319e-01 2.05916181e-01 -6.08443201e-01 -6.10335879e-02
-2.57520646e-01 3.01802993e-01 -9.11200225e-01 6.50553048e-01
6.45650566e-01 8.10231507e-01 -2.01802298e-01 7.82535017e-01
3.44562203e-01 2.31983691e-01 -3.86857390e-02 -3.26867521e-01
4.34762836e-01 -2.03748956e-01 -1.13733053e-01 1.53525162e+00
1.98481255e-03 -3.06278914e-02 2.29281723e-01 8.23207855e-01
1.23486273e-01 6.08513057e-01 -4.42311823e-01 -2.47478530e-01
-1.98352039e-01 1.20024788e+00 -4.87144887e-01 -7.25375056e-01
-4.87741441e-01 5.43957591e-01 4.45794970e-01 -4.41983975e-02
-4.87132013e-01 -6.09105408e-01 7.29552388e-01 -3.69208977e-02
5.93731165e-01 -1.67943880e-01 3.50873053e-01 -8.82269084e-01
-2.06536308e-01 -8.63305449e-01 1.57826424e-01 -9.41472530e-01
-1.32445419e+00 9.06782687e-01 -4.74921316e-02 -8.79004598e-01
-4.88860339e-01 -1.64503121e+00 -8.02090485e-03 1.29695737e+00
-1.29473376e+00 -1.01852763e+00 -4.13576625e-02 4.42589462e-01
1.81318685e-01 -8.11046243e-01 1.36211443e+00 2.47904584e-01
-9.30420905e-02 -8.38643312e-03 6.36424601e-01 -4.85392474e-02
6.27058983e-01 -1.54190302e+00 6.57273829e-01 2.11201813e-02
-8.20106566e-02 7.62711942e-01 9.74000752e-01 -3.87146145e-01
-1.28383636e+00 -9.56380129e-01 1.17128968e+00 -1.89232677e-01
6.62024915e-01 -4.54056144e-01 -1.26557934e+00 4.07023728e-01
2.98688948e-01 -1.21253192e-01 1.00386429e+00 -1.18456937e-01
-3.22529972e-01 1.53407633e-01 -1.04782748e+00 -1.94575757e-01
5.30050039e-01 -3.74140203e-01 -8.11403632e-01 7.76284695e-01
5.56842864e-01 -2.11361408e-01 -1.37431991e+00 3.55015367e-01
6.73696339e-01 -1.07549012e+00 9.97043788e-01 -1.47929132e+00
5.22030950e-01 -4.28674705e-02 4.92642168e-03 -9.70906496e-01
-4.35647577e-01 -2.35526651e-01 8.00219085e-03 3.59660774e-01
3.26846749e-01 -3.44299614e-01 9.03798103e-01 7.85441399e-02
-3.20135355e-01 -6.56550050e-01 -9.39483643e-01 -6.22659504e-01
2.13457614e-01 -1.50591731e-01 8.40659976e-01 6.58788085e-01
-5.36571164e-03 2.75313973e-01 -1.64058670e-01 -1.08508810e-01
1.08058646e-01 3.73889476e-01 5.44699550e-01 -1.67078483e+00
-1.81390643e-01 -2.28711888e-01 -9.42195594e-01 -6.00476563e-01
4.18559402e-01 -9.02570784e-01 -1.97626159e-01 -1.32744122e+00
7.92730004e-02 -8.53783488e-02 -6.05379164e-01 4.45990384e-01
6.02614403e-01 -9.87308919e-02 -3.13470930e-01 2.65168428e-01
-6.10682428e-01 7.59587740e-04 5.78933835e-01 -1.66592613e-01
3.12779635e-01 -7.57480264e-02 -4.37619574e-02 3.73539448e-01
8.26463461e-01 -2.53905267e-01 1.27866089e-01 2.08319649e-01
3.88407260e-01 -2.45749250e-01 1.69756562e-01 -1.32345700e+00
-4.90858741e-02 -1.40009999e-01 6.55325234e-01 -8.31559300e-01
7.74089396e-02 -1.16953051e+00 4.75133210e-01 8.96492124e-01
-6.93626940e-01 1.12387344e-01 6.09873794e-02 2.56021410e-01
-2.11355880e-01 -5.72115302e-01 7.54300535e-01 -3.15137923e-01
-8.90956521e-01 3.56767535e-01 -7.43446052e-01 -8.73538435e-01
9.94787395e-01 -2.18098775e-01 -5.29515259e-02 3.53152245e-01
-1.18815196e+00 -3.95000309e-01 6.86494052e-01 2.13890314e-01
4.33446884e-01 -1.03136003e+00 -5.28055251e-01 5.01777828e-01
5.74143946e-01 -6.55940413e-01 5.59095629e-02 1.54032633e-01
-1.49392295e+00 1.03557527e+00 -1.03247428e+00 -3.85694027e-01
-1.62507117e+00 1.22379231e+00 7.02952683e-01 -1.49575666e-01
-6.03491366e-01 2.28232086e-01 -9.76278186e-02 -3.82005602e-01
2.02376932e-01 -5.64122081e-01 -5.82298338e-01 -1.61484659e-01
7.08391070e-01 -2.29060799e-01 6.39038026e-01 -6.48383379e-01
-2.97559172e-01 3.48278672e-01 -1.14699453e-01 6.03401601e-01
1.66902602e+00 4.01143253e-01 -3.79380018e-01 2.37553939e-01
1.50652313e+00 -5.79016507e-01 -9.41231787e-01 -5.63257672e-02
4.13795203e-01 1.06063060e-01 -2.68878371e-01 -8.41121912e-01
-5.50565720e-01 8.03238690e-01 9.80382979e-01 3.28191489e-01
6.15763843e-01 2.23691706e-02 6.06989801e-01 1.27033806e+00
2.44642958e-01 -8.03202152e-01 -6.37065828e-01 8.30262780e-01
5.37714422e-01 -1.23492730e+00 1.24926724e-01 -6.41690120e-02
-5.52397110e-02 2.10757995e+00 7.42404684e-02 -2.76918523e-02
5.77097952e-01 2.99980104e-01 1.37189263e-02 -5.36059618e-01
-9.95416820e-01 -3.12667638e-01 1.64545551e-01 7.75304496e-01
1.02371132e+00 9.29048809e-04 -8.77954841e-01 4.31116283e-01
5.15997782e-02 4.48675781e-01 2.43863449e-01 8.32492888e-01
-8.89140248e-01 -1.61612129e+00 -8.49351436e-02 1.40355140e-01
-6.24924183e-01 -1.20646022e-01 -7.37241864e-01 7.97524035e-01
6.91127896e-01 2.84300983e-01 8.68850574e-02 -1.84112087e-01
2.69634396e-01 3.61549795e-01 4.31186199e-01 -5.91285408e-01
-8.01596999e-01 -2.90212452e-01 1.25214070e-01 -5.24717927e-01
-7.91141331e-01 -3.59356552e-01 -1.44639242e+00 -2.48546094e-01
2.85386592e-01 3.90195996e-01 1.17888308e+00 9.76569176e-01
3.01216125e-01 3.33386153e-01 2.59520471e-01 -7.51669347e-01
-6.02305233e-01 -1.09298158e+00 -6.65614665e-01 8.09032500e-01
3.66694748e-01 -4.68131840e-01 2.72468567e-01 2.54022181e-01] | [4.72590446472168, 5.595959663391113] |
91cd8822-4ce9-4b8e-bb69-9ae7afc76a80 | give-me-more-feedback-ii-annotating-thesis | null | null | https://aclanthology.org/P19-1390 | https://aclanthology.org/P19-1390.pdf | Give Me More Feedback II: Annotating Thesis Strength and Related Attributes in Student Essays | While the vast majority of existing work on automated essay scoring has focused on holistic scoring, researchers have recently begun work on scoring specific dimensions of essay quality. Nevertheless, progress on dimension-specific essay scoring is limited in part by the lack of annotated corpora. To facilitate advances in this area, we design a scoring rubric for scoring a core, yet unexplored dimension of persuasive essay quality, thesis strength, and annotate a corpus of essays with thesis strength scores. We additionally identify the attributes that could impact thesis strength and annotate the essays with the values of these attributes, which, when predicted by computational models, could provide further feedback to students on why her essay receives a particular thesis strength score. | ['Vincent Ng', 'Zixuan Ke', 'Hui Lin', 'Hrishikesh Inamdar'] | 2019-07-01 | null | null | null | acl-2019-7 | ['automated-essay-scoring'] | ['natural-language-processing'] | [-6.77982420e-02 4.06699777e-01 -6.41349137e-01 -3.99687052e-01
-9.89233017e-01 -7.08985865e-01 2.74119735e-01 5.93113005e-01
-1.89379379e-01 9.69294667e-01 8.30344617e-01 -7.23045051e-01
-5.58562040e-01 -7.20521152e-01 1.73596352e-01 -2.91177407e-02
7.27886438e-01 4.10974622e-01 -2.61103719e-01 -2.03903198e-01
1.21002150e+00 1.42810047e-01 -1.28023767e+00 2.11676598e-01
1.13845766e+00 2.36576051e-01 -4.56472747e-02 8.04839969e-01
-2.22267091e-01 1.00221789e+00 -1.14707553e+00 -8.68677914e-01
-2.61379331e-01 -8.83253396e-01 -1.00680029e+00 -8.24377835e-02
4.74891454e-01 -3.53105724e-01 5.96081764e-02 7.94128537e-01
2.04005525e-01 1.65182456e-01 8.73533010e-01 -8.87515187e-01
-9.58181143e-01 7.97776878e-01 -2.88817883e-01 3.86432052e-01
8.25910389e-01 -4.62934375e-02 1.56960523e+00 -5.82349479e-01
6.81933820e-01 7.35160530e-01 6.81026995e-01 5.60052454e-01
-1.15307665e+00 -6.30608618e-01 -1.45463675e-01 1.63436145e-01
-4.96507764e-01 -3.11551690e-01 9.20679927e-01 -5.99022567e-01
3.97822052e-01 2.19891176e-01 9.09528196e-01 7.85658717e-01
1.87017918e-01 7.31026530e-01 1.47606385e+00 -5.10287583e-01
1.66611075e-01 1.70795962e-01 7.38231838e-01 5.93578637e-01
4.77312356e-01 -3.71925145e-01 -8.10678244e-01 -1.59095764e-01
5.11083364e-01 -4.45005953e-01 5.96365519e-02 9.50115398e-02
-1.07578135e+00 9.35806692e-01 -1.48423567e-01 3.12715620e-01
-6.68048039e-02 -1.85523406e-01 2.38289386e-01 6.30679369e-01
3.07350844e-01 1.10817480e+00 -4.32489634e-01 -9.03546453e-01
-1.42694318e+00 6.87624931e-01 1.16348839e+00 2.31601074e-01
5.46680033e-01 -1.58443630e-01 -3.08168948e-01 9.02109444e-01
3.33613962e-01 2.01076433e-01 4.63104993e-01 -1.32165444e+00
3.88112187e-01 1.06140959e+00 3.16346109e-01 -9.65887725e-01
-4.62288231e-01 -4.45183963e-01 1.06907509e-01 3.19095522e-01
5.88069677e-01 -1.99681103e-01 -1.02709807e-01 1.32960427e+00
-1.38167128e-01 -5.76347888e-01 -2.31971275e-02 9.13927734e-01
9.74423826e-01 1.92324311e-01 1.92523718e-01 -1.07437424e-01
1.41363442e+00 -5.44438481e-01 -7.25798309e-01 -1.63224757e-01
9.40033913e-01 -9.84189570e-01 1.46957469e+00 5.55416405e-01
-1.59144032e+00 -3.58427048e-01 -1.26175642e+00 -1.35527685e-01
2.85476483e-02 2.09674418e-01 6.42851830e-01 1.10610282e+00
-8.05802822e-01 5.26897848e-01 -4.41499591e-01 -8.50592032e-02
2.56506950e-01 6.77991807e-02 3.13400291e-02 1.97658747e-01
-1.24764919e+00 1.18887377e+00 -1.48435652e-01 -6.46949351e-01
-7.88912922e-02 -9.03333426e-01 -5.31315148e-01 2.25994751e-01
-3.41608468e-03 -7.02010453e-01 1.41739786e+00 -6.17783189e-01
-1.50358629e+00 9.07259107e-01 -1.64184216e-02 2.40395382e-01
1.38542831e-01 2.86317430e-02 -1.52586386e-01 2.44838800e-02
3.55912775e-01 2.88202614e-01 5.37524186e-02 -7.02199876e-01
-6.80632591e-01 -4.91698414e-01 4.43599075e-01 5.70131838e-01
-1.08108377e+00 3.25712204e-01 2.36793786e-01 -4.89502072e-01
5.21898270e-04 -5.58373868e-01 1.53956354e-01 -3.89627934e-01
1.89549886e-02 -7.70788789e-01 4.59169716e-01 -5.48498690e-01
1.59402466e+00 -1.59624815e+00 -1.63589679e-02 5.82189895e-02
2.24589765e-01 -4.52601258e-03 -5.89316674e-02 5.28312147e-01
4.42108780e-01 3.91531795e-01 2.62596965e-01 -5.69585972e-02
3.03931773e-01 -2.17657983e-01 9.22243856e-03 1.31530672e-01
-5.96799515e-02 6.85101748e-01 -1.13619661e+00 -7.90448308e-01
-1.35354891e-01 -1.39505327e-01 -5.92467189e-01 6.14373609e-02
7.70009458e-02 -1.00813083e-01 -6.30603135e-01 7.55226016e-01
1.25881180e-01 -2.49492437e-01 2.46598884e-01 5.37452996e-01
-3.17168385e-01 1.05320084e+00 -8.39226484e-01 1.19395483e+00
-2.53705263e-01 8.26723576e-01 -1.57437041e-01 -5.35143614e-01
1.24706316e+00 2.89000273e-01 3.25510859e-01 -5.41938782e-01
-1.03154019e-01 3.28381270e-01 1.79134399e-01 -4.77402329e-01
1.08519053e+00 -3.83177429e-01 -2.66600966e-01 1.44348085e+00
-2.29120702e-01 -4.70061421e-01 6.03599131e-01 5.09717286e-01
1.39940310e+00 1.05625406e-01 4.55649376e-01 -3.02129716e-01
4.16683197e-01 2.62437880e-01 2.51719356e-01 6.60127640e-01
-4.84465390e-01 4.42477047e-01 7.63565838e-01 -2.05613285e-01
-1.22111690e+00 -7.48966753e-01 -2.88919061e-01 1.36778200e+00
-2.38277465e-01 -6.50382698e-01 -6.25054061e-01 -6.69088900e-01
-1.87474694e-02 7.95879662e-01 -5.53395271e-01 1.49498552e-01
-4.11346287e-01 -5.98596454e-01 6.96849704e-01 5.51621854e-01
1.31262958e-01 -1.10554886e+00 -7.46095359e-01 3.07567954e-01
-3.76185894e-01 -5.16130030e-01 -1.24017306e-01 1.33450404e-01
-9.02645350e-01 -1.35490167e+00 -4.58237767e-01 -4.96949315e-01
3.46825331e-01 3.84775102e-01 1.39926589e+00 6.43786311e-01
1.59765884e-01 4.94107783e-01 -4.44348276e-01 -5.16527653e-01
-4.76362646e-01 4.66355681e-01 -2.52974987e-01 -9.84385490e-01
8.69432807e-01 -2.91077822e-01 -3.94264221e-01 9.64173004e-02
-5.04519224e-01 -7.73654953e-02 4.37914103e-01 6.74792826e-01
-5.31366728e-02 -4.81957436e-01 1.06185853e+00 -9.34766114e-01
1.51695192e+00 -5.35797954e-01 -1.59551680e-01 -2.38413345e-02
-1.03954411e+00 -1.14682488e-01 3.57659966e-01 -7.22062811e-02
-1.01789069e+00 -6.30204201e-01 -1.76789731e-01 7.30891407e-01
-9.02416185e-02 1.04189384e+00 2.41284639e-01 -1.38953149e-01
9.29695964e-01 -2.17322513e-01 -7.44656846e-02 -8.14039707e-02
-1.31555021e-01 6.50221467e-01 3.02356392e-01 -9.49061096e-01
5.94707489e-01 -3.04550350e-01 9.93402302e-02 -4.16800231e-01
-1.48604751e+00 -4.61976469e-01 -5.66841364e-01 -5.40893316e-01
3.37105542e-01 -6.19836271e-01 -9.16177452e-01 -2.86643535e-01
-6.94406927e-01 -2.41081879e-01 -6.22798055e-02 5.64946592e-01
-5.16689837e-01 4.81370807e-01 -7.49419093e-01 -7.92019129e-01
-1.69815466e-01 -8.65182459e-01 4.62920308e-01 6.18777573e-01
-1.25900924e+00 -1.34815884e+00 5.86464405e-01 1.18513656e+00
2.82012582e-01 -7.78416023e-02 1.18137479e+00 -6.01539195e-01
-1.30882025e-01 -3.68935466e-01 6.99914433e-03 -1.54284518e-02
-4.12703931e-01 1.46396399e-01 -4.76536125e-01 2.60516047e-01
-5.05716279e-02 -8.92525911e-01 5.26572943e-01 3.49042267e-01
8.20335269e-01 -5.11878729e-02 9.67808291e-02 7.76474774e-02
1.06407356e+00 -2.44156331e-01 4.40738440e-01 1.07244432e+00
1.28241628e-01 5.68256557e-01 8.44493449e-01 5.69291234e-01
7.95573354e-01 5.66480637e-01 7.95046613e-02 4.13276464e-01
-4.68420461e-02 -4.35755067e-02 5.17774165e-01 9.35901880e-01
-2.54065990e-01 -1.52146416e-02 -9.50019598e-01 6.51180565e-01
-1.52982497e+00 -1.19818544e+00 -6.61392331e-01 1.67562294e+00
1.10204458e+00 3.54389608e-01 3.39067191e-01 5.16328096e-01
3.06570858e-01 2.16582954e-01 -9.12197679e-02 -8.40294957e-01
-5.46384975e-02 3.74651432e-01 1.85768194e-02 6.38459444e-01
-5.23889363e-01 6.98699415e-01 7.78083038e+00 2.33771637e-01
-4.85440582e-01 -2.29756042e-01 5.75423121e-01 -2.13435769e-01
-7.38745987e-01 1.14807516e-01 -8.06704283e-01 3.45110983e-01
1.02272248e+00 -5.57862043e-01 -1.10105574e-02 7.75981009e-01
4.32608306e-01 -3.00325781e-01 -9.96961355e-01 2.36752436e-01
1.88307285e-01 -1.15360916e+00 -3.13552439e-01 2.98495561e-01
9.52044785e-01 -7.97462404e-01 1.65833548e-01 4.59227562e-01
7.20311046e-01 -1.02681267e+00 5.65577209e-01 5.53978860e-01
3.84489417e-01 -8.90294731e-01 8.63327026e-01 3.06136131e-01
-4.10384059e-01 -1.10414252e-01 -5.46027720e-01 -1.12475443e+00
-2.64589965e-01 6.94915652e-01 -8.74434471e-01 2.55809035e-02
1.04691871e-01 5.04565001e-01 -7.85917997e-01 8.67682636e-01
-7.08829224e-01 1.26698303e+00 3.92038137e-01 -5.52055657e-01
2.44391263e-01 -2.15176478e-01 3.01499307e-01 1.03932428e+00
3.45472217e-01 4.45602298e-01 2.11228263e-02 8.68321836e-01
-3.84833276e-01 3.14139485e-01 -2.61177063e-01 -1.21718273e-01
9.23130691e-01 1.52823019e+00 -6.37282312e-01 -3.23669642e-01
-3.42043281e-01 3.73794615e-01 4.90982056e-01 -7.52717257e-02
-3.42742085e-01 -3.43102038e-01 4.84528214e-01 5.08213192e-02
-4.20836926e-01 -1.70529500e-01 -1.59200668e+00 -9.04982448e-01
-1.14573680e-01 -8.37027907e-01 3.61655593e-01 -7.57265747e-01
-1.03059244e+00 -2.75944173e-01 -5.55777550e-01 -5.91633856e-01
-1.24209061e-01 -4.73021835e-01 -1.01011550e+00 1.01945829e+00
-1.21118331e+00 -6.85002089e-01 -4.44510877e-01 -1.85531974e-02
4.00417268e-01 -2.54313141e-01 9.55544174e-01 -1.29384652e-01
-3.15808296e-01 7.87586868e-01 -7.79314339e-02 9.08698328e-03
1.20347738e+00 -1.64069295e+00 9.68004614e-02 3.97848457e-01
-3.11038326e-02 6.91192389e-01 8.55710328e-01 -7.55486608e-01
-8.93092990e-01 -2.00611427e-01 1.51109362e+00 -9.21162724e-01
7.99045205e-01 3.11852396e-01 -8.81752372e-01 2.61680186e-01
2.21680775e-01 -8.84927273e-01 1.51287353e+00 9.83349085e-01
-1.71051681e-01 4.43790644e-01 -9.30245936e-01 6.54941678e-01
6.09997511e-01 -5.23989081e-01 -1.02270508e+00 3.08676183e-01
-2.21548826e-01 -2.14536071e-01 -1.44007289e+00 -2.66739249e-01
8.67756307e-01 -1.01171672e+00 7.33359754e-01 -7.25853622e-01
1.45022011e+00 6.64167330e-02 3.04415494e-01 -1.38383174e+00
-7.77361035e-01 -2.40812749e-01 -1.10786229e-01 1.14414132e+00
3.98333490e-01 1.25863507e-01 1.42096841e+00 1.07697380e+00
-4.47531581e-01 -1.08115757e+00 -4.37658399e-01 -3.15371454e-01
6.34721577e-01 -1.35032520e-01 6.43060565e-01 1.26597071e+00
9.91076469e-01 4.98691827e-01 -2.06220433e-01 -5.68003356e-01
5.31794786e-01 2.27142975e-01 8.25865030e-01 -1.92080140e+00
2.56685205e-02 -1.15231752e+00 -2.46638462e-01 -7.55929947e-01
2.76295751e-01 -8.71203125e-01 -2.09266335e-01 -1.91662049e+00
7.23479509e-01 -3.43048036e-01 -1.60939679e-01 4.76469874e-01
-7.74178386e-01 4.61627275e-01 1.60406962e-01 3.00558150e-01
-7.99233437e-01 6.36432990e-02 1.22895813e+00 1.57924339e-01
-1.69457868e-01 -2.67243922e-01 -1.46525192e+00 6.62054479e-01
9.51732218e-01 -4.47384953e-01 -2.08995029e-01 -1.89931124e-01
7.08863914e-01 3.75622153e-01 -9.51331481e-02 -7.93412507e-01
3.73713553e-01 -8.94936979e-01 6.24265671e-01 -5.58599770e-01
-3.48744765e-02 -1.87715683e-02 -4.11412030e-01 2.40273878e-01
-8.89869571e-01 1.97137028e-01 -3.08212936e-01 -2.01390773e-01
-1.63911745e-01 -9.80294943e-01 6.15376532e-01 -1.02259703e-01
-1.51814103e-01 -3.00093591e-01 -6.49250031e-01 3.36017281e-01
7.19044209e-01 -6.62283778e-01 -4.87767428e-01 -5.56473434e-01
-4.30573672e-01 6.00672364e-02 8.55376959e-01 2.09739193e-01
4.81624722e-01 -1.33520806e+00 -1.02977943e+00 -4.60892946e-01
7.94030502e-02 -7.57844925e-01 -1.78095084e-02 7.15446711e-01
-3.24936718e-01 6.95038617e-01 -6.32971346e-01 3.21900368e-01
-1.43631625e+00 -1.81153417e-01 -9.34755877e-02 -7.00593650e-01
-3.95879626e-01 6.72576666e-01 -3.90193164e-01 -2.85615832e-01
-6.01797290e-02 4.15948391e-01 -7.47450590e-01 2.69210219e-01
7.65611351e-01 8.74754608e-01 1.77925322e-02 -3.30077022e-01
-6.98527843e-02 1.69302881e-01 -1.07810169e-01 -3.11234236e-01
1.39054632e+00 1.98738500e-02 3.91475707e-02 6.29965127e-01
4.81579781e-01 5.04552782e-01 -7.49499500e-01 1.45456925e-01
3.84640813e-01 -6.71620905e-01 8.90140310e-02 -1.19901514e+00
-4.47433829e-01 6.71604872e-01 -2.25064009e-01 3.01976413e-01
6.29330933e-01 -3.08444977e-01 5.11663795e-01 4.47568357e-01
-7.91904926e-02 -1.58012629e+00 5.93632638e-01 8.71568382e-01
4.87193435e-01 -1.08950937e+00 4.68419909e-01 8.03591385e-02
-6.10208213e-01 1.60365593e+00 7.95754135e-01 -2.59425899e-04
1.69479787e-01 1.37474835e-01 1.85145751e-01 -4.49162275e-01
-9.60996628e-01 3.27416837e-01 3.69072855e-01 3.43012124e-01
1.47744572e+00 2.87471294e-01 -1.19646728e+00 8.27573061e-01
-8.92483711e-01 6.21635839e-02 1.45163214e+00 7.68108785e-01
-1.02754426e+00 -1.46449709e+00 -6.81435168e-01 1.03491294e+00
-8.33754122e-01 -1.61443308e-01 -1.10086691e+00 5.33576488e-01
-3.31920564e-01 1.07608461e+00 -2.69163549e-01 -2.15174034e-01
1.10067263e-01 3.72209191e-01 5.82804382e-01 -7.60429680e-01
-1.07118309e+00 -4.68681008e-01 4.26086634e-01 1.63524926e-01
-3.83703917e-01 -1.15258062e+00 -1.22733128e+00 -9.14259374e-01
-2.94937521e-01 4.36413854e-01 6.61977112e-01 1.19717979e+00
-7.18277991e-02 6.21165991e-01 2.28339300e-01 -2.56935328e-01
-7.75488973e-01 -1.11882591e+00 -6.17634118e-01 3.86070788e-01
-1.15184627e-01 -7.23787963e-01 -2.21656397e-01 -2.03439474e-01] | [11.308167457580566, 9.26874828338623] |
45f943f7-62e0-4162-a81c-65cc96695252 | globally-injective-and-bijective-neural | 2306.03982 | null | https://arxiv.org/abs/2306.03982v1 | https://arxiv.org/pdf/2306.03982v1.pdf | Globally injective and bijective neural operators | Recently there has been great interest in operator learning, where networks learn operators between function spaces from an essentially infinite-dimensional perspective. In this work we present results for when the operators learned by these networks are injective and surjective. As a warmup, we combine prior work in both the finite-dimensional ReLU and operator learning setting by giving sharp conditions under which ReLU layers with linear neural operators are injective. We then consider the case the case when the activation function is pointwise bijective and obtain sufficient conditions for the layer to be injective. We remark that this question, while trivial in the finite-rank case, is subtler in the infinite-rank case and is proved using tools from Fredholm theory. Next, we prove that our supplied injective neural operators are universal approximators and that their implementation, with finite-rank neural networks, are still injective. This ensures that injectivity is not `lost' in the transcription from analytical operators to their finite-rank implementation with networks. Finally, we conclude with an increase in abstraction and consider general conditions when subnetworks, which may be many layers deep, are injective and surjective and provide an exact inversion from a `linearization.' This section uses general arguments from Fredholm theory and Leray-Schauder degree theory for non-linear integral equations to analyze the mapping properties of neural operators in function spaces. These results apply to subnetworks formed from the layers considered in this work, under natural conditions. We believe that our work has applications in Bayesian UQ where injectivity enables likelihood estimation and in inverse problems where surjectivity and injectivity corresponds to existence and uniqueness, respectively. | ['Maarten V. de Hoop', 'Matti Lassas', 'Michael Puthawala', 'Takashi Furuya'] | 2023-06-06 | null | null | null | null | ['operator-learning'] | ['miscellaneous'] | [ 2.42492333e-01 5.97271979e-01 -1.26565546e-01 -6.76871464e-02
-3.73155683e-01 -6.24901950e-01 1.89256772e-01 -4.49647248e-01
-2.37300843e-01 8.55617642e-01 2.68038481e-01 -2.97743142e-01
-7.09485829e-01 -7.91097164e-01 -1.20403278e+00 -7.42607892e-01
-5.85850358e-01 3.29653561e-01 -3.70890237e-02 -4.89609033e-01
-3.26450884e-01 6.91441417e-01 -1.30731702e+00 1.98660284e-01
7.28692651e-01 9.46029365e-01 -3.05087626e-01 9.67002571e-01
6.86482638e-02 1.06344879e+00 4.90052998e-02 -2.67609537e-01
6.94490075e-01 -4.27695811e-01 -1.19784069e+00 -3.05802077e-01
7.21145272e-01 -2.22379282e-01 -3.20349038e-01 1.26459277e+00
2.87515223e-01 1.56615928e-01 1.00964749e+00 -1.15841925e+00
-9.52550709e-01 7.93544173e-01 -5.67741990e-02 1.07300334e-01
1.53537467e-01 -4.35750097e-01 1.29309142e+00 -9.80295837e-01
5.59427142e-01 1.39466178e+00 1.36677814e+00 4.37068164e-01
-1.90148079e+00 -4.42086935e-01 -3.15620273e-01 -2.67983586e-01
-1.34923553e+00 -4.60947901e-01 4.88892019e-01 -7.96414256e-01
6.97502434e-01 4.18127716e-01 3.90687227e-01 6.89558387e-01
1.42636895e-01 6.14048839e-01 9.97395098e-01 -5.99122047e-01
-7.92493671e-03 2.66458750e-01 2.00844109e-01 1.16226125e+00
1.59797277e-02 1.94542348e-01 -2.86454529e-01 6.74006194e-02
1.43516517e+00 -5.71076311e-02 -5.70935726e-01 -5.62445462e-01
-1.14108038e+00 1.14072061e+00 6.33609891e-01 5.73705018e-01
-1.95576668e-01 2.15845674e-01 3.42830181e-01 9.54208255e-01
3.66279632e-01 6.00456893e-01 -2.11870179e-01 4.15777028e-01
-6.60740972e-01 -1.87374875e-02 1.31150138e+00 9.93073642e-01
8.25221419e-01 1.24577560e-01 1.94711193e-01 7.45169699e-01
9.03346464e-02 3.74172837e-01 1.50726782e-02 -1.39524698e+00
1.61305740e-01 6.85289651e-02 -1.85929090e-01 -9.42490995e-01
-4.58068013e-01 -4.62636411e-01 -1.14984477e+00 3.78635287e-01
7.16724575e-01 -1.28040805e-01 -3.17051381e-01 2.29399085e+00
-3.59236598e-01 2.35108435e-01 2.88579434e-01 7.36343026e-01
2.93760687e-01 5.48302650e-01 -4.57722247e-01 -3.15517277e-01
1.02844620e+00 -2.36580029e-01 -6.61067724e-01 4.22532558e-01
6.98099911e-01 -5.53005755e-01 9.87309694e-01 4.08169448e-01
-1.46932518e+00 -3.90988618e-01 -1.28334200e+00 -2.54893273e-01
-3.21446747e-01 4.49892320e-02 6.83029175e-01 3.89970571e-01
-1.38964510e+00 1.05416310e+00 -5.20356357e-01 -3.49292219e-01
3.96131158e-01 5.33414066e-01 -5.83930194e-01 3.68800551e-01
-1.55044615e+00 1.00960374e+00 3.05885881e-01 2.34455138e-01
-6.56803489e-01 -9.15295780e-01 -7.55363345e-01 1.54797658e-01
1.78022102e-01 -5.33031046e-01 1.23007274e+00 -1.18927312e+00
-1.25442958e+00 7.46477783e-01 1.27350748e-01 -7.10202217e-01
6.26060128e-01 7.80920088e-02 -1.03573702e-01 2.02129200e-01
3.24025191e-02 3.07306707e-01 5.94487369e-01 -9.01871562e-01
-1.07528664e-01 -2.22968683e-01 6.93243027e-01 -3.57529484e-02
-3.19404662e-01 -2.17622027e-01 2.64364958e-01 -4.77844805e-01
3.63349050e-01 -8.05241525e-01 9.06286389e-02 4.39985991e-01
-1.79042488e-01 -6.51542768e-02 4.01359707e-01 -4.07028109e-01
7.89711595e-01 -2.18061376e+00 4.14386690e-01 6.30850494e-01
2.71177113e-01 -1.09796770e-01 1.16274487e-02 3.73217344e-01
-6.04950488e-01 2.04227731e-01 -5.66051006e-01 -4.95908558e-02
3.26979727e-01 3.81516844e-01 -5.90423524e-01 8.22625160e-01
2.74258465e-01 8.07114005e-01 -5.94904959e-01 -3.25072706e-01
-1.65696606e-01 5.18168986e-01 -9.14424062e-01 -1.03019483e-01
9.14365426e-03 2.46482939e-01 -8.26814920e-02 1.57628864e-01
6.46237850e-01 -4.39063609e-02 -3.56311575e-02 -4.17603016e-01
-2.43379742e-01 1.38868153e-01 -1.66490698e+00 1.43789613e+00
-7.56021500e-01 7.39277959e-01 6.70384526e-01 -1.54297757e+00
4.97229010e-01 4.03647959e-01 4.00803626e-01 -2.99101949e-01
1.16328292e-01 4.86893356e-01 4.48025428e-02 -4.13686305e-01
3.29181887e-02 -8.93814385e-01 3.50722000e-02 3.88088882e-01
3.91092300e-01 1.53219506e-01 2.46924385e-01 2.51568079e-01
9.73407447e-01 -1.42082408e-01 1.66812196e-01 -1.02878249e+00
7.64345646e-01 -4.92226422e-01 2.16701791e-01 8.98333728e-01
1.89968497e-01 4.68384653e-01 1.01665831e+00 -1.39146134e-01
-9.86974001e-01 -1.63679194e+00 -7.30267704e-01 1.13373530e+00
-3.33502322e-01 3.65431979e-02 -5.48003435e-01 -1.26275837e-01
4.90225926e-02 2.77273476e-01 -9.54186618e-01 -2.15177089e-01
-6.66269004e-01 -4.23126370e-01 1.06832302e+00 6.78058803e-01
6.46927178e-01 -8.72446716e-01 3.48792672e-02 3.32654826e-02
1.16002530e-01 -6.86925828e-01 -6.13646150e-01 5.95923483e-01
-9.79122102e-01 -9.49108124e-01 -9.31647480e-01 -1.10725212e+00
6.82770252e-01 -2.81287789e-01 8.23361099e-01 -2.19747141e-01
9.18669719e-03 5.14277995e-01 4.15502936e-01 -1.27183586e-01
-5.47098398e-01 4.56020534e-02 3.78270090e-01 5.93210757e-02
-2.26418421e-01 -9.36171174e-01 -9.44718271e-02 3.67909044e-01
-1.30307984e+00 -2.24944144e-01 3.98812503e-01 9.52210128e-01
3.31340790e-01 1.40392587e-01 5.17562151e-01 -9.75143611e-01
7.36100733e-01 -3.69997323e-01 -7.66723096e-01 2.36167744e-01
-1.91398993e-01 7.96307743e-01 8.59062433e-01 -4.72801894e-01
-9.76930201e-01 6.09189039e-03 -9.58230197e-02 -2.59375751e-01
4.97254759e-01 6.50896251e-01 -1.17010407e-01 -4.77084637e-01
1.24659586e+00 -2.83209622e-01 3.34980130e-01 -3.98369998e-01
4.29470778e-01 2.19695553e-01 7.16015458e-01 -8.85922313e-01
7.75466144e-01 6.98609591e-01 5.03406644e-01 -8.72841835e-01
-8.67377102e-01 -3.22993249e-01 -6.97640657e-01 3.02735809e-03
7.71527231e-01 -7.04998195e-01 -1.13916492e+00 1.51264027e-01
-1.22948456e+00 -4.33652252e-01 -8.01906407e-01 5.74313819e-01
-9.43588912e-01 1.56147540e-01 -9.97644246e-01 -9.81055796e-01
2.54388481e-01 -1.03440034e+00 4.82048720e-01 -2.84718961e-01
1.12292737e-01 -1.55007195e+00 3.22325118e-02 -3.74170035e-01
5.34844220e-01 -1.52753100e-01 8.93390179e-01 -3.96623462e-01
-5.79667151e-01 1.38642080e-02 -4.37307328e-01 7.80318856e-01
-3.07099760e-01 -5.63537896e-01 -9.15182233e-01 -6.50339620e-03
4.11234587e-01 -2.22989589e-01 1.06772721e+00 4.35581058e-01
8.26889515e-01 -6.99124634e-01 1.98830813e-01 7.61163116e-01
1.38333440e+00 -4.64459509e-01 5.36382854e-01 -1.58101052e-01
7.04159260e-01 7.69883513e-01 -2.72097647e-01 -1.00713268e-01
-6.09082542e-02 5.83305538e-01 5.89012131e-02 -2.09712282e-01
2.31761396e-01 -1.01562515e-01 6.96862519e-01 1.00826144e+00
-4.63898063e-01 4.99753833e-01 -5.96353352e-01 2.40466148e-01
-1.85658813e+00 -1.22807300e+00 -1.20094776e-01 2.42415595e+00
1.11240876e+00 1.04537226e-01 2.00141326e-01 1.17227025e-01
6.15732610e-01 -2.53165543e-01 -2.20610097e-01 -7.09829986e-01
-5.11637866e-01 5.88673830e-01 9.26486135e-01 1.22880006e+00
-1.07188284e+00 4.46810812e-01 6.79881048e+00 5.32776296e-01
-8.52889359e-01 3.60676616e-01 1.16188325e-01 2.38458812e-01
-4.09635365e-01 -7.22975731e-02 -3.53072554e-01 -1.34476244e-01
9.58224773e-01 8.11129659e-02 7.97616005e-01 6.18825018e-01
-5.03496975e-02 3.27981174e-01 -1.47680259e+00 7.18869269e-01
-2.62436830e-02 -1.30551887e+00 -2.67027974e-01 2.30402693e-01
7.92793214e-01 -1.70414776e-01 4.05116677e-02 2.96938092e-01
4.05534595e-01 -1.31579328e+00 4.69735205e-01 8.87817860e-01
7.00502694e-01 -7.87609279e-01 5.71521938e-01 2.37746015e-01
-1.20383096e+00 4.81521077e-02 -6.09159410e-01 -3.74245018e-01
1.33655161e-01 6.36122406e-01 -4.60301816e-01 3.78126740e-01
1.71493575e-01 7.27277637e-01 5.98734580e-02 7.37407148e-01
1.24084845e-01 3.78819704e-01 -5.26374102e-01 2.97286004e-01
1.60386249e-01 -5.68114340e-01 5.81457019e-01 1.33312118e+00
1.91360608e-01 1.75944000e-01 -7.83878490e-02 1.31799483e+00
-2.02514008e-01 1.57189116e-01 -9.47180629e-01 1.98489979e-01
-3.44504029e-01 9.64192867e-01 -4.24790353e-01 -8.64586458e-02
-4.30733323e-01 7.26519823e-01 2.93356329e-01 6.47232711e-01
-5.74566543e-01 -4.73717809e-01 7.22078025e-01 1.66355431e-01
2.01437101e-01 -2.49121234e-01 -2.55315363e-01 -1.28851378e+00
1.39590785e-01 -5.99003017e-01 2.65851676e-01 -4.89626765e-01
-1.35940087e+00 7.11715668e-02 3.09909552e-01 -8.85193229e-01
-1.94559589e-01 -1.18548155e+00 -4.07879353e-01 1.07278883e+00
-9.64228511e-01 -7.58451402e-01 4.76145715e-01 5.69945693e-01
-2.38115385e-01 6.03941306e-02 7.81004012e-01 4.98839587e-01
-7.33473375e-02 5.41002095e-01 1.49544954e-01 3.36225718e-01
1.62996605e-01 -1.57503402e+00 -2.21713379e-01 7.47097790e-01
2.60899484e-01 1.06514943e+00 6.18042469e-01 -3.14468294e-01
-1.32178605e+00 -6.87183321e-01 7.63996124e-01 -2.34932125e-01
1.06993890e+00 -7.47836292e-01 -1.21545923e+00 9.88822162e-01
-1.83247983e-01 1.60530716e-01 2.76416749e-01 3.46609861e-01
-3.29713255e-01 -1.26777261e-01 -1.02355385e+00 5.20362914e-01
1.25377321e+00 -1.09043670e+00 -4.36030149e-01 4.79990929e-01
4.42339659e-01 -3.72634023e-01 -1.14519417e+00 4.14227843e-01
7.27898240e-01 -8.59656155e-01 1.23201704e+00 -8.60331237e-01
1.96577311e-01 -1.94456950e-01 -3.27740192e-01 -9.40671802e-01
-2.01551825e-01 -6.99512064e-01 6.44927397e-02 7.80062795e-01
5.88874280e-01 -9.95311022e-01 2.73483843e-01 3.57178897e-01
-1.60893872e-01 -7.55022585e-01 -1.02502584e+00 -1.03027213e+00
6.34549141e-01 -7.23581254e-01 -5.79436822e-03 9.99504030e-01
3.48512411e-01 4.95855242e-01 -3.48664880e-01 1.32401260e-02
3.01745743e-01 -5.25906384e-01 1.81801870e-01 -1.44674480e+00
-7.03971982e-01 -7.03797221e-01 -6.21313930e-01 -1.31202185e+00
5.34103513e-01 -1.54855943e+00 1.63690910e-01 -1.03909326e+00
-1.61861237e-02 -3.98552001e-01 -3.28034639e-01 2.57010818e-01
6.63476467e-01 3.05466443e-01 -2.02520639e-02 3.33640754e-01
-1.23171017e-01 1.67403474e-01 1.15982866e+00 -1.93224140e-02
-1.92366496e-01 2.56225675e-01 -6.29658818e-01 9.60668385e-01
5.88631988e-01 -1.29706174e-01 -4.93723005e-01 -8.93717334e-02
7.48767972e-01 4.76754047e-02 7.30242133e-01 -9.25688326e-01
2.15572104e-01 3.20179015e-01 9.52495337e-02 2.76564002e-01
2.48194695e-01 -7.25381672e-01 4.84138243e-02 3.93697947e-01
-8.47311318e-01 -3.59069407e-02 9.37731862e-02 4.11680639e-01
8.81533846e-02 -6.93797231e-01 1.04863489e+00 -1.47102684e-01
-1.58586830e-01 1.16563000e-01 -3.10308695e-01 4.50348258e-01
3.13338816e-01 5.46546876e-02 1.39544383e-01 -6.95868611e-01
-1.26910210e+00 -1.64892048e-01 1.93113148e-01 -3.43142271e-01
3.12555760e-01 -1.53312695e+00 -5.63917756e-01 3.71701777e-01
-3.56881052e-01 -1.82086825e-01 2.60307491e-02 1.47823620e+00
-4.94609296e-01 4.18091387e-01 -6.67646378e-02 -4.94448245e-01
-7.10058689e-01 2.73907661e-01 8.16511333e-01 -1.23022325e-01
-5.60609221e-01 8.46401334e-01 5.66159427e-01 -6.77728713e-01
4.22343135e-01 -6.00344777e-01 1.18094482e-01 1.43083259e-01
2.44215906e-01 4.06828433e-01 -1.46291647e-02 -7.35132813e-01
2.78020054e-02 5.57466149e-01 3.04624438e-01 -6.41548455e-01
1.06443012e+00 2.71658987e-01 -4.55922991e-01 9.04569149e-01
1.86536682e+00 1.05439506e-01 -1.05610859e+00 -4.06441242e-01
-2.18108952e-01 2.46991247e-01 -1.09591372e-01 -3.00715387e-01
-1.08231723e+00 9.70840275e-01 4.23970252e-01 6.18711412e-01
8.81752312e-01 3.21526319e-01 2.14429364e-01 7.85278141e-01
-1.71765596e-01 -9.71320510e-01 -7.28230551e-02 8.12908173e-01
1.35102987e+00 -7.58814096e-01 -1.63482964e-01 -4.26763922e-01
-5.56457937e-02 1.30249476e+00 9.40590575e-02 -4.63881701e-01
9.83579397e-01 2.21548140e-01 -3.47724736e-01 -6.45920113e-02
-1.54344350e-01 -3.15382451e-01 5.58969498e-01 3.34231704e-01
5.84767997e-01 4.54804003e-02 -1.07230723e-01 1.71239614e-01
-4.46152747e-01 -1.32667929e-01 5.15076876e-01 4.49239820e-01
-4.06087875e-01 -8.98108840e-01 -2.93234587e-01 5.14075279e-01
-4.55854833e-01 -1.95715636e-01 -1.75511584e-01 1.00149143e+00
1.58467606e-01 2.59625047e-01 1.02429375e-01 5.70713058e-02
3.43447417e-01 2.47813836e-01 8.37160528e-01 -5.01000702e-01
-1.22938238e-01 -1.19967073e-01 1.88767407e-02 -3.65431130e-01
-6.01678371e-01 -7.54389465e-01 -1.44398105e+00 -3.28508675e-01
-3.99927169e-01 1.47046238e-01 3.91142547e-01 1.01271570e+00
-1.81456104e-01 5.29037476e-01 2.59964615e-01 -5.92823982e-01
-8.81878078e-01 -8.65849793e-01 -9.06336069e-01 5.94260320e-02
8.00544322e-01 -6.38326168e-01 -7.87067771e-01 -6.74419105e-02] | [7.5599446296691895, 3.739657163619995] |
1e7085ca-fc19-4ed7-8097-c0785cacc09b | skin-lesion-classification-with-ensemble-of | 1809.02568 | null | http://arxiv.org/abs/1809.02568v1 | http://arxiv.org/pdf/1809.02568v1.pdf | Skin lesion classification with ensemble of squeeze-and-excitation networks and semi-supervised learning | In this report, we introduce the outline of our system in Task 3: Disease
Classification of ISIC 2018: Skin Lesion Analysis Towards Melanoma Detection.
We fine-tuned multiple pre-trained neural network models based on
Squeeze-and-Excitation Networks (SENet) which achieved state-of-the-art results
in the field of image recognition. In addition, we used the mean teachers as a
semi-supervised learning framework and introduced some specially designed data
augmentation strategies for skin lesion analysis. We confirmed our data
augmentation strategy improved classification performance and demonstrated
87.2% in balanced accuracy on the official ISIC2018 validation dataset. | ['Shunsuke Kitada', 'Hitoshi Iyatomi'] | 2018-09-07 | null | null | null | null | ['skin-lesion-classification'] | ['medical'] | [ 1.03804874e+00 1.24702565e-01 -5.01323998e-01 -1.79619223e-01
-1.13525259e+00 -2.09753603e-01 7.36847281e-01 -6.56732768e-02
-7.60058165e-01 6.20432854e-01 -3.54445539e-02 -5.42931139e-01
-1.57678705e-02 -4.10251975e-01 -4.49042469e-01 -1.07260597e+00
-7.63378292e-02 -1.07201889e-01 -1.42311845e-02 -3.01783048e-02
6.48901388e-02 3.72974217e-01 -1.03080010e+00 9.20416951e-01
9.89716232e-01 1.14259505e+00 -3.37963551e-01 1.26884854e+00
1.98635757e-01 8.54745150e-01 -4.04441357e-01 -5.38089097e-01
5.01319580e-03 -2.36000881e-01 -1.05687284e+00 -2.64448766e-02
6.81537390e-01 -6.18646219e-02 -5.79581499e-01 8.90406966e-01
5.63132644e-01 -4.55904067e-01 9.39958334e-01 -8.30429614e-01
-5.72025836e-01 4.19462383e-01 -7.10230589e-01 2.66922683e-01
-9.50526297e-02 2.56805420e-01 3.13928902e-01 -6.31193280e-01
7.80113459e-01 4.52456087e-01 1.11692953e+00 1.34493947e+00
-9.68284249e-01 -3.40954095e-01 -2.29991749e-01 1.08662881e-01
-1.10979724e+00 -5.64098179e-01 4.78840888e-01 -3.20134163e-01
1.05236197e+00 8.51032257e-01 5.20285368e-01 1.84425104e+00
2.14915514e-01 8.44616532e-01 1.40461779e+00 -7.27895260e-01
4.44512907e-03 3.15677226e-01 2.14832485e-01 7.95779705e-01
-6.18108213e-02 2.42765337e-01 -3.07972103e-01 -6.61766827e-02
7.15328693e-01 -2.51908839e-01 -2.26814337e-02 4.15167630e-01
-1.05318284e+00 6.58383191e-01 5.56444526e-01 2.49248579e-01
-4.58880961e-01 2.22498357e-01 6.45836532e-01 -6.49559945e-02
6.38899326e-01 3.96383673e-01 -4.04698908e-01 4.40605870e-03
-7.57404685e-01 -3.74131799e-01 3.30970407e-01 3.80103290e-01
-3.49744976e-01 -1.78645849e-02 -5.63276112e-01 1.05476737e+00
-8.84429216e-02 5.09547055e-01 5.68560243e-01 -6.06786370e-01
2.95192510e-01 6.47380710e-01 -4.38435078e-01 -4.31808650e-01
-7.35868812e-01 -9.84388590e-01 -1.46185493e+00 1.87390726e-02
2.27509350e-01 -2.52850443e-01 -1.60257626e+00 1.32337368e+00
-4.49340492e-02 1.22640714e-01 2.37228885e-01 6.36839509e-01
1.14657581e+00 1.13642305e-01 4.15991306e-01 9.88049060e-02
1.34780014e+00 -1.18451846e+00 -7.20233142e-01 1.87097669e-01
1.03268838e+00 -4.85037565e-01 6.46766067e-01 4.54417855e-01
-8.47076297e-01 -3.31292212e-01 -7.79605269e-01 1.76042289e-01
-7.03938246e-01 6.35587871e-01 8.93951416e-01 1.00843370e+00
-1.11173797e+00 2.66076773e-01 -9.71980572e-01 -7.38120914e-01
8.80697489e-01 3.08236390e-01 -7.54419088e-01 1.54368475e-01
-9.58158135e-01 7.75472343e-01 2.39036694e-01 3.04398715e-01
-9.43440676e-01 -7.78483927e-01 -7.33731866e-01 -5.24468541e-01
1.09540485e-01 -4.84600127e-01 9.36638296e-01 -8.51568103e-01
-1.42715108e+00 1.52740657e+00 -2.87407748e-02 -9.22566295e-01
1.98640153e-01 -4.12561260e-02 -9.27483201e-01 3.69180620e-01
-3.94332349e-01 1.00059676e+00 5.98745346e-01 -9.02685642e-01
-6.81430280e-01 -1.96474656e-01 -3.32372248e-01 -1.15876131e-01
-6.99806631e-01 1.42441019e-01 -4.96017456e-01 -7.07060575e-01
-3.38581324e-01 -7.83803940e-01 -4.06133831e-01 -1.99568011e-02
-9.51946914e-01 1.11913495e-02 3.38788360e-01 -9.96495605e-01
1.29600763e+00 -2.02072477e+00 -6.79772124e-02 4.02832091e-01
2.44530886e-01 7.58974075e-01 -6.05150163e-01 1.93188816e-01
-6.75741792e-01 5.48583210e-01 -1.65106654e-01 -5.16055405e-01
-3.05910170e-01 -1.71529993e-01 8.86579901e-02 4.24846530e-01
6.45218134e-01 1.04431915e+00 -4.48192984e-01 -6.33923888e-01
2.91479677e-01 6.69077873e-01 -1.11526757e-01 -1.80155575e-01
3.00092876e-01 1.97834402e-01 -2.39408724e-02 1.16986096e+00
8.05371702e-01 -2.07721278e-01 -1.85697734e-01 -4.53766018e-01
1.21562392e-01 -1.85924560e-01 -4.56054389e-01 1.70228016e+00
-2.22223163e-01 7.66625106e-01 -2.64958758e-02 -9.13586736e-01
5.83288312e-01 3.84191573e-01 5.91557503e-01 -5.78988910e-01
2.07902625e-01 8.92544091e-02 -6.42650500e-02 -8.69165182e-01
1.37535721e-01 2.65176117e-01 6.30992725e-02 -1.88709885e-01
2.95353889e-01 3.25599670e-01 1.07302107e-01 6.88294098e-02
1.28211284e+00 -1.04497731e-01 2.91182339e-01 -2.23624289e-01
7.12853789e-01 1.18406355e-01 2.04677328e-01 7.46775866e-01
-5.91900229e-01 6.23695493e-01 4.74817038e-01 -4.56093609e-01
-6.15528822e-01 -9.11921620e-01 -6.32726967e-01 7.92347908e-01
-5.31412601e-01 -3.01287174e-01 -1.03000975e+00 -1.10160637e+00
-3.54934514e-01 3.81081372e-01 -1.43291676e+00 -1.33940041e-01
-4.72087599e-02 -1.42058933e+00 1.28788590e+00 7.60317922e-01
6.02766871e-01 -8.66093993e-01 -3.95121947e-02 -2.42105737e-01
1.11026570e-01 -1.21974874e+00 -3.11265230e-01 4.50781405e-01
-5.15847445e-01 -1.61806417e+00 -9.41000462e-01 -1.02908921e+00
9.92704988e-01 -3.55124295e-01 6.57252848e-01 -6.41665012e-02
-1.09919143e+00 3.14475805e-01 -2.29361057e-01 -5.75512528e-01
-4.21557426e-01 2.80797720e-01 -1.48963183e-01 1.06662720e-01
5.43744683e-01 1.43006414e-01 -4.43738401e-01 3.59903350e-02
-9.83957052e-01 3.11431050e-01 1.02992034e+00 1.09227502e+00
6.04618549e-01 -8.16346854e-02 4.51859802e-01 -1.01785707e+00
5.49944043e-01 -1.04125682e-02 -1.86267883e-01 5.21850288e-01
-2.66461581e-01 -3.36814940e-01 2.98418701e-01 -3.61140639e-01
-9.26501155e-01 3.98336142e-01 -5.47463179e-01 1.20563209e-02
-5.93706131e-01 4.32688057e-01 4.27106917e-01 -6.56406462e-01
9.63726103e-01 3.69287759e-01 2.37221032e-01 -2.13584870e-01
-1.23388153e-02 1.04780209e+00 7.92068362e-01 -5.20936064e-02
4.97963130e-01 3.92875552e-01 3.72455478e-01 -7.70882368e-01
-1.06287408e+00 -4.21611071e-01 -7.95771480e-01 -2.29382187e-01
1.16791189e+00 -8.00733805e-01 -6.52342141e-01 1.34804845e+00
-8.13906133e-01 -7.73134232e-01 -9.69844684e-02 2.59053797e-01
-1.11535117e-01 1.43864229e-01 -8.56898189e-01 -8.98582101e-01
-8.94120216e-01 -9.16848540e-01 1.16511285e+00 4.52450603e-01
-1.58030495e-01 -1.50667417e+00 1.93086475e-01 4.04227495e-01
8.07043970e-01 7.97096670e-01 4.11066473e-01 -7.26904988e-01
3.49782467e-01 -6.83174253e-01 -3.28944206e-01 4.98280704e-01
1.16418473e-01 4.29517984e-01 -1.29653537e+00 -2.30688900e-01
-6.75517440e-01 -6.11435175e-01 1.45810235e+00 4.11353886e-01
1.89355469e+00 4.89536375e-02 -6.94870770e-01 1.01500213e+00
1.40480399e+00 -7.04185516e-02 9.47180510e-01 2.03410193e-01
3.77312958e-01 4.33030307e-01 1.39073759e-01 -2.25223303e-02
-1.41108140e-01 3.25552702e-01 5.41084230e-01 -7.86198318e-01
-5.67168534e-01 -7.87587836e-02 -9.85527784e-02 2.22122088e-01
-4.28506136e-01 -9.08437669e-02 -1.04089546e+00 5.03997087e-01
-1.21153951e+00 -5.72160840e-01 -5.56129217e-01 1.72922206e+00
8.93323779e-01 1.84026062e-01 3.06923147e-02 1.74211651e-01
5.38601041e-01 -2.28928119e-01 -3.04836631e-01 -2.59815454e-01
-4.89557803e-01 5.32044291e-01 7.18748093e-01 3.32864583e-01
-1.81600714e+00 7.84395278e-01 7.58793020e+00 1.14858174e+00
-1.20947802e+00 -2.19068453e-02 1.04015088e+00 -5.91998175e-02
2.53571421e-01 -9.44229424e-01 -6.75704479e-01 3.84181112e-01
1.09475660e+00 7.99557745e-01 8.27097967e-02 4.68354404e-01
-2.77779490e-01 -2.02733114e-01 -7.38903284e-01 1.00561082e+00
2.69155949e-01 -1.53390741e+00 -2.33871266e-01 3.15165281e-01
7.79867411e-01 1.42220378e-01 5.84115505e-01 3.91934454e-01
-2.76316464e-01 -1.42561507e+00 -2.88994730e-01 8.98098290e-01
1.58428168e+00 -5.08347094e-01 1.27247453e+00 -1.24462493e-01
-5.69129944e-01 8.80934671e-02 4.11019623e-02 4.26118791e-01
-4.75193530e-01 5.00607371e-01 -9.18914318e-01 4.99282986e-01
5.81831336e-01 7.57059276e-01 -1.39194393e+00 1.06069231e+00
-4.69707549e-02 1.13768888e+00 -2.42282927e-01 -1.63277999e-01
3.36988658e-01 4.53102350e-01 2.26970047e-01 1.54893589e+00
-1.94586545e-01 -2.79503524e-01 -3.87536883e-01 4.94924098e-01
-2.53881067e-02 -1.81804717e-01 -2.14220405e-01 -2.97723264e-01
4.17865515e-02 1.63618135e+00 -4.43181962e-01 -2.03954518e-01
1.30163673e-02 9.74569857e-01 8.46751928e-02 2.88630843e-01
-8.43006134e-01 -6.86918199e-01 1.92522466e-01 -2.85211921e-01
-1.95385695e-01 4.59462345e-01 -4.82121438e-01 -9.68737185e-01
-2.47473404e-01 -7.61725128e-01 5.38935542e-01 -5.25674582e-01
-1.43263459e+00 7.19473183e-01 -5.49302638e-01 -8.83855402e-01
5.42152599e-02 -1.36606872e+00 -9.96510804e-01 7.88236797e-01
-1.62465882e+00 -1.73819113e+00 -4.98762906e-01 5.23028135e-01
1.75625086e-01 -7.63213575e-01 1.68835866e+00 1.21790767e-01
-1.14135730e+00 1.35014820e+00 8.83932412e-02 4.96964008e-01
8.84851277e-01 -1.35280395e+00 3.69890690e-01 7.29522645e-01
-2.64272124e-01 3.35110247e-01 6.44851401e-02 -4.35954630e-01
-1.20980525e+00 -1.33258271e+00 5.89304090e-01 -4.33477283e-01
8.30888510e-01 -2.56453663e-01 -5.20533919e-01 3.78094792e-01
4.46039915e-01 2.31647193e-01 1.29824138e+00 2.82963008e-01
-1.94338977e-01 -1.65782765e-01 -1.31836915e+00 4.20811296e-01
6.70303464e-01 -6.15028560e-01 1.18809111e-01 6.72523618e-01
2.53031492e-01 -5.58723569e-01 -9.67163801e-01 7.47149706e-01
4.88285482e-01 -3.51112783e-01 9.74958420e-01 -1.24970627e+00
7.04533041e-01 3.01284492e-01 -7.78424442e-02 -1.15556312e+00
-3.84361804e-01 -4.47348595e-01 -3.59926313e-01 9.96925831e-01
6.56861424e-01 -5.51126659e-01 1.18269289e+00 2.51621485e-01
-6.65056407e-02 -1.49747992e+00 -8.16021562e-01 -4.20334578e-01
1.90210804e-01 -4.71918225e-01 -9.24908742e-02 7.94382393e-01
2.43698601e-02 -9.40851867e-03 -2.23428965e-01 4.91230376e-02
7.91177511e-01 -7.59426832e-01 2.64783293e-01 -7.79747069e-01
-1.19823068e-02 -5.70989192e-01 -5.51450908e-01 -2.36809731e-01
-7.47253001e-02 -1.08838475e+00 -3.26682210e-01 -1.29844630e+00
5.92333138e-01 -9.69636738e-02 -7.42469192e-01 9.61545825e-01
-3.97738814e-01 8.83704782e-01 -1.78152099e-01 -2.45555088e-01
-6.92127764e-01 -1.00938335e-01 1.01718950e+00 -4.22736406e-01
5.32399788e-02 2.95959145e-01 -8.92405570e-01 4.82434541e-01
1.09965992e+00 -2.38310359e-02 2.21178904e-01 -3.94932657e-01
-1.25157461e-01 -4.52054352e-01 8.65324914e-01 -1.11283422e+00
4.36576515e-01 5.28247505e-02 8.53080988e-01 -7.36608148e-01
4.59266305e-01 -5.53967476e-01 -4.94043708e-01 8.29899788e-01
-8.34218025e-01 -8.78190756e-01 5.52560747e-01 2.19298556e-01
8.64589885e-02 -2.02149659e-01 8.23213935e-01 2.74401158e-01
-6.50976419e-01 2.28295684e-01 -4.99005198e-01 -4.12094533e-01
1.29033661e+00 -3.14672850e-02 -7.99580812e-01 1.68586701e-01
-9.97740269e-01 2.63537109e-01 6.71561854e-03 1.47999048e-01
7.25278378e-01 -1.34098363e+00 -9.97143865e-01 3.13454509e-01
4.35556591e-01 -4.04836506e-01 5.97221494e-01 1.28878021e+00
-4.17349607e-01 8.79436016e-01 -3.16498041e-01 -7.29633927e-01
-1.62107468e+00 2.77305067e-01 7.68722057e-01 -6.84068441e-01
-1.11788660e-01 1.21236467e+00 -3.33126485e-01 -7.64590621e-01
5.96175730e-01 -1.40603423e-01 -3.93002003e-01 -2.81313390e-01
8.03041935e-01 1.99961215e-01 1.80038944e-01 -5.88885024e-02
-3.70903790e-01 3.11200023e-01 -3.61423999e-01 3.18895280e-01
1.29529953e+00 4.60759103e-01 -3.24861109e-02 3.77024375e-02
1.36352563e+00 -5.04362464e-01 -8.65860999e-01 -2.42330134e-02
-3.67767394e-01 7.35036731e-02 4.25362140e-01 -1.87164259e+00
-1.01929593e+00 8.48552942e-01 1.23871732e+00 6.39697835e-02
1.49145222e+00 -3.13230962e-01 5.24407387e-01 4.40051407e-01
-2.01524436e-01 -1.16333127e+00 -9.03649032e-02 2.61508197e-01
7.35529244e-01 -1.53798282e+00 -2.37754464e-01 -6.57222271e-01
-7.02876568e-01 1.25545442e+00 7.14369774e-01 3.45311761e-02
5.71841598e-01 5.94393849e-01 1.47987634e-01 -1.71634495e-01
-8.08418691e-01 -2.08388314e-01 8.01006675e-01 1.12691617e+00
5.06994367e-01 2.34831959e-01 -2.71345954e-03 7.23895729e-01
2.16915041e-01 1.13568321e-01 2.82559097e-01 6.24846280e-01
1.19168639e-01 -7.74228454e-01 -1.25546560e-01 9.32011187e-01
-7.90912747e-01 -2.44101673e-01 -8.04737449e-01 1.07869589e+00
2.19035655e-01 6.20924771e-01 5.46964332e-02 -6.13529205e-01
5.09008765e-02 1.01878121e-01 5.38420916e-01 -2.33313590e-01
-6.10085607e-01 -9.81585309e-03 3.86409163e-01 -4.69170392e-01
-4.93950069e-01 -3.79246026e-01 -7.72614121e-01 2.45437950e-01
-2.33635888e-01 -2.57675201e-01 1.03215051e+00 6.91569030e-01
3.87444615e-01 1.17492080e+00 5.00030160e-01 -2.54526377e-01
-5.44535339e-01 -1.49542689e+00 -5.14739394e-01 2.80198544e-01
3.48892778e-01 -1.08056411e-01 -9.26573128e-02 2.96385020e-01] | [15.710053443908691, -2.98388671875] |
f2dc8496-2fec-4ca5-a597-1cb80ea88e01 | protein-folding-a-meeting-point-for-leibniz | 2208.03150 | null | https://arxiv.org/abs/2208.03150v2 | https://arxiv.org/pdf/2208.03150v2.pdf | Protein Folding: From Classical Issues to a New Perspective | The Levinthal paradox exposes many critical questions on the protein folding problem, among which we could point out why proteins can reach their native state in a biologically reasonable time. A proper answer to this question is of foremost importance for evolutive biology since it enables us to understand life as we know it. Preliminary results, based on the upper bound protein marginal-stability limit, together with transition state theory arguments, lead us to show that two-state proteins must reach their native state in, at most, seconds rather than ($\sim 10^{27}$) years -- as indicated by a naive solution of the Levinthal paradox. This outcome -- added to the amide hydrogen-exchange protection factors analysis -- makes it possible for us to suggest how a protein point mutations and/or post-translational modifications impact its folding time scales but not its upper bound limit that obeys the physics ruling the process. Noteworthy for almost 50 years, the protein folding problem --mas the Levinthal paradox -- has been a topic of passionate debate because Anfinsen's challenge -- how a sequence encodes its folding -- remains unsolved despite the smashing success of accurately predicting the protein tridimensional structures by state-of-the-art numerical-methods. Aimed to unlock this long-standing challenge, we propose a new perspective of protein folding, specifically, as a problem that should be devised as an 'analytic whole' -- a Leibniz & Kant's notion. This viewpoint might help us decode Anfinsen's challenge and, thus, open new avenues for future research in the protein folding field. | ['Jorge A. Vila'] | 2022-08-05 | null | null | null | null | ['protein-folding'] | ['natural-language-processing'] | [ 3.49687457e-01 3.22051316e-01 -6.96034878e-02 -3.24950546e-01
-8.73457119e-02 -6.53620124e-01 1.53597906e-01 3.64488155e-01
-5.07556617e-01 1.15690994e+00 -1.12888247e-01 -9.75978315e-01
-1.88962705e-02 -5.22167683e-01 -8.15964818e-01 -1.26373053e+00
-2.56570965e-01 3.28474492e-01 9.10898969e-02 -7.38139212e-01
3.06539029e-01 5.11637211e-01 -1.41370845e+00 -2.52961125e-02
1.09510779e+00 5.90054393e-01 3.48685026e-01 5.80428004e-01
-1.32186502e-01 3.65109771e-01 -1.30307034e-01 -5.12389064e-01
2.16612771e-01 -1.05118620e+00 -1.02929175e+00 -4.68517393e-01
-1.37585044e-01 2.71703482e-01 3.43067832e-02 9.60884809e-01
2.13730112e-01 -3.91247347e-02 6.05959415e-01 -4.00148064e-01
-8.68670523e-01 1.00859944e-02 -1.52033120e-01 2.37677485e-01
2.40935668e-01 5.48418880e-01 1.06535089e+00 -4.59290594e-01
9.10691500e-01 9.29767609e-01 6.37947261e-01 7.11242497e-01
-1.51963484e+00 -4.19156998e-02 -2.34105811e-02 3.59361053e-01
-8.91557693e-01 -2.07610816e-01 4.15537059e-01 -5.69263935e-01
1.20792079e+00 4.84122187e-01 9.78739262e-01 7.16787636e-01
7.78231323e-01 1.19536623e-01 1.44534624e+00 -4.58050460e-01
3.43148798e-01 -3.68677288e-01 3.27037185e-01 7.40946174e-01
5.29707253e-01 3.02437931e-01 -3.64934623e-01 -2.87518859e-01
4.68262613e-01 -1.31577656e-01 -5.83536267e-01 -3.17636341e-01
-1.17688215e+00 5.81013560e-01 2.06700757e-01 5.33420861e-01
-3.58309209e-01 1.24754608e-01 1.84342653e-01 5.50778151e-01
2.31457159e-01 5.73562920e-01 -8.60981286e-01 -2.11141333e-01
-7.45934367e-01 4.23897594e-01 1.05592406e+00 3.21268141e-01
8.31071198e-01 -5.19549787e-01 4.89389181e-01 1.39400765e-01
4.76689398e-01 3.25952083e-01 9.63299274e-02 -6.79660678e-01
-5.82366735e-02 6.58994377e-01 4.65575844e-01 -5.92651486e-01
-6.08938098e-01 -2.27820829e-01 -8.65944564e-01 3.96905541e-01
1.19691277e+00 -1.34534966e-02 -3.82395893e-01 2.08409238e+00
3.91333818e-01 -4.18768823e-01 1.02999084e-01 1.15101349e+00
5.51464744e-02 6.72072649e-01 2.20612302e-01 -1.06786788e+00
1.59723568e+00 -2.42811516e-01 -5.09785831e-01 1.75141364e-01
5.16909182e-01 -5.91994047e-01 1.10560846e+00 3.31147611e-01
-1.01605940e+00 -1.65520445e-01 -1.23590934e+00 -7.83239901e-02
-2.76100904e-01 -3.39287341e-01 7.62066245e-01 8.86753619e-01
-8.93090010e-01 1.21166849e+00 -9.12842095e-01 -9.89222288e-01
-1.27162397e-01 3.51419777e-01 -3.64400506e-01 5.36524653e-01
-1.16333771e+00 1.34391463e+00 1.86845034e-01 1.22819617e-01
-3.33915383e-01 -6.21205986e-01 1.12948669e-02 -1.51653081e-01
5.56159876e-02 -8.73553395e-01 9.26972508e-01 -8.21102262e-01
-1.50412571e+00 1.16299725e+00 -4.31075871e-01 -3.83594781e-01
2.74860442e-01 3.00587416e-01 -2.32285336e-01 1.95657372e-01
-3.86674792e-01 1.35703400e-01 2.75976032e-01 -9.90651429e-01
1.18912093e-01 -8.81971657e-01 -7.01987892e-02 -5.31847477e-02
4.39498395e-01 -5.87629303e-02 5.70070088e-01 -3.97743136e-01
6.53661609e-01 -1.05642390e+00 -5.67783833e-01 -2.83789374e-02
3.54160257e-02 -2.09520936e-01 1.74048394e-01 -4.87946838e-01
9.12366569e-01 -1.71149480e+00 4.90692407e-01 -5.64837679e-02
4.07254547e-01 3.35766166e-01 3.42749268e-01 8.45051169e-01
-5.62585354e-01 1.95413888e-01 -4.81741965e-01 5.38842678e-01
-1.07624508e-01 1.30034968e-01 -4.64301199e-01 8.90306175e-01
-1.02087334e-02 1.05590725e+00 -9.06743288e-01 -6.58965949e-03
3.17374282e-02 4.40726608e-01 -3.78786802e-01 7.80030526e-03
-5.19168258e-01 6.97430313e-01 -3.70308280e-01 3.08066100e-01
7.34673023e-01 -4.37491834e-01 7.97601223e-01 -2.25863203e-01
-4.94385451e-01 1.34292230e-01 -5.38446903e-01 1.54242349e+00
4.05687869e-01 1.81359172e-01 -3.59110571e-02 -1.27263117e+00
9.83217895e-01 1.61632106e-01 5.65351367e-01 -5.17976880e-01
2.01795071e-01 5.91674447e-01 4.20708895e-01 -6.02071643e-01
1.37632966e-01 -9.34746861e-01 1.91105798e-01 2.40613773e-01
-3.52750383e-02 2.53998548e-01 1.51870206e-01 -1.76965863e-01
1.01865828e+00 6.44340396e-01 4.32111114e-01 -8.79528403e-01
6.82362258e-01 2.80112803e-01 7.17155278e-01 3.29557002e-01
-5.83738804e-01 3.22814047e-01 8.95216346e-01 -8.45489025e-01
-1.37454510e+00 -8.75348687e-01 -2.32248619e-01 8.09699833e-01
2.77769625e-01 -4.30231184e-01 -1.26051092e+00 -1.91943482e-01
-3.56925167e-02 4.15000021e-01 -7.01870203e-01 -2.69392639e-01
-7.46575773e-01 -1.49851608e+00 2.87969530e-01 -2.31713161e-01
1.28982276e-01 -8.73574138e-01 -1.07933450e+00 3.87786716e-01
-1.76756114e-01 -3.61762732e-01 5.64019457e-02 6.76388443e-01
-1.12758958e+00 -1.35503018e+00 -9.07788575e-01 -3.76185089e-01
4.83355075e-01 -3.31244022e-02 9.58919168e-01 3.06550354e-01
-4.53181446e-01 -1.54259384e-01 -2.96387643e-01 -2.99616549e-02
-8.66850376e-01 -2.07557857e-01 4.00724173e-01 -3.01705778e-01
7.97856033e-01 -1.10972941e+00 -1.03418756e+00 2.60041177e-01
-7.19811857e-01 -4.14303653e-02 3.62198204e-01 6.06286466e-01
5.00418901e-01 -5.40474594e-01 6.96503043e-01 -5.67623854e-01
3.39073598e-01 -2.17766315e-01 -6.08419538e-01 6.19261861e-01
-8.69549692e-01 5.13198376e-01 8.38687241e-01 2.71560699e-02
-6.95262551e-01 -1.21626273e-01 -3.98868054e-01 4.89041001e-01
-2.06547230e-02 2.16988832e-01 -3.17213833e-01 -1.04556337e-01
8.58371019e-01 3.96058947e-01 4.73793119e-01 -7.35385180e-01
3.38527352e-01 3.27966005e-01 4.72947538e-01 -8.72449160e-01
4.29687768e-01 4.39278781e-01 5.25114954e-01 -9.55932617e-01
-5.44751465e-01 -1.09728158e-01 -6.40398622e-01 -5.24153374e-02
1.23280168e+00 -3.84479612e-01 -1.55241299e+00 2.29957178e-01
-1.35041106e+00 -1.16869755e-01 -3.14368039e-01 2.22896412e-01
-9.19683158e-01 7.89135516e-01 -6.68017149e-01 -9.02823329e-01
-1.57396093e-01 -1.16008234e+00 3.91724855e-01 2.83370972e-01
-2.24881768e-01 -7.28416562e-01 3.76880735e-01 3.48364323e-01
4.00887340e-01 4.22402889e-01 1.40279567e+00 -3.55729461e-01
-5.06125867e-01 4.58466440e-01 3.48412879e-02 5.01303338e-02
-1.06272809e-01 3.61223631e-02 -7.29763150e-01 -2.31397465e-01
3.97407115e-01 -3.52018070e-03 9.61631954e-01 3.14409494e-01
4.34624434e-01 -1.59745976e-01 -1.66600361e-01 3.91054958e-01
1.63457215e+00 1.93572879e-01 6.39170885e-01 2.86518216e-01
9.44499075e-02 7.71403015e-01 4.29507017e-01 3.07982415e-01
1.12623377e-02 5.69516659e-01 4.86650556e-01 2.86098123e-01
2.11482923e-02 -2.12844163e-01 3.03453118e-01 9.16525543e-01
-6.78702593e-01 7.74335936e-02 -7.88825989e-01 -8.35472718e-02
-1.85896885e+00 -1.18478310e+00 -5.55598497e-01 2.38529682e+00
9.91164327e-01 -1.05014872e-02 3.07836950e-01 -1.38772130e-01
4.31299627e-01 -1.00140348e-01 -8.46557617e-01 -5.84417045e-01
-5.14285266e-01 1.75972342e-01 4.75012273e-01 7.37029970e-01
-4.88031834e-01 8.26773226e-01 6.77540064e+00 4.23568904e-01
-1.22172725e+00 3.19920108e-02 7.13937938e-01 3.71229678e-01
-4.44743603e-01 6.58766031e-01 -5.84012687e-01 5.51284790e-01
1.24256206e+00 -2.81382501e-01 5.55955708e-01 5.10863185e-01
6.03405654e-01 -4.15074408e-01 -9.68301773e-01 6.65779233e-01
-2.51987308e-01 -1.39459908e+00 -2.87527233e-01 3.08084130e-01
2.69702286e-01 -6.02903552e-02 -2.40773991e-01 -2.59229153e-01
-5.92715219e-02 -9.30079222e-01 5.63047528e-01 7.77269244e-01
6.11079514e-01 -4.94413465e-01 2.95144081e-01 6.38067901e-01
-7.63878703e-01 1.73208550e-01 -8.04902136e-01 -4.01321769e-01
3.08421314e-01 8.54077458e-01 -2.06334546e-01 4.23506588e-01
2.49406070e-01 5.69364466e-02 -1.29862905e-01 6.59092128e-01
7.61849061e-02 4.42175627e-01 -2.93335587e-01 -2.83259839e-01
-6.62840456e-02 -9.20617223e-01 4.62172508e-01 8.76369119e-01
1.51659578e-01 7.74171174e-01 -3.86988103e-01 1.16500545e+00
3.37388486e-01 2.23301157e-01 -1.86960816e-01 -3.99217844e-01
-7.41571411e-02 9.15334702e-01 -1.04198647e+00 -3.67671102e-02
-1.07439123e-01 9.98334050e-01 3.67472798e-01 2.94664264e-01
-6.58886433e-01 -3.08729231e-01 9.95248675e-01 2.37334609e-01
-2.16496494e-02 -4.99235332e-01 -2.12742582e-01 -1.42826605e+00
8.79238695e-02 -7.92446852e-01 -1.44327387e-01 -3.53933692e-01
-9.44016933e-01 4.49450433e-01 -4.38968480e-01 -5.58880270e-01
8.19762126e-02 -7.86864042e-01 -1.73166573e-01 9.98947144e-01
-1.24718678e+00 -6.83740318e-01 3.92445475e-01 -3.85850109e-02
-7.52343461e-02 2.18556434e-01 9.14887726e-01 -1.23842262e-01
-3.61197323e-01 2.02197254e-01 5.36580980e-01 -7.40191519e-01
5.31670511e-01 -1.27684426e+00 5.82622528e-01 6.06854498e-01
-2.92522013e-01 9.79691625e-01 1.46873844e+00 -6.77386165e-01
-1.80914760e+00 -2.29264662e-01 1.40745258e+00 -7.15294123e-01
5.71572185e-01 -2.61498719e-01 -1.17826843e+00 1.63233370e-01
7.56789520e-02 -2.11378872e-01 8.12132180e-01 -7.55844451e-03
-2.78334588e-01 1.12817183e-01 -1.13917720e+00 4.55606997e-01
1.09517801e+00 -3.90062571e-01 -7.94204712e-01 4.84982073e-01
6.90445423e-01 1.27465874e-01 -9.97315168e-01 3.06198597e-01
7.53329754e-01 -1.49037921e+00 7.81307697e-01 -1.03496087e+00
2.38102078e-01 -3.95212173e-01 -2.44020373e-01 -7.93426692e-01
-4.25757557e-01 -1.16570520e+00 6.95499256e-02 5.66778243e-01
3.89004081e-01 -7.71087825e-01 6.96713209e-01 6.83626175e-01
-3.43551636e-02 -1.17918491e+00 -1.16765833e+00 -8.37849736e-01
6.91382885e-01 8.87000561e-02 5.13963401e-01 8.10839593e-01
6.98954284e-01 2.65900940e-01 -1.33140087e-01 -2.23308131e-01
5.91445684e-01 9.39785466e-02 3.77535194e-01 -1.20525420e+00
-3.90037209e-01 -3.32885742e-01 -2.12025195e-01 -1.15753961e+00
3.56070995e-02 -7.93744206e-01 -1.69026405e-01 -1.09534621e+00
2.92255729e-01 -1.81531474e-01 -2.61999369e-01 1.86926767e-01
-8.72666948e-03 -9.77659076e-02 1.79993734e-01 3.65168691e-01
-2.36685842e-01 3.08263272e-01 1.18135202e+00 4.64635938e-01
-3.10299117e-02 -2.99203247e-01 -8.44285846e-01 4.42036480e-01
7.43095577e-01 -3.33626032e-01 -1.08595766e-01 3.19572210e-01
8.63925219e-01 2.40609646e-01 2.80820996e-01 -6.97666049e-01
-1.50460169e-01 -4.53134537e-01 -7.40755945e-02 -2.72990227e-01
8.25575665e-02 -5.95543385e-01 4.92094934e-01 1.03778839e+00
-1.48298606e-01 1.69467658e-01 -2.60616660e-01 6.35652661e-01
4.62474674e-01 -2.65330464e-01 1.16612053e+00 -4.75798845e-01
-2.00716093e-01 -1.24154743e-02 -6.35893941e-01 -6.49647489e-02
1.13614249e+00 -3.17657590e-01 -6.17019773e-01 1.59667537e-01
-1.02856731e+00 -3.01348358e-01 1.24355662e+00 -2.85230838e-02
9.80943963e-02 -8.04721832e-01 -4.32983756e-01 1.57845661e-01
4.57997397e-02 -8.95595908e-01 4.17365670e-01 1.12543845e+00
-1.08846056e+00 7.31259108e-01 -2.16699690e-01 -4.21835750e-01
-1.02108824e+00 9.53399837e-01 6.11338794e-01 -7.41050346e-03
-5.72389781e-01 6.18993580e-01 7.59209469e-02 -1.90197915e-01
-5.70269465e-01 -3.52436185e-01 3.56770337e-01 -1.02642693e-01
4.20009613e-01 2.86364377e-01 -1.83922164e-02 -7.12802052e-01
-5.07829309e-01 6.50796890e-01 1.39395028e-01 2.77355909e-01
1.26251841e+00 -2.25664049e-01 -7.85212994e-01 5.30535936e-01
9.57016170e-01 -3.08933198e-01 -1.12662148e+00 2.15903029e-01
2.33009741e-01 -1.89471811e-01 -5.79857230e-01 -9.41259861e-01
-8.51068348e-02 9.03465986e-01 6.82354867e-01 3.61835241e-01
7.81230927e-01 8.09897780e-02 7.46339560e-01 6.23839378e-01
5.71905136e-01 -9.67683613e-01 -5.03969014e-01 3.17813784e-01
6.70733571e-01 -1.05131114e+00 -7.53433257e-02 -1.22386850e-01
-1.43484712e-01 1.36904824e+00 6.24622405e-02 -8.63948762e-02
5.56447983e-01 -1.27792448e-01 -7.91407302e-02 -8.64816532e-02
-9.93265450e-01 -9.35871452e-02 -1.94877282e-01 3.49348933e-01
9.64104950e-01 2.19173178e-01 -1.32143176e+00 5.61188996e-01
-1.62724271e-01 7.78539777e-02 4.41813618e-01 7.93738246e-01
-1.09067333e+00 -1.74997473e+00 -3.09666693e-01 7.76439309e-02
-3.45774889e-01 9.92712379e-02 -6.13857985e-01 4.59293365e-01
2.76719183e-01 8.69301021e-01 -5.33956826e-01 -1.62184194e-01
3.11589334e-02 5.26635110e-01 9.70964909e-01 -1.39172718e-01
-6.56763554e-01 -2.29752921e-02 -2.01972559e-01 -4.88440663e-01
-6.66094005e-01 -5.98495841e-01 -1.37609959e+00 -8.33263934e-01
-3.25691938e-01 8.79978299e-01 7.24526525e-01 1.13457870e+00
4.87134755e-01 6.46909102e-05 2.50128299e-01 -6.33561730e-01
-9.58625436e-01 -5.47858119e-01 -7.39235461e-01 2.48911306e-01
3.24247211e-01 -3.34651262e-01 -3.71838659e-01 5.39992191e-02] | [4.72942590713501, 5.26641845703125] |
89ec0cfe-71cd-47c1-9b43-e1dc71611c52 | positive-unlabeled-learning-for-cell | 2106.15918 | null | https://arxiv.org/abs/2106.15918v1 | https://arxiv.org/pdf/2106.15918v1.pdf | Positive-unlabeled Learning for Cell Detection in Histopathology Images with Incomplete Annotations | Cell detection in histopathology images is of great value in clinical practice. \textit{Convolutional neural networks} (CNNs) have been applied to cell detection to improve the detection accuracy, where cell annotations are required for network training. However, due to the variety and large number of cells, complete annotations that include every cell of interest in the training images can be challenging. Usually, incomplete annotations can be achieved, where positive labeling results are carefully examined to ensure their reliability but there can be other positive instances, i.e., cells of interest, that are not included in the annotations. This annotation strategy leads to a lack of knowledge about true negative samples. Most existing methods simply treat instances that are not labeled as positive as truly negative during network training, which can adversely affect the network performance. In this work, to address the problem of incomplete annotations, we formulate the training of detection networks as a positive-unlabeled learning problem. Specifically, the classification loss in network training is revised to take into account incomplete annotations, where the terms corresponding to negative samples are approximated with the true positive samples and the other samples of which the labels are unknown. To evaluate the proposed method, experiments were performed on a publicly available dataset for mitosis detection in breast cancer cells, and the experimental results show that our method improves the performance of cell detection given incomplete annotations for training. | ['Chuyang Ye', 'Zhiwen Liu', 'Fengqian Pang', 'Zipei Zhao'] | 2021-06-30 | null | null | null | null | ['cell-detection', 'mitosis-detection'] | ['computer-vision', 'medical'] | [ 3.41892064e-01 1.23811640e-01 -3.25924933e-01 -2.85284221e-01
-6.23158872e-01 -3.72326076e-01 4.59495708e-02 5.78128994e-01
-6.57517850e-01 1.17476249e+00 -3.79893392e-01 -2.66333461e-01
3.25273454e-01 -8.06828678e-01 -5.32930076e-01 -1.26682150e+00
2.19077930e-01 3.73555928e-01 3.72361422e-01 1.81653753e-01
-1.51938815e-02 2.97053427e-01 -9.65191901e-01 3.04100364e-01
9.26918685e-01 1.14368486e+00 3.89112271e-02 3.92006338e-01
-1.96297064e-01 9.54716504e-01 -6.80769205e-01 -2.55708694e-01
-1.83905084e-02 -2.19548658e-01 -6.93432808e-01 1.48635224e-01
-2.73364820e-02 -3.10523391e-01 -5.98748699e-02 1.41643417e+00
4.44718540e-01 -2.92756200e-01 7.26445556e-01 -1.13607049e+00
-1.15960389e-01 2.80915707e-01 -7.83088207e-01 2.36816064e-01
-3.74125808e-01 1.26127467e-01 7.85961509e-01 -8.26700687e-01
4.88207877e-01 6.51222348e-01 6.63019657e-01 6.48966908e-01
-1.04846656e+00 -9.29119647e-01 -2.32600849e-02 -6.62062764e-02
-1.69392884e+00 -3.04001659e-01 5.31052768e-01 -5.03283322e-01
2.22374335e-01 2.52946734e-01 7.86615849e-01 5.66741467e-01
2.38077845e-02 9.61418629e-01 7.22709417e-01 -2.35451311e-01
2.89976984e-01 3.54127258e-01 3.19304198e-01 5.08085668e-01
4.63105589e-01 -6.42885193e-02 1.73910230e-01 5.45798317e-02
7.83473372e-01 3.09792966e-01 -4.91009772e-01 -2.14192271e-01
-1.07414830e+00 6.67743087e-01 6.68154776e-01 3.67654800e-01
-8.72174501e-02 -2.44165704e-01 6.91673756e-01 -1.44528389e-01
5.85782230e-01 1.19774073e-01 -4.67509776e-01 3.26567709e-01
-6.54444098e-01 -1.10298656e-01 6.22354925e-01 6.64549768e-01
8.24034274e-01 -2.56494671e-01 -2.96493292e-01 9.16971087e-01
5.11929132e-02 1.42462224e-01 7.09894419e-01 -3.70135576e-01
2.92562336e-01 1.17078745e+00 1.85624093e-01 -9.07304168e-01
-4.10214812e-01 -8.22846830e-01 -1.37682724e+00 -2.32668663e-03
8.08157682e-01 -1.32642075e-01 -1.01778197e+00 1.47112107e+00
4.53395277e-01 3.24531257e-01 3.63248169e-01 1.06014562e+00
1.03197324e+00 4.55131322e-01 1.67929813e-01 -6.04627132e-01
1.25563252e+00 -7.47888803e-01 -8.79630685e-01 -2.17428342e-01
1.19253445e+00 -5.87252617e-01 6.77303612e-01 2.16046810e-01
-6.03081703e-01 -2.84494162e-01 -9.21700180e-01 1.97503164e-01
-3.71182412e-01 5.92019439e-01 5.72234273e-01 2.73694664e-01
-6.17653131e-01 1.26966953e-01 -7.72491097e-01 -1.63328782e-01
8.54461074e-01 6.30035460e-01 -3.52876127e-01 -2.11595789e-01
-9.82880294e-01 3.75017107e-01 7.76672125e-01 5.62947512e-01
-9.45683897e-01 -5.66207170e-01 -5.76239288e-01 7.56772235e-02
5.60984433e-01 -1.02225460e-01 1.34898078e+00 -1.23120046e+00
-5.21837234e-01 9.46486950e-01 -2.09271863e-01 -1.75510153e-01
6.71952963e-01 5.33745050e-01 -2.12225497e-01 -1.02060683e-01
3.43797617e-02 6.56928957e-01 1.50714785e-01 -1.25530922e+00
-1.00373280e+00 -3.11513752e-01 9.08121392e-02 -3.47035974e-02
-6.25398636e-01 -4.54625070e-01 -6.31635308e-01 -5.86383283e-01
2.11015567e-01 -8.65133643e-01 -3.38208467e-01 2.18432203e-01
-6.67074680e-01 -1.77684188e-01 9.56373096e-01 -5.30421972e-01
1.13692772e+00 -2.29419565e+00 -3.50427210e-01 2.31239513e-01
4.08950448e-01 5.09886265e-01 1.58605367e-01 -2.59573609e-01
8.67981985e-02 3.33007485e-01 -2.10955158e-01 -2.17490897e-01
-5.62906265e-01 3.26245755e-01 1.74150273e-01 3.93350244e-01
2.52872080e-01 6.13713682e-01 -9.71204281e-01 -9.84265745e-01
-7.21001625e-02 4.35527086e-01 -2.28815585e-01 2.49191627e-01
-2.13843495e-01 6.00910008e-01 -5.43428063e-01 9.19453084e-01
6.17991686e-01 -6.19617641e-01 1.36773914e-01 -3.21331739e-01
3.11014473e-01 -3.41463476e-01 -1.07220483e+00 7.42415249e-01
-9.99430642e-02 4.89631981e-01 1.17803507e-01 -1.03240812e+00
6.85720444e-01 5.07666528e-01 4.32274252e-01 -3.21292847e-01
3.86921167e-01 5.18662632e-01 2.77705133e-01 -5.96897423e-01
2.07370982e-01 -3.10729295e-01 2.90670395e-01 -4.22431119e-02
-3.31920624e-01 2.57609487e-01 3.69727105e-01 -2.37122141e-02
8.82319987e-01 -5.07973015e-01 3.65909398e-01 7.19802231e-02
7.56376088e-01 1.38337329e-01 1.26657045e+00 3.49514186e-01
-4.17599678e-01 5.40785313e-01 9.07603323e-01 -4.98837292e-01
-8.54252219e-01 -4.54562694e-01 -3.33375275e-01 6.07562780e-01
1.96245193e-01 6.39173090e-02 -5.93594193e-01 -9.37308788e-01
-2.82142609e-01 -4.27217968e-02 -7.30617940e-01 -1.83018342e-01
-4.05899674e-01 -1.27047324e+00 4.92971241e-01 6.83479786e-01
5.02592444e-01 -1.03882325e+00 -2.54432946e-01 1.17116831e-01
-3.08963954e-01 -1.06774735e+00 -3.09973806e-01 2.31904149e-01
-9.16996121e-01 -1.55975389e+00 -9.14448321e-01 -1.21638179e+00
1.49920595e+00 1.29659802e-01 7.28816092e-01 7.21129596e-01
-4.31287438e-01 -4.67698038e-01 -1.66832015e-01 -6.85172975e-01
-4.15112913e-01 7.44916946e-02 -2.87307739e-01 2.92897284e-01
4.07496989e-01 1.29497964e-02 -7.17530787e-01 4.21810955e-01
-9.52982306e-01 1.08998440e-01 7.10717320e-01 1.34287596e+00
9.93236423e-01 3.41414303e-01 8.38769555e-01 -1.32103503e+00
1.81480438e-01 -3.29759538e-01 -5.18850565e-01 1.74369633e-01
-2.40015924e-01 -4.64037389e-01 7.15560198e-01 -7.57496774e-01
-7.69953549e-01 3.24521959e-01 -5.54590635e-02 -3.05365175e-01
-4.84222993e-02 7.90805280e-01 -1.90465704e-01 -1.94628656e-01
4.14342046e-01 2.08645210e-01 1.44491032e-01 -7.14851096e-02
-5.96331537e-01 7.08278656e-01 4.40004736e-01 -1.13990128e-01
2.94191599e-01 4.42690969e-01 1.59888461e-01 -5.94458520e-01
-1.19995928e+00 -6.79964066e-01 -3.50347638e-01 -2.86294222e-01
6.14264309e-01 -8.01979065e-01 -6.14084125e-01 5.91622293e-01
-9.23585534e-01 -1.24187216e-01 -1.08981311e-01 6.71865344e-01
-4.24559414e-03 2.21371472e-01 -7.57926166e-01 -8.81067753e-01
-3.20319921e-01 -1.27111018e+00 8.53242397e-01 5.81906497e-01
-2.77036838e-02 -1.13590527e+00 -2.55486667e-01 2.08456606e-01
1.14718482e-01 3.17686200e-01 1.21626663e+00 -1.01342690e+00
-4.96042192e-01 -9.04024899e-01 -2.57600963e-01 4.37458992e-01
3.22469026e-01 2.31482998e-01 -9.59719598e-01 -4.47021246e-01
-2.38963187e-01 -5.12068510e-01 8.50620270e-01 4.50323433e-01
1.29662633e+00 -2.71567583e-01 -9.61255908e-01 4.13688600e-01
1.37528408e+00 4.31585759e-01 5.78994930e-01 2.93753352e-02
5.67200065e-01 5.06864548e-01 8.29531610e-01 2.17675716e-01
2.79195942e-02 1.39747277e-01 6.20959461e-01 -7.33153880e-01
1.96983501e-01 6.19314909e-02 -2.56798416e-01 4.94603217e-01
1.51644915e-01 -5.52323461e-01 -9.57540512e-01 8.08770120e-01
-1.87185955e+00 -5.92367291e-01 -3.74119103e-01 2.34756970e+00
1.02805173e+00 2.67521948e-01 -1.47481382e-01 5.85275114e-01
1.07065988e+00 -3.13155115e-01 -7.81497777e-01 4.72515941e-01
1.92291767e-03 -1.83875963e-01 1.88757449e-01 1.73413560e-01
-1.24957204e+00 4.80283171e-01 5.42356920e+00 1.00120437e+00
-1.34030664e+00 3.64240147e-02 1.32825255e+00 6.06930740e-02
2.42450267e-01 -1.76622033e-01 -9.46423829e-01 4.72668916e-01
1.03325509e-01 1.32824838e-01 -3.22485745e-01 7.10187078e-01
2.82007605e-01 -1.76000938e-01 -1.08424938e+00 7.95131028e-01
-1.81588337e-01 -1.17286074e+00 -1.54786333e-01 1.89173028e-01
8.15637708e-01 -4.32285100e-01 -1.23816013e-01 4.58094835e-01
-1.13536026e-02 -9.03194427e-01 1.89913496e-01 3.43888521e-01
8.83556366e-01 -8.30996394e-01 1.55256629e+00 7.85330772e-01
-1.10172701e+00 -1.11365385e-01 -5.86189866e-01 1.11490667e-01
-1.91110760e-01 8.07623267e-01 -1.09586668e+00 1.41429290e-01
3.27263534e-01 7.53399611e-01 -5.15536606e-01 1.45179951e+00
7.36179352e-02 6.30697966e-01 -3.07888806e-01 -2.86084741e-01
2.67056301e-02 2.08383441e-01 3.95292044e-02 9.05458927e-01
3.10433209e-01 1.59857556e-01 5.33007145e-01 6.68438971e-01
-3.87494445e-01 2.91679889e-01 -2.45035902e-01 -6.07035495e-02
3.76314551e-01 1.50749528e+00 -1.07458615e+00 -3.46406311e-01
-3.64706129e-01 3.80505502e-01 3.96698952e-01 3.57499272e-01
-6.40098691e-01 -2.47425616e-01 5.95305823e-02 1.46905884e-01
-9.75102112e-02 5.01361847e-01 -3.64128828e-01 -8.88911247e-01
-1.27926078e-02 -3.59258175e-01 5.09821177e-01 -3.98272604e-01
-1.41726458e+00 2.29558825e-01 -4.58814532e-01 -1.55373991e+00
2.47281268e-01 -4.81093973e-01 -6.44075036e-01 7.02738523e-01
-1.55626655e+00 -1.03155088e+00 -6.03523135e-01 1.00450210e-01
3.44715089e-01 4.61804681e-02 5.54521799e-01 5.74203253e-01
-8.76103044e-01 8.44274998e-01 -2.47569978e-02 5.60271263e-01
5.68073392e-01 -9.63421643e-01 -5.30886233e-01 4.77576315e-01
-5.55019379e-01 2.86566913e-01 3.52321386e-01 -6.96281016e-01
-7.61503994e-01 -1.41506064e+00 8.17718983e-01 1.03005521e-01
2.65822411e-01 -8.06203038e-02 -1.17447221e+00 5.50973654e-01
-3.98194760e-01 6.73345566e-01 9.60560083e-01 -2.32973039e-01
2.34710753e-01 -2.03419298e-01 -1.18965077e+00 4.63576078e-01
4.35701489e-01 -1.64117530e-01 9.06045362e-02 5.34033179e-01
3.98770571e-01 -5.24779618e-01 -7.34320343e-01 7.49052405e-01
4.22067761e-01 -5.24116635e-01 6.14594936e-01 -3.95009160e-01
3.35281819e-01 -5.48743844e-01 3.27706367e-01 -1.05203915e+00
1.23785704e-01 3.20672452e-01 2.68636197e-02 1.30339992e+00
5.28023899e-01 -5.60957134e-01 1.19142783e+00 4.44126904e-01
-1.33839488e-01 -1.14922547e+00 -9.14247394e-01 -4.82326269e-01
-4.53895852e-02 1.84867606e-01 3.48707408e-01 1.11800158e+00
3.66948396e-02 3.60007048e-01 -1.88523129e-01 1.99842200e-01
3.32834870e-01 -2.10150048e-01 5.95600069e-01 -1.42316616e+00
-2.66535580e-03 -2.90343791e-01 -5.77543259e-01 -7.92120874e-01
2.60720700e-02 -8.54851961e-01 2.65969455e-01 -1.46110034e+00
6.62543774e-01 -7.66104102e-01 -3.63761514e-01 7.84075320e-01
-4.56142157e-01 6.11773014e-01 -3.22403193e-01 4.82676744e-01
-6.14142478e-01 2.57778883e-01 1.48642886e+00 -5.10699511e-01
-5.09328432e-02 1.58677489e-01 -4.80772138e-01 8.18611622e-01
7.03468978e-01 -4.97352481e-01 -3.35036039e-01 -9.00603682e-02
1.44794583e-01 2.48733357e-01 3.91537130e-01 -1.01703000e+00
3.23571712e-01 -1.73338890e-01 7.31006145e-01 -7.39655077e-01
2.86976576e-01 -8.94094408e-01 5.71061671e-02 7.14235961e-01
-4.50574070e-01 -6.16359353e-01 -2.62337439e-02 6.31660521e-01
-5.01645923e-01 -5.34788907e-01 9.04296756e-01 -2.11256891e-01
-2.05506220e-01 5.50679922e-01 -1.77442431e-01 -7.92112947e-02
1.22254741e+00 -5.13504565e-01 -4.74438101e-01 -5.34936599e-02
-7.97940373e-01 6.59039736e-01 4.15237963e-01 -1.96923092e-01
4.78551537e-01 -1.35727870e+00 -5.32114506e-01 8.01303983e-02
5.08567810e-01 7.17137814e-01 3.44786882e-01 1.09980524e+00
-5.38457513e-01 2.17233941e-01 1.95235834e-01 -8.08193684e-01
-1.40389097e+00 5.48024595e-01 6.35828674e-01 -5.07127643e-01
-1.51446983e-01 7.33253956e-01 6.02268040e-01 -3.09114754e-01
4.93369430e-01 -3.73712808e-01 -7.13922918e-01 -2.85824649e-02
5.29392779e-01 3.04083079e-02 7.31383339e-02 -4.97035205e-01
-7.34670758e-02 2.54937291e-01 -4.87466097e-01 7.02853382e-01
9.68145251e-01 1.75897583e-01 -2.89137334e-01 5.03704309e-01
1.07620370e+00 -3.84279966e-01 -1.08823776e+00 -3.02112579e-01
-2.00769037e-01 -2.15428576e-01 5.33682108e-02 -7.66106427e-01
-1.54254055e+00 1.01669741e+00 7.65944064e-01 7.71140680e-02
1.08525288e+00 -1.96615070e-01 6.41277969e-01 3.92083287e-01
2.11109281e-01 -9.99417007e-01 8.76726466e-04 2.67688662e-01
2.19992727e-01 -1.52113843e+00 -2.06661567e-01 -7.82035232e-01
-2.42597923e-01 1.08955455e+00 1.09482837e+00 3.07606071e-01
4.88786817e-01 2.75930017e-01 4.09086831e-02 -1.07072666e-01
-5.61574042e-01 3.52425389e-02 -2.21458927e-01 3.28632057e-01
4.93132651e-01 1.57625616e-01 -5.79581857e-01 9.47591245e-01
3.74702394e-01 3.31352115e-01 5.29746950e-01 1.04916310e+00
-4.20941353e-01 -7.23923981e-01 -2.72356272e-01 8.90665531e-01
-8.00122917e-01 1.66170090e-01 -4.40279305e-01 7.97558308e-01
2.46765584e-01 7.11052299e-01 5.47014810e-02 -2.31540620e-01
1.24852121e-01 -1.43700361e-01 1.04424581e-02 -6.64438784e-01
-3.91615152e-01 2.31953114e-01 -1.55408736e-02 3.60674679e-01
-2.72920042e-01 -2.51019865e-01 -1.64968824e+00 7.20249638e-02
-8.82488370e-01 2.72357911e-01 6.82153106e-02 9.38097298e-01
-1.67762205e-01 6.93783700e-01 5.43470502e-01 -3.97818297e-01
-3.92144859e-01 -1.06587052e+00 -6.63988233e-01 3.30417871e-01
5.56252718e-01 -5.74674726e-01 -3.94327015e-01 3.59051526e-01] | [14.928933143615723, -3.0348830223083496] |
36d05ef6-ef8b-476f-8546-be54faf6b4fe | cocatt-a-cognitive-conditioned-driver-1 | 2207.04028 | null | https://arxiv.org/abs/2207.04028v1 | https://arxiv.org/pdf/2207.04028v1.pdf | CoCAtt: A Cognitive-Conditioned Driver Attention Dataset (Supplementary Material) | The task of driver attention prediction has drawn considerable interest among researchers in robotics and the autonomous vehicle industry. Driver attention prediction can play an instrumental role in mitigating and preventing high-risk events, like collisions and casualties. However, existing driver attention prediction models neglect the distraction state and intention of the driver, which can significantly influence how they observe their surroundings. To address these issues, we present a new driver attention dataset, CoCAtt (Cognitive-Conditioned Attention). Unlike previous driver attention datasets, CoCAtt includes per-frame annotations that describe the distraction state and intention of the driver. In addition, the attention data in our dataset is captured in both manual and autopilot modes using eye-tracking devices of different resolutions. Our results demonstrate that incorporating the above two driver states into attention modeling can improve the performance of driver attention prediction. To the best of our knowledge, this work is the first to provide autopilot attention data. Furthermore, CoCAtt is currently the largest and the most diverse driver attention dataset in terms of autonomy levels, eye tracker resolutions, and driving scenarios. CoCAtt is available for download at https://cocatt-dataset.github.io. | ['Katherine Driggs-Campbell', 'Peter Du', 'Aamir Hasan', 'Pranav Sriram', 'Niviru Wijayaratne', 'Yuan Shen'] | 2022-07-08 | null | null | null | null | ['driver-attention-monitoring'] | ['computer-vision'] | [-3.87821704e-01 -2.12100577e-02 -4.75760520e-01 -1.68175951e-01
-1.92140520e-01 -1.43632740e-01 4.97116536e-01 1.88608263e-02
-4.33413088e-01 2.40506381e-01 2.14433491e-01 -5.60784042e-01
-7.65756294e-02 -2.42625311e-01 -3.59207660e-01 -3.54937822e-01
6.43454731e-01 -1.43027892e-02 4.53511268e-01 -4.55822438e-01
3.14126968e-01 4.38517869e-01 -2.20350790e+00 -1.90865204e-01
7.74690449e-01 7.80478120e-01 6.08444631e-01 8.63750219e-01
2.82079637e-01 1.00853407e+00 -3.37659538e-01 -2.13595301e-01
2.30318587e-02 -2.63560265e-01 -3.98882180e-01 -1.09485947e-01
4.09876704e-01 -1.35007143e-01 -6.94135129e-01 7.42702305e-01
6.02773011e-01 2.74130911e-01 2.22557947e-01 -1.83560419e+00
-8.13977838e-01 -2.01804459e-01 -3.51832747e-01 8.16783488e-01
2.38149226e-01 6.53044105e-01 5.16553223e-01 -7.60369599e-01
1.62033319e-01 1.12877834e+00 3.35829258e-01 6.62697196e-01
-8.09686720e-01 -9.45827365e-01 2.60618597e-01 9.85254407e-01
-1.25847018e+00 -6.86514020e-01 7.63633788e-01 -6.85271323e-01
9.14439142e-01 3.34485501e-01 7.35301733e-01 1.06070936e+00
5.61698258e-01 1.03280950e+00 7.89078176e-01 -2.09827676e-01
5.47048822e-02 2.60961324e-01 5.95397651e-01 2.63488978e-01
1.12675287e-01 5.10243952e-01 -7.23179579e-01 4.27493721e-01
2.31452167e-01 2.49824122e-01 -1.34820402e-01 -5.26965596e-02
-9.57881629e-01 5.50254047e-01 3.19075793e-01 -1.38795540e-01
-6.07415915e-01 -4.80275564e-02 2.07435519e-01 9.20255557e-02
4.24952388e-01 1.58791631e-01 -1.41703442e-01 -5.56971729e-01
-2.56077647e-01 1.03651099e-01 4.56548572e-01 1.36285889e+00
7.58135974e-01 2.97709182e-02 -6.73955023e-01 6.21643126e-01
2.40941972e-01 6.69352591e-01 1.62418559e-01 -9.76424277e-01
1.71381727e-01 5.20667613e-01 2.18924969e-01 -7.92473972e-01
-5.62467039e-01 -3.23903292e-01 -1.90734684e-01 3.76466960e-01
1.70339659e-01 -2.01835632e-01 -7.16131330e-01 1.59118879e+00
2.36033380e-01 3.34228188e-01 -2.11114079e-01 1.24820077e+00
9.69426811e-01 2.44524077e-01 4.39454645e-01 -7.99959823e-02
1.70723319e+00 -1.06299031e+00 -1.31267500e+00 -8.33611548e-01
4.75737959e-01 -9.10634696e-01 1.06805491e+00 2.52560318e-01
-1.05983973e+00 -1.02407181e+00 -7.86280990e-01 -4.14446205e-01
-2.64355510e-01 1.32875666e-01 6.47711456e-01 4.90022153e-01
-8.40509772e-01 -3.65352362e-01 -7.08071768e-01 -5.24790823e-01
4.55609739e-01 9.02494937e-02 -1.84700876e-01 -1.29138499e-01
-1.13366473e+00 1.45308578e+00 -1.33889109e-01 1.40116304e-01
-7.99697518e-01 -7.92347372e-01 -9.50628936e-01 8.27446580e-04
7.31548488e-01 -3.83611679e-01 1.68984771e+00 -5.58240831e-01
-1.13746881e+00 5.83724082e-01 -6.94153249e-01 -5.66330612e-01
2.43241370e-01 -6.41844690e-01 -8.01630139e-01 -4.12221700e-01
1.06129646e-01 7.90098488e-01 7.08257914e-01 -7.72959888e-01
-9.77959692e-01 -2.11600393e-01 1.27477825e-01 7.73023441e-02
4.16517034e-02 2.35072091e-01 -7.89972425e-01 -2.88981274e-02
-5.16058743e-01 -1.13917184e+00 1.43117411e-02 -6.56155348e-02
-1.77943379e-01 -5.60523748e-01 1.10779059e+00 -3.43875617e-01
1.46970320e+00 -2.47045970e+00 -1.56235740e-01 -5.31672657e-01
6.23198271e-01 5.99614978e-01 -1.11431424e-02 6.40192851e-02
4.62801233e-02 -2.06094429e-01 5.97893655e-01 -5.64105988e-01
7.39465207e-02 4.27548826e-01 -1.41666889e-01 4.23565060e-01
3.75267416e-01 1.06364703e+00 -8.13946545e-01 -5.04524410e-01
8.26640129e-01 5.66438675e-01 -6.24428868e-01 4.39171493e-01
1.44427031e-01 6.64158881e-01 -2.22779483e-01 4.16875064e-01
4.95110691e-01 2.55946070e-01 -7.29827046e-01 -8.90345797e-02
-6.74999475e-01 3.79191011e-01 -5.93584776e-01 1.15978193e+00
-4.61823404e-01 1.36524892e+00 -1.47624120e-01 -3.13550442e-01
8.64140511e-01 2.63359338e-01 2.01027527e-01 -1.14204907e+00
5.44219732e-01 -3.95554870e-01 4.78871912e-01 -8.41208994e-01
9.35483396e-01 3.21901143e-01 -3.57820578e-02 2.60666728e-01
-2.26833791e-01 3.22070390e-01 3.02726120e-01 -2.98172124e-02
7.62939870e-01 -3.25256675e-01 2.73258984e-01 -1.47084653e-01
5.24265885e-01 -8.67331773e-02 7.05757022e-01 7.86052108e-01
-1.01528358e+00 3.04171950e-01 3.65464270e-01 -4.29026067e-01
-6.97901666e-01 -6.72004819e-01 -1.45031689e-02 1.46061301e+00
3.94677997e-01 -3.81135017e-01 -8.68900955e-01 -1.36960790e-01
-3.71350199e-02 1.14573991e+00 -7.47927070e-01 -7.03778863e-01
-2.98507571e-01 -3.72724682e-02 2.17498496e-01 7.19308615e-01
3.12581897e-01 -1.22946882e+00 -9.02668536e-01 -2.21476838e-01
-4.26886648e-01 -1.23985088e+00 -8.61254990e-01 -5.58796264e-02
-8.93081948e-02 -1.27664614e+00 -1.19342946e-01 -5.72965801e-01
3.33476871e-01 8.57433498e-01 9.28924441e-01 1.02791980e-01
-3.45443726e-01 5.67852139e-01 -4.40425985e-02 -1.29596925e+00
-1.73143223e-01 1.42465174e-01 1.38649896e-01 1.29635081e-01
1.33919597e+00 1.50311694e-01 -7.32766569e-01 6.60555661e-01
-2.05984026e-01 2.51285166e-01 5.23161590e-01 2.68552333e-01
4.99881923e-01 -1.35947511e-01 4.94650334e-01 -3.38926136e-01
7.05571890e-01 -6.42667174e-01 -6.19065225e-01 -1.94724977e-01
-6.49462998e-01 -1.74297765e-01 8.28212276e-02 -5.00419378e-01
-1.07016242e+00 -1.02886520e-01 4.21610065e-02 -6.90461338e-01
-6.63034797e-01 6.50035590e-02 -1.42433718e-01 1.88226759e-01
5.36424935e-01 -4.51749079e-02 2.45355606e-01 -3.24777693e-01
4.79875654e-02 8.24238837e-01 6.51691079e-01 9.63725075e-02
2.57404894e-01 1.73282459e-01 -9.25294682e-02 -9.12912488e-01
-1.00860763e+00 -8.65202010e-01 -5.42967200e-01 -7.27901340e-01
1.09522903e+00 -9.17535543e-01 -1.20030260e+00 3.55123311e-01
-9.94268775e-01 -2.88097113e-01 -2.82509625e-01 7.86615551e-01
-5.40735841e-01 -1.76794782e-01 -1.54729113e-01 -1.12467122e+00
2.49020979e-02 -1.41929972e+00 9.75120187e-01 3.94436777e-01
-3.57069522e-01 -5.67248762e-01 -2.25326285e-01 5.86304426e-01
5.55617154e-01 -2.44974583e-01 1.64306853e-02 -1.94892094e-01
-6.48431242e-01 -3.23036641e-01 -1.88499480e-01 1.40535101e-01
1.30839691e-01 -8.05766806e-02 -1.14426064e+00 1.78672895e-02
-4.45712246e-02 4.71559800e-02 6.84128463e-01 6.34270608e-01
9.69603777e-01 2.78874099e-01 -4.43468153e-01 2.60160029e-01
5.91830671e-01 3.49060118e-01 9.23723757e-01 2.88957864e-01
6.17047668e-01 6.18844151e-01 1.04841101e+00 3.51218611e-01
8.48735034e-01 8.38870287e-01 6.92130923e-01 -1.43359721e-01
-5.84558368e-01 -7.10649937e-02 2.70651698e-01 3.89270633e-01
-1.86625108e-01 -3.93306196e-01 -1.08377850e+00 7.95009255e-01
-1.93053436e+00 -1.14080930e+00 -5.61230779e-01 2.06310940e+00
2.64286429e-01 1.97338879e-01 2.41685778e-01 3.81108522e-02
8.16891491e-01 -1.05702847e-01 -8.06079924e-01 -5.97855151e-01
1.86591849e-01 -3.53226870e-01 5.91777861e-01 6.67406857e-01
-8.98203909e-01 8.54963124e-01 5.97724056e+00 4.02971536e-01
-9.88328040e-01 4.01652068e-01 1.00859657e-01 -6.04705453e-01
2.64030665e-01 -1.52779758e-01 -1.20668435e+00 6.94424331e-01
1.22530496e+00 -3.20260435e-01 2.63365239e-01 9.15905118e-01
6.34816229e-01 -4.01057690e-01 -1.08410430e+00 1.07367098e+00
2.05379963e-01 -6.43035710e-01 -7.08483756e-01 3.45547169e-01
-1.08470703e-02 3.29557568e-01 3.41499507e-01 7.14255095e-01
-5.39927483e-02 -8.70712519e-01 9.53250587e-01 9.37737644e-01
4.55961585e-01 -9.32470858e-01 7.97463477e-01 5.28308451e-01
-1.05811810e+00 -3.20294321e-01 -2.44748428e-01 -4.92630690e-01
5.10392010e-01 3.64677310e-02 -5.96967041e-01 -2.47091100e-01
7.53806531e-01 8.46613705e-01 -7.92034030e-01 1.09213686e+00
-2.33764023e-01 6.90240800e-01 4.30084094e-02 1.50088351e-02
4.17537009e-03 1.60714209e-01 6.41378164e-01 1.09457374e+00
1.38990492e-01 2.18485966e-01 8.35151821e-02 7.82339454e-01
3.08761895e-01 -5.09189248e-01 -7.00457036e-01 2.47138113e-01
5.25564909e-01 1.09492743e+00 -4.05017100e-03 -5.36211096e-02
-7.83513069e-01 3.55372965e-01 1.71265930e-01 4.32120591e-01
-1.27257037e+00 -2.69976705e-01 1.53817415e+00 5.36404610e-01
1.99920252e-01 -2.78982729e-01 -2.47184038e-01 -6.06106341e-01
-1.31511644e-01 -2.87801236e-01 1.41913354e-01 -1.18479133e+00
-6.94584966e-01 5.40434420e-01 1.10806741e-01 -1.37142408e+00
-9.55813080e-02 -5.95117688e-01 -4.85230893e-01 1.10481775e+00
-1.88427401e+00 -8.94220352e-01 -8.53008687e-01 6.19913995e-01
8.72493863e-01 -2.44253129e-01 2.97866881e-01 4.64431733e-01
-1.07699811e+00 6.45680428e-01 -7.13269234e-01 -2.37376913e-01
9.29048657e-01 -8.21799338e-01 3.92346680e-01 9.07650650e-01
-5.06838620e-01 3.29637587e-01 1.00435328e+00 -4.17917550e-01
-1.42831993e+00 -1.16861737e+00 9.95354831e-01 -1.05512571e+00
4.63247031e-01 -2.09429115e-01 -1.01028216e+00 9.80176270e-01
5.27738690e-01 6.44963533e-02 6.55404627e-01 1.03230782e-01
1.31729394e-01 -6.37534931e-02 -6.82850897e-01 8.50106418e-01
9.91080344e-01 -6.38249576e-01 -6.09833777e-01 9.12593827e-02
6.50705099e-01 -6.66733682e-01 -5.62931299e-01 6.42402470e-02
5.22780776e-01 -9.96355772e-01 8.65027428e-01 -4.74586904e-01
6.56870157e-02 -3.47247064e-01 2.50699699e-01 -1.20521557e+00
-8.37241709e-01 -4.96238053e-01 -3.88214052e-01 1.00250947e+00
1.07082620e-01 -8.36433053e-01 3.07335317e-01 1.05521715e+00
-6.37900472e-01 -4.46348310e-01 -6.33994162e-01 -6.49280310e-01
-3.93090814e-01 -1.02482951e+00 7.78677344e-01 3.66736919e-01
1.24109961e-01 5.15209079e-01 -6.27399564e-01 3.13009769e-01
3.60309094e-01 -3.08321089e-01 1.14775360e+00 -1.33621073e+00
3.47003639e-01 -4.63613629e-01 -6.22179806e-01 -1.15960968e+00
3.03264618e-01 -4.23423320e-01 3.33197147e-01 -1.29245067e+00
-3.50601152e-02 -1.58449814e-01 -4.16942149e-01 5.24309039e-01
-4.07367259e-01 5.59050478e-02 1.45003691e-01 1.84233695e-01
-6.84836805e-01 4.82897133e-01 1.34113741e+00 2.21682638e-01
-1.53340414e-01 3.19882631e-02 -1.04247880e+00 5.65225959e-01
1.45579731e+00 -2.69153595e-01 -5.63798189e-01 -6.32683039e-01
-1.44239709e-01 -8.80353674e-02 5.69695771e-01 -1.05887294e+00
8.18694949e-01 -4.94009972e-01 -8.60710293e-02 -8.50737870e-01
6.36119425e-01 -8.55762601e-01 8.25538412e-02 2.27609828e-01
-2.39488572e-01 1.06274806e-01 7.93408513e-01 5.62128186e-01
-1.05832472e-01 1.10649183e-01 6.23428583e-01 4.84924197e-01
-1.13011634e+00 4.08921868e-01 -9.10070300e-01 3.13430540e-02
1.37365198e+00 -3.44392896e-01 -7.47086763e-01 -3.94725800e-01
-5.72487235e-01 5.88598251e-01 2.22253516e-01 1.23328853e+00
5.07806420e-01 -1.29720497e+00 -5.72782338e-01 5.97009659e-01
6.20401740e-01 -9.27449018e-02 4.35389906e-01 1.45236254e+00
1.56593889e-01 9.30914581e-01 -2.86419988e-01 -9.03180838e-01
-1.32746959e+00 9.51305747e-01 2.69273907e-01 5.88312149e-01
-6.00167453e-01 3.38862538e-01 5.47888994e-01 8.50093514e-02
1.58133119e-01 -2.92541623e-01 -5.49968779e-01 -1.35087296e-01
9.04668808e-01 6.56511664e-01 4.17537689e-02 -1.20412171e+00
-3.73511583e-01 2.15673953e-01 -1.90738246e-01 3.62533867e-01
6.30662084e-01 -6.13156438e-01 4.38456953e-01 3.74610901e-01
5.76590121e-01 -1.85897365e-01 -1.59772801e+00 -1.17436953e-01
-3.26859415e-01 -6.31575942e-01 6.34536982e-01 -6.36400521e-01
-1.04842293e+00 1.00534856e+00 7.01434195e-01 1.91939369e-01
1.06482399e+00 2.92715132e-01 7.85370529e-01 9.08420905e-02
1.87279359e-01 -1.01440167e+00 -2.04279855e-01 6.07963860e-01
1.01578450e+00 -1.57964325e+00 -4.30584133e-01 -3.51065516e-01
-1.01691043e+00 3.46162915e-01 1.23401594e+00 2.96642452e-01
6.79330289e-01 2.21548587e-01 2.92284459e-01 -3.02632451e-01
-1.13571167e+00 -8.41660202e-01 6.43080652e-01 7.44692266e-01
3.40791970e-01 9.79688093e-02 4.87361997e-02 5.94710410e-01
-3.77249122e-01 2.07809642e-01 4.19555247e-01 8.55509460e-01
-5.66029608e-01 -5.34347296e-01 -3.74547780e-01 3.72994870e-01
-1.09822683e-01 1.41722172e-01 -1.72804296e-01 7.46149838e-01
3.70017350e-01 1.51803124e+00 4.51196313e-01 -4.55124050e-01
9.19247806e-01 3.00620007e-03 1.92891655e-03 -4.93572325e-01
-2.72445112e-01 -2.58406907e-01 -5.00412397e-02 -7.61275232e-01
-3.11114192e-01 -7.49094129e-01 -8.42461288e-01 -6.60025656e-01
-4.02763546e-01 -9.53397378e-02 5.19209802e-01 8.53174150e-01
8.22073042e-01 1.07128608e+00 1.84591785e-01 -9.89132047e-01
1.89028129e-01 -1.03583157e+00 -2.37322018e-01 1.38906315e-01
7.38718152e-01 -1.24829590e+00 -2.65472859e-01 -5.74467257e-02] | [7.59960412979126, -0.05926382541656494] |
455133e7-835a-4aa2-8ff0-fa2f37f6e7ab | lite-pose-efficient-architecture-design-for | 2205.01271 | null | https://arxiv.org/abs/2205.01271v4 | https://arxiv.org/pdf/2205.01271v4.pdf | Lite Pose: Efficient Architecture Design for 2D Human Pose Estimation | Pose estimation plays a critical role in human-centered vision applications. However, it is difficult to deploy state-of-the-art HRNet-based pose estimation models on resource-constrained edge devices due to the high computational cost (more than 150 GMACs per frame). In this paper, we study efficient architecture design for real-time multi-person pose estimation on edge. We reveal that HRNet's high-resolution branches are redundant for models at the low-computation region via our gradual shrinking experiments. Removing them improves both efficiency and performance. Inspired by this finding, we design LitePose, an efficient single-branch architecture for pose estimation, and introduce two simple approaches to enhance the capacity of LitePose, including Fusion Deconv Head and Large Kernel Convs. Fusion Deconv Head removes the redundancy in high-resolution branches, allowing scale-aware feature fusion with low overhead. Large Kernel Convs significantly improve the model's capacity and receptive field while maintaining a low computational cost. With only 25% computation increment, 7x7 kernels achieve +14.0 mAP better than 3x3 kernels on the CrowdPose dataset. On mobile platforms, LitePose reduces the latency by up to 5.0x without sacrificing performance, compared with prior state-of-the-art efficient pose estimation models, pushing the frontier of real-time multi-person pose estimation on edge. Our code and pre-trained models are released at https://github.com/mit-han-lab/litepose. | ['Song Han', 'Wei-Ming Chen', 'Han Cai', 'Muyang Li', 'Yihan Wang'] | 2022-05-03 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Wang_Lite_Pose_Efficient_Architecture_Design_for_2D_Human_Pose_Estimation_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Wang_Lite_Pose_Efficient_Architecture_Design_for_2D_Human_Pose_Estimation_CVPR_2022_paper.pdf | cvpr-2022-1 | ['2d-human-pose-estimation', 'multi-person-pose-estimation'] | ['computer-vision', 'computer-vision'] | [-5.62517941e-01 -1.62295520e-01 1.25884578e-01 -2.87202418e-01
-6.78580344e-01 -2.30806187e-01 6.98525161e-02 -3.14450651e-01
-7.38154709e-01 4.29154515e-01 3.36914808e-01 -1.14972740e-02
2.90747911e-01 -6.10504627e-01 -6.67071462e-01 -2.54700154e-01
-3.08245923e-02 2.44576797e-01 5.25605857e-01 -1.65287524e-01
-2.57404894e-01 3.71526152e-01 -1.54931569e+00 1.02951773e-01
5.80767274e-01 1.09791231e+00 1.05311982e-01 9.18934286e-01
4.55368817e-01 4.17693138e-01 -5.37862480e-01 -4.60471153e-01
3.11716348e-01 1.67473972e-01 -4.24581975e-01 -1.86577633e-01
9.40798819e-01 -7.66301453e-01 -7.73530185e-01 5.92016935e-01
1.16447318e+00 -7.52248392e-02 1.41321063e-01 -1.10195816e+00
-2.07097277e-01 1.07606106e-01 -8.23347926e-01 3.92394572e-01
2.93853641e-01 2.42121398e-01 5.23235083e-01 -1.12248921e+00
4.08986628e-01 1.03845894e+00 1.13507366e+00 5.27842760e-01
-7.58064151e-01 -5.44064224e-01 1.29137203e-01 8.64771679e-02
-1.84850109e+00 -7.21899331e-01 3.74659926e-01 -7.47893155e-02
1.18115270e+00 3.25799048e-01 8.85087729e-01 8.80262733e-01
3.01586360e-01 7.21787870e-01 7.91609287e-01 -5.12984060e-02
2.44861245e-02 -2.80193657e-01 3.08945384e-02 1.02624714e+00
4.59252179e-01 -3.72205861e-02 -9.80587363e-01 -8.15940574e-02
1.10722625e+00 9.05948579e-02 -5.21948397e-01 -1.77557275e-01
-9.81012106e-01 4.97740000e-01 6.67851031e-01 -5.29241934e-02
-4.78633553e-01 7.84034610e-01 4.71865892e-01 -2.00095966e-01
5.34725428e-01 9.32826996e-02 -5.45791268e-01 -3.71170580e-01
-9.80714738e-01 2.82858431e-01 4.03567582e-01 1.16671371e+00
4.31197435e-01 -6.91575110e-02 -2.50104100e-01 5.55742383e-01
2.11492509e-01 7.10124612e-01 3.36301103e-02 -8.17589104e-01
4.99942452e-01 4.02726859e-01 1.29264712e-01 -8.60753715e-01
-7.77453721e-01 -7.89047062e-01 -7.43460596e-01 -1.46700561e-01
5.31657517e-01 -5.56728899e-01 -8.49829197e-01 1.58032942e+00
5.11519790e-01 2.98308015e-01 -4.65193391e-01 1.32910132e+00
1.00747120e+00 3.65693778e-01 7.85097554e-02 1.17645502e-01
1.90303218e+00 -1.09343302e+00 -3.67658436e-01 -5.28998852e-01
6.51433170e-01 -6.57036960e-01 9.18837726e-01 2.07473531e-01
-1.06455398e+00 -6.43848777e-01 -9.59635854e-01 -3.40921581e-01
3.32817100e-02 6.67860508e-01 1.13270688e+00 8.19950998e-01
-1.27302897e+00 3.53445679e-01 -1.06439984e+00 -4.38113481e-01
5.70016086e-01 6.86502695e-01 -3.84783596e-01 -2.76116244e-02
-8.15832019e-01 5.92279911e-01 -1.81244463e-02 2.91566491e-01
-3.78155619e-01 -8.70435417e-01 -8.06748331e-01 1.38147116e-01
5.75224817e-01 -1.05951452e+00 1.24452209e+00 -3.57730776e-01
-1.32483137e+00 5.82996726e-01 -2.73050934e-01 -4.63688433e-01
6.28548503e-01 -7.54733622e-01 -2.94184685e-01 2.75456965e-01
-4.22981940e-02 8.70207608e-01 7.42968380e-01 -7.48254955e-01
-5.66689432e-01 -6.75423026e-01 -2.71584224e-02 5.09676516e-01
-5.93956470e-01 6.67575300e-02 -1.24628520e+00 -4.65184182e-01
8.66785794e-02 -1.13089371e+00 -3.39346558e-01 1.66080698e-01
-8.70398879e-02 -1.19542787e-02 6.94157064e-01 -7.54691720e-01
1.27455831e+00 -2.00351930e+00 -2.80267209e-01 -8.66225362e-02
6.30439639e-01 3.09365422e-01 1.11577988e-01 -3.96144390e-02
3.93471986e-01 -2.73944914e-01 4.09093350e-01 -7.37243295e-01
-2.00587630e-01 -2.14918047e-01 1.36817738e-01 6.99083745e-01
-2.58758396e-01 1.26965976e+00 -5.22946477e-01 -5.47981620e-01
3.46237093e-01 9.75044131e-01 -5.80480993e-01 -9.36252326e-02
3.73195648e-01 1.60599545e-01 -3.67285997e-01 7.14656770e-01
8.33266079e-01 -4.89405632e-01 1.23862505e-01 -5.94324529e-01
-1.03033110e-01 8.03605914e-02 -1.29186726e+00 1.85743225e+00
-4.11918968e-01 7.15196609e-01 8.05004686e-02 -2.47585833e-01
6.43714249e-01 6.38077259e-02 3.97684366e-01 -5.22610366e-01
4.42152679e-01 -4.90938406e-03 -2.66379744e-01 2.51067858e-02
8.44307065e-01 4.03910249e-01 -1.36353850e-01 3.96685004e-02
-2.47306321e-02 4.42872643e-01 -7.73908421e-02 1.05928183e-01
1.12844169e+00 1.51426479e-01 1.25905767e-01 -3.25217903e-01
1.06793597e-01 -2.76449889e-01 6.28695250e-01 6.17254078e-01
-4.11050111e-01 7.90653944e-01 4.14786069e-03 -7.50544488e-01
-8.06198180e-01 -8.99136424e-01 -3.36532034e-02 1.22473264e+00
2.61290222e-01 -9.80738461e-01 -9.52309549e-01 -3.75311434e-01
-8.95375460e-02 -7.03878850e-02 -3.80458862e-01 7.03710616e-02
-8.07431936e-01 -8.22916448e-01 7.69096792e-01 1.11324430e+00
1.03082943e+00 -5.43025076e-01 -1.21342635e+00 6.77263504e-03
-2.46877193e-01 -1.32049704e+00 -8.64695847e-01 -2.64701903e-01
-7.49458611e-01 -8.36720526e-01 -9.40591753e-01 -5.36684036e-01
6.37111485e-01 4.86875027e-01 1.19145358e+00 7.57433102e-02
-4.53438878e-01 3.73681754e-01 -1.14019029e-01 -3.54599178e-01
7.62304842e-01 2.69775838e-01 3.39867264e-01 -3.09032202e-01
4.37666386e-01 -3.56938630e-01 -1.17541885e+00 3.49417150e-01
-9.76517946e-02 1.62733957e-01 5.92150748e-01 5.71503222e-01
5.45607746e-01 -2.12525427e-01 1.42124787e-01 -3.04717839e-01
2.47715622e-01 2.02679917e-01 -5.94127476e-01 1.42304793e-01
-2.50989705e-01 -7.37756267e-02 3.72129261e-01 -4.16169614e-01
-9.76458371e-01 2.41564780e-01 -1.10668717e-02 -5.82674921e-01
1.27398625e-01 1.56134348e-02 -4.19988409e-02 -3.79731804e-01
7.72758186e-01 -2.42202245e-02 -2.43585229e-01 -3.54288280e-01
1.72419757e-01 5.98098457e-01 6.79021776e-01 -4.93378162e-01
4.84931380e-01 6.29799843e-01 4.34769690e-02 -1.04180968e+00
-6.18662775e-01 -6.24299526e-01 -4.19378936e-01 -3.63749057e-01
9.19323027e-01 -1.63407767e+00 -1.24061799e+00 5.95657527e-01
-1.07259357e+00 -3.41993868e-01 1.03767224e-01 7.15620160e-01
-3.27624828e-01 1.78068221e-01 -8.39180887e-01 -7.60808885e-01
-8.59508157e-01 -9.42342758e-01 1.48311198e+00 5.44507682e-01
-2.14498982e-01 -4.04380113e-01 -3.65246564e-01 5.90003729e-01
4.44452256e-01 -8.69369358e-02 -1.61720589e-01 1.11751169e-01
-6.84514165e-01 -1.83218941e-01 -4.37632054e-01 -2.50296980e-01
-4.74092931e-01 -6.00997269e-01 -1.12461114e+00 -6.23391569e-01
-2.84243673e-01 -1.77570626e-01 8.55418444e-01 7.84093618e-01
1.12894213e+00 1.69124320e-01 -5.08389831e-01 1.20120168e+00
1.26127601e+00 -3.49126846e-01 7.95484424e-01 8.55893269e-02
1.03859067e+00 7.81518444e-02 6.28235459e-01 6.42106175e-01
7.15628445e-01 9.93899882e-01 1.50189310e-01 -2.78711468e-01
-4.02160794e-01 -2.39099234e-01 2.89954662e-01 4.34075654e-01
-6.47190630e-01 -1.29999042e-01 -8.58996987e-01 3.36553752e-01
-1.95385432e+00 -7.27262974e-01 -1.57365635e-01 2.36914444e+00
3.04551274e-01 1.73215359e-01 4.12711233e-01 -4.20200884e-01
6.49815798e-01 1.67033494e-01 -4.76900429e-01 8.43152851e-02
9.81972292e-02 2.40800396e-01 9.95820642e-01 3.89840066e-01
-1.36634898e+00 1.17514789e+00 5.49546051e+00 8.96768868e-01
-1.06756616e+00 2.86113828e-01 7.38192260e-01 -6.55189812e-01
3.74882340e-01 -2.15759516e-01 -1.34093094e+00 4.08358276e-01
8.29702854e-01 3.33561242e-01 3.12552571e-01 1.26462376e+00
-3.41859870e-02 -2.87288755e-01 -8.47295523e-01 1.71037173e+00
2.67369807e-01 -1.38782823e+00 -5.23763835e-01 4.15854931e-01
3.56694996e-01 2.77144730e-01 -1.47667915e-01 3.49250436e-01
-2.13900909e-01 -9.34175313e-01 6.57956183e-01 3.70668590e-01
1.15693390e+00 -9.72253323e-01 8.84730637e-01 1.25420883e-01
-1.74516320e+00 3.05849984e-02 -5.90604305e-01 -2.20780671e-01
3.16952199e-01 5.40070295e-01 -5.15358806e-01 3.16537052e-01
1.14739704e+00 2.73460597e-01 -8.47774863e-01 8.79273593e-01
-3.29108188e-05 2.24039868e-01 -6.51101410e-01 -1.25685602e-01
-1.94011793e-01 2.08338290e-01 3.22412640e-01 1.27596915e+00
4.08511937e-01 3.57734501e-01 3.03776622e-01 3.22225243e-01
-1.56367719e-01 -2.04898685e-01 -3.29307228e-01 5.49280584e-01
6.03238881e-01 1.37363052e+00 -8.97281885e-01 -3.38243395e-01
-4.52530444e-01 1.35093534e+00 4.32005674e-01 1.67327717e-01
-1.30998850e+00 -2.66580909e-01 8.14078033e-01 4.95492518e-01
4.16080147e-01 -5.36283493e-01 -4.01295811e-01 -1.34279037e+00
3.07295799e-01 -6.27583683e-01 1.55032381e-01 -6.50170803e-01
-7.28545845e-01 6.81098759e-01 -9.90592167e-02 -1.11378634e+00
-8.08970258e-02 -6.42767966e-01 -3.69645417e-01 6.98524535e-01
-1.05102134e+00 -1.48643672e+00 -7.75067329e-01 6.62706912e-01
4.87276971e-01 -9.25080676e-04 6.55390322e-01 4.97817844e-01
-6.48364007e-01 1.16514242e+00 -4.72953409e-01 3.50804299e-01
6.84333444e-01 -9.14379954e-01 8.04191411e-01 1.00049913e+00
4.15463373e-03 7.70345032e-01 4.32500482e-01 -8.07365417e-01
-1.72138762e+00 -9.63069677e-01 7.04688013e-01 -5.12213767e-01
1.76112745e-02 -6.96766019e-01 -3.20495814e-01 6.31623685e-01
-9.13255066e-02 5.93432248e-01 4.15449113e-01 4.27405745e-01
-2.04490185e-01 -1.40071824e-01 -8.99786472e-01 9.19505239e-01
1.58871591e+00 -3.84005696e-01 -6.95866719e-02 1.87177122e-01
5.62924266e-01 -7.71991849e-01 -8.15335631e-01 4.24124360e-01
7.19746351e-01 -1.00418890e+00 1.26132739e+00 -1.16429232e-01
-5.78332990e-02 -4.31474328e-01 -6.76102787e-02 -7.93921053e-01
-6.56336427e-01 -7.27398574e-01 -6.17502749e-01 7.33342588e-01
7.87418485e-02 -5.84435761e-01 1.23109353e+00 7.78354526e-01
-1.53352574e-01 -1.16693902e+00 -9.56758142e-01 -7.02552438e-01
-5.60747206e-01 -4.28805441e-01 4.18489665e-01 3.55825096e-01
-1.94951192e-01 5.80274284e-01 -6.35064304e-01 3.41268659e-01
6.42344296e-01 -1.98431090e-01 1.16628277e+00 -8.75970781e-01
-4.65576470e-01 -9.46164876e-02 -6.85979068e-01 -1.44621027e+00
-4.01330531e-01 -2.13925779e-01 -2.55158156e-01 -1.42603385e+00
2.38966614e-01 -3.07303905e-01 2.07857043e-01 5.16124785e-01
-2.31847540e-01 7.53535032e-01 5.13063192e-01 5.42389527e-02
-8.25136900e-01 5.53633213e-01 1.01799977e+00 3.74947757e-01
-2.67868102e-01 -2.36763611e-01 -6.03688180e-01 1.03648388e+00
7.59920895e-01 -1.15804404e-01 -3.96383464e-01 -5.79845965e-01
1.70269489e-01 -1.70500092e-02 6.13976061e-01 -1.62077928e+00
5.24241388e-01 2.93143898e-01 1.04331994e+00 -6.61910236e-01
9.44573402e-01 -4.77743596e-01 1.43095911e-01 5.26093721e-01
3.92040104e-01 2.77113467e-01 3.25174600e-01 2.52636790e-01
2.33588815e-01 4.38888013e-01 6.29088163e-01 -9.50424224e-02
-9.15127337e-01 6.47479892e-01 1.91486448e-01 -2.26270184e-02
9.47908998e-01 -3.57016295e-01 -5.67270935e-01 -4.92767423e-01
-5.37937105e-01 1.78510323e-01 4.90044743e-01 3.33468318e-01
7.50008106e-01 -1.24708736e+00 -6.25333071e-01 2.05698863e-01
-8.45739096e-02 3.95026915e-02 7.32811749e-01 1.00119400e+00
-8.19632053e-01 5.06899297e-01 -8.31401870e-02 -6.28773630e-01
-1.53454471e+00 2.31206670e-01 4.21548039e-01 -2.22159714e-01
-8.80019963e-01 1.12076020e+00 2.88961589e-01 -5.96480928e-02
1.27251416e-01 -1.00776032e-01 1.06352590e-01 -2.48177081e-01
8.72072041e-01 6.68667257e-01 7.92118460e-02 -6.91960216e-01
-7.20031083e-01 8.24657023e-01 -1.77202687e-01 -7.49024227e-02
1.09497964e+00 -9.33045819e-02 3.94632548e-01 -3.11928749e-01
1.00022733e+00 1.10285714e-01 -1.56146395e+00 6.02384694e-02
-6.62338376e-01 -6.32070720e-01 4.42477226e-01 -5.18335044e-01
-1.12093365e+00 4.74022210e-01 9.36762869e-01 -5.74565649e-01
1.14718986e+00 1.17540225e-01 1.08564639e+00 2.99160838e-01
8.78738761e-01 -1.30454814e+00 9.65236202e-02 3.89849365e-01
6.84252560e-01 -1.11685050e+00 3.71265441e-01 -9.48363304e-01
-5.80331266e-01 6.99171841e-01 1.11915684e+00 -1.79736406e-01
5.31436384e-01 4.93229061e-01 -1.14466473e-01 -5.01115143e-01
-3.49128872e-01 -3.93717319e-01 4.52854991e-01 5.99049509e-01
4.55411494e-01 2.61675835e-01 -2.94777215e-01 6.46984816e-01
-4.57146555e-01 1.24988891e-02 1.39173925e-01 8.56243968e-01
-2.54517525e-01 -5.77081919e-01 -5.04918337e-01 4.25768495e-01
-3.86180878e-01 -1.79054573e-01 8.63783807e-03 8.22995842e-01
1.84519157e-01 8.52925241e-01 2.21317530e-01 -6.07099593e-01
2.72483200e-01 -2.96392530e-01 6.20731354e-01 -3.68065029e-01
-5.58547318e-01 2.95644581e-01 2.46599168e-01 -1.04237342e+00
-4.15876806e-02 -4.97713685e-01 -1.15772152e+00 -7.64787674e-01
-2.94787318e-01 -3.22281599e-01 5.72112322e-01 6.63835585e-01
9.57067370e-01 5.69327414e-01 -3.90865728e-02 -1.10240066e+00
-3.57812822e-01 -9.08960283e-01 -3.61702651e-01 -4.78112698e-02
-7.41937682e-02 -6.82907701e-01 1.87932938e-01 -2.65230894e-01] | [7.214210033416748, -0.7235434055328369] |
5a98875f-8cf7-4c4c-a7fe-b6d0276274e1 | toward-fair-facial-expression-recognition | 2306.06696 | null | https://arxiv.org/abs/2306.06696v1 | https://arxiv.org/pdf/2306.06696v1.pdf | Toward Fair Facial Expression Recognition with Improved Distribution Alignment | We present a novel approach to mitigate bias in facial expression recognition (FER) models. Our method aims to reduce sensitive attribute information such as gender, age, or race, in the embeddings produced by FER models. We employ a kernel mean shrinkage estimator to estimate the kernel mean of the distributions of the embeddings associated with different sensitive attribute groups, such as young and old, in the Hilbert space. Using this estimation, we calculate the maximum mean discrepancy (MMD) distance between the distributions and incorporate it in the classifier loss along with an adversarial loss, which is then minimized through the learning process to improve the distribution alignment. Our method makes sensitive attributes less recognizable for the model, which in turn promotes fairness. Additionally, for the first time, we analyze the notion of attractiveness as an important sensitive attribute in FER models and demonstrate that FER models can indeed exhibit biases towards more attractive faces. To prove the efficacy of our model in reducing bias regarding different sensitive attributes (including the newly proposed attractiveness attribute), we perform several experiments on two widely used datasets, CelebA and RAF-DB. The results in terms of both accuracy and fairness measures outperform the state-of-the-art in most cases, demonstrating the effectiveness of the proposed method. | ['Ali Etemad', 'Mojtaba Kolahdouzi'] | 2023-06-11 | null | null | null | null | ['facial-expression-recognition'] | ['computer-vision'] | [ 8.05656835e-02 2.57983208e-01 -7.20316321e-02 -7.28111804e-01
-3.23826730e-01 -1.93540588e-01 3.83503407e-01 2.48025015e-01
-6.70827329e-01 6.60497725e-01 2.00569391e-01 2.90844262e-01
-1.11887693e-01 -7.38027334e-01 -5.73052227e-01 -8.87424707e-01
-8.86291713e-02 6.80955127e-02 -4.23613608e-01 -2.68510669e-01
2.38862306e-01 6.13257945e-01 -1.63661361e+00 -6.73217848e-02
1.03497756e+00 1.27617347e+00 -6.90533400e-01 -1.40072741e-02
2.38499418e-01 5.96019924e-01 -5.88682532e-01 -1.06788063e+00
2.92814523e-01 -2.00752750e-01 -4.31305736e-01 -1.47227615e-01
5.89209557e-01 -3.18824768e-01 2.82792863e-03 1.14066768e+00
5.28218389e-01 1.93934768e-01 9.28517520e-01 -1.43334985e+00
-7.18657613e-01 2.80203760e-01 -7.91505218e-01 -1.13363974e-01
1.39596641e-01 -1.85242862e-01 9.89083111e-01 -1.04118299e+00
4.48408276e-01 1.35886967e+00 7.02707112e-01 7.17367649e-01
-1.27625942e+00 -1.08434081e+00 1.75048307e-01 1.16027735e-01
-1.53019834e+00 -7.34583974e-01 9.14148390e-01 -4.21908468e-01
2.86107492e-02 3.21708739e-01 4.37111884e-01 9.93415892e-01
1.04701333e-01 4.68691826e-01 1.26956952e+00 -4.86889720e-01
2.58185595e-01 6.21928811e-01 2.75805537e-02 6.51786208e-01
8.32415372e-02 1.27430871e-01 -4.73860264e-01 -3.42415154e-01
2.56497085e-01 -1.10054187e-01 -1.35481268e-01 -5.13069689e-01
-5.75335205e-01 1.11105394e+00 3.29894036e-01 2.42437366e-02
-4.69639987e-01 -1.37546018e-01 3.52120906e-01 5.03869802e-02
9.05891478e-01 3.00722986e-01 -4.67755497e-02 2.17107162e-01
-5.28740644e-01 2.68673211e-01 5.52291989e-01 3.70809257e-01
7.94942975e-01 -9.55711529e-02 -4.63285983e-01 1.10822499e+00
2.69556582e-01 5.85863411e-01 2.64608026e-01 -9.38028395e-01
1.46024957e-01 4.66354966e-01 -1.25151668e-02 -1.39932489e+00
-1.11635327e-01 -3.53171527e-01 -8.56261969e-01 3.93443704e-01
4.56661642e-01 1.71091575e-02 -4.53218937e-01 2.28319716e+00
5.16906977e-01 1.81739614e-03 -9.43344682e-02 7.94472456e-01
4.17118579e-01 2.05316499e-01 6.35770738e-01 -2.67742217e-01
1.31710982e+00 -5.19810140e-01 -6.86366081e-01 -3.76297310e-02
4.88562435e-01 -5.67214906e-01 9.89353895e-01 2.24214777e-01
-9.97143567e-01 -4.26051259e-01 -8.88295829e-01 1.69299141e-01
-2.71018118e-01 3.45007807e-01 6.33263767e-01 9.22795296e-01
-6.91695511e-01 7.46483028e-01 -2.99144357e-01 -1.02267727e-01
8.20665121e-01 4.47098196e-01 -7.12176442e-01 9.31522697e-02
-1.37505198e+00 9.04168189e-01 -2.11421087e-01 3.88845243e-02
-4.50982839e-01 -9.86271024e-01 -8.68967950e-01 9.47635770e-02
-3.47114354e-02 -3.39673817e-01 7.01207399e-01 -1.41216290e+00
-1.38464844e+00 1.16653013e+00 2.02783551e-02 -2.67593741e-01
5.86614251e-01 -9.35317427e-02 -3.83495569e-01 -1.01857707e-01
-6.34497330e-02 6.81405783e-01 1.19681692e+00 -1.16156733e+00
-3.57681543e-01 -7.74845362e-01 -2.48154383e-02 1.52947456e-01
-8.56740713e-01 6.00920357e-02 2.04754904e-01 -4.96392101e-01
-4.31700677e-01 -8.27051163e-01 1.77098393e-01 4.53650326e-01
-1.73410401e-01 -1.75105378e-01 5.19544482e-01 -7.70626962e-01
1.34441292e+00 -2.61160612e+00 8.08392763e-02 4.84358460e-01
1.85103789e-01 2.90773034e-01 -1.22853793e-01 -1.62372321e-01
-2.22015679e-01 7.69934431e-02 -1.78124487e-01 -5.48115253e-01
1.56252235e-01 -8.41541123e-03 -2.55419731e-01 6.54589355e-01
3.84180188e-01 4.04865921e-01 -6.14771724e-01 -4.50779110e-01
-1.55117232e-02 7.86682487e-01 -6.11265481e-01 2.32080787e-01
3.84794384e-01 2.04668820e-01 -2.74702936e-01 4.40206796e-01
1.06017625e+00 5.27067721e-01 6.49487004e-02 -3.80497873e-01
1.39561519e-01 -1.81961641e-01 -9.21495855e-01 9.55878019e-01
-3.56164455e-01 3.89205068e-01 4.96204756e-02 -9.65109050e-01
1.34484744e+00 -1.26402546e-02 4.40506220e-01 -7.88157403e-01
2.65517443e-01 1.81476742e-01 -5.54038063e-02 -7.43274614e-02
3.90963852e-01 -3.74903917e-01 -3.73455253e-03 1.71274051e-01
-6.96433708e-02 3.09597075e-01 -2.05142364e-01 -1.19060315e-01
4.36417788e-01 -1.65241182e-01 2.39825189e-01 -4.09546256e-01
8.87190580e-01 -7.85859227e-01 5.79948008e-01 2.47975037e-01
-6.79157197e-01 2.89710075e-01 7.97540903e-01 -3.29121470e-01
-9.63653684e-01 -1.04782009e+00 -4.07526165e-01 1.16298866e+00
-2.41492521e-02 -1.51091859e-01 -9.73554671e-01 -9.36616063e-01
4.22690153e-01 7.77848780e-01 -1.12877238e+00 -7.67428577e-01
-1.45401120e-01 -9.08491373e-01 6.45665169e-01 4.07929987e-01
4.69611585e-01 -7.02259719e-01 -3.50310385e-01 -3.61423492e-01
2.81473063e-02 -8.09054136e-01 -4.61399406e-01 -2.60606349e-01
-4.91418242e-01 -8.42664540e-01 -5.89386702e-01 -4.28298026e-01
8.60212147e-01 -2.88187504e-01 8.22769821e-01 6.75526867e-03
-2.49060482e-01 3.30328315e-01 -1.06465213e-01 -6.47703588e-01
-2.62013555e-01 -2.55383044e-01 2.10648790e-01 7.12460339e-01
5.53041756e-01 -3.50614965e-01 -7.38164485e-01 4.07283038e-01
-6.83983803e-01 -4.57916290e-01 3.50884825e-01 7.97351003e-01
3.95193040e-01 -1.66284859e-01 7.15191424e-01 -9.30953622e-01
7.16073453e-01 -4.56024855e-01 -2.84092903e-01 1.50236145e-01
-8.21813345e-01 -2.58990675e-02 4.80553895e-01 -5.58150411e-01
-1.12092650e+00 -1.89464852e-01 -5.68808429e-02 -4.35073733e-01
5.26797809e-02 1.18456110e-02 -2.40438670e-01 -2.72806972e-01
5.74910045e-01 -2.38237187e-01 4.46409047e-01 -2.42548376e-01
2.93792486e-01 7.46099949e-01 2.51449317e-01 -6.00723445e-01
6.61655724e-01 6.00600183e-01 2.24030271e-01 -6.31174982e-01
-8.41474712e-01 7.94629008e-03 -4.25934553e-01 -2.78451532e-01
6.10544443e-01 -6.91956341e-01 -8.90991330e-01 5.13335824e-01
-8.70913565e-01 1.89437121e-01 -4.68686879e-01 3.98359507e-01
-3.89587343e-01 2.54559964e-01 -2.83315599e-01 -1.14063132e+00
-4.74974483e-01 -8.38478267e-01 8.28655541e-01 3.44638139e-01
-3.70512962e-01 -9.30108249e-01 -7.73705170e-02 2.60438800e-01
5.11064231e-01 4.51784939e-01 1.11131060e+00 -7.62575984e-01
2.65406549e-01 -3.20915610e-01 -2.79489011e-01 7.40757585e-01
1.30835235e-01 1.92663819e-01 -1.24274921e+00 -1.42693907e-01
-9.05827582e-02 -3.16029489e-01 7.97976196e-01 2.13985622e-01
1.31447411e+00 -2.90790886e-01 3.85589190e-02 5.14240623e-01
1.13542438e+00 3.16771381e-02 8.64046991e-01 1.49311885e-01
4.35151219e-01 1.04803538e+00 8.46044421e-01 8.72322023e-01
2.22604856e-01 7.11448908e-01 4.59761590e-01 -2.91327894e-01
2.33109012e-01 -2.44302332e-01 4.27384973e-01 3.25377971e-01
-1.50641635e-01 1.22794218e-01 -4.29001391e-01 2.83922315e-01
-1.53232229e+00 -8.91530871e-01 2.13889927e-01 2.52535057e+00
9.21772480e-01 -2.19116479e-01 1.88506901e-01 1.07638858e-01
7.81108379e-01 1.38977721e-01 -3.62154305e-01 -8.70592058e-01
-9.85667184e-02 4.43790942e-01 3.70239675e-01 4.47711051e-01
-1.13520920e+00 7.16380894e-01 6.05475378e+00 8.28193307e-01
-1.13694847e+00 -1.11486845e-01 1.10085499e+00 -1.85233265e-01
-3.40258688e-01 -1.82099223e-01 -5.46931863e-01 5.86786866e-01
6.65752590e-01 -3.20728153e-01 4.01192874e-01 1.02688205e+00
1.46058068e-01 2.75765080e-02 -1.01499343e+00 8.08052361e-01
3.49042982e-01 -6.82240188e-01 -1.16319386e-02 1.83356851e-01
4.76337433e-01 -7.15178132e-01 6.31792724e-01 3.78515661e-01
-2.26240084e-01 -1.15851355e+00 6.70613289e-01 7.21318722e-01
9.10302103e-01 -1.15893972e+00 8.30045879e-01 -1.43653423e-01
-6.57847166e-01 -2.07072198e-01 -5.55973887e-01 -1.90502573e-02
-4.53555375e-01 6.61762595e-01 -4.22775149e-01 2.12775603e-01
4.96425599e-01 5.09090722e-01 -6.00995600e-01 5.29665232e-01
-6.23688735e-02 2.46022180e-01 -3.77292447e-02 1.38059869e-01
-2.97906458e-01 -4.49046493e-01 3.35346013e-01 7.76659250e-01
3.34188223e-01 -1.84698120e-01 -3.88530225e-01 9.88478422e-01
-3.80456984e-01 4.94677126e-01 -5.43039203e-01 1.53293554e-02
5.17227292e-01 1.40301275e+00 -6.60622865e-02 -5.08891381e-02
-1.23261146e-01 8.44913960e-01 4.54331547e-01 2.11830810e-01
-8.95259380e-01 -4.40724254e-01 1.11151195e+00 1.59322545e-01
2.07916163e-02 3.62603992e-01 -2.28503898e-01 -8.17984641e-01
-2.29386818e-02 -9.67144847e-01 4.20350909e-01 -4.38292623e-01
-1.48764932e+00 4.19384331e-01 -9.26528350e-02 -9.62488890e-01
1.33942738e-02 -5.20609140e-01 -5.13445079e-01 9.95157242e-01
-1.63224924e+00 -1.04838896e+00 -2.60838538e-01 5.51345348e-01
-2.34693751e-01 -3.15217108e-01 9.76356626e-01 5.96528769e-01
-6.40949726e-01 1.28290462e+00 2.11517140e-03 2.14278102e-01
1.17097807e+00 -8.93574655e-01 -1.73362836e-01 3.78055841e-01
-3.10954958e-01 7.84112334e-01 5.82111180e-01 -2.28126809e-01
-8.75235617e-01 -9.66067493e-01 8.42528105e-01 -3.20915699e-01
3.89113605e-01 -3.83502871e-01 -8.51206005e-01 3.41950387e-01
-2.49663800e-01 1.42532393e-01 1.03550339e+00 2.30391830e-01
-7.82124162e-01 -6.34635568e-01 -1.65037525e+00 4.53163415e-01
7.27570713e-01 -4.70046937e-01 -1.23862579e-01 -8.63509998e-02
2.20718041e-01 5.58884665e-02 -1.06738377e+00 4.72433627e-01
1.01097798e+00 -1.17251956e+00 8.29614282e-01 -6.85580313e-01
5.61452508e-01 1.37926385e-01 -1.88331455e-01 -1.46989036e+00
-3.78268778e-01 -3.08977604e-01 1.31415009e-01 1.65443110e+00
2.04681918e-01 -7.61134744e-01 7.46628582e-01 8.93628657e-01
4.70494121e-01 -1.02268910e+00 -1.02099955e+00 -4.89812642e-01
2.62926817e-01 5.83670626e-04 8.09543967e-01 8.95241261e-01
-1.35786414e-01 -6.22364283e-02 -5.61173379e-01 -1.28452972e-01
7.63916492e-01 -2.81007826e-01 7.08152831e-01 -1.44724762e+00
1.43946409e-01 -4.65692371e-01 -6.56424820e-01 -8.81256014e-02
7.26428509e-01 -6.59165323e-01 -1.62329257e-01 -5.82602859e-01
2.47344494e-01 -5.02635002e-01 -5.14470994e-01 3.97752315e-01
-3.07633847e-01 3.10683846e-01 2.27977127e-01 -7.31093735e-02
-2.59598672e-01 9.53497112e-01 1.02133656e+00 -1.06378190e-01
5.40618598e-03 -3.76706682e-02 -1.08811378e+00 7.76480496e-01
6.65684700e-01 -4.72736537e-01 -2.46443078e-01 1.89978145e-02
2.22902489e-03 -5.01584053e-01 4.30592984e-01 -7.66564608e-01
-3.64229977e-01 -1.22658849e-01 7.00241268e-01 2.98759460e-01
4.04447287e-01 -7.95158207e-01 -2.36921117e-01 3.06940138e-01
-6.50809348e-01 -2.88763195e-01 1.32341668e-01 3.68545353e-01
-1.02198370e-01 -8.05508196e-02 1.20553124e+00 4.51153517e-01
-3.12430292e-01 3.57488722e-01 8.39994624e-02 6.10756278e-02
1.11808324e+00 -6.07935637e-02 -3.18817526e-01 -4.29706126e-01
-4.52160031e-01 4.81698243e-03 4.65953678e-01 3.63529533e-01
4.98022377e-01 -1.54823887e+00 -8.62381399e-01 5.16254246e-01
4.12068903e-01 -7.17181444e-01 3.06866407e-01 1.03331423e+00
-8.52258727e-02 -2.36145005e-01 -5.80800056e-01 -1.70643598e-01
-1.52897811e+00 4.46589470e-01 4.81714427e-01 -2.56988332e-02
9.01094824e-02 8.69337499e-01 3.36415648e-01 -3.78113240e-01
2.15140402e-01 2.33480319e-01 -3.69788170e-01 4.24108535e-01
4.55356807e-01 5.98243773e-01 -6.36081249e-02 -9.43127215e-01
-5.58245361e-01 6.44330919e-01 -1.38975471e-01 -2.13364903e-02
1.14279330e+00 -6.41930550e-02 -2.28602722e-01 1.12319484e-01
1.39327002e+00 2.80472934e-01 -1.15059745e+00 4.78950217e-02
-2.50222772e-01 -8.85844827e-01 -4.33021151e-02 -6.05236113e-01
-1.15092015e+00 9.24386978e-01 1.00859582e+00 -2.05943897e-01
1.28029275e+00 -2.67153203e-01 5.25115848e-01 -2.60603011e-01
3.33482139e-02 -1.32994449e+00 -9.42126513e-02 9.41154882e-02
8.25994730e-01 -1.12415278e+00 5.48163988e-03 -4.92247730e-01
-8.23489726e-01 9.48077619e-01 5.93658149e-01 -1.78614929e-01
5.97637653e-01 -4.01186459e-02 1.77982554e-01 1.14577822e-01
-4.58662719e-01 1.53918818e-01 3.22231591e-01 7.61631131e-01
4.06164944e-01 6.57172874e-02 -6.34678960e-01 7.47269630e-01
-1.46359950e-01 -1.34781584e-01 2.94335157e-01 4.20215696e-01
-1.10304698e-01 -9.77004707e-01 -1.93332627e-01 3.80201429e-01
-5.98518908e-01 1.69388980e-01 -5.10287046e-01 5.89872479e-01
3.44988644e-01 6.83989763e-01 2.58260995e-01 -5.12534320e-01
5.07180929e-01 1.75832853e-01 4.94091898e-01 -1.55149624e-01
-5.16912699e-01 -4.81520772e-01 -3.97992134e-03 -4.78949010e-01
-2.58687198e-01 -5.74001372e-01 -8.54733884e-01 -5.53989887e-01
-1.86795786e-01 1.18257277e-01 6.68971598e-01 6.58717453e-01
3.78418118e-01 9.05933529e-02 1.16048336e+00 -5.57323039e-01
-1.05857277e+00 -6.93308949e-01 -8.27720702e-01 1.07629931e+00
2.18257532e-01 -9.24951553e-01 -7.20661283e-01 -3.54027092e-01] | [13.111485481262207, 1.343023419380188] |
e0a3d2c9-f21a-4c5b-82d9-da3d94755d43 | stardata-a-starcraft-ai-research-dataset | 1708.02139 | null | http://arxiv.org/abs/1708.02139v1 | http://arxiv.org/pdf/1708.02139v1.pdf | STARDATA: A StarCraft AI Research Dataset | We release a dataset of 65646 StarCraft replays that contains 1535 million
frames and 496 million player actions. We provide full game state data along
with the original replays that can be viewed in StarCraft. The game state data
was recorded every 3 frames which ensures suitability for a wide variety of
machine learning tasks such as strategy classification, inverse reinforcement
learning, imitation learning, forward modeling, partial information extraction,
and others. We use TorchCraft to extract and store the data, which standardizes
the data format for both reading from replays and reading directly from the
game. Furthermore, the data can be used on different operating systems and
platforms. The dataset contains valid, non-corrupted replays only and its
quality and diversity was ensured by a number of heuristics. We illustrate the
diversity of the data with various statistics and provide examples of tasks
that benefit from the dataset. We make the dataset available at
https://github.com/TorchCraft/StarData . En Taro Adun! | ['Jonas Gehring', 'Gabriel Synnaeve', 'Zeming Lin', 'Vasil Khalidov'] | 2017-08-07 | null | null | null | null | ['real-time-strategy-games'] | ['playing-games'] | [-1.90630168e-01 -1.71192423e-01 -4.65891093e-01 -2.60684118e-02
-6.80059552e-01 -8.82510900e-01 7.12062180e-01 -1.79277167e-01
-7.39756286e-01 8.01159501e-01 4.93493110e-01 -1.28670067e-01
-6.35392293e-02 -5.63135624e-01 -5.84538400e-01 -5.64054430e-01
-2.95077175e-01 3.30187321e-01 6.02209032e-01 -5.12523055e-01
4.37442988e-01 1.94422141e-01 -1.73314393e+00 4.31154788e-01
7.39980265e-02 9.36831236e-01 2.88406104e-01 1.18722963e+00
4.51135904e-01 1.32671714e+00 -8.58121455e-01 -4.54039395e-01
6.35281920e-01 -4.59510267e-01 -7.53464937e-01 -2.22291559e-01
-1.37932554e-01 -6.09385490e-01 -8.80931735e-01 7.00986862e-01
7.64364183e-01 3.33863437e-01 6.78529814e-02 -1.52685094e+00
5.37054203e-02 3.82199615e-01 -7.03620240e-02 6.71965957e-01
6.73998773e-01 7.72539020e-01 9.11899328e-01 -3.27763140e-01
9.40286875e-01 7.61651397e-01 4.23248440e-01 6.96956754e-01
-6.61971807e-01 -6.39565349e-01 -4.35183465e-01 5.22012115e-01
-1.00200570e+00 -8.78729582e-01 3.79056424e-01 -4.91724283e-01
1.15531421e+00 2.83847213e-01 1.02771115e+00 1.43634915e+00
3.46243709e-01 9.97740209e-01 7.86024690e-01 -4.79560345e-02
2.98822582e-01 -6.67523921e-01 -1.87681526e-01 8.02909434e-01
-1.53767914e-01 5.34332395e-01 -1.02059197e+00 -1.49171442e-01
9.89582658e-01 -2.04349771e-01 -8.13308284e-02 -4.20170963e-01
-1.51981652e+00 7.35395014e-01 -1.63477331e-01 -3.88007969e-01
-3.04522723e-01 2.97252864e-01 8.47472310e-01 6.08316481e-01
-2.69262381e-02 4.44387138e-01 -5.46152771e-01 -1.43122804e+00
-6.05157912e-01 6.61805809e-01 8.40290964e-01 9.95682180e-01
4.71871972e-01 2.04827785e-01 2.31818289e-01 6.42111897e-01
-1.47483677e-01 4.65570748e-01 7.84408867e-01 -1.76403344e+00
4.79192734e-01 7.75105953e-02 8.73337165e-02 -7.09701300e-01
-4.62500513e-01 8.70600268e-02 -2.39650637e-01 4.65282083e-01
6.92863226e-01 -3.57109696e-01 -4.81569231e-01 1.72298646e+00
2.29115665e-01 2.76792794e-01 1.31719887e-01 1.03645813e+00
1.07958341e+00 4.31279838e-01 -4.54499602e-01 3.86414975e-02
1.16983557e+00 -8.76507461e-01 -5.32542229e-01 -3.74311656e-01
6.61708891e-01 -5.42330742e-01 1.14068878e+00 7.19198346e-01
-1.23988461e+00 -1.84803516e-01 -9.53190565e-01 -1.24369062e-01
-8.03760737e-02 -2.10052088e-01 6.87849760e-01 3.27352494e-01
-7.83572435e-01 6.72661126e-01 -1.33361638e+00 -2.37260997e-01
1.46712199e-01 3.07160527e-01 -5.51165044e-01 2.88143188e-01
-1.06837869e+00 7.32633710e-01 5.95313966e-01 -4.95244145e-01
-1.03767061e+00 -3.74146938e-01 -1.01755953e+00 -3.48542422e-01
8.04654717e-01 -2.56725520e-01 1.66132987e+00 -5.39032757e-01
-1.92682064e+00 7.40936160e-01 2.14215308e-01 -5.87785006e-01
5.63531935e-01 -3.36530805e-01 -3.13801140e-01 1.77148253e-01
6.76846877e-02 3.57237160e-01 7.13245511e-01 -4.14006412e-01
-9.18119669e-01 -2.06753954e-01 3.77955586e-01 4.97035027e-01
3.27486932e-01 1.60091549e-01 -8.36780548e-01 -6.14171386e-01
-3.37564170e-01 -1.18951178e+00 1.73526764e-01 -2.81962305e-01
-2.39080250e-01 2.77752697e-01 6.69124722e-01 -6.51905000e-01
1.09287596e+00 -2.16814017e+00 3.50988895e-01 -3.19921941e-01
2.46359348e-01 1.76444858e-01 -1.82331264e-01 6.09410703e-01
8.04610625e-02 -2.52384245e-01 1.60096481e-01 -1.68737143e-01
-5.31460680e-02 4.57712799e-01 -2.06570283e-01 7.13994682e-01
-2.36130968e-01 9.42171752e-01 -1.05219519e+00 -3.88261437e-01
3.74227732e-01 -1.22517772e-01 -8.17843795e-01 3.30921978e-01
-1.41459882e-01 5.58821678e-01 -2.66843379e-01 6.39801145e-01
-1.85845848e-02 1.41931564e-01 6.97113723e-02 1.60038173e-01
-1.00029103e-01 7.60246694e-01 -1.15209723e+00 2.15146947e+00
-1.23990215e-01 8.61912012e-01 1.70082431e-02 -5.86617351e-01
6.11925006e-01 1.98921502e-01 7.71929085e-01 -7.65104413e-01
2.41339847e-01 -1.33289605e-01 3.21562327e-02 -5.98228872e-01
1.19684041e+00 2.46093363e-01 -5.67799926e-01 7.25414574e-01
1.56691045e-01 -2.46091768e-01 6.86392665e-01 2.86004424e-01
1.56191170e+00 2.94441074e-01 4.97683376e-01 1.35565221e-01
-8.81647170e-02 4.54500198e-01 7.81440020e-01 6.67983055e-01
-5.76323748e-01 5.13913453e-01 7.58470118e-01 -4.89333808e-01
-1.17928493e+00 -1.07730699e+00 3.98863494e-01 1.21784592e+00
8.57537314e-02 -1.02353406e+00 -5.13220668e-01 -3.32454324e-01
-2.49012634e-01 3.99569720e-01 -4.58337992e-01 -4.98114377e-01
-5.79311788e-01 -2.68291742e-01 9.26168144e-01 3.41598362e-01
6.34244859e-01 -1.45821297e+00 -1.14903307e+00 -3.36593091e-02
-4.63570863e-01 -1.09923768e+00 -4.86340672e-01 1.15266405e-01
-6.88506365e-01 -1.43521821e+00 -1.29142061e-01 -1.20898321e-01
-2.62463361e-01 7.94280767e-02 1.29024720e+00 5.22297882e-02
-1.93750560e-01 3.92514557e-01 -7.23725855e-01 -1.81179449e-01
-6.16314769e-01 2.41541222e-01 4.19376433e-01 -6.42535627e-01
2.14095414e-01 -5.91805339e-01 -2.54173964e-01 3.14857960e-01
-7.31974721e-01 2.65912265e-01 2.17970103e-01 7.52083063e-01
4.02320802e-01 -1.04367442e-01 9.42013636e-02 -5.27063489e-01
7.81090617e-01 -7.19750524e-01 -6.72721803e-01 -3.77513856e-01
-1.08384378e-01 -1.32265776e-01 5.04826665e-01 -4.11196411e-01
-5.24570227e-01 -9.15513560e-02 -6.36485070e-02 -5.10862231e-01
-3.30257416e-01 3.55240494e-01 9.06450376e-02 4.68784034e-01
8.08164477e-01 3.95519853e-01 4.53586042e-01 -4.94668365e-01
4.07822460e-01 7.32057273e-01 1.11578572e+00 -5.43438196e-01
2.76524723e-01 2.69864678e-01 -3.78906310e-01 -8.16211760e-01
-3.24552536e-01 -3.06683838e-01 -4.42807138e-01 -3.59103352e-01
6.15354896e-01 -9.09357786e-01 -1.20079529e+00 8.70328963e-01
-6.20626032e-01 -1.09082782e+00 -6.01306379e-01 4.77712333e-01
-1.33138776e+00 3.31945449e-01 -9.09352541e-01 -6.58784449e-01
-1.74198002e-01 -1.24035907e+00 7.68364310e-01 2.20042631e-01
-3.33846837e-01 -5.81109881e-01 3.12666386e-01 3.47757787e-01
2.84346063e-02 2.03946814e-01 2.35912248e-01 -6.27299845e-01
-3.17526072e-01 -1.03415437e-01 3.86059225e-01 -4.07709740e-03
1.98407322e-01 6.02180101e-02 -4.27478045e-01 -4.80450213e-01
-2.10020244e-01 -7.07272530e-01 6.02411032e-01 3.29112351e-01
8.01633716e-01 -3.13633323e-01 2.31775925e-01 8.19303095e-01
9.66195405e-01 2.34142751e-01 7.99136996e-01 7.83728242e-01
4.63062257e-01 2.84067690e-01 8.38425875e-01 9.15143609e-01
4.76732373e-01 9.08372283e-01 8.09647918e-01 3.83778214e-01
1.29413391e-02 -3.97233278e-01 8.02510142e-01 7.05744207e-01
-2.30410919e-01 -2.70996183e-01 -7.22405195e-01 4.55411136e-01
-1.96951425e+00 -1.49802470e+00 2.19115481e-01 2.24413896e+00
7.29518950e-01 2.19942749e-01 7.96578109e-01 2.17220142e-01
4.10546571e-01 4.64191258e-01 -6.50045812e-01 -4.73758727e-01
-9.73642990e-02 1.36182338e-01 7.79416144e-01 3.92910957e-01
-1.03543687e+00 1.20409775e+00 7.02784014e+00 9.74438071e-01
-9.88371611e-01 -5.29657416e-02 2.08896361e-02 -8.83407593e-01
3.00127566e-01 -1.25404701e-01 -5.15716016e-01 7.72219598e-01
1.22476554e+00 -2.49724075e-01 9.50302184e-01 7.46406138e-01
5.05785346e-01 -5.38279772e-01 -6.18484855e-01 1.27437592e+00
-3.74343365e-01 -1.53464115e+00 -4.46453989e-01 2.80651271e-01
3.17680687e-01 5.31688392e-01 -1.25338346e-01 4.61750627e-01
9.77654397e-01 -8.61007810e-01 1.18020177e+00 4.48489130e-01
9.97498453e-01 -8.30296576e-01 4.55073982e-01 5.66352487e-01
-1.09679806e+00 -1.64069653e-01 -1.82474777e-01 -6.21173143e-01
2.32489586e-01 2.29773819e-02 -5.55373430e-01 4.67480153e-01
9.32367444e-01 1.17225516e+00 -5.71389019e-01 7.83504426e-01
-1.95417732e-01 4.77941841e-01 -4.36420798e-01 7.94819593e-02
9.34195518e-02 -2.11876646e-01 7.74427295e-01 7.75541604e-01
9.81063917e-02 2.24860355e-01 2.05413252e-01 3.35817397e-01
1.99534476e-01 -3.69659066e-01 -7.46155977e-01 -2.00886548e-01
6.64309442e-01 1.05893660e+00 -5.84507644e-01 -2.46333227e-01
-3.08580130e-01 1.02734900e+00 2.24809170e-01 1.63510684e-02
-9.97886062e-01 -3.12262952e-01 1.41451943e+00 8.20634067e-02
1.97905123e-01 -7.36364782e-01 2.08335727e-01 -1.40879440e+00
-1.59841061e-01 -1.31916118e+00 7.07966864e-01 -8.06493640e-01
-6.61609650e-01 4.75179017e-01 2.65408933e-01 -1.37328160e+00
-9.97349024e-01 -5.40616155e-01 -4.19131070e-01 2.37104475e-01
-7.13700593e-01 -4.90330100e-01 -1.90807566e-01 8.60286415e-01
7.11548209e-01 -5.97718775e-01 5.31778455e-01 1.77991211e-01
-7.82936335e-01 2.70198345e-01 2.49185991e-02 4.17958319e-01
7.01447725e-01 -1.18574214e+00 9.27300453e-01 7.96876788e-01
1.75903812e-01 1.90480635e-01 8.41995776e-01 -6.37667239e-01
-1.69373584e+00 -6.36378050e-01 -4.95359395e-03 -5.46107471e-01
8.23750377e-01 -2.57525474e-01 -4.69587833e-01 8.63078952e-01
-4.31575403e-02 1.17043845e-01 4.79291141e-01 -2.38603294e-01
4.86024804e-02 3.61230969e-01 -7.94464111e-01 7.99820364e-01
1.22284520e+00 -4.61825579e-01 -4.61199433e-01 -1.32246479e-01
3.12357366e-01 -1.11408937e+00 -7.74678290e-01 -1.13379247e-01
6.55515611e-01 -1.42631173e+00 8.12300801e-01 -8.31600904e-01
4.70773041e-01 -1.51293740e-01 -3.02719176e-01 -1.29720688e+00
-9.15956348e-02 -9.34682429e-01 -4.79807079e-01 7.90333509e-01
1.03094526e-01 -4.32373941e-01 1.09118080e+00 5.58774948e-01
-1.51183203e-01 -6.38831735e-01 -1.11369765e+00 -9.32087183e-01
-7.94822276e-02 -7.59884894e-01 7.02380419e-01 6.24941826e-01
4.08937514e-01 -5.54873404e-05 -7.35626936e-01 -3.73989910e-01
2.97072262e-01 -4.31653336e-02 1.38161004e+00 -7.32467651e-01
-5.63820779e-01 -2.50364959e-01 -6.51393473e-01 -1.29041553e+00
1.54154852e-01 -7.47626960e-01 -5.93309328e-02 -1.12057006e+00
-5.97044863e-02 -2.20761746e-01 1.46396846e-01 7.27992117e-01
2.11616039e-01 2.20210046e-01 3.71098936e-01 4.78524864e-01
-8.43539238e-01 6.28433824e-01 9.44674432e-01 3.12910378e-01
-2.43902534e-01 -7.22594466e-03 -3.99367332e-01 6.84691370e-01
1.11908007e+00 -4.56801265e-01 -3.11057538e-01 -3.37279171e-01
6.15624823e-02 4.75224644e-01 4.44701940e-01 -1.14393973e+00
1.20572448e-01 -4.54239160e-01 -4.12057824e-02 -1.78837001e-01
8.13701749e-01 -3.58535647e-01 3.36694032e-01 5.40172696e-01
-1.97751224e-01 4.37740266e-01 3.10809880e-01 3.75385076e-01
6.55280519e-03 -3.46178561e-02 6.16742194e-01 -2.96084225e-01
-1.27608299e+00 2.32516736e-01 -6.78818047e-01 6.25380576e-01
9.72814202e-01 -2.74359643e-01 -5.07520914e-01 -9.37392354e-01
-6.34543180e-01 3.86176892e-02 9.88761306e-01 7.14992523e-01
4.99107420e-01 -1.28015089e+00 -5.80323696e-01 2.29440227e-01
-6.17837310e-02 -3.20594430e-01 1.91404745e-01 7.74503827e-01
-7.25314975e-01 2.98284423e-02 -7.20264435e-01 -4.16780174e-01
-1.35059381e+00 4.58528310e-01 2.57227123e-01 -1.08509213e-01
-1.04171038e+00 3.41576278e-01 -5.32710910e-01 -3.29667628e-01
-1.57791361e-01 -1.09281845e-01 1.07577145e-02 -1.97208509e-01
6.54911280e-01 5.50403714e-01 4.54597622e-02 -8.27233553e-01
-3.49831164e-01 -1.33457445e-02 1.80677444e-01 -5.36910236e-01
1.26470804e+00 -1.61851883e-01 3.44870567e-01 5.88231623e-01
8.26515257e-01 -4.16958779e-02 -1.80754972e+00 -1.89301386e-01
-2.03743368e-01 -6.27402067e-01 -5.94152659e-02 -4.66810018e-01
-1.21878254e+00 4.27671582e-01 2.58700103e-01 8.69024619e-02
8.88540566e-01 -7.84309655e-02 8.96614313e-01 3.59237790e-01
8.04159462e-01 -9.63573337e-01 6.06700219e-02 9.23907936e-01
6.33811831e-01 -1.09440899e+00 8.41954648e-02 1.93877548e-01
-9.63477850e-01 8.88941228e-01 3.69329095e-01 -2.42749766e-01
2.62976795e-01 7.42111087e-01 6.00271970e-02 -1.34894878e-01
-1.21325815e+00 -2.84039855e-01 -2.78228164e-01 7.24405825e-01
9.12674367e-02 -7.72501156e-02 -8.09839293e-02 8.70180011e-01
-1.01172960e+00 4.54306677e-02 1.13921404e+00 1.40699363e+00
-3.76752853e-01 -1.12277412e+00 -3.27253640e-01 4.62668777e-01
-5.00041246e-01 3.25587280e-02 -3.11106145e-01 8.34307313e-01
-3.06028575e-01 1.01990497e+00 1.14489757e-01 -7.83717096e-01
3.18925500e-01 -1.45747408e-01 4.36720341e-01 -5.33580780e-01
-6.09169245e-01 -2.78595030e-01 4.17759478e-01 -1.16305184e+00
-6.75159320e-02 -7.70473599e-01 -1.35274243e+00 -1.12391460e+00
1.92924052e-01 9.31230038e-02 4.33639169e-01 7.12634683e-01
5.28052032e-01 3.24549168e-01 1.17770471e-01 -1.19585299e+00
-2.58980006e-01 -7.66703546e-01 -6.56515718e-01 3.13705862e-01
2.78855294e-01 -8.61881375e-01 -2.76268810e-01 1.59530982e-01] | [3.8884732723236084, 1.5185869932174683] |
438c63a7-11e7-4734-8fcf-29d9f3fe3d0d | real-time-and-robust-3d-object-detection-with | 2207.05200 | null | https://arxiv.org/abs/2207.05200v1 | https://arxiv.org/pdf/2207.05200v1.pdf | Real-Time And Robust 3D Object Detection with Roadside LiDARs | This work aims to address the challenges in autonomous driving by focusing on the 3D perception of the environment using roadside LiDARs. We design a 3D object detection model that can detect traffic participants in roadside LiDARs in real-time. Our model uses an existing 3D detector as a baseline and improves its accuracy. To prove the effectiveness of our proposed modules, we train and evaluate the model on three different vehicle and infrastructure datasets. To show the domain adaptation ability of our detector, we train it on an infrastructure dataset from China and perform transfer learning on a different dataset recorded in Germany. We do several sets of experiments and ablation studies for each module in the detector that show that our model outperforms the baseline by a significant margin, while the inference speed is at 45 Hz (22 ms). We make a significant contribution with our LiDAR-based 3D detector that can be used for smart city applications to provide connected and automated vehicles with a far-reaching view. Vehicles that are connected to the roadside sensors can get information about other vehicles around the corner to improve their path and maneuver planning and to increase road traffic safety. | ['Alois C. Knoll', 'Xingcheng Zhou', 'Jialong Wu', 'Walter Zimmer'] | 2022-07-11 | null | null | null | null | ['robust-3d-object-detection'] | ['computer-vision'] | [-1.80526838e-01 8.92572328e-02 -3.07737857e-01 -5.79186559e-01
-7.72466540e-01 -5.04315972e-01 5.88596940e-01 -1.96022719e-01
-6.86733186e-01 2.53517658e-01 -3.23911935e-01 -9.69928682e-01
2.14963511e-01 -1.16388750e+00 -9.05563951e-01 -3.34964097e-01
-8.54691863e-02 6.09038413e-01 8.20251346e-01 -2.72559136e-01
-3.32164462e-03 9.94767368e-01 -2.02976179e+00 -4.96806167e-02
9.61918771e-01 8.92347932e-01 5.61091483e-01 5.76117873e-01
1.94601178e-01 1.65014252e-01 -2.72008389e-01 -4.04381007e-03
5.50328791e-01 6.36930823e-01 -2.15178475e-01 -1.59158587e-01
7.04748988e-01 -7.22957492e-01 -3.65665972e-01 7.26907790e-01
6.71945512e-01 -2.70209223e-01 6.56950831e-01 -1.68319595e+00
-4.27862536e-03 3.95949651e-03 -3.85209799e-01 2.73770899e-01
1.18111245e-01 5.07331908e-01 5.35945773e-01 -1.03198993e+00
3.10723931e-01 1.44154370e+00 6.78877592e-01 3.75577182e-01
-9.10658896e-01 -1.09600735e+00 1.53453141e-01 7.65415430e-01
-1.50163352e+00 -5.43178320e-01 5.28551877e-01 -3.98418635e-01
9.80603516e-01 -8.85368809e-02 2.95231342e-01 1.12456524e+00
1.73208974e-02 8.70141089e-01 8.41024578e-01 3.25186737e-02
2.13453263e-01 4.87157613e-01 1.63866565e-01 7.50781298e-01
4.62147295e-01 5.62804997e-01 -3.05871487e-01 4.46096331e-01
3.11754495e-02 -1.87416375e-01 3.99858773e-01 -4.87928957e-01
-1.01800716e+00 1.01102412e+00 6.07312143e-01 -1.88765213e-01
-2.56285310e-01 1.10342167e-01 3.00828308e-01 6.45390302e-02
3.07985097e-01 -3.66037101e-01 -5.14814794e-01 -7.92110935e-02
-5.04680216e-01 3.41463804e-01 4.94068891e-01 1.38660109e+00
1.04126108e+00 1.44539317e-02 -1.77597776e-01 4.58226860e-01
4.38468933e-01 1.41987264e+00 -2.77103066e-01 -1.18377256e+00
7.66907632e-01 5.64636171e-01 6.82064593e-02 -5.60587883e-01
-6.81372821e-01 4.25760746e-02 -3.99917483e-01 5.46170950e-01
2.14765295e-01 -3.59617025e-01 -9.55359638e-01 1.43510759e+00
5.79352617e-01 5.21331191e-01 -1.14177689e-01 7.80199826e-01
9.52906013e-01 5.27131677e-01 1.40936136e-01 3.48978281e-01
1.46104896e+00 -6.39093280e-01 -3.30767184e-01 -4.82667357e-01
9.62588251e-01 -3.28784287e-01 9.71750379e-01 4.31726053e-02
-6.14712894e-01 -9.81907368e-01 -1.14476335e+00 -9.40104201e-02
-7.74773300e-01 2.09063828e-01 2.44875610e-01 9.22575831e-01
-1.02834833e+00 -1.18036054e-01 -5.96698642e-01 -7.00750053e-01
5.77495337e-01 3.10145795e-01 -1.25009179e-01 -1.44260600e-01
-1.20871055e+00 1.43620062e+00 1.80589810e-01 -1.87488109e-01
-1.08654594e+00 -8.26323390e-01 -9.22836185e-01 -2.09207639e-01
4.57380861e-01 -5.78850389e-01 1.38161767e+00 2.80695200e-01
-1.23580217e+00 9.54844832e-01 -4.94224012e-01 -7.81841159e-01
5.70428669e-01 -1.36496380e-01 -6.40700877e-01 -4.27668653e-02
5.13528943e-01 1.12360346e+00 4.30168271e-01 -1.07063890e+00
-1.38548541e+00 -4.99258667e-01 -6.99912757e-02 -6.44646138e-02
6.52733743e-02 -4.05322552e-01 -4.95917082e-01 3.28743786e-01
-1.62784070e-01 -1.13604224e+00 -1.12790011e-01 2.32853256e-02
-1.80207804e-01 -5.05622149e-01 1.40936196e+00 -1.54739201e-01
6.44570887e-01 -2.03424621e+00 -7.64191270e-01 2.09563315e-01
-8.30639154e-02 5.52976251e-01 -3.23511302e-01 1.72988325e-01
4.76886481e-01 -3.67135950e-03 1.99072771e-02 -2.93501139e-01
1.07952327e-01 4.52897996e-01 -2.55703002e-01 2.97987968e-01
4.52052832e-01 1.11130059e+00 -9.17976439e-01 -5.11446476e-01
8.15527022e-01 5.71707666e-01 -4.76169378e-01 4.42712568e-02
1.99901000e-01 3.92723829e-01 -5.15115738e-01 6.05922818e-01
1.14799929e+00 4.13397551e-01 -3.11677575e-01 1.00383140e-01
-5.56202948e-01 7.42292464e-01 -1.15048611e+00 1.26362157e+00
-6.09472573e-01 8.34785104e-01 9.07461792e-02 -1.05347753e+00
1.07299685e+00 -5.98842204e-02 2.15897202e-01 -1.35618591e+00
-4.21859473e-02 1.86234444e-01 -2.11277843e-01 -7.33295202e-01
5.38568199e-01 3.57699692e-01 -2.28768468e-01 1.32615402e-01
-3.44521075e-01 -3.58794421e-01 1.88285097e-01 1.20329812e-01
1.12848532e+00 -7.20167682e-02 -9.80262831e-02 -1.40679777e-01
8.06923449e-01 5.96949682e-02 2.69912213e-01 9.80150998e-01
-6.80338144e-01 -1.45969102e-02 1.91763323e-02 -3.98703992e-01
-9.59541976e-01 -1.35017610e+00 -4.23569232e-01 9.01107788e-01
2.61806160e-01 -5.69575429e-02 -4.19225216e-01 -9.27501738e-01
6.49851203e-01 1.10833454e+00 -3.43643129e-01 -1.06665581e-01
-5.69384634e-01 -2.34771699e-01 6.12422049e-01 7.59808600e-01
8.92178535e-01 -5.96489727e-01 -8.46916437e-01 3.93569097e-02
-1.29631355e-01 -1.77891171e+00 5.51737994e-02 -1.53411692e-03
-5.38965046e-01 -9.80798125e-01 -5.27964644e-02 -6.93678796e-01
1.67570427e-01 1.12983406e+00 8.12758684e-01 -2.33965963e-01
3.72637287e-02 3.66622597e-01 -2.92344447e-02 -1.28065562e+00
-4.71300155e-01 3.98928314e-01 2.66734898e-01 -3.71048450e-01
1.08946979e+00 -3.24592590e-01 -5.55425525e-01 6.99186027e-01
-1.04690187e-01 -1.53644189e-01 5.76247633e-01 7.60897323e-02
2.61947542e-01 -7.38063306e-02 7.53899038e-01 -5.71553051e-01
2.24166304e-01 -6.33056223e-01 -1.17448449e+00 -4.41837043e-01
-8.70501339e-01 -2.17558354e-01 8.22958350e-02 2.85132807e-02
-8.05234134e-01 3.64272147e-01 -3.97713065e-01 -1.33488566e-01
-5.44255257e-01 -6.10652268e-02 -3.82401586e-01 6.73293918e-02
5.76168656e-01 -5.24650561e-03 2.23932371e-01 -3.10815722e-01
6.18466437e-01 1.10664117e+00 4.85145807e-01 -1.94904834e-01
1.25882077e+00 8.60516131e-01 3.07493925e-01 -9.81896222e-01
-6.11461461e-01 -7.26020277e-01 -8.21780980e-01 -3.43179166e-01
6.85873985e-01 -1.46197271e+00 -1.28264260e+00 5.74938804e-02
-1.17139876e+00 -3.60863805e-01 -1.48608863e-01 4.42819029e-01
-5.22651613e-01 1.73835233e-02 -5.34972399e-02 -1.00070918e+00
5.28287776e-02 -1.10752761e+00 1.44820499e+00 -8.06715060e-03
1.62943184e-01 -5.76563001e-01 -1.14327401e-01 3.96586776e-01
3.98799539e-01 -1.12462655e-01 5.18738747e-01 -3.87739956e-01
-9.39264119e-01 -2.94382006e-01 -6.63609326e-01 2.51595140e-01
-2.25855395e-01 -2.64144570e-01 -1.39516985e+00 9.74408351e-03
-4.96801585e-01 1.81885615e-01 1.15599835e+00 2.14783385e-01
1.04736531e+00 2.28949696e-01 -8.65959346e-01 2.62637794e-01
1.00554395e+00 1.40926361e-01 6.21952474e-01 5.25268793e-01
4.58137989e-01 6.77634418e-01 9.02909815e-01 1.24143988e-01
1.15265512e+00 8.20428431e-01 8.11699569e-01 -2.82182425e-01
-1.97949767e-01 -4.65306252e-01 5.40326059e-01 2.57787615e-01
1.25564322e-01 -3.15620899e-02 -1.04125214e+00 7.40747273e-01
-1.78284764e+00 -1.02951097e+00 -6.36004925e-01 2.20081186e+00
7.03112483e-02 6.12751901e-01 4.68632400e-01 3.04258019e-01
5.69956601e-01 -2.61079401e-01 -4.88427073e-01 -4.68695581e-01
2.52669394e-01 2.60994714e-02 1.10455060e+00 6.79837286e-01
-1.13523149e+00 1.09167218e+00 6.64212084e+00 5.14223695e-01
-1.03595579e+00 2.83537298e-01 3.62335406e-02 -3.34564038e-02
1.59158036e-01 -2.89885134e-01 -1.49504817e+00 4.80233282e-01
1.41042686e+00 5.38616218e-02 1.74049631e-01 1.05083513e+00
9.04535592e-01 -2.75341362e-01 -1.16350865e+00 7.45101810e-01
-2.37597048e-01 -1.18822908e+00 -1.50054261e-01 3.81960541e-01
3.58455926e-01 8.53572130e-01 6.95125759e-02 8.19034696e-01
3.67293060e-01 -8.13437879e-01 6.80134535e-01 2.28730381e-01
6.79264545e-01 -9.35393870e-01 6.95248127e-01 8.26434731e-01
-1.41944230e+00 -3.20900798e-01 -4.60347593e-01 -2.19199911e-01
3.24301690e-01 4.79267299e-01 -1.65286255e+00 2.67854661e-01
8.55596304e-01 6.54463530e-01 -6.87368572e-01 1.04181659e+00
-4.36942250e-01 5.65234900e-01 -5.05682647e-01 -1.72426105e-01
3.42613816e-01 -7.69845247e-02 6.47504032e-01 1.37469935e+00
4.95705247e-01 -2.19623625e-01 2.44192749e-01 9.23335314e-01
4.52977233e-02 -4.30505842e-01 -1.38718081e+00 7.84882009e-01
9.36893225e-01 1.23683572e+00 -1.13806821e-01 -3.28834862e-01
-5.89485645e-01 2.48845726e-01 5.93707934e-02 3.49217057e-01
-1.23305976e+00 -3.91563296e-01 1.03272378e+00 5.79503775e-01
6.17085040e-01 -5.81404388e-01 -3.44867915e-01 -3.51621836e-01
-7.40609765e-02 -2.42970645e-01 3.21649164e-02 -7.99444675e-01
-7.69308746e-01 2.19237379e-04 2.33074427e-01 -1.28062570e+00
-2.03827873e-01 -8.67199123e-01 -5.16621888e-01 7.41132379e-01
-2.33624434e+00 -1.23064220e+00 -5.91763973e-01 6.31664813e-01
4.70465302e-01 -1.54846355e-01 3.92751604e-01 6.81188107e-01
-3.81574333e-01 4.41378742e-01 -3.48288983e-01 4.89672460e-02
5.10336518e-01 -8.96807909e-01 9.35816467e-01 6.89780772e-01
-7.16964230e-02 -3.66842523e-02 4.37879592e-01 -4.57751036e-01
-1.46259987e+00 -1.67256939e+00 1.09432411e+00 -1.13991725e+00
3.77704620e-01 -6.58681214e-01 -5.04653394e-01 8.12061965e-01
-1.35197476e-01 -1.27939656e-01 1.90079406e-01 1.03233926e-01
-2.57932901e-01 -6.31673753e-01 -1.27389777e+00 5.17435670e-01
1.39177096e+00 -4.02231693e-01 -4.75078881e-01 1.38991207e-01
7.43008792e-01 -2.62450397e-01 -1.68676645e-01 6.89250231e-01
4.95649606e-01 -9.00710523e-01 1.12862647e+00 -2.27434993e-01
-3.66546035e-01 -5.38525522e-01 -3.26986134e-01 -1.04887736e+00
-2.17140839e-01 -2.21166208e-01 -1.92728266e-01 8.92105043e-01
4.18907255e-01 -1.01002431e+00 7.87048578e-01 1.27533525e-01
-4.20850396e-01 -1.59518093e-01 -1.36767983e+00 -1.08027315e+00
3.57382745e-02 -1.25913608e+00 9.10409689e-01 9.18145776e-02
-4.33779508e-01 6.81358218e-01 1.25573486e-01 6.78487539e-01
6.35429502e-01 -2.20927149e-01 1.46266973e+00 -1.52731717e+00
7.31398582e-01 -2.23532453e-01 -5.78637838e-01 -1.48202085e+00
2.22431049e-01 -9.87382293e-01 1.53798550e-01 -1.59232724e+00
-1.05582252e-01 -6.59620166e-01 6.08361401e-02 3.40761304e-01
3.02872688e-01 3.02455902e-01 4.59841415e-02 -2.18735710e-01
-5.39144039e-01 4.21042591e-01 1.02586484e+00 -4.16162789e-01
-2.88020857e-02 4.49783325e-01 -6.42613053e-01 7.68472433e-01
7.59695828e-01 -4.38921809e-01 -2.90932775e-01 -5.43818235e-01
-4.41286899e-02 -4.72942352e-01 7.80763209e-01 -1.12714612e+00
3.35395545e-01 4.09165658e-02 1.74410850e-01 -1.50830495e+00
4.78395015e-01 -1.09876823e+00 -4.94563580e-01 6.39438152e-01
3.42908889e-01 -6.28359094e-02 5.14993489e-01 4.81265455e-01
3.84280354e-01 2.19275802e-01 6.75329387e-01 2.54422814e-01
-1.10441720e+00 4.55575377e-01 -8.41465414e-01 -2.51411140e-01
1.38364375e+00 -4.75704998e-01 -4.28396016e-01 -3.18440318e-01
-8.20876285e-02 7.96421111e-01 1.77377373e-01 8.14502060e-01
6.14054859e-01 -1.45372641e+00 -8.35197806e-01 6.21473253e-01
5.17943084e-01 -5.54079264e-02 -2.53157765e-01 7.21646428e-01
-9.79953259e-02 1.08175790e+00 -2.80033704e-02 -1.29045987e+00
-1.07331777e+00 6.45663857e-01 2.40196913e-01 3.96848947e-01
-5.82916498e-01 1.58525839e-01 -1.27491996e-01 -8.91978085e-01
2.88811177e-01 -5.90137959e-01 -2.53387004e-01 -6.56844233e-04
4.91584688e-01 7.26352453e-01 2.90306270e-01 -8.48265707e-01
-7.10788667e-01 7.27995634e-01 3.54266375e-01 -7.15004951e-02
9.87316310e-01 -4.84099954e-01 7.22068250e-01 2.50314713e-01
1.00338113e+00 -9.10854191e-02 -1.38747716e+00 -1.53453335e-01
-1.69203393e-02 -3.62195402e-01 2.56881654e-01 -5.41268885e-01
-7.47784138e-01 1.15163863e+00 1.12227893e+00 2.04603031e-01
6.02702558e-01 1.41797364e-01 8.30838263e-01 7.58628726e-01
6.98713064e-01 -1.15803862e+00 -3.29435587e-01 8.54205608e-01
4.83368903e-01 -1.76232338e+00 -3.36813956e-01 -5.14422297e-01
-3.36196780e-01 6.78552568e-01 6.94432557e-01 1.37604877e-01
7.75770366e-01 3.52901727e-01 2.10433990e-01 -1.62739262e-01
-6.49947464e-01 -1.10379326e+00 1.28588244e-01 1.31588173e+00
-2.97831893e-01 2.05603242e-01 1.10482968e-01 -4.17958274e-02
-2.71445841e-01 -3.55750136e-02 2.71590382e-01 5.65285683e-01
-8.65291119e-01 -8.93460989e-01 -1.77428067e-01 1.51201442e-01
3.23227316e-01 2.61589438e-01 -5.92356883e-02 1.16511858e+00
5.05481064e-01 1.45724964e+00 3.25727969e-01 -5.15060186e-01
9.60678041e-01 6.52697384e-02 3.07567686e-01 -5.16119421e-01
1.21757790e-01 -5.28347611e-01 2.76078045e-01 -8.38862956e-01
-1.84982896e-01 -7.91929483e-01 -1.39945579e+00 -6.24038577e-01
-1.03862412e-01 -2.86085844e-01 1.19204390e+00 7.81154454e-01
7.59611726e-01 3.09411705e-01 9.57105815e-01 -1.04639196e+00
-4.87867028e-01 -7.89625108e-01 -1.90466017e-01 2.38104332e-02
4.51241642e-01 -9.64926064e-01 -2.13686302e-01 -4.21114326e-01] | [7.8606462478637695, -1.7277733087539673] |
c4e5e2b5-cd5b-4814-9387-ac40315e6306 | 3d-context-enhanced-region-based | 1806.09648 | null | http://arxiv.org/abs/1806.09648v2 | http://arxiv.org/pdf/1806.09648v2.pdf | 3D Context Enhanced Region-based Convolutional Neural Network for End-to-End Lesion Detection | Detecting lesions from computed tomography (CT) scans is an important but
difficult problem because non-lesions and true lesions can appear similar. 3D
context is known to be helpful in this differentiation task. However, existing
end-to-end detection frameworks of convolutional neural networks (CNNs) are
mostly designed for 2D images. In this paper, we propose 3D context enhanced
region-based CNN (3DCE) to incorporate 3D context information efficiently by
aggregating feature maps of 2D images. 3DCE is easy to train and end-to-end in
training and inference. A universal lesion detector is developed to detect all
kinds of lesions in one algorithm using the DeepLesion dataset. Experimental
results on this challenging task prove the effectiveness of 3DCE. We have
released the code of 3DCE in
https://github.com/rsummers11/CADLab/tree/master/lesion_detector_3DCE. | ['Mohammadhadi Bagheri', 'Ke Yan', 'Ronald M. Summers'] | 2018-06-25 | null | null | null | null | ['medical-object-detection'] | ['computer-vision'] | [-5.53677976e-02 -1.69937566e-01 -2.57574648e-01 -4.43048090e-01
-1.01732278e+00 -3.79139245e-01 3.86097699e-01 6.80407584e-02
-4.24969047e-01 2.59681582e-01 1.83680817e-01 -4.43313748e-01
1.30988017e-01 -7.62767017e-01 -4.79761422e-01 -6.41688287e-01
-2.07516611e-01 3.95996183e-01 7.36845136e-01 1.87986091e-01
-1.40658826e-01 8.53404105e-01 -9.99818265e-01 6.57505691e-01
4.82551098e-01 1.08895135e+00 5.09777308e-01 1.02100837e+00
-4.26703412e-03 6.65371954e-01 -1.00629106e-01 -3.85600328e-02
4.38111365e-01 -2.13629514e-01 -8.43207479e-01 -7.33336955e-02
4.79781717e-01 -7.01974928e-01 -3.96683276e-01 8.56714249e-01
8.86176586e-01 -4.00823325e-01 8.38573456e-01 -9.32158828e-01
-4.29606676e-01 3.02319705e-01 -4.81784075e-01 8.54519010e-01
-2.28907049e-01 2.68123388e-01 5.79407990e-01 -1.17028975e+00
5.53015828e-01 9.82784212e-01 9.09081578e-01 5.99013031e-01
-6.04505837e-01 -5.82406938e-01 -2.78818727e-01 7.36150444e-02
-1.32302809e+00 3.45472619e-02 4.75436300e-01 -5.13758004e-01
7.78068423e-01 4.35323328e-01 5.94562948e-01 1.12009382e+00
4.17497367e-01 1.06729269e+00 9.29084420e-01 -1.44085065e-01
-7.83425346e-02 -2.19884217e-01 -3.82064208e-02 1.11433995e+00
1.58309266e-01 3.45136613e-01 1.83823138e-01 1.28236301e-02
1.35621881e+00 3.86601716e-01 -3.53166223e-01 -3.82790804e-01
-1.21095550e+00 8.47115159e-01 1.05727649e+00 2.89987981e-01
-4.34615195e-01 3.57340276e-01 6.09963655e-01 -1.38443783e-01
5.18171906e-01 -7.08857784e-03 -2.61442035e-01 2.62428105e-01
-6.01470113e-01 7.71996677e-02 3.43983829e-01 6.00717008e-01
3.38761248e-02 -3.74912173e-01 -4.75662768e-01 1.04256892e+00
-4.57296744e-02 1.24403834e-01 6.53247476e-01 -5.49123049e-01
1.63817704e-01 6.62700355e-01 -3.58715594e-01 -3.93425763e-01
-9.51422989e-01 -8.62943530e-01 -9.64274585e-01 4.70091909e-01
4.80598360e-01 -2.25507900e-01 -1.33394074e+00 1.28902638e+00
5.43407083e-01 3.80107202e-02 -4.57289010e-01 1.23518598e+00
1.25014293e+00 1.92971125e-01 1.44602254e-01 4.54876423e-01
1.46195900e+00 -1.20550179e+00 -1.19242609e-01 -1.80350140e-01
9.24451888e-01 -6.26016200e-01 8.69831383e-01 7.29121491e-02
-1.09580362e+00 -3.40217084e-01 -8.07603836e-01 -2.86544919e-01
-3.96855801e-01 4.47817117e-01 5.23632824e-01 3.79372984e-01
-9.55938816e-01 2.77348280e-01 -1.18898916e+00 -4.92606372e-01
8.98360074e-01 2.50831366e-01 -3.72605145e-01 -3.08090359e-01
-8.51913989e-01 1.13496590e+00 4.46752906e-01 3.40912007e-02
-1.08310509e+00 -8.90007079e-01 -4.60308045e-01 -2.19507784e-01
3.61873299e-01 -8.06561768e-01 1.59243453e+00 -6.57750487e-01
-9.69093382e-01 1.10276461e+00 2.21263126e-01 -3.10467333e-01
9.99203384e-01 4.37369645e-02 -2.54679564e-02 4.66031611e-01
2.61667132e-01 7.66655147e-01 3.44792873e-01 -9.62298036e-01
-7.49380410e-01 -2.55261421e-01 -6.54229671e-02 1.08337067e-01
1.49598017e-01 6.65939301e-02 -5.01685917e-01 -6.86114609e-01
2.64636070e-01 -7.31625736e-01 -3.19328845e-01 8.50339472e-01
-7.06859589e-01 -2.68816769e-01 8.69242549e-01 -7.73767710e-01
7.14541256e-01 -1.86482537e+00 -3.61893773e-01 1.74799878e-02
5.68547547e-01 3.81050378e-01 7.82201719e-03 -1.31677523e-01
-3.45002264e-01 2.39442065e-02 -3.98841500e-01 -7.81252980e-02
-1.22491576e-01 4.38937359e-02 4.14832652e-01 5.65941930e-01
4.24370855e-01 1.17918384e+00 -9.42338586e-01 -7.64156818e-01
3.90453488e-01 6.30095482e-01 -4.15189832e-01 8.04553479e-02
1.20087273e-01 4.49742556e-01 -5.31296730e-01 1.11639285e+00
6.90610945e-01 -5.70746005e-01 -3.05616319e-01 -2.92527109e-01
3.14350463e-02 7.96342790e-02 -6.90677404e-01 1.54406583e+00
-5.59137940e-01 6.22877777e-01 1.66859388e-01 -9.74656165e-01
3.91674906e-01 4.23378497e-01 5.56471407e-01 -6.48899198e-01
4.43532735e-01 3.61765862e-01 1.37797162e-01 -7.30205894e-01
-1.46272048e-01 -1.30791277e-01 1.61438644e-01 3.59230310e-01
-8.06281492e-02 -1.21516444e-01 3.27982120e-02 2.43254621e-02
1.47609270e+00 -2.29724407e-01 5.75624645e-01 -9.91387069e-02
2.20839247e-01 1.04495101e-01 4.03479964e-01 7.62613237e-01
-6.45483553e-01 1.03960907e+00 3.06344420e-01 -6.58432841e-01
-9.90511417e-01 -1.38571095e+00 -5.02806187e-01 8.48775566e-01
-1.30708337e-01 -1.67660054e-03 -4.60902840e-01 -1.01470518e+00
-2.77512763e-02 2.75881082e-01 -8.50195110e-01 2.00877279e-01
-6.60223961e-01 -7.43266284e-01 6.65275931e-01 9.67104316e-01
7.55384803e-01 -7.31729507e-01 -9.65458691e-01 2.25917816e-01
3.50672789e-02 -9.21837211e-01 -5.70138931e-01 3.34415615e-01
-8.68371904e-01 -1.46527970e+00 -1.12455857e+00 -9.15432453e-01
7.89356589e-01 2.69953936e-01 1.00866246e+00 3.83841902e-01
-1.02268279e+00 2.17119709e-01 -3.16103041e-01 -5.03800273e-01
-4.41705227e-01 2.18061022e-02 -5.45037150e-01 -4.89239693e-01
1.35441363e-01 -5.13272643e-01 -1.10173273e+00 3.72965544e-01
-9.95113134e-01 2.67760336e-01 1.04080248e+00 7.96281517e-01
6.64851069e-01 -1.89462826e-01 4.49456781e-01 -8.47216010e-01
4.69150513e-01 -5.90467751e-01 -2.61733383e-01 2.88779169e-01
-7.79393613e-02 -2.99854338e-01 2.79244334e-01 -2.77057022e-01
-8.12072575e-01 4.20966089e-01 -5.29366672e-01 -4.02217656e-01
-5.79021633e-01 4.21875149e-01 4.58366305e-01 -8.72992203e-02
8.78767848e-01 2.04740033e-01 -2.41061766e-03 -3.49198550e-01
6.62512630e-02 7.78627515e-01 6.17650628e-01 -2.10488960e-01
3.76625359e-01 6.98708296e-01 9.15344581e-02 -3.81897122e-01
-8.48452091e-01 -6.62131310e-01 -8.38373303e-01 -3.42303663e-01
8.41104865e-01 -8.52466404e-01 -1.15613334e-01 5.03477514e-01
-1.00074172e+00 -4.81230378e-01 -5.47958501e-02 4.99530762e-01
-3.18651408e-01 1.39768034e-01 -9.14363325e-01 -2.80203760e-01
-7.87325025e-01 -1.48777843e+00 1.32282114e+00 1.49989098e-01
6.73678294e-02 -1.02807498e+00 -5.30418307e-02 6.84326068e-02
5.59214175e-01 6.30375504e-01 8.10168028e-01 -5.70317388e-01
-5.09679675e-01 -5.36288202e-01 -6.69657767e-01 3.45709413e-01
2.65104562e-01 1.84902921e-02 -8.49402547e-01 -1.51073396e-01
-3.52981418e-01 -2.93822378e-01 1.22021341e+00 7.25869834e-01
1.37851155e+00 1.12775214e-01 -5.86159647e-01 7.27700770e-01
1.43061447e+00 -3.15638743e-02 3.72840375e-01 2.16521323e-01
6.37773633e-01 1.67378336e-01 2.60752946e-01 3.17121118e-01
2.40894765e-01 4.07509267e-01 7.37805545e-01 -5.28388143e-01
-7.27748215e-01 2.29950607e-01 4.25094962e-02 3.25959295e-01
5.96792400e-02 -2.61135027e-02 -1.30759823e+00 8.33256125e-01
-1.45324051e+00 -6.02624476e-01 -3.41332883e-01 1.50118029e+00
8.14437926e-01 8.09671506e-02 9.57650319e-02 -1.64482445e-01
9.93206739e-01 -1.50324270e-01 -5.89863539e-01 -4.51662112e-03
1.42978430e-01 3.44562173e-01 5.89178622e-01 1.92844108e-01
-1.54768825e+00 6.04174018e-01 5.19227076e+00 1.01517057e+00
-1.56326807e+00 4.59651262e-01 7.39663839e-01 -2.97664344e-01
2.29005396e-01 -3.92342657e-01 -5.20513535e-01 3.88146490e-01
2.96537846e-01 1.76764533e-01 -4.49698865e-01 1.03706920e+00
1.78550377e-01 -3.92826706e-01 -1.12711298e+00 6.53214037e-01
-2.23199219e-01 -1.49978375e+00 -1.40428334e-01 -6.78561926e-02
5.36300421e-01 6.98800981e-01 1.80958688e-01 2.42759570e-01
1.66031510e-01 -9.42787290e-01 5.51671863e-01 1.70759320e-01
1.00448263e+00 -5.45918643e-01 8.58393848e-01 2.99705744e-01
-1.14797759e+00 4.87976559e-02 -3.49000692e-01 4.87474710e-01
1.04096547e-01 7.60394871e-01 -1.63683140e+00 3.29480022e-01
8.51152539e-01 5.91258407e-01 -8.53051722e-01 1.69305074e+00
-2.18992442e-01 4.96986240e-01 -5.32163978e-01 1.02937698e-01
5.08719206e-01 5.84852636e-01 5.87209225e-01 1.55119145e+00
4.35662895e-01 1.15924738e-01 1.57171190e-01 1.00198269e+00
4.28361408e-02 -1.49629354e-01 -4.17339742e-01 3.97419125e-01
2.69553870e-01 1.51424170e+00 -1.16057086e+00 -2.45640859e-01
-2.95429140e-01 8.69132280e-01 1.80760965e-01 -9.17527154e-02
-9.04217720e-01 -2.74056882e-01 2.69726962e-01 1.45365089e-01
3.52246046e-01 -1.54914707e-02 -9.22865123e-02 -8.59158278e-01
-6.11381568e-02 -2.57897079e-01 6.19703889e-01 -8.66459727e-01
-1.57230556e+00 5.91191649e-01 -1.70427173e-01 -1.43849170e+00
-5.81511017e-03 -1.02707160e+00 -9.68742728e-01 5.71710408e-01
-1.73215091e+00 -1.20561862e+00 -6.17815435e-01 7.44315386e-01
6.21324539e-01 2.98640162e-01 6.14528656e-01 3.96286398e-01
-5.83608687e-01 6.74117565e-01 -8.41110125e-02 7.34071136e-01
6.26728058e-01 -1.25415289e+00 3.48696142e-01 7.58163631e-01
-4.11591053e-01 1.81642935e-01 1.28769904e-01 -6.94249213e-01
-9.55519676e-01 -1.53224206e+00 4.26084906e-01 -5.06042004e-01
5.55190265e-01 -5.43470196e-02 -7.75026917e-01 6.64081991e-01
-3.54814716e-02 7.65036464e-01 5.76177895e-01 -5.49285531e-01
-4.06743348e-01 3.33215743e-01 -1.34654093e+00 5.07760704e-01
9.86668169e-01 -3.87615651e-01 -3.24346781e-01 6.98008120e-01
4.99812543e-01 -8.33362222e-01 -9.50008631e-01 6.35954201e-01
4.35623825e-01 -1.03241932e+00 1.20218503e+00 -2.89105177e-01
6.60776377e-01 -1.35495707e-01 -5.28039783e-03 -1.13744211e+00
-1.91530883e-01 2.15629920e-01 2.01945499e-01 3.40962023e-01
4.02308971e-01 -5.61205983e-01 9.21074629e-01 1.67394519e-01
-7.05892324e-01 -1.12403548e+00 -1.13857257e+00 -7.30930626e-01
4.61331427e-01 -5.31703889e-01 1.75939485e-01 8.68190944e-01
-2.90107906e-01 -1.70462027e-01 2.87208080e-01 4.22066778e-01
5.40661693e-01 2.95550395e-02 2.79943764e-01 -9.33564484e-01
-1.00489996e-01 -8.39330971e-01 -4.80946511e-01 -7.33204186e-01
-4.93261963e-01 -1.38517523e+00 -9.22804475e-02 -1.75664496e+00
3.46837550e-01 -4.77889240e-01 -3.21238041e-01 6.22842789e-01
-1.20526075e-01 3.88139516e-01 1.07157361e-02 5.90794720e-02
-4.94984984e-01 1.36271924e-01 1.71342850e+00 -1.88543051e-01
3.26500028e-01 1.27183303e-01 -2.50299573e-01 7.45225847e-01
1.04724848e+00 -4.53226149e-01 -9.67445001e-02 -4.31236237e-01
-2.70710349e-01 1.79092169e-01 1.01641989e+00 -1.17316210e+00
3.65319490e-01 1.94868207e-01 7.49140680e-01 -1.06584907e+00
3.11571717e-01 -5.79973459e-01 -2.89468348e-01 7.38290548e-01
-2.71287173e-01 -9.70075503e-02 1.90908328e-01 3.11678082e-01
-1.43505752e-01 -2.57854939e-01 1.05482352e+00 -6.01968229e-01
-7.59487629e-01 7.48603165e-01 -3.81990314e-01 -5.54666519e-02
1.09772265e+00 -2.69363314e-01 -4.10738707e-01 -9.47960094e-02
-8.27966690e-01 1.90257683e-01 4.16804962e-02 2.00367600e-01
9.42533374e-01 -1.33284938e+00 -8.84895027e-01 -5.85365966e-02
3.03570151e-01 4.37883884e-01 4.51610625e-01 1.28646278e+00
-1.02638340e+00 4.69146639e-01 -2.87849963e-01 -9.87572193e-01
-1.33859551e+00 2.02221319e-01 9.97689903e-01 -3.12234432e-01
-7.96084046e-01 1.26434958e+00 3.63087058e-01 -6.16460264e-01
2.83247739e-01 -8.39127600e-01 9.70805436e-02 -1.95443839e-01
5.37816525e-01 -5.46377450e-02 4.94980812e-01 -2.92557448e-01
-4.37725425e-01 4.00675803e-01 -4.62290406e-01 2.52090573e-01
1.31985903e+00 2.74148792e-01 3.91863771e-02 -2.21725270e-01
1.55553615e+00 -5.96910834e-01 -1.23535788e+00 -4.14856911e-01
-1.38462275e-01 -3.07936072e-01 5.29085338e-01 -1.07084513e+00
-1.48314464e+00 1.04877937e+00 1.11559582e+00 -1.09071054e-01
1.10123181e+00 2.86181092e-01 1.01865733e+00 2.97860086e-01
1.28756359e-01 -6.33523762e-01 2.23480299e-01 3.33721012e-01
1.00006735e+00 -1.47894037e+00 -1.49618462e-02 -4.75194246e-01
-3.88443470e-01 1.29747999e+00 7.75806248e-01 -2.95835614e-01
9.76942062e-01 3.90051544e-01 1.50126785e-01 -5.04194021e-01
-6.62842929e-01 -4.69120651e-01 2.09747314e-01 5.99143922e-01
4.33064699e-01 2.65423059e-01 -7.66984895e-02 5.04109204e-01
1.90779552e-01 8.45600069e-02 4.53791648e-01 1.31866610e+00
-4.84519005e-01 -8.94392431e-01 -3.60278904e-01 5.43659031e-01
-5.61160207e-01 -8.19093958e-02 -3.85130674e-01 1.17318952e+00
3.00909013e-01 3.23750675e-01 9.76832062e-02 -2.46338233e-01
1.79235533e-01 -2.53051668e-01 3.74513179e-01 -7.26330757e-01
-7.03768194e-01 2.24503785e-01 -1.11664206e-01 -6.05275214e-01
-1.62607074e-01 -4.86240536e-01 -1.46119559e+00 -5.76789640e-02
-2.24198118e-01 -4.28538293e-01 6.61945462e-01 6.37224972e-01
2.78963506e-01 7.51570702e-01 4.68596935e-01 -8.70402396e-01
-3.51686597e-01 -8.59530091e-01 -4.12322313e-01 5.86465262e-02
4.60842550e-01 -8.15685749e-01 -1.20442830e-01 -1.99022666e-01] | [15.058865547180176, -2.350785732269287] |
7513ebde-b465-4f70-bdc4-b75b8b4612f0 | gennet-framework-interpretable-deep-learning | null | null | https://www.nature.com/articles/s42003-021-02622-z | https://www.nature.com/articles/s42003-021-02622-z.pdf | GenNet framework: interpretable deep learning for predicting phenotypes from genetic data | Applying deep learning in population genomics is challenging because of computational issues and lack of interpretable models. Here, we propose GenNet, a novel open-source deep learning framework for predicting phenotypes from genetic variants. In this framework, interpretable and memory-efficient neural network architectures are constructed by embedding biologically knowledge from public databases, resulting in neural networks that contain only biologically plausible connections.
We applied the framework to seventeen phenotypes and found well-replicated genes such as HERC2 and OCA2 for hair and eye color, and novel genes such as ZNF773 and PCNT for schizophrenia. Additionally, the framework identified ubiquitin mediated proteolysis, endocrine system and viral infectious diseases as most predictive biological pathways for schizophrenia.
GenNet is a freely available, end-to-end deep learning framework that allows researchers to develop and use interpretable neural networks to obtain novel insights into the genetic architecture of complex traits and diseases. | ['Gennady V. Roshchupkin', 'Wiro J. Niessen', 'Caroline C. W. Klaver', 'Hieab H. H. Adams', 'M. Arfan Ikram', 'Manfred Kayser', 'Seven A. Kushner', 'Arno van Hilten'] | 2021-09-17 | null | null | null | nature-communications-biology-2021-9 | ['genetic-risk-prediction', 'medical-genetics'] | ['medical', 'miscellaneous'] | [-3.82686891e-02 4.00548428e-01 2.42590513e-02 -6.01760805e-01
-3.18765819e-01 -2.92149127e-01 -2.38623307e-03 4.20735657e-01
3.43454666e-02 1.03775132e+00 3.33976567e-01 -2.79910862e-01
-4.83124703e-01 -6.26738429e-01 -8.18109095e-01 -4.67610598e-01
-4.40577269e-01 6.01810277e-01 -4.80844826e-01 1.38860270e-02
-2.93959171e-01 1.76274046e-01 -1.17091393e+00 4.61400419e-01
1.10314345e+00 6.76082134e-01 -2.86786973e-01 5.05773783e-01
3.59826505e-01 5.04662357e-02 -3.04125369e-01 -7.13072717e-01
-3.44797410e-02 -5.76181173e-01 -5.17395496e-01 -5.89135945e-01
1.38389871e-01 -2.60607421e-01 -1.07812986e-01 7.78263986e-01
9.30102825e-01 -4.78825986e-01 5.68759441e-01 -8.27793896e-01
-1.19921088e+00 7.48171866e-01 -8.73012617e-02 -1.09506369e-01
2.23840967e-01 5.02843797e-01 1.12017584e+00 -6.46256030e-01
8.44575405e-01 1.72501421e+00 8.80393505e-01 7.31208324e-01
-1.54319370e+00 -4.96456921e-01 -1.97782263e-01 2.04171911e-01
-1.12212360e+00 -4.35276031e-01 -1.15594275e-01 -4.68977720e-01
1.22002780e+00 3.66080314e-01 9.20583785e-01 1.28968048e+00
5.42785704e-01 4.34048206e-01 6.58814132e-01 -5.45346439e-02
2.39137590e-01 -8.72758031e-01 6.76962733e-02 1.19889843e+00
7.44667709e-01 2.71090139e-02 -5.88534832e-01 -6.13701522e-01
4.33467299e-01 1.34695649e-01 1.24637000e-01 2.84410626e-01
-1.15457523e+00 4.81569588e-01 2.72670329e-01 -3.11365873e-01
-4.02850270e-01 4.35566038e-01 3.41157973e-01 2.13429868e-01
6.87204540e-01 7.03540564e-01 -1.19610560e+00 1.41322181e-01
-3.11574012e-01 3.23412508e-01 4.63589638e-01 3.66178751e-01
5.34018338e-01 9.81332511e-02 -1.33045137e-01 9.33911204e-01
6.76826537e-01 3.05035502e-01 3.58039230e-01 -8.81225169e-01
-1.29404113e-01 1.00264454e+00 -3.70327413e-01 -8.57399881e-01
-1.04139876e+00 -2.62838125e-01 -9.49455619e-01 -4.33941856e-02
3.90282005e-01 -5.91010034e-01 -9.14400041e-01 1.98214710e+00
4.09124047e-01 4.58453476e-01 2.11121179e-02 7.28364587e-01
9.18524683e-01 -2.62739789e-02 3.57849836e-01 3.48319262e-01
1.52464592e+00 -4.87715513e-01 -4.12976682e-01 5.15184179e-02
6.21585190e-01 -9.50116012e-03 4.61069852e-01 4.87628490e-01
-1.20297372e+00 -1.35351196e-01 -7.98136115e-01 -4.43661392e-01
-7.39258826e-01 2.16877311e-01 1.34537125e+00 4.36101377e-01
-9.98217046e-01 7.95130372e-01 -1.05291581e+00 -5.85303485e-01
7.86833704e-01 6.12850130e-01 -4.34769064e-01 6.09781221e-02
-1.21730793e+00 6.42531753e-01 5.65689266e-01 4.34196085e-01
-1.03015316e+00 -1.07133436e+00 -5.56304574e-01 4.77553904e-01
-2.79711206e-02 -1.35826457e+00 6.09034181e-01 -5.77013552e-01
-1.18391812e+00 7.39810348e-01 -1.14862755e-01 -4.47097063e-01
-1.42018884e-01 -2.38212749e-01 -7.54406035e-01 1.13391139e-01
-1.75914377e-01 7.93523371e-01 1.27918780e-01 -6.16990440e-02
-1.08169429e-01 -8.96359503e-01 -1.15979612e-01 -3.95121455e-01
-6.05761297e-02 9.04659182e-02 5.33330999e-02 -6.13432646e-01
-1.45105878e-02 -9.74208832e-01 -2.21105620e-01 5.96178114e-01
-8.00343990e-01 -2.07971618e-01 1.42771944e-01 -1.04348445e+00
6.10358596e-01 -1.72814715e+00 3.09135646e-01 -2.42810659e-02
7.08256900e-01 2.92257100e-01 -4.85343546e-01 1.70555741e-01
-5.81394434e-01 6.04658306e-01 3.48180905e-02 4.50519323e-01
9.54906866e-02 -1.96408078e-01 3.45748037e-01 5.84650099e-01
1.02208257e+00 1.46962392e+00 -8.58203411e-01 3.78763199e-01
-4.18666825e-02 7.36622512e-01 -9.81590450e-01 -8.09698403e-02
-7.31245220e-01 3.86911809e-01 -6.00526214e-01 1.12187016e+00
6.17859185e-01 -3.56352180e-01 5.77683628e-01 9.97767448e-02
3.01196158e-01 1.97826058e-01 -4.19810921e-01 1.76696944e+00
1.02435254e-01 3.70966136e-01 9.04343873e-02 -6.56512976e-01
9.19984043e-01 2.27937862e-01 1.36796385e-01 -3.92308027e-01
1.94063500e-01 -7.85927027e-02 4.77202833e-01 -8.76419067e-01
-5.80559671e-01 2.50678092e-01 4.44545634e-02 2.09874153e-01
1.78543776e-01 6.35595679e-01 4.56006676e-02 -3.39617103e-01
1.62358630e+00 2.74101913e-01 3.26584339e-01 -1.74043059e-01
1.23044536e-01 -3.25504839e-01 1.17156518e+00 2.67121077e-01
6.80607706e-02 4.25223798e-01 1.09079373e+00 -9.76321697e-01
-1.13408351e+00 -9.16266382e-01 -2.12794155e-01 1.00040841e+00
-7.78250158e-01 -4.27804381e-01 -4.96623814e-01 -2.79261619e-01
3.65401149e-01 2.98658699e-01 -7.78331399e-01 -3.59290421e-01
7.84584060e-02 -1.31324077e+00 1.26113343e+00 2.85907984e-01
1.62776306e-01 -8.35407495e-01 -3.26972157e-02 2.92241454e-01
1.21153690e-01 -5.38310528e-01 1.76370293e-01 1.69818878e-01
-7.61960089e-01 -1.55026674e+00 -3.30680102e-01 -3.23836297e-01
6.55856729e-01 -6.34807467e-01 9.62216556e-01 5.11620529e-02
-9.75474417e-01 -3.66561145e-01 -1.78708881e-02 -1.03439426e+00
-3.85035187e-01 -8.53345245e-02 3.35700870e-01 -4.21565622e-01
5.12523830e-01 -5.38131893e-01 -6.82345688e-01 -1.99632153e-01
-8.53651941e-01 9.09721404e-02 8.28597248e-01 1.12600362e+00
5.15083432e-01 -3.28239471e-01 1.11810946e+00 -7.97968745e-01
6.89378321e-01 -9.46470261e-01 -6.26002371e-01 4.15587634e-01
-6.41659021e-01 1.92125544e-01 4.71029788e-01 5.22165447e-02
-9.06943798e-01 1.29661202e-01 -4.49392557e-01 2.40965366e-01
-4.79918927e-01 9.87894893e-01 -4.34763819e-01 2.30709702e-01
3.15768450e-01 -1.55309483e-01 2.85946935e-01 -7.12888956e-01
3.03269416e-01 2.98890620e-01 2.49314502e-01 -3.67470920e-01
-1.71887860e-01 2.32119977e-01 4.73592550e-01 -8.73820424e-01
-7.91790664e-01 2.70244062e-01 -4.94827181e-01 3.53635222e-01
1.10292423e+00 -1.07823110e+00 -1.52471077e+00 6.23788655e-01
-1.02038836e+00 -1.34518415e-01 5.18660307e-01 4.91699845e-01
-1.90152541e-01 -2.06777789e-02 -8.42949808e-01 -4.71324533e-01
-7.45432913e-01 -9.18184340e-01 1.11628044e+00 9.27192047e-02
-4.53072518e-01 -1.03438413e+00 3.54083359e-01 4.34599698e-01
-3.64145860e-02 6.83507979e-01 1.72015607e+00 -8.64003360e-01
-3.64136070e-01 1.50852725e-01 -4.00641233e-01 1.32457837e-01
-2.72581391e-02 1.82889611e-01 -8.82316649e-01 7.46104792e-02
-1.04972851e+00 -6.99864745e-01 9.08260822e-01 5.05932391e-01
1.59050953e+00 -4.83930051e-01 -4.13773865e-01 1.06450582e+00
1.10787678e+00 2.07259983e-01 8.09569716e-01 -1.06722519e-01
6.83911145e-01 5.35529792e-01 1.09613396e-01 4.41895515e-01
3.48203421e-01 4.04626608e-01 4.22899663e-01 -3.32981050e-01
2.00525120e-01 -9.51443166e-02 1.86467648e-01 1.82453647e-01
-3.50634396e-01 -4.49553907e-01 -9.21299458e-01 4.09218222e-01
-1.80455482e+00 -7.76149929e-01 -3.43883097e-01 1.75068998e+00
7.27155924e-01 -2.72913009e-01 -6.35689124e-02 -6.97711289e-01
5.30399680e-01 -5.11219501e-01 -1.23807955e+00 -6.48704112e-01
-5.15681624e-01 5.87636828e-01 8.20983797e-02 -1.41364828e-01
-8.19228292e-01 8.76419902e-01 7.30758667e+00 1.75248817e-01
-7.82975376e-01 4.80499975e-02 1.11405957e+00 -3.81460011e-01
-4.20137703e-01 -2.63597280e-01 -5.31756282e-01 4.94368583e-01
1.45192301e+00 2.42458507e-01 3.76300633e-01 4.46351260e-01
6.82166636e-01 3.86327654e-01 -1.13112462e+00 5.16284525e-01
-4.66892570e-01 -1.77056515e+00 -1.48345724e-01 2.37385467e-01
5.42655706e-01 3.67250830e-01 -1.49145499e-01 6.71776757e-02
4.77587163e-01 -1.61917233e+00 -7.60018528e-02 1.21079552e+00
8.47625911e-01 -8.59739900e-01 6.23360038e-01 -3.59128386e-01
-5.10321975e-01 -2.08752632e-01 -7.67055154e-01 -1.74018472e-01
-4.12285924e-01 9.86249983e-01 -1.25755119e+00 3.83666158e-01
6.76424325e-01 7.28868067e-01 -5.41579843e-01 8.94406199e-01
-2.35833749e-01 7.38030493e-01 1.84853271e-01 2.27932073e-02
-1.41493967e-02 -1.56188726e-01 2.34983429e-01 1.07160163e+00
4.30937320e-01 1.17168821e-01 -1.89354643e-01 1.56402457e+00
-2.14738727e-01 -1.41156942e-01 -4.40197378e-01 -6.94639862e-01
3.08962166e-01 1.43876481e+00 -4.17143196e-01 -2.37622187e-02
-2.01779857e-01 8.04607153e-01 6.27381384e-01 3.07180971e-01
-5.60615063e-01 -3.40687752e-01 1.54475439e+00 -2.93450385e-01
2.34797411e-02 1.90432668e-01 -2.90748607e-02 -1.03445148e+00
-3.51241320e-01 -9.47108388e-01 4.29424584e-01 -7.59969771e-01
-1.53393900e+00 -1.31368339e-01 -6.01344407e-01 -5.19622982e-01
-2.01677740e-01 -9.55042064e-01 -6.18979514e-01 9.89380777e-01
-1.12543976e+00 -1.29741633e+00 1.53562874e-01 -3.33523639e-02
1.28416955e-01 -4.23142463e-01 1.32370281e+00 1.33519560e-01
-1.32857382e+00 4.47719187e-01 5.30904293e-01 -3.80543098e-02
7.39990652e-01 -1.25392926e+00 8.82725120e-01 2.51484841e-01
-5.95243812e-01 1.08152187e+00 2.55470783e-01 -1.08426952e+00
-1.42251503e+00 -1.67488873e+00 6.88211024e-01 -2.92775124e-01
6.86047912e-01 -5.63795924e-01 -8.07684302e-01 9.18175101e-01
-3.10260653e-02 -7.88663849e-02 1.26409435e+00 5.95689833e-01
-3.43558401e-01 -1.00886010e-01 -1.37776530e+00 5.34724534e-01
1.09900534e+00 -2.16321111e-01 -3.03608507e-01 3.92077476e-01
9.27286208e-01 -1.00204520e-01 -1.20645428e+00 1.35101542e-01
1.07693255e+00 -7.10020125e-01 1.11215997e+00 -1.42963028e+00
7.50060201e-01 -7.88135976e-02 3.57672304e-01 -1.65233445e+00
-5.90907931e-01 -4.26901609e-01 -2.11302757e-01 1.03652751e+00
7.93568015e-01 -7.54088223e-01 4.92634743e-01 6.37842357e-01
-3.70244235e-01 -8.86899292e-01 -8.31839025e-01 -3.73719543e-01
-8.02568868e-02 -2.75268883e-01 1.04730511e+00 9.10714626e-01
3.01851034e-02 2.56172091e-01 -2.66173661e-01 2.33005241e-01
4.22734976e-01 -3.77206981e-01 2.77544677e-01 -1.63592958e+00
-2.68702775e-01 -1.61468729e-01 -6.94309294e-01 1.59604363e-02
2.79727817e-01 -1.10165000e+00 -4.51835215e-01 -1.40261805e+00
4.97663707e-01 -1.86164945e-01 -5.30304193e-01 9.53845203e-01
-1.36733904e-01 1.61630362e-01 -5.29740274e-01 -6.52159154e-01
-3.08292657e-01 2.84727484e-01 1.06301904e+00 -4.02274668e-01
4.38086763e-02 -3.32539171e-01 -1.30530798e+00 9.73395467e-01
8.41206431e-01 -2.56356418e-01 -2.32892707e-01 -6.52442634e-01
6.86652362e-01 -9.09728557e-02 8.01235855e-01 -5.20800769e-01
-4.88500595e-01 -2.31924519e-01 1.21825790e+00 -4.44206059e-01
3.91430885e-01 -2.85155118e-01 3.36858213e-01 7.98298120e-01
-5.40538132e-01 -6.27301261e-02 3.89252692e-01 6.86191440e-01
4.38517869e-01 2.04629615e-01 3.57341051e-01 -1.92629963e-01
-3.63745153e-01 3.96227062e-01 -4.85045820e-01 -3.01100731e-01
8.73442948e-01 2.32602611e-01 -7.16622412e-01 -6.72069415e-02
-1.18157017e+00 2.81885922e-01 -5.33647127e-02 4.49627429e-01
7.84914076e-01 -1.08116341e+00 -1.12052929e+00 2.19018653e-01
1.51203305e-01 -2.74748892e-01 5.37408769e-01 7.30344713e-01
-7.15073466e-01 7.96562672e-01 -6.02115989e-01 -1.94105640e-01
-1.28073037e+00 3.64771456e-01 3.38777035e-01 1.99296087e-01
-2.59932160e-01 9.12666798e-01 1.61566630e-01 -8.29285741e-01
-5.63327968e-02 -6.14317417e-01 -1.12587616e-01 -1.73332796e-01
7.10015476e-01 6.54464781e-01 4.61358055e-02 2.19926634e-03
-4.63942617e-01 -1.68335170e-01 1.78285632e-02 7.30567753e-01
1.96647894e+00 1.11446224e-01 -8.47256362e-01 2.25550011e-01
9.91188705e-01 -5.06945789e-01 -6.48981154e-01 3.18845421e-01
-5.86523488e-03 -5.53790890e-02 -1.98689550e-01 -1.45125020e+00
-9.77734625e-01 5.85472465e-01 8.74609351e-01 -3.72175217e-01
9.18995976e-01 -8.48534182e-02 7.28782177e-01 7.49331415e-01
1.23635024e-01 -8.36236358e-01 -5.31566441e-01 5.28577745e-01
7.60745943e-01 -9.93282259e-01 5.39817810e-02 -3.33193094e-01
7.93616697e-02 1.33499312e+00 9.11558807e-01 6.62375465e-02
2.82796860e-01 2.33666897e-02 -2.44273603e-01 -4.27437961e-01
-1.48412955e+00 1.98012725e-01 6.47448897e-01 7.26437867e-01
9.56199169e-01 4.18767840e-01 -5.99219024e-01 1.13684738e+00
-6.47552833e-02 1.87665030e-01 4.07548338e-01 1.78779483e-01
-1.09036759e-01 -1.18367279e+00 -9.37332064e-02 1.01555121e+00
-7.87203014e-01 -4.38823372e-01 -8.91378224e-01 5.00137568e-01
7.49438763e-01 5.21260977e-01 1.90687388e-01 -2.14177132e-01
-1.32248342e-01 3.47767800e-01 2.42508844e-01 -6.98971927e-01
-4.13349390e-01 2.46354744e-01 6.43064737e-01 -7.66034186e-01
-6.61352137e-03 -5.63296080e-01 -1.32131577e+00 -1.12474501e-01
7.52627477e-02 -4.22177315e-01 5.97786605e-01 9.79377747e-01
1.27826905e+00 9.74733293e-01 -1.32367015e-01 -2.73072153e-01
-6.75963983e-02 -9.99843121e-01 -7.34047949e-01 -1.12837233e-01
1.64795265e-01 -3.89162242e-01 4.67044264e-01 6.60549477e-02] | [6.123726844787598, 5.678106307983398] |
567f67ae-db80-44c5-8c5d-1c64d426ae16 | optimal-low-rank-matrix-completion | 2305.12292 | null | https://arxiv.org/abs/2305.12292v1 | https://arxiv.org/pdf/2305.12292v1.pdf | Optimal Low-Rank Matrix Completion: Semidefinite Relaxations and Eigenvector Disjunctions | Low-rank matrix completion consists of computing a matrix of minimal complexity that recovers a given set of observations as accurately as possible, and has numerous applications such as product recommendation. Unfortunately, existing methods for solving low-rank matrix completion are heuristics that, while highly scalable and often identifying high-quality solutions, do not possess any optimality guarantees. We reexamine matrix completion with an optimality-oriented eye, by reformulating low-rank problems as convex problems over the non-convex set of projection matrices and implementing a disjunctive branch-and-bound scheme that solves them to certifiable optimality. Further, we derive a novel and often tight class of convex relaxations by decomposing a low-rank matrix as a sum of rank-one matrices and incentivizing, via a Shor relaxation, that each two-by-two minor in each rank-one matrix has determinant zero. In numerical experiments, our new convex relaxations decrease the optimality gap by two orders of magnitude compared to existing attempts. Moreover, we showcase the performance of our disjunctive branch-and-bound scheme and demonstrate that it solves matrix completion problems over 150x150 matrices to certifiable optimality in hours, constituting an order of magnitude improvement on the state-of-the-art for certifiably optimal methods. | ['Jean Pauphilet', 'Sean Lo', 'Ryan Cory-Wright', 'Dimitris Bertsimas'] | 2023-05-20 | null | null | null | null | ['low-rank-matrix-completion', 'matrix-completion', 'product-recommendation'] | ['methodology', 'methodology', 'miscellaneous'] | [ 4.06324208e-01 3.35217237e-01 -2.19800949e-01 -3.55976894e-02
-1.08480310e+00 -1.00573087e+00 2.59244628e-03 2.75358707e-02
-1.68316402e-02 6.50335610e-01 2.38318831e-01 -7.28286505e-01
-7.59670258e-01 -5.10999143e-01 -1.00937819e+00 -7.12414980e-01
-5.01926780e-01 9.14710522e-01 -5.73323011e-01 -2.66526699e-01
1.90154865e-01 3.72022539e-01 -9.92402494e-01 4.15406823e-01
8.91560793e-01 9.28935647e-01 -1.80916709e-03 6.00248635e-01
4.46914166e-01 4.62407917e-01 4.62531531e-03 -5.19596696e-01
1.13773143e+00 2.41742618e-02 -8.87064874e-01 5.23547471e-01
6.66684151e-01 -3.02044958e-01 -5.23828447e-01 1.26689661e+00
-7.94841424e-02 -3.38237849e-03 4.43398431e-02 -1.26511347e+00
-4.96462673e-01 6.56469166e-01 -1.14374983e+00 6.96552917e-02
7.30807662e-01 -2.44041413e-01 1.64860678e+00 -8.47413599e-01
6.03461385e-01 1.12472641e+00 4.23185825e-01 -1.16168186e-02
-1.72024679e+00 -6.58669829e-01 2.34201759e-01 5.74081251e-03
-1.43934608e+00 -4.54822838e-01 2.88626164e-01 -3.37348461e-01
7.09136367e-01 8.46772552e-01 2.85735905e-01 3.85554016e-01
-1.75158218e-01 8.87265921e-01 1.28696048e+00 -8.29467773e-02
7.61798397e-02 -9.06406045e-02 4.11098659e-01 5.90381742e-01
1.02640140e+00 7.94944912e-02 -4.78186160e-01 -6.81264281e-01
3.85396659e-01 2.05917999e-01 -3.92092913e-01 -5.56244791e-01
-1.34961760e+00 8.02576780e-01 1.05287224e-01 -2.37268254e-01
-5.38125575e-01 2.46150479e-01 8.60286057e-02 4.03023452e-01
-7.18062297e-02 5.93630075e-01 -4.73862797e-01 7.48414099e-02
-1.03238869e+00 3.12829494e-01 1.24490309e+00 1.07536662e+00
9.42145467e-01 -2.28022829e-01 -9.37381685e-02 2.60377824e-01
1.37204947e-02 9.08213258e-01 -3.63793850e-01 -1.06124413e+00
9.20597613e-01 3.19351435e-01 2.08629891e-01 -1.48312235e+00
-2.55250901e-01 -6.62150502e-01 -9.26027119e-01 -2.23727331e-01
3.10555965e-01 -9.82528031e-02 -4.90690172e-01 1.55436814e+00
3.38448197e-01 1.31551102e-01 8.89341012e-02 1.18457329e+00
1.20161101e-01 8.96009564e-01 -6.97910249e-01 -6.73139393e-01
1.35318685e+00 -9.84146357e-01 -5.11603713e-01 -4.03393954e-01
5.13954759e-01 -7.62932956e-01 7.05789268e-01 9.93236661e-01
-1.22903728e+00 8.80037323e-02 -1.14347446e+00 8.79700258e-02
3.13539714e-01 7.54310563e-02 1.21466613e+00 8.41641903e-01
-8.98285747e-01 5.68629563e-01 -6.78046167e-01 1.95856407e-01
1.07091516e-01 7.57218719e-01 -8.44269931e-01 -5.88602364e-01
-5.25187433e-01 3.29176724e-01 1.95663020e-01 2.30139092e-01
-1.00347340e+00 -9.55409706e-01 -6.41292572e-01 2.31989294e-01
1.06726050e+00 -5.92128873e-01 9.58136320e-01 -5.01396060e-01
-9.41507518e-01 7.04401910e-01 -3.30828696e-01 -3.94410163e-01
1.57363638e-01 -2.98166990e-01 -2.75188446e-01 2.08010390e-01
1.90368488e-01 -1.93620577e-01 8.41313303e-01 -1.37366486e+00
-6.51186049e-01 -4.64697987e-01 5.05198717e-01 1.18178248e-01
-3.75668436e-01 -2.05526248e-01 -7.79551327e-01 -2.38190547e-01
6.77968621e-01 -1.33885086e+00 -9.87391174e-01 -4.14851457e-01
-7.25452662e-01 3.39154631e-01 1.24115728e-01 -7.14259267e-01
1.41876423e+00 -2.10361218e+00 4.69641238e-01 7.75594652e-01
8.77567887e-01 -2.87482645e-02 -3.84431660e-01 8.20818722e-01
-2.57916480e-01 7.66765848e-02 -1.50106773e-01 -3.43047231e-01
1.04325727e-01 3.29789400e-01 -6.53281212e-01 8.78504574e-01
-2.92586535e-01 5.27981937e-01 -1.10493815e+00 7.53997713e-02
-1.58511475e-01 -8.57292637e-02 -1.03352165e+00 -3.13022844e-02
1.98266506e-01 -6.54142275e-02 -4.44356233e-01 6.07345104e-01
1.18080604e+00 -6.03001595e-01 8.36309850e-01 -3.33390415e-01
9.92676765e-02 1.11709699e-01 -1.84404850e+00 1.57870734e+00
-2.59573281e-01 3.43391597e-01 7.39766479e-01 -1.01021862e+00
3.39671850e-01 8.08064416e-02 8.12391520e-01 -3.49171817e-01
1.51524106e-02 3.80896360e-01 -1.34807974e-02 -6.41389564e-02
7.56414235e-01 1.73966307e-02 -2.01625109e-01 6.31465554e-01
-4.10553783e-01 2.94056356e-01 6.31898522e-01 8.09137285e-01
1.47813618e+00 -6.69800997e-01 1.05421528e-01 -3.97417277e-01
5.29418349e-01 -9.63185281e-02 7.60813594e-01 8.75953078e-01
3.66902292e-01 5.57375669e-01 8.54601324e-01 -2.90010095e-01
-9.13222194e-01 -9.38214123e-01 1.34881109e-01 9.71009910e-01
-2.29290873e-02 -9.32804644e-01 -4.95014668e-01 -2.76964605e-01
4.09886479e-01 3.81624341e-01 -6.21978581e-01 2.57171303e-01
-2.06484288e-01 -1.01403296e+00 -3.46695222e-02 1.46101519e-01
6.94980100e-02 8.08740929e-02 1.17788009e-01 1.34142965e-01
-2.20884960e-02 -1.43271434e+00 -6.85902536e-01 2.36233160e-01
-9.69536304e-01 -1.32673550e+00 -2.53585994e-01 -4.17617887e-01
1.26747334e+00 1.00109684e+00 1.05654001e+00 6.05286956e-02
-2.26631910e-02 3.57222378e-01 -1.86366364e-01 1.92032397e-01
3.06338500e-02 -6.35410398e-02 4.32375222e-01 2.51467735e-01
1.46459052e-02 -7.27198422e-01 -4.19157445e-01 1.98042199e-01
-9.43450093e-01 -9.54426751e-02 8.86195004e-01 7.90150821e-01
8.54914904e-01 4.18416917e-01 -1.47303134e-01 -1.39726067e+00
7.90807009e-01 -5.29545546e-01 -1.10203922e+00 1.93988830e-01
-9.17127907e-01 3.70079219e-01 7.09267199e-01 -2.11205631e-01
-5.29306710e-01 6.05599880e-01 4.46005613e-01 -6.84252203e-01
6.58454061e-01 8.42153966e-01 -1.40400693e-01 -2.58821756e-01
5.81739902e-01 2.71753937e-01 -8.06705430e-02 -3.58978152e-01
7.35760152e-01 2.39093617e-01 7.42619634e-01 -9.59153056e-01
1.53153741e+00 7.97723651e-01 5.16906023e-01 -5.62496066e-01
-1.17730391e+00 -9.50200260e-01 -2.49372110e-01 3.15395713e-01
4.26953845e-03 -1.24031734e+00 -1.27295053e+00 -3.27526808e-01
-6.48956954e-01 1.70756370e-01 -9.96005237e-02 4.70403910e-01
-4.89307970e-01 7.61223555e-01 -5.84975004e-01 -8.02711368e-01
-2.51432240e-01 -1.00754392e+00 8.27930212e-01 -2.37417161e-01
1.11374371e-01 -5.20084858e-01 1.66657418e-01 7.69947290e-01
1.01396382e-01 1.39552370e-01 6.00066900e-01 -5.02031922e-01
-1.02037418e+00 -4.66725767e-01 -4.55454141e-01 5.58805764e-02
-4.45034325e-01 -5.04010797e-01 -3.93610865e-01 -8.73348951e-01
-2.19166219e-01 -9.87316594e-02 6.81031644e-01 7.61021152e-02
1.04539788e+00 -8.40552032e-01 -4.17165279e-01 8.64013433e-01
1.58796573e+00 -4.66586143e-01 6.57947063e-01 1.51695624e-01
7.08954811e-01 2.86532789e-01 8.19529772e-01 1.01883352e+00
2.91940540e-01 3.78336906e-01 6.14062309e-01 -1.03146903e-01
5.87121189e-01 -2.24062338e-01 2.15356752e-01 8.78619373e-01
-1.17434673e-01 1.08406387e-01 -4.62916851e-01 6.22252941e-01
-2.05901814e+00 -9.06408370e-01 -5.95650852e-01 2.70677185e+00
7.50286937e-01 -2.58146349e-04 1.06688313e-01 2.22990379e-01
3.78886282e-01 1.19989482e-03 -3.90765399e-01 -3.46722394e-01
-1.16347961e-01 2.40276605e-01 1.27817523e+00 6.07329011e-01
-7.92521536e-01 7.53403425e-01 6.69496155e+00 6.57483935e-01
-5.51186085e-01 -2.19297037e-03 3.03463608e-01 -4.27620590e-01
-5.98509610e-01 6.69508398e-01 -8.17921519e-01 -3.21102515e-02
8.23030055e-01 -4.38083082e-01 1.32351029e+00 1.01638937e+00
1.84177086e-01 6.80675432e-02 -1.41761744e+00 1.34775758e+00
1.10824920e-01 -1.40997362e+00 -1.31015658e-01 7.60505676e-01
1.27426839e+00 -2.73501277e-01 2.48324424e-01 1.16290428e-01
5.33181071e-01 -9.86021340e-01 4.32166129e-01 4.50690798e-02
6.45032823e-01 -6.89546168e-01 4.17364597e-01 2.45538741e-01
-1.17577314e+00 -3.86613905e-01 -6.78583264e-01 -3.75351846e-01
1.95763052e-01 1.14701283e+00 -5.92761517e-01 6.45555615e-01
5.36374092e-01 5.13195753e-01 -1.96046755e-01 1.00219607e+00
-9.79461148e-02 6.61769986e-01 -6.66554034e-01 3.45849365e-01
2.13963687e-01 -7.23948956e-01 6.60498440e-01 9.61647511e-01
2.02890918e-01 6.53921604e-01 5.75647891e-01 5.69556713e-01
-3.02648723e-01 2.27049828e-01 -3.62347752e-01 -4.47781473e-01
4.52899516e-01 1.52171135e+00 -4.57073689e-01 -1.72683105e-01
-3.75885487e-01 9.02841747e-01 3.62145782e-01 3.43303174e-01
-6.42730117e-01 -2.13123500e-01 9.77180362e-01 1.59739330e-01
5.37311494e-01 -4.85240996e-01 -5.22383690e-01 -1.26278508e+00
3.59023809e-01 -1.38862705e+00 5.18313348e-01 -3.08707356e-01
-1.21137917e+00 2.77521700e-01 -2.22093344e-01 -1.03624403e+00
-6.12652451e-02 -6.15821302e-01 -1.97117799e-03 6.68095589e-01
-1.23738301e+00 -7.62350440e-01 1.17560066e-02 6.30360126e-01
-1.54315189e-01 4.38690111e-02 5.13351083e-01 5.05937815e-01
-6.86643898e-01 7.26200342e-01 5.08132398e-01 -2.67640531e-01
2.72057265e-01 -1.27307451e+00 -1.26331430e-02 1.18698788e+00
1.59299418e-01 1.01276302e+00 9.37560678e-01 -4.99396712e-01
-2.59989357e+00 -6.95000410e-01 6.70006573e-01 -2.32237309e-01
1.03924549e+00 -3.23027045e-01 -4.18976456e-01 8.72808278e-01
-7.94890076e-02 5.72007746e-02 7.21775830e-01 8.42330217e-01
-6.22083127e-01 -3.75647694e-01 -8.84897172e-01 5.48732400e-01
1.12938821e+00 -5.64489722e-01 -2.80388325e-01 1.06271112e+00
4.64117885e-01 -6.78062916e-01 -9.11117315e-01 1.62064150e-01
3.14543635e-01 -7.29223609e-01 1.10342753e+00 -9.01672125e-01
3.72726083e-01 -5.39691925e-01 -5.47587395e-01 -7.73319125e-01
-7.77209759e-01 -1.41343701e+00 -4.94247019e-01 6.33655131e-01
5.36759496e-01 -5.11588037e-01 1.15023720e+00 7.10232496e-01
-2.18675677e-02 -7.89667308e-01 -6.26686931e-01 -1.01386631e+00
-3.68932903e-01 -4.64614749e-01 4.84342635e-01 8.24854553e-01
2.25430608e-01 4.89212155e-01 -9.19209361e-01 7.83767939e-01
8.59657526e-01 5.63370705e-01 1.19806719e+00 -9.69156086e-01
-8.71217191e-01 -9.74654555e-02 -1.80391341e-01 -1.53169239e+00
-4.05411273e-02 -1.13489676e+00 -1.82972714e-01 -1.33756399e+00
7.99845278e-01 -6.31565034e-01 -1.38367012e-01 5.83436310e-01
-8.44948515e-02 2.20476434e-01 4.21706468e-01 2.92943150e-01
-9.76073265e-01 3.70336920e-02 1.03899956e+00 -1.46117106e-01
-1.95048034e-01 -9.09103155e-02 -1.40420187e+00 2.73437023e-01
2.03459665e-01 -5.97104490e-01 -4.15760756e-01 -3.20644975e-01
9.36792314e-01 5.38751900e-01 -1.30934164e-01 -7.09280491e-01
3.54687303e-01 -4.50050980e-01 -2.34619930e-01 -7.19131231e-01
3.64836901e-01 -8.43015671e-01 5.38369000e-01 5.25410354e-01
-2.15528235e-01 1.77736938e-01 -9.48480517e-02 9.15445030e-01
1.20367341e-01 -1.84263781e-01 3.30284536e-01 1.98712438e-01
-2.32295826e-01 5.49490631e-01 -1.01280808e-01 2.03746989e-01
1.02664077e+00 -6.63751513e-02 -3.31059039e-01 -5.70703208e-01
-5.40136993e-01 4.79461789e-01 3.61320019e-01 3.85566354e-02
4.25105929e-01 -1.14563060e+00 -9.08592403e-01 1.43823056e-02
-3.48738506e-02 -2.16869473e-01 3.27607542e-01 1.15018666e+00
-4.58905011e-01 7.36588478e-01 1.91575855e-01 -4.53056067e-01
-1.25862467e+00 1.03385437e+00 -2.76496112e-01 -8.98818970e-01
-3.27226341e-01 8.02763343e-01 2.89645374e-01 -3.50742906e-01
-9.11944080e-03 -6.14344478e-02 4.78991061e-01 -3.62144530e-01
6.88517928e-01 6.06046557e-01 1.03880450e-01 -4.43044275e-01
-2.76957303e-01 2.36351013e-01 -3.87630165e-01 -4.27960493e-02
1.34970260e+00 -7.89179206e-02 -5.78186870e-01 -3.10259163e-01
1.24534082e+00 6.92408025e-01 -9.81532454e-01 -3.35571855e-01
-1.85587108e-01 -8.92848611e-01 7.02793300e-02 -4.75360781e-01
-1.44478047e+00 3.38491261e-01 -6.11246154e-02 1.72095671e-01
1.29928017e+00 -1.99732587e-01 8.89662862e-01 9.46316719e-01
6.88639522e-01 -8.54523480e-01 -1.88557073e-01 3.69320601e-01
7.74706364e-01 -9.47537959e-01 7.80095756e-01 -1.05374837e+00
-5.13337314e-01 8.72597337e-01 -3.55698802e-02 -3.49392563e-01
5.79246879e-01 2.20681384e-01 -5.51976085e-01 -1.98858023e-01
-9.15669680e-01 -1.47478208e-01 3.44845772e-01 2.55569458e-01
-4.90966141e-02 5.20382941e-01 -5.77277005e-01 7.80012846e-01
-3.76529276e-01 -2.61407405e-01 8.23299289e-01 7.82633841e-01
-3.04048896e-01 -1.22681892e+00 -7.39530504e-01 7.97213733e-01
-4.08652395e-01 -3.29636782e-01 -1.03554480e-01 3.63399625e-01
-4.36061502e-01 1.25168872e+00 -3.97584975e-01 -6.91558361e-01
2.50971287e-01 -8.01239133e-01 4.79492813e-01 -7.03698039e-01
-3.20650816e-01 3.32646221e-02 2.95214266e-01 -8.22427690e-01
1.90752238e-01 -8.01748455e-01 -1.08916879e+00 -9.52135146e-01
-3.48839402e-01 3.28341395e-01 4.83503848e-01 8.61238539e-01
5.65491080e-01 -2.80993804e-03 1.07103157e+00 -4.97614533e-01
-1.20576942e+00 -3.55416268e-01 -8.32656741e-01 4.36722845e-01
5.67212962e-02 -3.00026119e-01 -4.84272510e-01 -2.59014875e-01] | [6.85284423828125, 4.810906410217285] |
69e81f72-6132-4d4a-a77b-3fe2aa44a9e7 | learning-semantic-relatedness-in-community | null | null | https://aclanthology.org/W16-1616 | https://aclanthology.org/W16-1616.pdf | Learning Semantic Relatedness in Community Question Answering Using Neural Models | null | ['Mitra Mohtarami', 'James Glass', 'Henry Nassif'] | 2016-08-01 | null | null | null | ws-2016-8 | ['question-similarity'] | ['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] | [-7.355537414550781, 3.8220536708831787] |
e46222c1-85ad-4598-a369-33a914523bc7 | modeling-generalized-rate-distortion | 1906.05178 | null | http://arxiv.org/abs/1906.05178v1 | http://arxiv.org/pdf/1906.05178v1.pdf | Modeling Generalized Rate-Distortion Functions | Many multimedia applications require precise understanding of the
rate-distortion characteristics measured by the function relating visual
quality to media attributes, for which we term it the generalized
rate-distortion (GRD) function. In this study, we explore the GRD behavior of
compressed digital videos in a three-dimensional space of bitrate, resolution,
and viewing device/condition. Our analysis on a large-scale video dataset
reveals that empirical parametric models are systematically biased while
exhaustive search methods require excessive computation time to depict the GRD
surfaces. By exploiting the properties that all GRD functions share, we develop
an Robust Axial-Monotonic Clough-Tocher (RAMCT) interpolation method to model
the GRD function. This model allows us to accurately reconstruct the complete
GRD function of a source video content from a moderate number of measurements.
To further reduce the computational cost, we present a novel sampling scheme
based on a probabilistic model and an information measure. The proposed
sampling method constructs a sequence of quality queries by minimizing the
overall informativeness in the remaining samples. Experimental results show
that the proposed algorithm significantly outperforms state-of-the-art
approaches in accuracy and efficiency. Finally, we demonstrate the usage of the
proposed model in three applications: rate-distortion curve prediction,
per-title encoding profile generation, and video encoder comparison. | [] | 2019-06-12 | null | null | null | null | ['pgtask'] | ['natural-language-processing'] | [ 4.09169197e-01 -5.61531603e-01 -3.39160234e-01 -4.33932513e-01
-1.03619230e+00 -4.21254724e-01 2.90779561e-01 -6.44285530e-02
-3.51166390e-02 5.16382515e-01 2.49684095e-01 -2.23780423e-01
-5.21633148e-01 -5.67312241e-01 -7.13526785e-01 -6.42306030e-01
-2.36431196e-01 -1.01104915e-01 4.12867397e-01 2.41369918e-01
5.78181446e-01 4.45854574e-01 -1.62870193e+00 5.41849807e-02
8.21430981e-01 1.49938953e+00 5.03052592e-01 6.66266739e-01
3.92863840e-01 5.33018947e-01 -4.69464213e-01 -4.58964676e-01
3.39687496e-01 -5.03829718e-01 -2.83359051e-01 4.18244720e-01
1.07970044e-01 -7.80272543e-01 -3.55273664e-01 1.19636381e+00
5.25432110e-01 5.47322817e-02 9.57008183e-01 -1.07141113e+00
-4.08894897e-01 5.64609878e-02 -5.46696544e-01 3.20538878e-01
6.21228635e-01 -7.77022764e-02 9.83482897e-01 -8.43068659e-01
5.72369456e-01 1.01241314e+00 4.69588310e-01 9.29945856e-02
-1.06392431e+00 -3.43586951e-01 -3.08695316e-01 3.97585303e-01
-1.85364771e+00 -5.41742861e-01 8.31617475e-01 -5.71482122e-01
5.15116870e-01 2.82787144e-01 4.46791381e-01 6.82349622e-01
2.91111231e-01 5.12149334e-01 8.05453420e-01 -4.31435168e-01
3.77828002e-01 7.53536448e-02 -4.56145257e-01 7.79063880e-01
2.74550259e-01 -1.05155362e-02 -5.25325954e-01 -3.02798599e-01
1.10940218e+00 -2.98605025e-01 -6.59491360e-01 -7.61464298e-01
-7.14870214e-01 5.29465735e-01 -1.66058406e-01 -9.76792797e-02
-3.72310966e-01 6.08448274e-02 2.04769433e-01 1.53160721e-01
5.28007507e-01 7.32454529e-04 -8.32018852e-02 -4.25624490e-01
-1.29423261e+00 3.44357818e-01 6.84034705e-01 1.28065276e+00
4.72753316e-01 -1.10302940e-01 -1.38774067e-01 1.05053294e+00
4.97345150e-01 4.08346862e-01 2.87782401e-01 -1.31216836e+00
6.84229136e-01 1.65146112e-01 3.09184223e-01 -1.23269129e+00
1.07094087e-01 -2.45564044e-01 -7.26040423e-01 1.00903064e-01
1.85236633e-01 2.87861407e-01 -3.30873966e-01 1.22531092e+00
1.58546031e-01 -9.33602601e-02 -2.08702549e-01 9.92682040e-01
3.25314611e-01 7.69104064e-01 -3.40217054e-01 -7.29610741e-01
1.08368504e+00 -3.03299665e-01 -7.53384769e-01 4.23670083e-01
2.28954852e-01 -7.65437841e-01 7.87286639e-01 6.59842253e-01
-1.54228532e+00 -5.08054197e-01 -1.36296141e+00 4.78246510e-02
3.73531967e-01 3.01159233e-01 1.08178854e-02 8.93276632e-01
-9.62032855e-01 6.92737401e-01 -5.61663032e-01 4.01089108e-03
4.06925648e-01 1.47260148e-02 1.22775465e-01 -3.86219919e-01
-8.47305715e-01 3.57408077e-01 2.25266054e-01 -2.25154355e-01
-1.09685886e+00 -6.96134567e-01 -6.39732778e-01 1.52811497e-01
4.51040864e-01 -4.22987252e-01 1.02200258e+00 -6.62019253e-01
-1.55591142e+00 5.96568406e-01 -2.34112605e-01 -3.64196301e-01
6.34806454e-01 -1.95346296e-01 -3.78718257e-01 6.72166467e-01
-3.50023508e-01 1.71741247e-01 1.02993250e+00 -1.40337288e+00
-4.85692590e-01 -1.14941172e-01 -1.82339460e-01 2.26381734e-01
-1.90092057e-01 1.64582729e-01 -8.30325961e-01 -7.94577718e-01
2.93730587e-01 -4.13728714e-01 2.35265732e-01 3.11583638e-01
-1.94849759e-01 9.71386880e-02 4.82696563e-01 -7.88388312e-01
1.76956928e+00 -2.24383235e+00 1.75119117e-01 3.26321781e-01
1.29241109e-01 6.77393377e-02 1.12956524e-01 3.91916364e-01
2.62057126e-01 2.36324564e-01 -2.15516344e-01 -9.36932340e-02
-8.56864378e-02 -3.71520780e-02 6.74655363e-02 7.05574453e-01
-4.97928448e-02 2.96223789e-01 -6.92466617e-01 -6.64196432e-01
1.53417915e-01 5.02436280e-01 -8.96877825e-01 5.51335990e-01
-1.28441423e-01 3.42184342e-02 -3.49807382e-01 7.32273281e-01
7.95969188e-01 -2.57588536e-01 1.78274289e-01 -4.65249389e-01
-7.65216798e-02 1.17458403e-01 -1.30692565e+00 1.58741140e+00
-3.56647968e-01 6.64613903e-01 -4.14118208e-02 -8.47041965e-01
1.07201040e+00 2.44721279e-01 5.77017546e-01 -5.74430823e-01
6.03233539e-02 3.11418086e-01 -3.49695444e-01 -7.52021790e-01
5.30320287e-01 1.62078515e-01 3.16048056e-01 1.91528708e-01
-1.93477813e-02 -7.33493716e-02 1.36353150e-01 8.63457769e-02
7.46268868e-01 7.14319348e-02 5.51754892e-01 -2.39931718e-01
4.25913781e-01 -6.63966298e-01 4.43100154e-01 5.95438302e-01
-1.71070978e-01 9.21613932e-01 7.30174184e-01 -5.01811467e-02
-1.58426940e+00 -1.02391398e+00 -3.14289004e-01 5.53871393e-01
2.88059533e-01 -4.63354260e-01 -9.86289382e-01 -1.27489895e-01
-3.20558667e-01 6.13241673e-01 -2.20244214e-01 6.38381317e-02
-3.50751221e-01 -5.17065704e-01 4.44918454e-01 6.75848648e-02
6.23274148e-01 -3.34383637e-01 -6.44145846e-01 3.78341079e-02
-2.94219047e-01 -1.28752697e+00 -5.67074180e-01 -4.63268697e-01
-8.24778795e-01 -1.00635183e+00 -8.33756924e-01 -3.38192225e-01
3.34247947e-01 2.98899174e-01 1.08250594e+00 -3.24181244e-02
-2.50132859e-01 5.41568100e-01 -5.11359632e-01 1.65854588e-01
-5.67055523e-01 -4.30586159e-01 -1.29807815e-01 3.29332948e-01
-7.65286833e-02 -6.77795887e-01 -9.52289224e-01 6.87593758e-01
-9.80915844e-01 5.95523184e-03 2.71180421e-01 4.45036948e-01
9.00054455e-01 4.64054614e-01 3.19740832e-01 -3.12874556e-01
5.75840175e-01 -7.80134916e-01 -8.62627506e-01 1.94773614e-01
-7.75457799e-01 7.16200843e-02 4.75146741e-01 -3.06580931e-01
-1.11598229e+00 -3.30676317e-01 -5.85950259e-03 -6.84297502e-01
1.35634884e-01 3.01955253e-01 -3.81495237e-01 8.65920167e-03
4.91084814e-01 5.04850984e-01 -6.13192432e-02 -6.45353615e-01
-6.14366680e-03 8.62534225e-01 5.15299737e-01 -6.40654624e-01
5.49512386e-01 3.15173298e-01 3.17324132e-01 -9.68210638e-01
-4.07378376e-01 -4.30995464e-01 -1.93425104e-01 -4.48049515e-01
6.02154493e-01 -9.04672742e-01 -6.16761088e-01 3.29065323e-01
-9.71968830e-01 7.55397826e-02 -4.51121107e-02 5.71699500e-01
-9.70306277e-01 7.08452821e-01 -4.09242839e-01 -1.19082022e+00
3.21077220e-02 -1.34527695e+00 1.03645074e+00 -9.44742262e-02
2.48189658e-01 -8.06459963e-01 -1.27518982e-01 2.39484593e-01
2.68073589e-01 1.37456998e-01 9.49126959e-01 -1.38160363e-01
-9.84617174e-01 -7.36328587e-02 -4.28464085e-01 6.77168489e-01
7.06864744e-02 1.18289776e-01 -6.75878048e-01 -2.25381792e-01
3.09916854e-01 1.36846909e-02 4.72989440e-01 7.02095270e-01
1.68710291e+00 -5.23681402e-01 5.61766550e-02 1.05053079e+00
1.78921211e+00 3.59881490e-01 8.87032032e-01 1.00278206e-01
4.17465955e-01 3.95474523e-01 7.96662867e-01 1.11721289e+00
1.83814242e-01 1.01864290e+00 5.77670217e-01 5.82543671e-01
-1.56251222e-01 -4.18710560e-01 3.43023866e-01 7.85079002e-01
-3.46282989e-01 -6.28248930e-01 -4.82766658e-01 3.45494896e-01
-1.37254143e+00 -9.26854312e-01 1.12913966e-01 2.53243780e+00
7.62501657e-01 1.62831292e-01 2.06111744e-01 4.22950208e-01
7.08188951e-01 2.30290592e-01 -3.39991748e-01 -2.25741923e-01
-1.08820431e-01 -2.11718231e-01 6.66126490e-01 5.15703797e-01
-7.50506341e-01 2.68174037e-02 6.55066776e+00 1.34695148e+00
-6.70271933e-01 -2.10042924e-01 7.21852541e-01 -2.08288096e-02
-5.04858017e-01 -2.73348778e-01 -8.07875276e-01 7.44604647e-01
1.08948123e+00 -3.83435488e-01 6.67435884e-01 7.22581148e-01
6.49834454e-01 -2.23023430e-01 -1.03929484e+00 1.36034048e+00
2.55322695e-01 -1.24931240e+00 2.84856677e-01 1.65659562e-01
5.44226348e-01 -5.64694345e-01 2.01278076e-01 -4.75941926e-01
-4.55022275e-01 -9.04977083e-01 9.36644912e-01 7.32263863e-01
1.22000253e+00 -5.88545084e-01 4.76829588e-01 2.21088812e-01
-1.15517867e+00 -1.73552126e-01 -3.46568286e-01 3.38019788e-01
2.64811367e-01 5.46634674e-01 -6.50500834e-01 6.06358767e-01
5.24619222e-01 5.17362833e-01 -2.89105088e-01 1.48757887e+00
3.33206892e-01 6.17501855e-01 -1.57044441e-01 1.73013378e-02
-2.69280016e-01 -2.24349320e-01 7.12614655e-01 1.11488569e+00
9.51139152e-01 2.84565926e-01 -3.25793713e-01 7.58741260e-01
-4.92696203e-02 1.67094380e-01 -3.56884331e-01 1.40084118e-01
7.29485095e-01 6.73907518e-01 -2.69795448e-01 -6.13784268e-02
-3.65080655e-01 7.98649669e-01 -2.44724959e-01 5.01097739e-01
-7.94548273e-01 -1.22813262e-01 5.41444123e-01 6.00344121e-01
5.53980947e-01 -2.98399955e-01 -3.93743254e-02 -8.50865722e-01
3.49409282e-01 -8.29093516e-01 6.99018762e-02 -9.19197083e-01
-9.30867076e-01 4.03058469e-01 5.54417849e-01 -1.78331733e+00
-5.56974649e-01 -4.20645177e-01 -3.08917277e-03 7.42623031e-01
-1.51282990e+00 -3.28464597e-01 -2.78172702e-01 6.67773783e-01
8.81336808e-01 -2.55476177e-01 3.86404514e-01 5.16838193e-01
-2.32824117e-01 7.05816865e-01 4.16472256e-01 -2.56363899e-01
2.12718293e-01 -7.30077565e-01 -1.08658716e-01 6.96719050e-01
-2.44855523e-01 2.01581419e-01 1.00127220e+00 -3.64802241e-01
-1.46346903e+00 -7.68818557e-01 5.07385254e-01 1.32025421e-01
2.96515197e-01 -1.39930680e-01 -9.08079684e-01 1.05125591e-01
-1.42011479e-01 1.55240476e-01 5.12158215e-01 -5.80782473e-01
-3.02887648e-01 -2.77642012e-01 -1.42026985e+00 2.97622114e-01
1.03309870e+00 -4.51403677e-01 -6.77224770e-02 8.70706365e-02
5.91180980e-01 -3.24207902e-01 -1.12627375e+00 4.25855160e-01
6.38695002e-01 -1.50167429e+00 9.76038694e-01 1.72662586e-01
7.37451851e-01 -2.94593006e-01 -8.59119534e-01 -9.17813003e-01
-1.87176287e-01 -8.47858489e-01 -5.37095845e-01 1.11309266e+00
8.36359430e-03 -1.02007426e-01 4.95913655e-01 4.17333007e-01
1.43410079e-02 -1.07490242e+00 -1.02879214e+00 -8.79510581e-01
-4.80777979e-01 -6.08690977e-01 4.50528115e-01 2.63714015e-01
-2.12513566e-01 -3.16399336e-01 -5.54723680e-01 2.60744542e-01
1.08353257e+00 -2.72757858e-01 3.80418450e-01 -7.61354506e-01
-5.82277298e-01 -2.13919267e-01 -6.85585201e-01 -1.61923695e+00
-4.67609018e-01 -3.77586037e-01 -1.51982591e-01 -1.11491585e+00
3.83285403e-01 -5.02069533e-01 -5.56123555e-02 -7.35454917e-01
1.52570829e-01 1.93196945e-02 8.37293789e-02 3.81743670e-01
-6.36904895e-01 6.10715151e-01 1.27607334e+00 2.45905831e-01
-9.11090598e-02 1.50609583e-01 -3.46573502e-01 6.97091103e-01
5.03068566e-01 -2.93101251e-01 -7.30878472e-01 -3.45205724e-01
3.75095189e-01 6.08988106e-01 4.45354134e-01 -1.02695429e+00
-2.15806775e-02 -1.71853244e-01 2.72812128e-01 -5.54518104e-01
5.37236750e-01 -9.07503784e-01 3.17388147e-01 -9.79426131e-02
-3.74440104e-01 -4.62546460e-02 -2.57857174e-01 7.19361961e-01
-2.61054277e-01 -5.43360710e-01 7.99041033e-01 2.05502018e-01
-3.95419389e-01 2.87725538e-01 -1.92282811e-01 6.83701411e-02
8.71677160e-01 -5.10147393e-01 -6.91505671e-02 -6.81573331e-01
-2.18432724e-01 -3.40361863e-01 5.68925500e-01 4.05512899e-02
1.03998053e+00 -1.49657714e+00 -7.68318772e-01 3.00830692e-01
-1.90590136e-02 -3.39969635e-01 3.12632173e-01 4.75362211e-01
-8.01763952e-01 2.98074991e-01 9.14283469e-02 -7.56172419e-01
-1.15298045e+00 5.45083344e-01 3.05798262e-01 -1.48104250e-01
-3.46287489e-01 5.32422423e-01 6.78538904e-02 5.17156959e-01
3.09341729e-01 -3.63856375e-01 -1.43032208e-01 -1.33221343e-01
5.61966896e-01 7.18950152e-01 -2.32661217e-02 -6.94096506e-01
-1.12230487e-01 7.14711010e-01 2.71570295e-01 -2.88750798e-01
1.09139562e+00 -6.68032706e-01 3.30696672e-01 3.82113934e-01
1.55078781e+00 7.65376315e-02 -1.68858409e+00 -1.21773332e-01
-3.66334200e-01 -1.15421677e+00 1.53602511e-01 -3.99647057e-01
-9.67965007e-01 6.81642890e-01 7.69365430e-01 5.84228098e-01
1.64270270e+00 -1.30782381e-01 5.30525148e-01 -1.87847659e-01
5.13918638e-01 -9.54136670e-01 7.77764013e-04 -9.65614766e-02
1.02655280e+00 -8.57262075e-01 3.22086394e-01 -6.88126445e-01
-4.67113763e-01 1.14827383e+00 -8.00734684e-02 -2.37648562e-03
7.92794049e-01 1.17454179e-01 -5.22983730e-01 1.22438990e-01
-7.89136291e-01 3.62041503e-01 5.11481702e-01 6.37817979e-01
3.24220210e-01 -2.09460799e-02 -3.64452064e-01 3.64505827e-01
-4.55917940e-02 2.27063566e-01 6.23516679e-01 5.55171132e-01
-5.25886476e-01 -8.73536110e-01 -3.86042774e-01 3.37758571e-01
-6.79770947e-01 -1.69044107e-01 4.64941770e-01 4.99164313e-01
-1.75140440e-01 9.55293894e-01 4.49573621e-02 -3.33642840e-01
3.13078851e-01 -2.97098845e-01 6.83474958e-01 -9.75471921e-03
2.68283218e-01 3.15912366e-01 6.30075857e-02 -7.14331329e-01
-4.94996279e-01 -7.36614943e-01 -6.56096518e-01 -2.73066580e-01
-1.74933776e-01 -8.96452367e-02 8.77159655e-01 6.53065860e-01
2.56547481e-01 2.30502978e-01 1.09923160e+00 -7.57054090e-01
-7.23682761e-01 -6.50447845e-01 -7.72148132e-01 4.00984943e-01
5.12403488e-01 -6.83296204e-01 -4.63937461e-01 3.63876641e-01] | [11.469207763671875, -1.839739441871643] |
0fdd11bf-b765-42c2-8bdc-fd36d630ab06 | rich-feature-distillation-with-feature | 2207.11250 | null | https://arxiv.org/abs/2207.11250v1 | https://arxiv.org/pdf/2207.11250v1.pdf | Rich Feature Distillation with Feature Affinity Module for Efficient Image Dehazing | Single-image haze removal is a long-standing hurdle for computer vision applications. Several works have been focused on transferring advances from image classification, detection, and segmentation to the niche of image dehazing, primarily focusing on contrastive learning and knowledge distillation. However, these approaches prove computationally expensive, raising concern regarding their applicability to on-the-edge use-cases. This work introduces a simple, lightweight, and efficient framework for single-image haze removal, exploiting rich "dark-knowledge" information from a lightweight pre-trained super-resolution model via the notion of heterogeneous knowledge distillation. We designed a feature affinity module to maximize the flow of rich feature semantics from the super-resolution teacher to the student dehazing network. In order to evaluate the efficacy of our proposed framework, its performance as a plug-and-play setup to a baseline model is examined. Our experiments are carried out on the RESIDE-Standard dataset to demonstrate the robustness of our framework to the synthetic and real-world domains. The extensive qualitative and quantitative results provided establish the effectiveness of the framework, achieving gains of upto 15\% (PSNR) while reducing the model size by $\sim$20 times. | ['Varun P. Gopi', 'Nisha J. S.', 'Anushri Suresh', 'Sai Mitheran'] | 2022-07-13 | null | null | null | null | ['image-dehazing', 'single-image-haze-removal'] | ['computer-vision', 'computer-vision'] | [ 4.20375824e-01 1.09652117e-01 1.55403381e-02 -1.56827956e-01
-7.55908847e-01 -1.20310307e-01 4.77214456e-01 -1.76612601e-01
-4.93682265e-01 5.02988398e-01 -9.11303684e-02 -1.52367100e-01
-4.93052155e-02 -7.32966900e-01 -8.18767846e-01 -9.64613438e-01
1.44090533e-01 -1.12167656e-01 6.59345865e-01 -2.96097040e-01
2.02114657e-01 3.63279849e-01 -1.96915877e+00 1.71293408e-01
1.27316749e+00 8.05086792e-01 4.51232612e-01 4.90674287e-01
2.99430974e-02 8.64440441e-01 -6.47982895e-01 -4.41301495e-01
4.00086492e-01 -2.25793988e-01 -5.07053375e-01 3.00986886e-01
8.83445203e-01 -4.15890723e-01 -1.82795465e-01 1.30510819e+00
5.03698528e-01 3.04605097e-01 4.10151362e-01 -9.10281897e-01
-6.44271135e-01 1.41641498e-01 -6.97857559e-01 5.27375638e-01
8.68918374e-03 2.25772798e-01 6.73425257e-01 -1.05169082e+00
5.12145936e-01 9.52595294e-01 4.59795058e-01 3.81641448e-01
-1.05234623e+00 -8.32863510e-01 2.84503818e-01 5.01207471e-01
-1.44174838e+00 -4.16676849e-01 7.29099870e-01 -3.75347883e-01
7.07439542e-01 3.37266475e-02 5.08988023e-01 7.18973041e-01
-7.56541416e-02 8.16163003e-01 1.33285928e+00 -5.49279094e-01
2.67019600e-01 5.71489751e-01 4.05375324e-02 7.22398758e-01
4.60164607e-01 2.08681095e-02 -6.78353190e-01 4.12932903e-01
7.42502630e-01 -1.65469751e-01 -3.61408263e-01 -2.43931919e-01
-3.89125884e-01 8.28105867e-01 6.05197847e-01 1.89367384e-01
-3.28571022e-01 -1.55020401e-01 -2.43836865e-01 1.67328089e-01
8.56833696e-01 2.26241782e-01 -5.57600409e-02 2.75249660e-01
-1.16073978e+00 2.31137365e-01 5.18982708e-01 8.20087790e-01
8.99697483e-01 3.14493150e-01 -4.56742346e-02 7.53409624e-01
1.05632409e-01 4.38538998e-01 1.90988138e-01 -1.09939873e+00
2.72940606e-01 5.29554665e-01 1.00129418e-01 -9.91323173e-01
5.06838271e-03 -7.66037285e-01 -6.60166025e-01 5.78480005e-01
2.10780740e-01 -2.48904582e-02 -1.24037051e+00 1.42478061e+00
7.61871576e-01 6.89406037e-01 -2.21488881e-03 1.06957698e+00
8.63273621e-01 7.14716017e-01 -3.35791744e-02 -1.78973600e-01
1.31020260e+00 -1.30612338e+00 -5.55509329e-01 -3.36580515e-01
3.06461632e-01 -8.06895196e-01 1.00553834e+00 4.87832218e-01
-1.31423783e+00 -5.40120602e-01 -1.05968249e+00 -1.95769623e-01
-4.96955335e-01 -1.78780466e-01 3.98019552e-01 7.19446242e-01
-1.11736548e+00 4.18510497e-01 -6.23122454e-01 -2.89875716e-01
7.85323381e-01 2.71729261e-01 -1.64898708e-01 -4.33013111e-01
-9.69306350e-01 8.80524993e-01 4.13668245e-01 -2.01683100e-02
-1.07602274e+00 -1.05295563e+00 -6.29421115e-01 1.23585269e-01
6.32491469e-01 -8.03389966e-01 9.15881813e-01 -1.01338434e+00
-1.43927157e+00 7.40081847e-01 1.45272883e-02 -4.93535638e-01
2.13871151e-01 -5.79322934e-01 -2.00098187e-01 3.84954751e-01
6.70819432e-02 5.32382488e-01 1.15486348e+00 -1.53566515e+00
-1.06560886e+00 -2.73640305e-01 3.21408182e-01 4.82151091e-01
-6.15590870e-01 -1.43203908e-03 -5.33664346e-01 -7.66612649e-01
-8.38703886e-02 -7.66863406e-01 -1.46837845e-01 -1.15137234e-01
-1.00676201e-01 1.73770070e-01 9.28181469e-01 -8.01669359e-01
1.12992144e+00 -2.15518332e+00 1.26970420e-02 -4.95271310e-02
2.34358788e-01 6.74032092e-01 -1.18766613e-01 -3.44874635e-02
7.50474557e-02 -2.41728406e-02 -3.41348827e-01 -2.94504106e-01
-4.49306667e-01 9.55241844e-02 -2.38468200e-01 4.49480087e-01
3.80611777e-01 6.40360653e-01 -7.84424067e-01 -6.35354996e-01
3.71808589e-01 9.30023313e-01 -5.23424923e-01 4.31568503e-01
-1.61634102e-01 4.49189961e-01 -3.36506963e-01 7.17118323e-01
8.09821606e-01 -1.31008968e-01 -3.24755669e-01 -8.15265477e-02
-3.17092955e-01 -2.60947291e-02 -1.20994985e+00 1.47681868e+00
-4.30320114e-01 5.35236955e-01 4.01508063e-01 -8.56824219e-01
4.26669419e-01 1.01553813e-01 2.86768645e-01 -9.89143014e-01
1.18564621e-01 -2.88135614e-02 -3.19603711e-01 -5.82813025e-01
5.39963841e-01 -7.80269429e-02 4.76666838e-01 1.52655050e-01
1.09632507e-01 -1.32501394e-01 2.15825364e-01 2.52908915e-01
9.87501800e-01 1.70953408e-01 8.41311142e-02 -2.69051492e-01
5.62237680e-01 2.24357508e-02 3.38915527e-01 6.57253087e-01
-1.49903670e-01 6.71121538e-01 -2.19096262e-02 -1.96790904e-01
-7.27610111e-01 -1.09270406e+00 2.71433126e-02 1.33249974e+00
3.66690874e-01 -7.97739401e-02 -1.10935748e+00 -4.96453285e-01
-2.98940748e-01 6.66067004e-01 -5.24434745e-01 -1.40550360e-01
-5.50711572e-01 -8.84946644e-01 2.41607338e-01 1.63526356e-01
9.47146595e-01 -8.09360266e-01 -8.51477325e-01 -5.60385957e-02
-1.77397713e-01 -1.51024401e+00 -7.88419321e-02 1.58206057e-02
-7.48261750e-01 -8.98384333e-01 -7.06163704e-01 -8.80924702e-01
5.87481081e-01 9.09717500e-01 9.77555752e-01 1.34354040e-01
-4.80967730e-01 3.45607996e-01 -3.24970752e-01 -4.85647380e-01
-2.59819701e-02 1.36300415e-01 -2.84464687e-01 1.83984175e-01
1.01825982e-01 -7.51641691e-01 -1.07297075e+00 2.10378796e-01
-1.23012376e+00 2.33618826e-01 8.35783362e-01 6.90311909e-01
5.61329603e-01 3.79658252e-01 2.75442004e-01 -9.93138969e-01
2.45623097e-01 -4.04418647e-01 -7.76766181e-01 5.28270220e-06
-8.48363638e-01 -1.52780682e-01 3.56784582e-01 -2.58214325e-01
-1.59030640e+00 -2.47158436e-03 9.21614170e-02 -5.81022203e-01
-2.17732608e-01 2.85064816e-01 -2.24996388e-01 -4.60630476e-01
5.36730587e-01 3.51750940e-01 -2.38489032e-01 -4.98962373e-01
7.08677709e-01 4.89835829e-01 8.24251950e-01 -4.83713299e-01
1.27576721e+00 8.81833315e-01 -3.27341445e-02 -1.05605614e+00
-1.25915360e+00 -6.06501758e-01 -5.80082417e-01 -1.68950737e-01
9.46431339e-01 -1.33239627e+00 -3.44339192e-01 4.51414853e-01
-8.45109046e-01 -2.48173714e-01 -1.73112035e-01 1.64989471e-01
-1.91911668e-01 3.06779206e-01 -3.67543340e-01 -8.19561720e-01
-3.11301321e-01 -9.93669569e-01 8.93207371e-01 5.98329544e-01
5.20532906e-01 -8.95888865e-01 -1.51729599e-01 9.35019314e-01
6.37081325e-01 1.53442129e-01 8.43552172e-01 -2.30456471e-01
-1.03196609e+00 1.21200301e-01 -5.52150249e-01 5.80039680e-01
-1.01736739e-01 -2.21272349e-01 -1.50359297e+00 -2.69261241e-01
1.07822552e-01 -1.46755531e-01 1.18709528e+00 1.48528591e-01
1.03867972e+00 -1.85050383e-01 -6.80665672e-02 7.99240828e-01
1.63466084e+00 -1.49540722e-01 6.36864066e-01 5.24173200e-01
9.08095121e-01 7.14954734e-01 6.01799846e-01 2.09906965e-01
4.58569527e-01 6.17357314e-01 8.53630066e-01 -4.29324150e-01
-6.58340812e-01 3.75072695e-02 3.51584345e-01 6.52920842e-01
-2.97417283e-01 -1.41475245e-01 -6.82957292e-01 8.83032322e-01
-1.59977961e+00 -8.05693209e-01 5.36382161e-02 2.14731169e+00
8.85809600e-01 1.36980042e-01 -4.75698821e-02 -8.46274663e-03
5.89524329e-01 1.83321312e-01 -1.96480438e-01 -1.89649284e-01
-3.12290452e-02 4.41834033e-01 3.92817229e-01 5.42361975e-01
-1.28457367e+00 1.28169990e+00 5.34850979e+00 1.06700110e+00
-1.20370758e+00 3.13864797e-01 5.18415689e-01 -2.57692695e-01
-8.93431306e-02 -1.80067241e-01 -8.07849944e-01 3.93538535e-01
8.62096727e-01 5.55710085e-02 4.82323110e-01 6.67012870e-01
2.36950696e-01 -3.09441805e-01 -6.50288820e-01 7.36792088e-01
2.97675729e-01 -1.26823914e+00 7.88248405e-02 -5.07358182e-03
1.10075819e+00 2.69582141e-02 4.48508352e-01 2.93127477e-01
2.26370588e-01 -9.61155772e-01 6.33903503e-01 2.22181678e-01
4.45301861e-01 -8.10577631e-01 6.46163046e-01 2.37249956e-01
-1.18937290e+00 -1.39919341e-01 -3.08019221e-01 -3.80059928e-02
-5.30650765e-02 7.70851851e-01 -8.55867743e-01 7.09445834e-01
1.03145790e+00 4.27180231e-01 -8.08093488e-01 1.05353343e+00
-3.46743315e-01 7.11046040e-01 -2.48288363e-01 6.02630198e-01
2.12984264e-01 -1.30972177e-01 4.04158682e-01 1.39717746e+00
1.49356440e-01 2.87437201e-01 1.00502051e-01 8.09608579e-01
-6.10851124e-02 -2.46018395e-02 -4.61103827e-01 2.54294425e-01
2.44140774e-01 1.38739383e+00 -7.85992503e-01 -2.71828860e-01
-7.07816422e-01 1.00733411e+00 2.18885943e-01 4.96724814e-01
-1.00648355e+00 -3.48477781e-01 5.49270511e-01 4.09550101e-01
7.05862820e-01 -2.21487973e-02 -2.91489691e-01 -9.85605478e-01
-1.88488457e-02 -7.28167415e-01 2.09721133e-01 -7.54217207e-01
-1.07706690e+00 4.52156931e-01 2.02893704e-01 -8.45016837e-01
1.43709511e-01 -5.19682467e-01 -5.09140790e-01 7.46543050e-01
-2.23757696e+00 -1.33397222e+00 -8.12857449e-01 7.55775750e-01
6.90455198e-01 8.68697912e-02 4.25557971e-01 6.00698411e-01
-7.88969576e-01 6.48459971e-01 7.76585266e-02 -6.36069179e-02
5.65228879e-01 -1.10917282e+00 -6.86943904e-02 1.34645128e+00
1.52923062e-01 4.40789819e-01 8.93083572e-01 -4.04797107e-01
-1.23453522e+00 -1.40973413e+00 3.78640085e-01 -4.21574086e-01
4.30665851e-01 -2.45640606e-01 -1.21749926e+00 3.62371981e-01
4.27805245e-01 1.55646265e-01 3.51383269e-01 -2.33342960e-01
-4.43639010e-01 -2.26271316e-01 -1.17238462e+00 6.79734349e-01
8.29968214e-01 -2.91113973e-01 -4.57318753e-01 2.39448082e-02
7.35398710e-01 -3.42864215e-01 -7.77849793e-01 4.90103632e-01
3.24854225e-01 -1.17625010e+00 1.16697001e+00 -8.36236104e-02
5.30350804e-01 -4.78654176e-01 -5.39933257e-02 -1.15392303e+00
-1.87772393e-01 -6.08943164e-01 -2.12430641e-01 1.23569214e+00
1.74913257e-01 -2.93124288e-01 1.00544107e+00 3.44981849e-01
-2.37149283e-01 -7.39702880e-01 -7.89673269e-01 -5.61034024e-01
2.47142706e-02 -2.72430420e-01 2.43127853e-01 9.19036508e-01
-6.40794337e-01 3.57525408e-01 -3.07693332e-01 6.29416347e-01
8.98566365e-01 -8.62383004e-03 8.03981006e-01 -1.08874428e+00
-2.66662568e-01 -2.23635003e-01 -2.78416663e-01 -8.74069035e-01
-2.02039815e-02 -6.50356710e-01 1.13413066e-01 -1.50212860e+00
3.31181854e-01 -3.13850529e-02 -4.31375742e-01 3.45710218e-01
-4.87380415e-01 5.97765803e-01 2.88369626e-01 1.39833512e-02
-7.90874720e-01 5.32865226e-01 1.24504650e+00 -1.26687884e-01
-5.44021912e-02 -1.45837605e-01 -7.82336712e-01 8.30046356e-01
6.37891710e-01 -5.63562095e-01 -7.20451951e-01 -5.98433673e-01
6.87814653e-02 -3.29413235e-01 5.21366239e-01 -1.01827967e+00
3.75329733e-01 -2.22104833e-01 1.34073138e-01 -1.98560387e-01
5.10534704e-01 -9.16765630e-01 -1.85625568e-01 1.45767912e-01
4.39574495e-02 -3.08613867e-01 3.38699579e-01 6.19305193e-01
-2.47299299e-01 -1.61947951e-01 9.92405653e-01 -1.65022388e-01
-1.04259861e+00 1.04794905e-01 -8.38100016e-02 2.12819755e-01
1.14890242e+00 -4.43665832e-01 -4.88407791e-01 -2.66460359e-01
-4.84039307e-01 3.78363840e-02 4.38812315e-01 2.61045963e-01
8.23468626e-01 -6.60872519e-01 -6.57622814e-01 2.50033528e-01
5.91006503e-02 3.75780672e-01 3.50294352e-01 7.88102686e-01
-5.62074363e-01 1.75291672e-01 -1.42068014e-01 -5.19515693e-01
-1.46794057e+00 5.58454156e-01 3.44819456e-01 -6.06890582e-02
-7.82837927e-01 9.88890469e-01 5.85609317e-01 2.73721255e-02
4.25164372e-01 -8.83465931e-02 -2.27805883e-01 -1.32482186e-01
5.82082033e-01 6.54874623e-01 1.54986367e-01 -4.45378035e-01
-1.00018628e-01 5.73468864e-01 -1.55738860e-01 -3.38779949e-02
1.39026022e+00 -4.81291384e-01 -2.82316860e-02 2.91166380e-02
7.63907790e-01 -5.87466210e-02 -1.42492425e+00 -4.24722165e-01
-2.70338710e-02 -7.17785299e-01 3.89748603e-01 -7.16177285e-01
-1.35056531e+00 8.76385331e-01 9.10777450e-01 -7.06603471e-03
1.40033793e+00 -1.07467666e-01 6.54337168e-01 1.32389158e-01
2.93796182e-01 -9.16660011e-01 3.12004656e-01 2.35424891e-01
3.83413196e-01 -1.37634373e+00 2.20010087e-01 -7.68435597e-01
-5.90837181e-01 4.88910168e-01 8.77674222e-01 -1.52884766e-01
7.55509555e-01 1.46064878e-01 2.14775458e-01 -2.32766435e-01
-6.95210218e-01 -4.21614915e-01 2.63001293e-01 6.46334231e-01
2.57070661e-01 -2.45940700e-01 1.23664506e-01 2.74908632e-01
-3.23181413e-03 2.36225165e-02 6.93207502e-01 8.85929286e-01
-8.73475432e-01 -7.06014693e-01 -4.00841117e-01 2.46644616e-02
-6.85965300e-01 -1.42872363e-01 -9.33279842e-02 1.01619315e+00
5.16612947e-01 1.14042389e+00 -2.26146862e-01 -1.54716000e-01
2.59785205e-01 -1.95904255e-01 5.28050065e-01 -7.08043516e-01
-5.99960208e-01 1.23374768e-01 -2.38756508e-01 -5.28422773e-01
-6.05342746e-01 -2.58007079e-01 -1.05817139e+00 -1.87086523e-01
-4.41460758e-01 -2.27901265e-02 6.51286423e-01 9.32905793e-01
3.76201749e-01 7.76590049e-01 5.03762424e-01 -8.71354640e-01
-2.59331346e-01 -6.95645392e-01 -5.05999982e-01 3.86228919e-01
3.77600938e-01 -6.74002349e-01 -3.43708187e-01 3.35811436e-01] | [10.901259422302246, -2.9497783184051514] |
1b5f18f2-068f-49d6-932b-26c8760947ad | structural-models-for-policy-making-coping | 2103.01115 | null | https://arxiv.org/abs/2103.01115v4 | https://arxiv.org/pdf/2103.01115v4.pdf | Structural models for policy-making: Coping with parametric uncertainty | The ex-ante evaluation of policies using structural econometric models is based on estimated parameters as a stand-in for the true parameters. This practice ignores uncertainty in the counterfactual policy predictions of the model. We develop a generic approach that deals with parametric uncertainty using uncertainty sets and frames model-informed policy-making as a decision problem under uncertainty. The seminal human capital investment model by Keane and Wolpin (1997) provides a well-known, influential, and empirically-grounded test case. We document considerable uncertainty in the models's policy predictions and highlight the resulting policy recommendations obtained from using different formal rules of decision-making under uncertainty. | ['Christopher Walsh', 'Lena Janys', 'Janoś Gabler', 'Philipp Eisenhauer'] | 2021-03-01 | null | null | null | null | ['decision-making-under-uncertainty', 'decision-making-under-uncertainty'] | ['medical', 'reasoning'] | [-3.27566504e-01 8.25747490e-01 -9.47013140e-01 -2.74155617e-01
-3.34862143e-01 -6.73433363e-01 7.40774214e-01 -9.91702676e-02
-5.81026912e-01 1.22910309e+00 7.56354749e-01 -1.42852747e+00
-5.88384569e-01 -6.59625769e-01 -4.99989033e-01 -2.94430465e-01
2.92236209e-01 6.62832558e-01 -3.37219924e-01 2.84540713e-01
4.27958965e-01 2.14774907e-01 -9.30457652e-01 -7.46388435e-02
8.71760368e-01 9.30789232e-01 -2.39205323e-02 3.74986321e-01
3.01267412e-02 1.10094965e+00 -5.49305320e-01 -8.92114043e-01
5.47728717e-01 9.88823920e-02 -4.36868727e-01 -1.78730249e-01
-6.00687444e-01 -7.57548571e-01 -1.55844837e-01 9.10990715e-01
3.37894261e-01 -2.12139897e-02 9.00375545e-01 -1.08709216e+00
-7.33351588e-01 1.26909971e+00 -1.63310006e-01 1.59208670e-01
7.13238716e-02 4.44983512e-01 6.66324914e-01 -1.12951390e-01
4.17783260e-01 1.74170482e+00 4.60986853e-01 1.02655627e-01
-1.35313535e+00 -6.05403364e-01 3.20596009e-01 -1.94755346e-01
-1.01095676e+00 -4.59557742e-01 4.03734237e-01 -8.04895520e-01
8.98700893e-01 1.54373914e-01 7.40990639e-01 1.03593981e+00
6.01861417e-01 2.97478199e-01 1.55174887e+00 -6.94111109e-01
4.42929864e-01 2.45295674e-01 -1.84640765e-01 -2.72691429e-01
1.23564255e+00 1.05947292e+00 3.92417997e-01 -4.10569757e-01
9.23329592e-01 -9.05358866e-02 7.28177726e-02 -6.24146461e-02
-9.59760368e-01 9.08394277e-01 -1.70034558e-01 2.51480658e-02
-8.43122184e-01 5.05559981e-01 2.96437502e-01 5.13416767e-01
2.71121919e-01 6.95122719e-01 -7.42251158e-01 -2.43493259e-01
-6.08497381e-01 4.94822651e-01 9.16881919e-01 5.21679163e-01
1.44308925e-01 3.53178412e-01 -5.81994712e-01 4.96309884e-02
6.30906761e-01 7.89019644e-01 4.19469535e-01 -1.79017508e+00
3.04171294e-01 -2.85751879e-01 1.09028804e+00 -7.31203675e-01
-1.24053828e-01 -5.17689764e-01 -9.38132480e-02 4.42212373e-01
4.40247506e-01 -9.06435192e-01 -6.35634422e-01 1.53582716e+00
-1.02911554e-01 -1.42849967e-01 3.51894468e-01 2.89995462e-01
-1.80271253e-01 5.70554435e-01 3.61806363e-01 -8.00884128e-01
7.98497319e-01 -4.58402336e-01 -9.03353691e-01 7.88657553e-03
5.73403180e-01 -4.34454113e-01 2.40707234e-01 4.21355098e-01
-1.19791687e+00 -2.48426959e-01 -5.46784818e-01 7.83156514e-01
7.65984729e-02 -3.58091027e-01 9.50433195e-01 8.55834186e-01
-8.64490449e-01 7.73032367e-01 -4.67726320e-01 2.55439341e-01
2.18247041e-01 8.67578909e-02 3.99277359e-01 4.22779590e-01
-1.14497423e+00 1.47237289e+00 6.51318610e-01 1.13642789e-01
-8.11333060e-01 -5.82589805e-01 -4.40455258e-01 2.65181124e-01
8.99661779e-01 -8.52036655e-01 1.74749005e+00 -1.07615983e+00
-1.73893929e+00 1.48434326e-01 4.68902826e-01 -8.52216899e-01
8.86946321e-01 -1.32777527e-01 -5.70639908e-01 -3.10036808e-01
8.34188312e-02 3.95302474e-02 4.71872032e-01 -1.29687154e+00
-5.61709523e-01 -1.76129803e-01 2.64487147e-01 6.74194917e-02
6.26411676e-01 4.86900508e-01 7.59451926e-01 -8.34338248e-01
-3.76351893e-01 -8.54124963e-01 -7.37975419e-01 -7.83989966e-01
-3.39721441e-01 2.19797879e-01 -1.61048457e-01 -5.30703545e-01
1.58563530e+00 -1.53187883e+00 -6.50988281e-01 5.17597258e-01
-1.87808886e-01 -1.37522459e-01 4.36020017e-01 4.39716667e-01
-2.33631685e-01 8.06289673e-01 -1.09879933e-01 1.75091863e-01
5.94484389e-01 2.47915477e-01 -5.91555715e-01 2.51823097e-01
-1.35475576e-01 1.01238203e+00 -6.00498259e-01 -1.30427837e-01
3.61171633e-01 -1.87937126e-01 -2.63995558e-01 6.50354400e-02
-1.56358883e-01 1.47645190e-01 -5.82152903e-01 4.88080978e-01
3.96756709e-01 1.80283815e-01 6.07745767e-01 4.41908121e-01
-7.65420139e-01 4.92092133e-01 -1.15348566e+00 8.66354048e-01
-3.32794011e-01 3.96842390e-01 -1.14403404e-01 -1.14517355e+00
4.89186257e-01 5.12333810e-01 2.18062103e-01 -4.79303062e-01
2.89723843e-01 3.39186370e-01 4.12628412e-01 -6.05929911e-01
1.00262284e-01 -5.58582485e-01 -2.24978417e-01 7.85780132e-01
-3.28463018e-01 -4.17115062e-01 -1.68011803e-02 -2.49636248e-01
5.77684581e-01 3.36079329e-01 8.32542598e-01 -8.77185941e-01
-2.58814663e-01 -6.67256117e-02 9.48475122e-01 9.99675632e-01
-5.88991418e-02 1.35468856e-01 8.80110085e-01 -2.49139085e-01
-9.33824003e-01 -7.23960161e-01 -3.73632342e-01 5.00531197e-01
-8.58877659e-01 3.27861965e-01 -4.84246880e-01 -4.29372340e-01
6.56907380e-01 1.58844185e+00 -9.19629991e-01 2.50012428e-01
-3.22998241e-02 -7.39443064e-01 1.40064210e-01 4.56990629e-01
2.22431228e-01 -7.09250569e-01 -1.07371604e+00 4.12490487e-01
3.89320999e-01 -5.81336439e-01 -5.82335750e-03 -1.25598237e-01
-8.47224712e-01 -9.59272444e-01 -8.45632255e-01 4.51613635e-01
3.70444328e-01 -3.29640150e-01 9.92011309e-01 -3.40218008e-01
6.72302604e-01 3.97633940e-01 -1.87779851e-02 -1.31444740e+00
-6.15719080e-01 -7.28353798e-01 1.93705037e-01 -2.64199555e-01
2.08697945e-01 -3.17645252e-01 -4.96063590e-01 5.93095608e-02
-4.89614606e-01 -3.04749042e-01 4.85908628e-01 8.16418648e-01
2.31158063e-01 1.37056991e-01 9.38938439e-01 -9.70259905e-01
9.44818676e-01 -6.13917470e-01 -1.14988780e+00 6.84872270e-01
-1.27317560e+00 2.16872007e-01 -1.03807621e-01 -3.68415803e-01
-1.81050384e+00 -8.27885807e-01 2.87571698e-01 -8.34455937e-02
-1.99884459e-01 9.52662408e-01 -6.14660159e-02 2.84799159e-01
4.44418579e-01 -5.93413889e-01 -7.93207320e-04 -5.47327101e-01
1.32340267e-01 5.78000367e-01 4.47524339e-01 -1.01014745e+00
3.24168652e-01 1.99143782e-01 -9.60848182e-02 1.43273314e-03
-7.05736816e-01 2.84570783e-01 -1.76242024e-01 -1.01448856e-01
3.93303066e-01 -1.05633569e+00 -5.85473359e-01 -1.34503730e-02
-7.67353415e-01 -7.11051702e-01 -7.98417389e-01 1.14790463e+00
-1.07327843e+00 -6.19126596e-02 -7.06454515e-02 -1.65684068e+00
1.89670742e-01 -1.31905591e+00 1.95875958e-01 1.10319994e-01
-4.55389500e-01 -1.22886097e+00 2.83737004e-01 4.57955301e-02
6.21768534e-01 6.08816326e-01 8.47539008e-01 -7.34959602e-01
-3.01534712e-01 -3.63513641e-02 5.33118397e-02 1.25363886e-01
-2.12667942e-01 6.13352001e-01 -6.32306516e-01 -1.00288257e-01
2.75393814e-01 6.00571781e-02 5.85855186e-01 1.16075599e+00
7.60018170e-01 -9.09016252e-01 -2.33515888e-01 2.58285820e-01
1.40819597e+00 8.21430802e-01 6.31394684e-01 6.39339745e-01
-1.66313931e-01 1.05176985e+00 7.13636577e-01 9.08313155e-01
3.27740043e-01 4.28451836e-01 1.38905302e-01 4.45667028e-01
7.89981246e-01 -5.97842216e-01 1.43850759e-01 2.47109737e-02
-5.22481859e-01 -2.31143370e-01 -9.95036960e-01 5.03520846e-01
-2.09398556e+00 -1.35947001e+00 4.39505577e-01 2.46465302e+00
6.73307240e-01 6.52579546e-01 4.24024224e-01 -9.52248275e-02
4.61921871e-01 -1.94809273e-01 -3.67666513e-01 -8.54073167e-01
-5.91416992e-02 -2.16513142e-01 9.41223145e-01 7.19211519e-01
-4.51595813e-01 5.45028389e-01 8.47356510e+00 3.97262543e-01
-5.74133098e-01 -2.04806507e-01 1.26562178e+00 -1.27949968e-01
-1.02999139e+00 5.59484899e-01 -5.18592775e-01 6.94263756e-01
1.47454166e+00 -8.89754772e-01 -6.69733360e-02 5.87091148e-01
7.41446435e-01 -6.91118479e-01 -7.03432024e-01 2.81787127e-01
-7.99813211e-01 -1.44845772e+00 -2.76265174e-01 3.59957129e-01
1.14673567e+00 -3.74831975e-01 3.58274937e-01 4.23102349e-01
1.44952857e+00 -1.04452217e+00 1.36826837e+00 1.01380086e+00
7.42777824e-01 -1.10916114e+00 1.09474027e+00 4.60173398e-01
-1.21479861e-01 -8.68226230e-01 -3.87092412e-01 -7.43213296e-01
3.42552483e-01 7.62589455e-01 -6.28041685e-01 2.78136969e-01
3.58424842e-01 8.06819871e-02 6.49699420e-02 8.55118394e-01
-3.39598954e-01 9.71441567e-01 -1.02724984e-01 2.98775196e-01
3.39670300e-01 -4.95672226e-01 1.50508896e-01 1.01397085e+00
4.42631066e-01 5.99171937e-01 -1.81210116e-01 9.62489128e-01
2.90959954e-01 -3.65320683e-01 -8.16696644e-01 -4.23089117e-01
7.28727520e-01 2.04077512e-01 -4.71816570e-01 -3.61276060e-01
-4.81766492e-01 -2.95262009e-01 -2.60506183e-01 7.93608308e-01
-2.92082191e-01 1.19872853e-01 6.53344452e-01 8.19309521e-03
-5.98107800e-02 -7.72021487e-02 -6.03335083e-01 -1.01253998e+00
-3.29481184e-01 -8.02495420e-01 3.23886991e-01 -6.06109262e-01
-8.03658366e-01 -3.95335257e-01 8.35213840e-01 -5.41949272e-01
-8.94955635e-01 -5.84731698e-01 -8.42096925e-01 1.12944591e+00
-1.37644827e+00 -4.04969037e-01 8.84316862e-01 -1.86634377e-01
1.10292114e-01 6.09783679e-02 6.07211471e-01 -7.72311211e-01
-7.06043601e-01 2.70170838e-01 7.86333740e-01 -2.58431911e-01
4.80630755e-01 -1.32771111e+00 3.77241254e-01 7.57725596e-01
-5.39310217e-01 6.85189247e-01 9.27980721e-01 -9.29014862e-01
-6.83085561e-01 -4.83258665e-01 6.90904915e-01 -4.89496529e-01
8.31669152e-01 3.36867601e-01 -5.25271237e-01 1.14338946e+00
4.73571301e-01 -7.03461647e-01 6.37248635e-01 6.34172112e-02
1.80038914e-01 9.25557315e-02 -1.36456931e+00 4.79803383e-01
5.85027933e-01 -2.71553814e-01 -9.19666409e-01 -1.33476377e-01
9.78574574e-01 9.28468108e-02 -1.00485003e+00 4.23262894e-01
9.44548547e-01 -1.04620683e+00 8.19588006e-01 -9.57339287e-01
-1.51506484e-01 3.56792152e-01 -3.29405576e-01 -1.45882761e+00
-5.92236400e-01 -1.24239922e+00 1.37708619e-01 1.07313991e+00
5.86563230e-01 -1.11560833e+00 1.25476748e-01 1.27719581e+00
9.13543031e-02 -5.63220620e-01 -8.49243045e-01 -7.51467705e-01
5.59996605e-01 -5.04820287e-01 1.25560641e+00 1.11213756e+00
1.37979507e-01 -6.03136420e-01 -1.46861434e-01 -2.74219871e-01
7.72383273e-01 1.24580704e-01 5.33146858e-01 -1.35354078e+00
-4.07726347e-01 -6.45954370e-01 1.33595973e-01 -4.15149570e-01
3.98935258e-01 -9.23582539e-02 -3.28376532e-01 -1.32388580e+00
-1.66003015e-02 -3.68745327e-01 -4.93292481e-01 2.53287405e-02
-1.76622067e-02 -8.80414784e-01 4.80930448e-01 -9.70153196e-04
1.41477473e-02 2.88054913e-01 9.75636125e-01 1.74992278e-01
-1.84370935e-01 4.80954558e-01 -1.33790493e+00 9.97134209e-01
9.52716231e-01 -3.86801511e-01 -5.14149427e-01 -2.63135344e-01
5.97484231e-01 7.35776722e-01 2.99221337e-01 -3.34352076e-01
-1.08096577e-01 -1.36648989e+00 2.19391003e-01 -3.24686915e-01
-3.67552698e-01 -7.12069869e-01 9.28469360e-01 6.53043091e-01
-6.41420126e-01 2.94175178e-01 3.28095824e-01 4.67731178e-01
1.61463395e-01 -4.87776548e-01 3.67369831e-01 -4.85225320e-01
-3.51085246e-01 -7.74391219e-02 -5.64641654e-01 1.02178531e-03
1.04524720e+00 -1.38151005e-01 -4.25641537e-01 -6.57632411e-01
-8.12941134e-01 2.72934407e-01 6.07628822e-01 -1.52361810e-01
5.21269776e-02 -1.14627469e+00 -9.75873768e-01 -3.36308241e-01
-4.21690911e-01 -3.53949279e-01 -6.22414835e-02 5.31284153e-01
-2.31861487e-01 1.23908257e+00 -2.58547753e-01 2.93313175e-01
-2.06768528e-01 6.61132097e-01 7.06829071e-01 -3.48382264e-01
3.76207419e-02 4.43795174e-01 1.85008049e-01 1.62508681e-01
-1.27118483e-01 -4.08792526e-01 -1.59632295e-01 1.41384795e-01
5.05742431e-01 9.05053675e-01 -5.42701125e-01 -3.61965805e-01
8.68806094e-02 1.61283314e-02 4.46503192e-01 -1.00349867e+00
1.14534378e+00 -5.63057780e-01 7.78834894e-02 7.97214627e-01
3.14999849e-01 -1.30838320e-01 -1.57723486e+00 -2.74955064e-01
3.74630094e-01 -6.66610062e-01 3.85939926e-01 -1.10823107e+00
-5.61042905e-01 4.91107941e-01 1.22273386e-01 9.48910117e-02
8.18916023e-01 -5.16728401e-01 -3.47128093e-01 4.86516893e-01
5.23204863e-01 -1.47656846e+00 -7.90181518e-01 -2.18208492e-01
1.22208059e+00 -1.12889826e+00 3.56368095e-01 2.94908702e-01
-8.03749919e-01 7.06272006e-01 5.73572107e-02 1.58287555e-01
1.01879478e+00 1.13578953e-01 -8.53688791e-02 3.43189806e-01
-1.19769478e+00 -7.00863749e-02 4.65982892e-02 7.30035424e-01
1.63321599e-01 8.80627394e-01 -8.52563381e-01 1.03178191e+00
-2.46312663e-01 2.80462772e-01 8.01718950e-01 9.47027981e-01
-4.56807584e-01 -8.59674394e-01 -8.12125087e-01 7.11050749e-01
-1.03678310e+00 1.47694461e-02 -1.82947278e-01 1.05141640e+00
-8.56627971e-02 1.04827631e+00 1.92535013e-01 9.36596617e-02
3.45975339e-01 1.22762896e-01 3.51345874e-02 -3.97840738e-01
-3.75321120e-01 4.02981550e-01 5.62875390e-01 -3.78462166e-01
-3.52117419e-01 -1.08698821e+00 -8.49830151e-01 -5.04770339e-01
-2.71031469e-01 3.60878855e-01 5.46747267e-01 1.02256060e+00
3.02184701e-01 2.99822092e-01 7.79270351e-01 -5.26679575e-01
-1.52249908e+00 -9.75748956e-01 -8.09022605e-01 -2.43208334e-01
2.09126323e-01 -9.31178391e-01 -6.08792067e-01 -5.14154017e-01] | [7.906651973724365, 5.1545939445495605] |
f53ab91a-7f55-41de-992f-45391d757b9f | pups-point-cloud-unified-panoptic | 2302.06185 | null | https://arxiv.org/abs/2302.06185v2 | https://arxiv.org/pdf/2302.06185v2.pdf | PUPS: Point Cloud Unified Panoptic Segmentation | Point cloud panoptic segmentation is a challenging task that seeks a holistic solution for both semantic and instance segmentation to predict groupings of coherent points. Previous approaches treat semantic and instance segmentation as surrogate tasks, and they either use clustering methods or bounding boxes to gather instance groupings with costly computation and hand-crafted designs in the instance segmentation task. In this paper, we propose a simple but effective point cloud unified panoptic segmentation (PUPS) framework, which use a set of point-level classifiers to directly predict semantic and instance groupings in an end-to-end manner. To realize PUPS, we introduce bipartite matching to our training pipeline so that our classifiers are able to exclusively predict groupings of instances, getting rid of hand-crafted designs, e.g. anchors and Non-Maximum Suppression (NMS). In order to achieve better grouping results, we utilize a transformer decoder to iteratively refine the point classifiers and develop a context-aware CutMix augmentation to overcome the class imbalance problem. As a result, PUPS achieves 1st place on the leader board of SemanticKITTI panoptic segmentation task and state-of-the-art results on nuScenes. | ['Xi Li', 'Dayang Hao', 'Xin Zhan', 'Zhenwei Miao', 'Huanyu Wang', 'Jianyun Xu', 'Shihao Su'] | 2023-02-13 | null | null | null | null | ['panoptic-segmentation'] | ['computer-vision'] | [ 3.05278122e-01 -3.75372507e-02 -4.73022491e-01 -5.29234231e-01
-9.65085626e-01 -6.83926880e-01 3.74768078e-01 3.27997245e-02
1.30409464e-01 2.79117227e-02 -2.28181675e-01 -2.79129475e-01
-2.20468324e-02 -1.00209641e+00 -8.31221759e-01 -6.18206024e-01
1.62990421e-01 8.25743675e-01 3.39616209e-01 -1.96330383e-01
4.01284009e-01 4.66133773e-01 -1.63647091e+00 5.22247493e-01
1.19163513e+00 1.06756568e+00 4.22902644e-01 3.47278565e-01
-4.53669012e-01 3.03704217e-02 -3.98293138e-01 -4.20138210e-01
8.13199043e-01 -2.06447378e-01 -6.34997427e-01 3.82976651e-01
7.54334807e-01 6.99373567e-03 1.58263028e-01 1.11407340e+00
9.97457579e-02 3.36025353e-03 7.50405610e-01 -1.14322734e+00
-1.62579507e-01 5.35080254e-01 -1.32997894e+00 -2.40146771e-01
-3.09674263e-01 2.90836841e-01 1.58336484e+00 -8.59259546e-01
3.83843780e-01 1.19208372e+00 7.32619882e-01 9.87739414e-02
-1.19746780e+00 -1.00233972e+00 4.98753935e-01 -6.95898086e-02
-1.41355383e+00 -1.79769933e-01 9.92666781e-01 -3.55025917e-01
5.83388269e-01 5.50181806e-01 8.70618463e-01 4.50929254e-01
-2.79376954e-01 9.66980577e-01 8.22063982e-01 -2.06774399e-01
2.40593120e-01 -1.58424571e-01 2.79668123e-01 5.35956979e-01
7.24310204e-02 -1.08401805e-01 -2.88519353e-01 3.71172465e-02
6.38282776e-01 2.01651797e-01 -1.31898642e-01 -3.80538315e-01
-1.11011529e+00 7.93005109e-01 1.09147465e+00 6.15806952e-02
-3.70655060e-01 9.54112709e-02 3.80282216e-02 -1.70677215e-01
7.86648870e-01 6.09778404e-01 -6.30673766e-01 4.76663947e-01
-1.62264991e+00 3.71334642e-01 5.02144694e-01 1.27884114e+00
1.21279609e+00 -4.50011432e-01 -3.10790241e-01 1.11021912e+00
3.74748170e-01 3.77555728e-01 -1.24237299e-01 -7.53575265e-01
8.21166277e-01 1.13506377e+00 -1.72337338e-01 -9.46387589e-01
-4.17945176e-01 -8.48952115e-01 -5.56047320e-01 7.15417564e-02
1.07596286e-01 9.44351256e-02 -1.53260064e+00 1.32097244e+00
5.78422785e-01 3.72313797e-01 -2.43511036e-01 1.20807910e+00
5.90511203e-01 9.66518164e-01 8.66409205e-03 1.64855063e-01
1.57565916e+00 -1.33303046e+00 -1.82487234e-01 -4.51678455e-01
4.69001770e-01 -6.88214600e-01 9.53447461e-01 3.68441403e-01
-7.31144607e-01 -6.37800753e-01 -1.24912751e+00 -2.39207447e-01
-2.53286213e-01 3.05271775e-01 6.05590045e-01 5.24518967e-01
-6.54016614e-01 6.45231783e-01 -8.82764697e-01 2.02418882e-02
7.42351651e-01 4.82192308e-01 2.04012081e-01 -4.31422815e-02
-7.01050878e-01 3.54470879e-01 4.47954744e-01 1.58179820e-01
-5.14601588e-01 -1.02065098e+00 -6.67146444e-01 4.37589809e-02
6.97821200e-01 -7.79646158e-01 9.72199500e-01 -9.87545252e-01
-1.24512351e+00 9.23559546e-01 -6.04259297e-02 -3.54265034e-01
1.58264413e-01 -7.85563812e-02 -5.06901816e-02 2.89632350e-01
2.49240205e-01 1.40121400e+00 9.40491498e-01 -1.43967974e+00
-1.14291227e+00 -6.85258985e-01 -1.17933750e-01 3.70772928e-01
-9.82879195e-03 -2.16409981e-01 -9.12663817e-01 -6.98037148e-01
8.76395404e-01 -1.05542779e+00 -5.02465725e-01 -1.15884624e-01
-8.12563837e-01 -2.07654089e-01 8.04651022e-01 -6.92499757e-01
9.56286609e-01 -2.07371283e+00 2.13944331e-01 4.64341104e-01
2.86572129e-01 4.98823309e-03 -7.31611252e-02 1.92155000e-02
4.24842909e-02 1.99827835e-01 -7.58372605e-01 -6.29474521e-01
2.14378573e-02 1.48908332e-01 -4.22519505e-01 3.13629150e-01
4.72384959e-01 7.93133140e-01 -6.56904995e-01 -5.17378867e-01
3.95050675e-01 8.50967690e-02 -8.60694230e-01 1.60911113e-01
-8.32489669e-01 5.99148512e-01 -5.87995410e-01 1.03235090e+00
1.17708409e+00 -1.62242755e-01 -1.16200019e-02 -3.60255182e-01
-4.11987394e-01 4.20571506e-01 -1.09877431e+00 2.05716562e+00
-2.60975242e-01 2.95546442e-01 1.07890971e-01 -9.20729220e-01
1.06427276e+00 -1.58808619e-01 5.86831152e-01 -4.95767117e-01
9.84361842e-02 4.35363740e-01 -2.90892959e-01 -1.80071816e-01
6.96934640e-01 -1.53361708e-01 -1.69385567e-01 -1.28151014e-01
-8.11965838e-02 -5.88284612e-01 5.29321879e-02 -6.45061582e-02
7.57264435e-01 2.64731377e-01 -2.06401691e-01 -3.64015400e-01
2.99209654e-01 6.75484598e-01 6.54366910e-01 4.45937008e-01
3.42936784e-01 1.26472163e+00 4.49749500e-01 -1.53387845e-01
-1.06526494e+00 -7.35583842e-01 -2.11020917e-01 9.11849320e-01
5.53314865e-01 -3.98810893e-01 -7.93276787e-01 -6.69759572e-01
-1.62490681e-02 8.07167232e-01 -2.68516034e-01 1.28076032e-01
-6.13819659e-01 -9.43384886e-01 1.87652782e-01 4.72232372e-01
5.01964271e-01 -6.37840450e-01 -4.55465883e-01 1.96228206e-01
-1.75982744e-01 -1.07819557e+00 -5.18473744e-01 3.43879789e-01
-8.85078192e-01 -1.01839232e+00 -7.90345371e-01 -7.69349933e-01
7.34702289e-01 5.11380136e-01 9.92542922e-01 7.59168640e-02
-1.99478850e-01 -3.71421784e-01 -3.65555584e-01 -2.23718345e-01
5.83651662e-02 4.78517622e-01 -5.02033949e-01 1.25295803e-01
2.99503386e-01 -6.06213629e-01 -9.18438792e-01 6.30269945e-01
-6.70326889e-01 5.48280597e-01 7.43490994e-01 4.61893141e-01
9.80894327e-01 -6.99438062e-03 2.47173104e-03 -7.84656763e-01
-1.24411590e-01 -3.52654070e-01 -9.75877047e-01 2.14124143e-01
-3.05709988e-01 -6.59785196e-02 4.94479179e-01 -8.30329359e-02
-7.86242604e-01 3.20036024e-01 -2.92159140e-01 -7.65427351e-01
-1.14044145e-01 3.26991707e-01 -4.75288779e-01 -6.34129345e-02
4.07117546e-01 2.08798964e-02 -4.22259003e-01 -7.71517694e-01
6.87582552e-01 8.87891412e-01 5.73384702e-01 -6.70447946e-01
7.37361848e-01 5.33164978e-01 1.49316853e-02 -6.92095697e-01
-9.20476854e-01 -9.33230340e-01 -7.56145656e-01 -6.37951642e-02
9.56954658e-01 -1.14743066e+00 -2.81084746e-01 2.78671324e-01
-9.90143180e-01 -2.49951810e-01 -1.84121162e-01 4.54482064e-02
-2.83214629e-01 2.46920526e-01 -3.19706351e-01 -5.92020035e-01
-4.14761424e-01 -1.31375051e+00 1.69089234e+00 2.30279416e-01
6.55654445e-02 -3.10100496e-01 -4.24622186e-02 7.28085279e-01
-1.73173085e-01 3.14273566e-01 8.28791320e-01 -4.98098046e-01
-1.09594071e+00 -5.40068373e-02 -6.10061228e-01 2.76587933e-01
-2.02205405e-01 1.86102554e-01 -8.58518183e-01 -7.25374371e-02
4.39958908e-02 2.16132179e-01 1.09566081e+00 3.83396357e-01
1.33269775e+00 -2.63947845e-01 -7.62646556e-01 1.11891246e+00
1.52131474e+00 1.56283081e-01 5.47095478e-01 2.25618184e-02
1.12936413e+00 9.80104566e-01 7.66301274e-01 1.47660509e-01
4.74066734e-01 8.78680944e-01 5.32257318e-01 -2.86522210e-01
-2.66503841e-01 -4.14949089e-01 -9.60331857e-02 4.68162388e-01
1.72656924e-01 -3.64728481e-01 -1.12340653e+00 7.27455974e-01
-1.79313004e+00 -3.71508271e-01 -7.50383973e-01 1.91827750e+00
6.67714834e-01 1.78902298e-01 4.66366252e-03 1.15312906e-02
8.99828315e-01 2.98395604e-01 -4.43874747e-01 9.92001817e-02
1.08546749e-01 3.01897734e-01 9.10150468e-01 1.84655920e-01
-1.24940276e+00 1.22159064e+00 4.92562580e+00 1.29527211e+00
-1.11127543e+00 1.09268807e-01 8.19448888e-01 2.27186866e-02
-3.18755656e-01 4.30798948e-01 -1.02919734e+00 6.34094059e-01
4.70042557e-01 6.63682163e-01 2.10035384e-01 6.93033874e-01
3.05763520e-02 -6.47831932e-02 -9.58559692e-01 9.86021161e-01
-2.80742705e-01 -1.41464686e+00 1.01914689e-01 1.55391082e-01
8.98924470e-01 2.60902941e-01 -9.58716646e-02 8.79211351e-02
2.52453443e-02 -8.32008600e-01 9.94219840e-01 1.40492991e-01
5.60236573e-01 -7.00292945e-01 2.36249343e-01 4.13310945e-01
-1.27432001e+00 -4.41814438e-02 -3.68132055e-01 2.10083723e-01
1.67052329e-01 9.47042406e-01 -6.71666205e-01 9.12696183e-01
4.83860075e-01 6.40738666e-01 -4.23140794e-01 1.31824422e+00
-2.99029291e-01 7.34370053e-01 -8.94665897e-01 3.29039693e-01
7.33187318e-01 -5.57786286e-01 6.11881375e-01 1.06689143e+00
3.80334973e-01 2.87976146e-01 4.46757823e-01 1.20398068e+00
-1.47049502e-01 9.67115611e-02 1.15296243e-04 2.53317177e-01
2.69561499e-01 1.42541015e+00 -1.27122140e+00 -2.01174498e-01
-7.18993694e-02 7.04246163e-01 7.28634968e-02 1.03797516e-04
-9.57736135e-01 -2.33777240e-01 6.11611426e-01 3.02507013e-01
5.89692771e-01 -2.37385258e-01 -1.08217514e+00 -1.07954454e+00
-2.21703164e-02 -5.97776771e-01 2.48989865e-01 -7.34938920e-01
-1.00657630e+00 3.08543146e-01 1.62702892e-02 -1.47714365e+00
2.52251953e-01 -4.21520352e-01 -6.82339311e-01 7.45509982e-01
-1.48300767e+00 -1.45625508e+00 -4.15822804e-01 2.00620741e-01
7.69019902e-01 3.66331160e-01 9.19905379e-02 4.41239506e-01
-6.52593970e-01 2.51277775e-01 -1.01212502e-01 -1.43548110e-02
3.54434520e-01 -1.14795029e+00 6.09256506e-01 9.13816392e-01
3.68053436e-01 2.46089056e-01 6.74336255e-01 -7.59012461e-01
-1.09002340e+00 -1.27624166e+00 3.20519388e-01 -3.25345069e-01
3.18764806e-01 -6.03646398e-01 -8.66892099e-01 3.69533390e-01
-1.52656898e-01 -3.32052201e-01 3.07805121e-01 2.03910992e-01
-1.96681559e-01 -2.47565582e-01 -8.97876441e-01 7.34312832e-01
1.24778914e+00 -1.63542271e-01 -5.75255573e-01 5.28246999e-01
9.97503519e-01 -5.60992956e-01 -4.42499727e-01 7.83940852e-01
1.89293325e-01 -7.77358890e-01 1.00357771e+00 3.35573927e-02
6.31612778e-01 -8.08860064e-01 -5.49469329e-02 -1.13635015e+00
-1.26912758e-01 -4.76099133e-01 2.96113938e-01 1.26803184e+00
6.16027057e-01 -2.81395972e-01 1.05394697e+00 3.62644732e-01
-7.80876815e-01 -8.72246325e-01 -8.38319242e-01 -4.98386115e-01
-7.90028274e-02 -6.23124719e-01 9.77167189e-01 9.11433220e-01
-3.68842781e-01 3.94880414e-01 -7.45059550e-02 4.94276375e-01
6.50548398e-01 7.52807915e-01 7.97875822e-01 -1.18455160e+00
-5.11966310e-02 -6.16909921e-01 -1.02577366e-01 -1.54615283e+00
-1.29618675e-01 -1.06682599e+00 2.81065166e-01 -1.50573909e+00
-1.23430751e-01 -1.04198074e+00 1.66364256e-02 5.83851218e-01
-2.26478234e-01 4.59583282e-01 2.29957014e-01 5.08202851e-01
-4.68643099e-01 5.33467948e-01 1.28445017e+00 -4.01878297e-01
-5.57653666e-01 2.72897780e-01 -8.55086207e-01 5.93788147e-01
6.95621133e-01 -6.64001584e-01 -3.18679154e-01 -5.51461041e-01
8.67823437e-02 -1.09151937e-01 4.57950920e-01 -1.20491374e+00
3.30856383e-01 -9.01560783e-02 2.04312012e-01 -1.35878396e+00
3.95960748e-01 -9.38852668e-01 3.24862123e-01 8.24915916e-02
8.73296056e-03 -5.10178149e-01 1.08785845e-01 4.25851524e-01
-1.20271049e-01 -2.94629037e-01 7.09964395e-01 -1.09690987e-01
-6.55834198e-01 4.72309023e-01 2.94626057e-01 -2.69112766e-01
1.03536868e+00 -5.40957570e-01 -2.09518224e-01 3.32423329e-01
-4.20810312e-01 8.55344415e-01 5.87843776e-01 3.52835834e-01
3.16714764e-01 -8.37448776e-01 -6.32280588e-01 2.51827210e-01
1.72646716e-01 9.32040274e-01 3.00267339e-01 9.29819465e-01
-7.64103055e-01 2.49811113e-01 7.51476437e-02 -1.08799005e+00
-8.59502792e-01 4.39749509e-01 1.49730340e-01 -2.19223440e-01
-7.18580246e-01 9.57933545e-01 4.66890156e-01 -5.68666697e-01
-1.60688892e-01 -6.98236644e-01 3.68241332e-02 4.16171432e-01
1.76021154e-03 6.95876777e-02 2.31577590e-01 -4.04453337e-01
-2.12085918e-01 8.34825516e-01 -4.37049195e-02 4.31154185e-04
1.29476345e+00 6.44534379e-02 -2.27844700e-01 5.59365936e-02
1.23766363e+00 -1.86541751e-01 -1.57272065e+00 -2.99112010e-03
2.54794266e-02 -4.59781170e-01 1.85115814e-01 -7.74932504e-01
-1.36400592e+00 9.82749283e-01 3.16276878e-01 7.96630085e-02
1.13220704e+00 1.67494714e-01 1.11903739e+00 -1.35434344e-01
3.21280539e-01 -1.11095893e+00 -3.99880946e-01 2.06545189e-01
7.15413511e-01 -9.23733473e-01 1.05448931e-01 -1.10760391e+00
-5.37338018e-01 1.00421858e+00 6.28736436e-01 -2.37137094e-01
5.17122209e-01 -2.32341187e-03 -2.34139770e-01 -4.19086248e-01
-4.36747104e-01 -4.84108776e-01 6.24864042e-01 1.93571672e-01
-1.75241217e-01 3.14357102e-01 -1.24606259e-01 6.71028376e-01
-3.86086792e-01 -3.26643765e-01 -2.19449997e-01 5.78178227e-01
-6.76608205e-01 -9.29588854e-01 -6.61059439e-01 6.38517737e-01
-1.34381637e-01 -1.89539433e-01 -3.00006479e-01 5.30141652e-01
5.48307538e-01 7.40057111e-01 4.79808003e-01 -5.82986593e-01
4.80444402e-01 -1.14065953e-01 1.64616644e-01 -7.56867111e-01
-8.15099299e-01 4.46163177e-01 -2.16353778e-02 -3.35940748e-01
-3.80406380e-01 -4.80363101e-01 -1.33725727e+00 -1.36133833e-02
-6.19009912e-01 5.19974530e-03 7.93600678e-01 8.70285571e-01
3.92202675e-01 5.04212976e-01 7.36093938e-01 -1.18585527e+00
-1.95525214e-01 -8.39901865e-01 -3.76877159e-01 5.53974770e-02
7.55727589e-02 -4.74540502e-01 -1.99478522e-01 -1.42102882e-01] | [7.937295913696289, -3.1003193855285645] |
5f5e19a8-45b5-418e-a1a1-c21c7af185a7 | the-impact-of-spelling-correction-and-task | null | null | https://aclanthology.org/W19-6310 | https://aclanthology.org/W19-6310.pdf | The Impact of Spelling Correction and Task Context on Short Answer Assessment for Intelligent Tutoring Systems | null | ['Bj{\\"o}rn Rudzewitz', 'Florian Nuxoll', 'Detmar Meurers', 'Ramon Ziai', 'Kordula De Kuthy'] | 2019-09-01 | null | null | null | ws-2019-9 | ['spelling-correction'] | ['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] | [-7.441454887390137, 3.667285442352295] |
b72477d7-25e6-47b5-8b09-34910df6818d | multi-frame-joint-enhancement-for-early | 2109.14151 | null | https://arxiv.org/abs/2109.14151v1 | https://arxiv.org/pdf/2109.14151v1.pdf | Multi-frame Joint Enhancement for Early Interlaced Videos | Early interlaced videos usually contain multiple and interlacing and complex compression artifacts, which significantly reduce the visual quality. Although the high-definition reconstruction technology for early videos has made great progress in recent years, related research on deinterlacing is still lacking. Traditional methods mainly focus on simple interlacing mechanism, and cannot deal with the complex artifacts in real-world early videos. Recent interlaced video reconstruction deep deinterlacing models only focus on single frame, while neglecting important temporal information. Therefore, this paper proposes a multiframe deinterlacing network joint enhancement network for early interlaced videos that consists of three modules, i.e., spatial vertical interpolation module, temporal alignment and fusion module, and final refinement module. The proposed method can effectively remove the complex artifacts in early videos by using temporal redundancy of multi-fields. Experimental results demonstrate that the proposed method can recover high quality results for both synthetic dataset and real-world early interlaced videos. | ['Xiaoping Liu', 'Ronggang Wang', 'Wei Jia', 'Yuan Chen', 'Yanbo Ma', 'Yang Zhao'] | 2021-09-29 | null | null | null | null | ['video-reconstruction'] | ['computer-vision'] | [ 9.56175253e-02 -7.53113449e-01 -7.85788987e-03 9.68539622e-03
-1.12409338e-01 -1.44805387e-02 3.18209976e-01 -2.40645871e-01
-5.12135029e-01 5.74384332e-01 3.00906926e-01 1.08913206e-01
-1.51912078e-01 -4.09287333e-01 -6.67395294e-01 -7.22809136e-01
-3.37994881e-02 -5.69524109e-01 6.71207786e-01 -8.48911852e-02
1.48023829e-01 1.79056272e-01 -1.32254255e+00 4.89060611e-01
9.52475309e-01 9.19919312e-01 4.47368830e-01 6.38999701e-01
1.44533485e-01 9.95579064e-01 -5.08524597e-01 -6.86735213e-02
2.20702305e-01 -4.68913704e-01 -2.74986714e-01 2.82534957e-01
7.30850035e-03 -8.33927274e-01 -7.56411552e-01 1.07327819e+00
5.31039178e-01 1.14286728e-01 2.60887653e-01 -1.26629770e+00
-4.72663850e-01 3.82265091e-01 -1.11485553e+00 5.40914834e-01
3.79407465e-01 -7.18878955e-02 4.41086411e-01 -1.21804404e+00
6.64219737e-01 1.08384430e+00 9.32798803e-01 8.69452953e-02
-1.03410637e+00 -9.95137513e-01 -1.04835719e-01 6.72334433e-01
-1.52342880e+00 -5.78365624e-01 8.58315706e-01 -3.00818533e-01
3.15800160e-01 3.43529731e-02 7.16502011e-01 6.56188607e-01
4.07000035e-01 8.35983276e-01 9.94215608e-01 -1.40162736e-01
-3.36148471e-01 -4.58863676e-01 -1.14350833e-01 2.84401953e-01
2.82248497e-01 5.37694335e-01 -5.53737938e-01 2.86462516e-01
1.12923431e+00 4.55005318e-01 -7.96853721e-01 -7.61097893e-02
-1.45770800e+00 1.94829911e-01 1.96421698e-01 4.86229956e-01
-3.46508652e-01 -5.02117127e-02 5.36375046e-01 3.37101728e-01
4.84312683e-01 -1.60774037e-01 -1.57697111e-01 -1.60970733e-01
-1.39865732e+00 1.77717060e-01 1.15749702e-01 1.18550122e+00
6.00426495e-01 1.06649876e-01 -7.20814839e-02 8.48616362e-01
3.84596348e-01 1.49396479e-01 2.81536937e-01 -9.94677484e-01
5.90310454e-01 4.83073182e-02 3.59874219e-01 -1.34917939e+00
-3.33650440e-01 -5.46920538e-01 -1.39887786e+00 1.73673660e-01
1.18823618e-01 -3.39498490e-01 -5.74003100e-01 1.20565164e+00
2.88577557e-01 5.82188070e-01 -6.27978519e-02 1.16143668e+00
9.97286975e-01 1.22077155e+00 -1.07149117e-01 -8.87837350e-01
9.72226858e-01 -1.02362335e+00 -1.29601419e+00 2.88509130e-01
1.45931855e-01 -1.35041833e+00 4.34176028e-01 6.08030915e-01
-1.57091451e+00 -9.71637070e-01 -1.28040433e+00 -1.10926948e-01
4.60005969e-01 1.87627256e-01 2.76896715e-01 -1.29805699e-01
-7.82307267e-01 5.95555902e-01 -6.25526905e-01 2.12537199e-01
2.00342745e-01 4.78645712e-02 -5.03386855e-01 -5.96410334e-01
-1.23404384e+00 6.31400883e-01 6.22775495e-01 5.02069831e-01
-5.94360173e-01 -8.96796405e-01 -7.33950496e-01 -1.25813723e-01
4.98846292e-01 -3.57313216e-01 1.04143667e+00 -1.13169825e+00
-1.12080646e+00 1.70419425e-01 9.10596363e-03 -2.33033016e-01
4.15633857e-01 -2.76257873e-01 -9.03060555e-01 2.79832602e-01
-6.27059713e-02 3.09984356e-01 1.00931501e+00 -1.26439357e+00
-8.69100869e-01 6.91147447e-02 -5.19624949e-01 3.33157122e-01
-1.87734455e-01 2.89193898e-01 -7.37554371e-01 -1.26013899e+00
2.44145408e-01 -4.37180996e-01 -5.80222672e-03 4.36239243e-01
5.43996468e-02 3.28222394e-01 1.32528734e+00 -1.11914301e+00
1.68209803e+00 -2.38893175e+00 1.76196307e-01 -2.32958525e-01
3.79479736e-01 4.48616534e-01 -4.70856987e-02 2.36523181e-01
-3.60262603e-01 -4.06821519e-01 -1.84181064e-01 -1.73329666e-01
-5.83887458e-01 1.44553810e-01 -2.07260251e-01 6.76162660e-01
-1.67719275e-01 4.77311462e-01 -1.03087056e+00 -7.92537868e-01
4.86753404e-01 6.82356954e-01 -4.46423799e-01 3.91504079e-01
2.95057714e-01 4.36375707e-01 -4.75380644e-02 5.00608861e-01
1.34616423e+00 -2.42442871e-03 -2.62173086e-01 -7.83046186e-01
-5.45932114e-01 -2.81557173e-01 -1.44321239e+00 1.93549252e+00
-1.62718460e-01 6.90685511e-01 6.05231464e-01 -7.54187465e-01
6.48279130e-01 4.60825533e-01 8.77559960e-01 -7.71201849e-01
1.78797200e-01 2.93265373e-01 -7.37889409e-02 -7.61947274e-01
7.12707520e-01 -2.04816520e-01 5.21677256e-01 1.10562310e-01
-1.45692348e-01 3.61621261e-01 3.46671611e-01 2.84753680e-01
7.44314075e-01 4.69652772e-01 3.11562978e-03 2.63997111e-02
6.52033567e-01 -3.25141788e-01 1.08247387e+00 9.55891907e-02
-1.70609310e-01 1.08341467e+00 -3.08709349e-02 -2.93872535e-01
-1.29845870e+00 -1.08086085e+00 -7.45005086e-02 4.82442230e-01
9.41426158e-01 -7.19968200e-01 -3.25132608e-01 -8.17847177e-02
-5.33376575e-01 4.43459637e-02 -1.59753829e-01 -1.43157423e-01
-8.94207716e-01 -7.11875021e-01 1.57058969e-01 4.74638075e-01
1.18816316e+00 -7.38937914e-01 -3.69547099e-01 5.16290009e-01
-5.88459074e-01 -1.19633245e+00 -9.36579525e-01 -5.01955390e-01
-9.62382138e-01 -1.00792921e+00 -1.30613589e+00 -9.92797554e-01
4.57109451e-01 9.12899494e-01 9.60628152e-01 4.37665492e-01
-1.18486926e-01 -6.01688214e-02 -4.95253444e-01 2.63214350e-01
-2.51678020e-01 -6.10045016e-01 -3.24712425e-01 3.44570637e-01
-9.48919207e-02 -7.60202527e-01 -8.95370483e-01 7.64248133e-01
-1.15461934e+00 6.94616675e-01 7.04002857e-01 1.04730904e+00
5.26195407e-01 5.53445101e-01 2.71652609e-01 -1.84995502e-01
3.04425776e-01 -4.20225114e-01 -5.14173090e-01 1.09977089e-01
-5.82057774e-01 -3.13989401e-01 8.73457432e-01 -6.13666117e-01
-1.38902426e+00 -3.06947917e-01 -1.87346727e-01 -8.03746402e-01
1.66731685e-01 3.83812696e-01 -1.92495093e-01 -8.60669538e-02
1.72248080e-01 5.16970098e-01 -9.50735509e-02 -5.98225832e-01
-1.27334788e-01 6.46014690e-01 9.10841763e-01 -3.50856036e-01
8.08298767e-01 3.68019819e-01 -3.15168835e-02 -8.14404070e-01
-2.98703492e-01 -5.24892032e-01 -2.07658306e-01 -6.39184356e-01
9.10112083e-01 -1.22348154e+00 -7.12251425e-01 9.19458926e-01
-1.32163513e+00 -1.74472388e-02 -4.08849567e-02 9.28142846e-01
-7.55939856e-02 9.90269899e-01 -1.05842078e+00 -2.06568569e-01
-1.91370174e-01 -1.20792210e+00 8.02879453e-01 4.30622756e-01
3.34594041e-01 -5.85565805e-01 -5.98971322e-02 5.09772338e-02
4.27845448e-01 7.07384720e-02 2.70031065e-01 6.42063439e-01
-8.70173573e-01 2.00717837e-01 -5.84846079e-01 4.34501320e-01
-3.95458192e-02 1.30216032e-01 -3.97192776e-01 -3.90804589e-01
1.75352395e-01 1.61684275e-01 7.27285147e-01 5.69132507e-01
1.25745535e+00 -7.54427910e-02 -1.41814277e-01 1.00108993e+00
1.36855304e+00 4.73296463e-01 1.15108740e+00 3.62333775e-01
7.18339026e-01 3.18080634e-01 1.09366667e+00 4.89018440e-01
1.70157272e-02 8.00318778e-01 4.31608528e-01 -3.70714247e-01
-2.93085694e-01 -1.12938039e-01 3.77901405e-01 1.21531808e+00
-4.33333755e-01 -2.80518234e-01 -3.82227302e-01 6.19754255e-01
-1.94747269e+00 -1.20638537e+00 -4.80762929e-01 1.96752667e+00
7.94711292e-01 -4.23551239e-02 -9.88263935e-02 4.16425407e-01
1.10117757e+00 3.15453410e-01 -3.14590305e-01 1.04308039e-01
-3.33796948e-01 -1.61136404e-01 4.46670294e-01 1.44443572e-01
-8.88673365e-01 2.27311894e-01 5.24764729e+00 1.23368871e+00
-9.37251210e-01 2.16745973e-01 4.56598043e-01 1.25717223e-01
-1.28212050e-01 1.38190344e-01 -4.11699355e-01 7.84505427e-01
4.10292059e-01 -7.48418421e-02 1.86299980e-01 3.76157552e-01
4.78190094e-01 -3.58872980e-01 -5.43585420e-01 1.47952986e+00
-1.26209697e-02 -1.52149057e+00 -1.84876069e-01 -2.03419894e-01
8.23106825e-01 -4.61699605e-01 -4.78338376e-02 -1.32083386e-01
-3.37186486e-01 -6.88730896e-01 7.81391859e-01 6.82730019e-01
9.34469521e-01 -9.54823375e-01 7.50596106e-01 2.62826443e-01
-1.64697993e+00 2.42746733e-02 -2.90183485e-01 4.60992493e-02
5.76805890e-01 7.54631221e-01 1.09468564e-01 9.34699655e-01
9.00309682e-01 1.53811264e+00 -2.12153748e-01 1.46298444e+00
-1.05435625e-02 3.80041927e-01 -1.93661094e-01 6.97644651e-01
1.90479252e-02 -4.61621553e-01 6.57377958e-01 1.11876345e+00
5.64883471e-01 4.17839378e-01 3.31411026e-02 4.59607095e-01
1.86528638e-01 -6.57975748e-02 -2.34886080e-01 2.40946069e-01
2.49686629e-01 1.00054717e+00 -4.57293689e-01 -3.97276551e-01
-7.91685343e-01 1.12889087e+00 -5.83764017e-01 5.38110077e-01
-1.10821819e+00 -5.07327139e-01 2.91996717e-01 2.83973843e-01
2.51445472e-01 -4.72622514e-01 6.81012645e-02 -1.53049099e+00
2.10183710e-01 -7.82602012e-01 4.27204341e-01 -1.31031609e+00
-8.80235136e-01 2.98953891e-01 9.73917991e-02 -1.91947091e+00
1.33877203e-01 -2.34253421e-01 -6.69333220e-01 6.62217677e-01
-1.44150615e+00 -8.78880501e-01 -6.47938371e-01 9.01708186e-01
7.98651218e-01 8.54405761e-02 1.79166868e-02 1.07703876e+00
-5.99367738e-01 3.26816678e-01 4.41065133e-01 -8.63102265e-03
1.01381016e+00 -4.82239604e-01 -2.82782555e-01 1.46446800e+00
-4.63875651e-01 5.03098965e-01 7.39190400e-01 -6.90151393e-01
-1.24640739e+00 -1.09729803e+00 4.95962113e-01 8.46379280e-01
2.33989149e-01 -1.34217059e-02 -1.13322592e+00 4.60413843e-01
4.36231256e-01 4.01018351e-01 4.25193273e-03 -6.95462346e-01
1.58865169e-01 -3.84141445e-01 -9.07552779e-01 6.04820132e-01
1.00249350e+00 -1.65519625e-01 -3.39433283e-01 -8.75606984e-02
6.67898595e-01 -5.43346703e-01 -9.93932486e-01 6.61586165e-01
6.28333151e-01 -1.30494332e+00 1.24906468e+00 3.05861682e-01
8.14190745e-01 -8.82412910e-01 2.35795170e-01 -8.77922535e-01
-4.28517103e-01 -9.11646426e-01 -3.11051488e-01 1.34772217e+00
-2.94622302e-01 -2.51090169e-01 5.43165088e-01 2.36277329e-03
-4.19985682e-01 -7.14592218e-01 -6.60480380e-01 -5.49057126e-01
-6.89522982e-01 -2.39360139e-01 3.48358750e-01 9.17614758e-01
-3.74952286e-01 2.05533192e-01 -1.05006778e+00 8.02996978e-02
8.35650563e-01 1.74963221e-01 5.86214900e-01 -7.86684275e-01
-3.49347234e-01 -3.15690860e-02 -3.82696360e-01 -1.51932764e+00
-3.14839602e-01 -4.11322296e-01 7.75181055e-02 -1.46593976e+00
1.38690799e-01 -4.70038980e-01 -1.53346226e-01 -1.47714451e-01
-1.15968771e-01 3.79027486e-01 -1.59190800e-02 4.02106375e-01
-4.78575438e-01 8.02049577e-01 1.65659285e+00 -4.86436263e-02
-6.52520210e-02 -3.13124031e-01 -4.13748100e-02 8.58116925e-01
3.66213322e-01 -3.29231381e-01 -4.85143512e-01 -6.25458837e-01
1.21512644e-01 6.80835843e-01 4.76575315e-01 -1.28962088e+00
5.23914576e-01 -2.18069807e-01 7.05913186e-01 -8.79547358e-01
3.17815453e-01 -1.04479289e+00 5.72937906e-01 3.79550457e-01
1.08541995e-01 3.29141796e-01 2.36323342e-01 5.88827312e-01
-7.62210071e-01 -9.98237729e-02 1.01365697e+00 1.45855024e-01
-1.04897749e+00 5.59091270e-01 -4.31629211e-01 -7.56447986e-02
9.18613851e-01 -4.41614747e-01 -2.42062494e-01 -3.09301585e-01
-6.16376460e-01 1.10779643e-01 5.38941026e-01 2.25369319e-01
1.07581389e+00 -1.37718236e+00 -8.22664678e-01 4.63671565e-01
-3.13468754e-01 1.15451738e-01 9.52528477e-01 1.13381886e+00
-8.89759302e-01 -2.47348577e-01 -5.04341006e-01 -6.36587143e-01
-1.42502332e+00 6.36840999e-01 1.94269434e-01 -1.33590057e-01
-9.89959359e-01 6.52134359e-01 3.08383733e-01 4.59304512e-01
5.98808490e-02 -1.41182765e-01 -2.64160842e-01 -1.58177212e-01
7.29291499e-01 5.64658344e-01 -2.30363622e-01 -9.45352435e-01
-6.92660958e-02 9.11332011e-01 -1.37475699e-01 -3.85724902e-02
1.38340545e+00 -4.95913148e-01 -8.64438042e-02 1.21357389e-01
1.36811423e+00 6.70751650e-03 -1.42237663e+00 -2.25579411e-01
-6.44019604e-01 -1.01685321e+00 3.80746424e-01 -1.96547017e-01
-1.38113570e+00 8.81492376e-01 5.50123930e-01 -2.83045799e-01
1.75834501e+00 -7.00905085e-01 1.56871617e+00 -2.01959372e-01
2.77657241e-01 -1.11224258e+00 2.30345353e-01 1.36992380e-01
8.80179882e-01 -8.18805754e-01 4.90000099e-01 -5.55968046e-01
-1.24199115e-01 1.30073857e+00 5.61397433e-01 -1.10789411e-01
5.62874854e-01 3.91897768e-01 -4.35975760e-01 8.20957050e-02
-5.69755077e-01 2.92280257e-01 1.28404230e-01 3.47089708e-01
3.20976138e-01 -6.19505644e-01 -7.66706347e-01 3.11682761e-01
5.76943219e-01 3.47585499e-01 7.38992810e-01 9.90205348e-01
-2.83002228e-01 -9.82849777e-01 -8.38475406e-01 1.73105299e-01
-6.20059967e-01 -5.09887002e-03 5.92411935e-01 8.50928843e-01
4.84243780e-01 9.79879439e-01 9.05026495e-02 -5.50556600e-01
2.21277401e-01 -6.82394028e-01 5.28817415e-01 1.33144990e-01
-3.13141942e-01 5.98663568e-01 -2.96039209e-02 -5.74378490e-01
-7.89161503e-01 -4.54994202e-01 -1.16726506e+00 -4.97867554e-01
-4.67127025e-01 2.68194169e-01 2.15239838e-01 5.95133126e-01
2.71890938e-01 6.36666656e-01 5.77319384e-01 -1.11499333e+00
1.50321826e-01 -7.65988052e-01 -7.62495756e-01 5.27377665e-01
5.05324841e-01 -7.38040745e-01 -3.75483721e-01 4.79709417e-01] | [11.040820121765137, -1.8521862030029297] |
9f4b3a35-3ccb-4ee2-bb65-b4d91592bac1 | learning-versatile-3d-shape-generation-with | 2303.14700 | null | https://arxiv.org/abs/2303.14700v1 | https://arxiv.org/pdf/2303.14700v1.pdf | Learning Versatile 3D Shape Generation with Improved AR Models | Auto-Regressive (AR) models have achieved impressive results in 2D image generation by modeling joint distributions in the grid space. While this approach has been extended to the 3D domain for powerful shape generation, it still has two limitations: expensive computations on volumetric grids and ambiguous auto-regressive order along grid dimensions. To overcome these limitations, we propose the Improved Auto-regressive Model (ImAM) for 3D shape generation, which applies discrete representation learning based on a latent vector instead of volumetric grids. Our approach not only reduces computational costs but also preserves essential geometric details by learning the joint distribution in a more tractable order. Moreover, thanks to the simplicity of our model architecture, we can naturally extend it from unconditional to conditional generation by concatenating various conditioning inputs, such as point clouds, categories, images, and texts. Extensive experiments demonstrate that ImAM can synthesize diverse and faithful shapes of multiple categories, achieving state-of-the-art performance. | ['xiangyang xue', 'Chengjie Wang', 'Zhenyu Zhang', 'Ying Tai', 'yinda zhang', 'Yanwei Fu', 'Xuelin Qian', 'Simian Luo'] | 2023-03-26 | null | null | null | null | ['3d-shape-generation'] | ['computer-vision'] | [ 8.98507386e-02 5.46373501e-02 1.73347294e-01 -4.72899228e-02
-1.02932644e+00 -5.70888638e-01 9.89252806e-01 -3.33574682e-01
2.25524515e-01 7.09997535e-01 2.28389546e-01 -2.82632619e-01
8.27442482e-02 -1.23632646e+00 -8.22345018e-01 -7.23365963e-01
2.97998816e-01 8.15784395e-01 -1.52885541e-01 -1.36664957e-01
4.71650623e-02 8.85784388e-01 -1.48806953e+00 -2.82280203e-02
8.43601644e-01 8.57748389e-01 1.05695546e-01 5.29477954e-01
-2.89373279e-01 4.63675141e-01 -3.67390275e-01 -3.27253312e-01
3.00855666e-01 -2.79233217e-01 -4.45268929e-01 4.40702230e-01
4.62828934e-01 -3.49791467e-01 -3.03932935e-01 5.76090038e-01
4.45288837e-01 1.19656548e-02 1.47897053e+00 -1.09170544e+00
-1.23714745e+00 2.03868434e-01 -6.79641664e-01 -6.02835774e-01
4.26040217e-02 1.26332164e-01 9.48145211e-01 -1.32331967e+00
5.54667592e-01 1.39727390e+00 6.62367702e-01 5.64932466e-01
-1.78931904e+00 -6.15994036e-01 6.95441961e-02 -4.47434485e-01
-1.43277884e+00 -6.77281767e-02 1.08828866e+00 -7.38464057e-01
8.83525670e-01 1.15497492e-01 7.15840161e-01 1.15890920e+00
2.06952214e-01 7.63469517e-01 1.05461645e+00 -4.09487426e-01
2.62694895e-01 -1.43594593e-02 -6.21190965e-01 5.96349776e-01
9.82837901e-02 2.18570590e-01 -1.21944450e-01 -4.17367548e-01
1.55119944e+00 -7.55845159e-02 1.25862077e-01 -9.45976794e-01
-1.14203072e+00 1.14267254e+00 4.65368181e-01 1.93454936e-01
-4.92209375e-01 1.96004108e-01 -7.05899373e-02 -4.51369584e-02
8.76912594e-01 3.86735022e-01 -2.88608015e-01 2.79078066e-01
-1.01486957e+00 6.04150414e-01 4.44200218e-01 1.14374352e+00
7.75586963e-01 6.72032893e-01 -2.13495910e-01 9.16338027e-01
3.45952213e-01 8.75816643e-01 5.38907535e-02 -8.91826391e-01
2.57179111e-01 5.12896180e-01 1.36859104e-01 -8.93625379e-01
-4.78228852e-02 -3.85321558e-01 -1.26577091e+00 6.10342979e-01
5.36499219e-03 -1.68993883e-03 -1.27632380e+00 1.60674250e+00
3.28849137e-01 2.07173422e-01 -3.45228724e-02 6.48126721e-01
8.52793038e-01 8.29164624e-01 1.89225912e-01 -2.34960467e-02
8.18717778e-01 -6.35974228e-01 -3.64731431e-01 2.99656153e-01
1.31523564e-01 -8.84430468e-01 1.12379456e+00 2.11218819e-01
-1.36438453e+00 -6.81560099e-01 -7.67233253e-01 -1.94074050e-01
-2.48962790e-01 1.65555522e-01 8.36482465e-01 3.88429105e-01
-1.38628960e+00 4.32005137e-01 -7.75986135e-01 -1.25095723e-02
5.81321001e-01 1.90850213e-01 -2.12945625e-01 1.30960613e-01
-7.40222692e-01 6.53217554e-01 -1.71506524e-01 -4.70833957e-01
-8.70111406e-01 -1.01904118e+00 -1.16147900e+00 -1.91427678e-01
-1.68496892e-01 -1.35718024e+00 1.06762958e+00 -3.47490400e-01
-1.73510957e+00 7.15282679e-01 -1.54244363e-01 -2.18924329e-01
4.65745807e-01 -1.12138204e-02 1.21522434e-01 -1.08564161e-01
1.28780857e-01 1.10671043e+00 1.17705595e+00 -1.73164320e+00
-2.39438653e-01 -2.73928165e-01 -2.67920464e-01 2.51914829e-01
-1.27572834e-01 -5.49190342e-01 -1.77850679e-01 -9.54653680e-01
1.94462657e-01 -9.65791643e-01 -6.56617284e-01 9.18231979e-02
-2.93396294e-01 -2.76206851e-01 7.22908795e-01 -3.66650403e-01
5.97768605e-01 -1.97104788e+00 4.78077233e-01 5.58763631e-02
1.65068805e-01 -2.26158142e-01 -3.35449755e-01 2.91418344e-01
1.88606177e-02 3.30916554e-01 -4.26907122e-01 -6.78507268e-01
1.08868042e-02 2.21375957e-01 -9.15954292e-01 1.47045389e-01
6.75008714e-01 1.20790875e+00 -6.55034363e-01 -4.67535675e-01
6.15566432e-01 9.60556030e-01 -7.43858814e-01 2.73014665e-01
-4.58296031e-01 7.68294811e-01 -4.72835004e-01 5.39865375e-01
8.04163098e-01 -4.53056335e-01 -4.00908291e-01 -1.81807578e-01
-1.15366817e-01 -5.31929657e-02 -9.50347483e-01 1.75483584e+00
-7.33533919e-01 9.29178298e-02 -2.63022333e-01 -7.31041551e-01
1.28633726e+00 3.35387468e-01 8.23062658e-01 -3.91372442e-01
-1.54067114e-01 1.27638265e-01 -6.74043477e-01 1.76172525e-01
6.14398003e-01 -4.50843096e-01 -2.24494576e-01 3.56834382e-01
1.96171686e-01 -1.16382635e+00 -2.48936862e-01 3.37798931e-02
5.55873990e-01 5.74863672e-01 1.21425958e-02 -1.05910994e-01
2.83049703e-01 -5.81675209e-02 2.29091212e-01 5.41807234e-01
6.60278499e-01 1.05226076e+00 1.20147511e-01 -4.80257571e-01
-1.66436982e+00 -1.60534501e+00 -2.49281287e-01 5.31201243e-01
-1.05168305e-01 -8.69107842e-02 -3.09848398e-01 -4.22200114e-01
2.82046497e-01 9.85786676e-01 -6.38247252e-01 4.50272188e-02
-6.08364165e-01 -5.79932690e-01 3.47915500e-01 7.25860298e-01
1.48741201e-01 -1.01920104e+00 -1.85020253e-01 2.91198134e-01
-4.49295994e-03 -9.61754799e-01 -3.32836151e-01 -3.22915018e-01
-1.15936458e+00 -5.55616796e-01 -1.15468037e+00 -6.67250574e-01
8.11016619e-01 1.99611545e-01 1.34694564e+00 -3.45624804e-01
-4.36089456e-01 6.48296654e-01 -7.21464911e-03 -4.51414764e-01
-4.71893221e-01 -9.39221010e-02 -9.74401534e-02 -4.68440577e-02
-2.90135473e-01 -9.47618186e-01 -4.21668649e-01 2.45842710e-02
-8.79212379e-01 3.26031327e-01 6.63727760e-01 9.68735218e-01
1.18332767e+00 -6.51098788e-02 5.55517554e-01 -7.30112374e-01
4.43229824e-01 -3.17188799e-01 -6.71376109e-01 -6.61498457e-02
-4.84914541e-01 2.12171510e-01 5.35109818e-01 -4.00868744e-01
-1.19256532e+00 1.58127591e-01 -2.97622979e-01 -9.14502978e-01
-2.57343590e-01 1.55715346e-01 9.93068814e-02 -3.89802940e-02
5.02813399e-01 5.91951549e-01 1.48024648e-01 -4.81923550e-01
8.85295868e-01 2.26995856e-01 4.38063294e-01 -7.69848645e-01
1.27407134e+00 6.59059823e-01 4.53930378e-01 -8.39737236e-01
-5.18972576e-01 -2.01612897e-02 -7.86175728e-01 2.97082495e-02
8.86710346e-01 -1.19334173e+00 -4.64288294e-01 3.66987824e-01
-1.23522913e+00 -3.41339618e-01 -6.91478789e-01 1.80668011e-01
-1.04568136e+00 2.29466304e-01 -5.76481938e-01 -9.26082850e-01
-5.19830048e-01 -8.88583899e-01 1.58600974e+00 -1.28763318e-01
-8.55154321e-02 -1.04769778e+00 5.83045743e-02 -3.31134945e-02
5.20567298e-01 6.62410319e-01 1.36419678e+00 4.51786518e-02
-7.76767015e-01 -1.32932931e-01 -2.20103860e-01 1.06736019e-01
2.91711569e-01 6.56915233e-02 -7.26157069e-01 -2.47933492e-01
-1.23203605e-01 -4.05673057e-01 8.14810276e-01 6.87272131e-01
1.23407459e+00 -3.46176058e-01 -2.87286907e-01 6.06387138e-01
1.46416831e+00 -1.27176762e-01 7.29776263e-01 -1.92057744e-01
8.31227064e-01 3.21972638e-01 2.21658513e-01 4.36639309e-01
4.57277417e-01 6.28379762e-01 3.82936746e-01 -4.62914944e-01
-4.75775301e-01 -6.12278163e-01 4.48050313e-02 9.60215330e-01
-3.90951127e-01 -7.64025599e-02 -6.63505256e-01 6.33547008e-01
-1.55577087e+00 -7.50393152e-01 -3.47819999e-02 2.15885496e+00
7.24516749e-01 -1.76606059e-01 1.74263000e-01 2.66520958e-03
4.35281903e-01 8.19069222e-02 -5.20551205e-01 -1.57899320e-01
-2.70835340e-01 4.99427378e-01 1.77767321e-01 5.53598881e-01
-9.58903551e-01 1.01332724e+00 7.16560698e+00 9.86972332e-01
-1.11676145e+00 -1.55778766e-01 8.65455449e-01 2.46807203e-01
-1.01272631e+00 -1.55286148e-01 -7.04137385e-01 9.06501487e-02
3.28382343e-01 -8.92009065e-02 2.97318608e-01 9.63323176e-01
-3.88856083e-02 4.81631488e-01 -8.54376495e-01 1.05444658e+00
8.17610100e-02 -1.60018730e+00 7.90162683e-01 3.81280929e-01
1.06703711e+00 -1.90284654e-01 5.05939543e-01 2.97089487e-01
6.49919510e-01 -1.26548648e+00 7.31494367e-01 7.58954287e-01
1.41316831e+00 -8.79340351e-01 2.24349082e-01 3.31311554e-01
-1.16211021e+00 3.62412542e-01 -5.82906246e-01 1.03188567e-01
2.45242804e-01 6.57675505e-01 -9.50714588e-01 6.20875061e-01
3.31309885e-01 6.98847950e-01 -2.06306353e-01 6.32176161e-01
-1.00467704e-01 3.44254017e-01 -2.51346558e-01 1.33638442e-01
4.14702073e-02 -4.04663801e-01 6.78636849e-01 9.20982838e-01
7.11533308e-01 1.76027492e-01 2.17635795e-01 1.50204253e+00
1.39512375e-01 6.14857785e-02 -1.12076271e+00 9.10931826e-02
3.69809598e-01 1.17776012e+00 -5.06963611e-01 -1.54385969e-01
-2.37917334e-01 7.35791683e-01 2.59352475e-01 4.45924878e-01
-6.36987984e-01 -4.81984094e-02 4.92933661e-01 3.23374808e-01
5.47578216e-01 -7.50998497e-01 -7.66460359e-01 -9.56477046e-01
-3.20549101e-01 -4.45218682e-01 -3.20680663e-02 -8.92485619e-01
-1.67780292e+00 7.53755391e-01 4.26829010e-02 -1.36338401e+00
-5.83097219e-01 -5.22744536e-01 -4.40573752e-01 1.02284467e+00
-1.43732405e+00 -1.84106994e+00 -1.76141798e-01 6.40698493e-01
6.00568473e-01 -2.79022187e-01 1.22990227e+00 -1.44511595e-01
1.78036824e-01 4.99211222e-01 -2.62366205e-01 -3.89954261e-02
3.30627531e-01 -1.39859462e+00 6.79394782e-01 3.57299685e-01
3.97183329e-01 2.54021943e-01 4.56366241e-01 -7.75744200e-01
-1.30522859e+00 -1.29619360e+00 6.95101976e-01 -6.56004131e-01
2.65034229e-01 -3.27619195e-01 -7.78198719e-01 5.59228480e-01
8.47790018e-02 -1.10660553e-01 5.13512731e-01 3.33517194e-02
-3.72203588e-01 2.17930689e-01 -1.03465629e+00 8.98000598e-01
1.05776930e+00 -3.38329315e-01 -3.80620241e-01 2.18856379e-01
9.14939284e-01 -4.95184213e-01 -1.14658618e+00 8.58335912e-01
1.83110058e-01 -6.84420228e-01 1.37670708e+00 -3.96021634e-01
7.48632908e-01 -2.72714853e-01 -2.39395648e-01 -1.41600442e+00
-7.36857355e-01 -5.08487999e-01 -2.62007684e-01 1.22872221e+00
3.78385127e-01 -3.86447996e-01 8.12903285e-01 4.05352414e-01
-1.38202950e-01 -1.03820765e+00 -8.44670892e-01 -6.39876008e-01
7.08832860e-01 -3.88603747e-01 1.01700068e+00 6.23024464e-01
-6.13750041e-01 4.84407455e-01 -4.90056723e-01 4.11193334e-02
9.08591509e-01 5.96665621e-01 9.67616975e-01 -1.32385850e+00
-1.67006597e-01 -5.20120740e-01 -2.18434632e-01 -1.28823388e+00
1.66813344e-01 -9.37941194e-01 -6.49867430e-02 -1.72232306e+00
1.68449447e-01 -8.30426037e-01 9.16149020e-02 3.62616867e-01
2.34688055e-02 5.02501547e-01 2.34745353e-01 3.48122418e-01
-1.80667471e-02 1.08214688e+00 1.78797853e+00 -2.64481336e-01
-1.78080767e-01 7.52068460e-02 -6.84805393e-01 6.61800921e-01
5.72863519e-01 -1.13945283e-01 -4.67488825e-01 -5.15114248e-01
1.26560360e-01 2.95098603e-01 5.91610253e-01 -8.30896020e-01
-7.11253136e-02 -3.29635888e-01 8.89488518e-01 -9.76996541e-01
6.71449184e-01 -5.95265388e-01 3.73790532e-01 -1.98562164e-03
-2.11902291e-01 -1.05348513e-01 2.73397237e-01 5.82566917e-01
-1.35351315e-01 2.37951979e-01 7.51807868e-01 -2.31095523e-01
-1.01807475e-01 8.17769706e-01 -1.65267944e-01 -2.14811757e-01
8.12677026e-01 -2.94367433e-01 1.93937451e-01 -5.66794276e-01
-7.69511223e-01 -1.15948498e-01 6.91244781e-01 3.75039548e-01
9.13840771e-01 -1.85231829e+00 -1.11155307e+00 5.47033668e-01
-1.49879185e-02 4.26972240e-01 2.71941662e-01 2.36691430e-01
-1.59075275e-01 4.76630181e-01 -1.19250715e-01 -9.19599354e-01
-7.23412395e-01 5.53798020e-01 -6.65873894e-03 -4.02517676e-01
-8.16040516e-01 7.16819942e-01 6.33216560e-01 -7.95216918e-01
-2.16863573e-01 -8.54029134e-02 -1.71322525e-01 -2.84973860e-01
6.98385239e-02 4.15831916e-02 -1.44091845e-01 -6.93690538e-01
8.95818397e-02 9.99691427e-01 1.57810420e-01 -2.26566166e-01
1.44258940e+00 4.74826396e-02 5.49213551e-02 4.50843662e-01
9.16275024e-01 1.05662465e-01 -1.60514939e+00 -1.16200186e-01
-5.95387042e-01 -4.89808381e-01 -5.92074841e-02 -5.53783894e-01
-9.64344978e-01 9.62482750e-01 1.61375761e-01 1.01932980e-01
8.28692555e-01 1.72010303e-01 7.72678375e-01 -7.01275542e-02
6.10607326e-01 -5.49634635e-01 4.89184737e-01 6.02472067e-01
1.52009380e+00 -8.16873491e-01 9.02502462e-02 -6.02532744e-01
-6.78446293e-01 9.66322362e-01 5.07175505e-01 -5.61161578e-01
8.39311182e-01 3.12155694e-01 -4.62058708e-02 -6.92892820e-02
-7.99147010e-01 -1.43149480e-01 5.78595281e-01 9.28410411e-01
4.12879169e-01 2.08505392e-01 9.07126814e-02 3.50259423e-01
-4.50641036e-01 -1.64614275e-01 2.73423672e-01 4.52137291e-01
-8.57069120e-02 -1.14737153e+00 -3.15542459e-01 4.07927811e-01
-7.16382265e-02 -4.92255017e-02 -1.13739558e-01 7.33235717e-01
1.15619460e-02 4.97022420e-01 2.84808993e-01 -2.56484360e-01
3.47292155e-01 -2.37528961e-02 8.08931053e-01 -6.63402796e-01
6.90426826e-02 4.38562363e-01 -3.96424741e-01 -2.32680872e-01
-1.90156832e-01 -7.21346557e-01 -1.06232285e+00 -8.70886073e-02
-2.63306975e-01 -1.58314835e-02 5.49267650e-01 5.83883166e-01
6.66052997e-01 5.17084956e-01 9.17001605e-01 -1.33741879e+00
-7.19910920e-01 -8.88152361e-01 -5.39407670e-01 4.72116739e-01
1.11762911e-01 -9.44867432e-01 -1.80619538e-01 2.49293625e-01] | [8.916244506835938, -3.6184093952178955] |
a0c7efa9-0c41-4374-8e67-a79344ac7abf | recurrent-video-restoration-transformer-with | 2206.02146 | null | https://arxiv.org/abs/2206.02146v3 | https://arxiv.org/pdf/2206.02146v3.pdf | Recurrent Video Restoration Transformer with Guided Deformable Attention | Video restoration aims at restoring multiple high-quality frames from multiple low-quality frames. Existing video restoration methods generally fall into two extreme cases, i.e., they either restore all frames in parallel or restore the video frame by frame in a recurrent way, which would result in different merits and drawbacks. Typically, the former has the advantage of temporal information fusion. However, it suffers from large model size and intensive memory consumption; the latter has a relatively small model size as it shares parameters across frames; however, it lacks long-range dependency modeling ability and parallelizability. In this paper, we attempt to integrate the advantages of the two cases by proposing a recurrent video restoration transformer, namely RVRT. RVRT processes local neighboring frames in parallel within a globally recurrent framework which can achieve a good trade-off between model size, effectiveness, and efficiency. Specifically, RVRT divides the video into multiple clips and uses the previously inferred clip feature to estimate the subsequent clip feature. Within each clip, different frame features are jointly updated with implicit feature aggregation. Across different clips, the guided deformable attention is designed for clip-to-clip alignment, which predicts multiple relevant locations from the whole inferred clip and aggregates their features by the attention mechanism. Extensive experiments on video super-resolution, deblurring, and denoising show that the proposed RVRT achieves state-of-the-art performance on benchmark datasets with balanced model size, testing memory and runtime. | ['Luc van Gool', 'Radu Timofte', 'Kai Zhang', 'JieZhang Cao', 'Simon Green', 'Eddy Ilg', 'Rakesh Ranjan', 'Xiaoyu Xiang', 'Yuchen Fan', 'Jingyun Liang'] | 2022-06-05 | null | null | null | null | ['video-super-resolution', 'video-denoising', 'video-restoration'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 1.78596318e-01 -5.73949635e-01 -1.23648778e-01 -1.46364212e-01
-9.56010818e-01 -7.60659501e-02 2.44426370e-01 -2.94909954e-01
-1.16758063e-01 4.89783287e-01 6.92437351e-01 2.72296846e-01
-1.48768499e-01 -5.94919026e-01 -6.90183282e-01 -9.95781779e-01
2.87903905e-01 -2.05194116e-01 5.84420562e-01 -8.08444172e-02
2.89727747e-01 2.65532851e-01 -1.54091501e+00 2.18128458e-01
9.88508224e-01 1.13769329e+00 6.35093153e-01 4.53418404e-01
-3.34682898e-03 1.09784710e+00 -3.04165512e-01 -8.02596956e-02
1.44445240e-01 -4.41917181e-01 -4.17171657e-01 2.31681332e-01
4.00719076e-01 -6.36803985e-01 -6.60497546e-01 1.11271238e+00
5.80399215e-01 2.90709138e-01 -2.61507034e-02 -8.81704569e-01
-9.35915291e-01 4.31625217e-01 -9.03820753e-01 5.87719262e-01
5.29261887e-01 3.57065834e-02 7.10424662e-01 -9.26805496e-01
4.59836155e-01 1.25065398e+00 5.04356086e-01 4.37770694e-01
-1.04901934e+00 -6.42229021e-01 4.87401128e-01 5.32226264e-01
-1.46203446e+00 -7.38331556e-01 7.05741823e-01 -1.97268948e-01
9.03580785e-01 2.93648273e-01 6.46253884e-01 6.55304551e-01
2.45862365e-01 6.78959668e-01 7.48407423e-01 2.61500906e-02
8.12253058e-02 -5.27921796e-01 -2.19243448e-02 3.50964218e-01
-1.20838359e-01 -1.06756963e-01 -6.41303480e-01 -5.35071902e-02
1.00991666e+00 4.56858456e-01 -8.72475922e-01 4.41830829e-02
-1.32618821e+00 3.64972472e-01 2.25147232e-01 3.24736446e-01
-6.08078957e-01 -5.80022782e-02 4.83845800e-01 2.85195023e-01
5.46177566e-01 -2.76888341e-01 -2.17910647e-01 -6.36186302e-02
-1.00214314e+00 1.72135592e-01 2.55142272e-01 7.44627833e-01
6.80370152e-01 1.70556411e-01 -2.66203403e-01 9.39386249e-01
1.83900476e-01 3.46426934e-01 7.43778467e-01 -1.13222718e+00
6.49631679e-01 2.16077328e-01 1.76623106e-01 -1.28543293e+00
-5.31771220e-02 -1.10102646e-01 -1.08457446e+00 -5.13081513e-02
-1.66703641e-01 1.41838074e-01 -8.23338330e-01 1.68762922e+00
3.52517337e-01 7.88152099e-01 1.93729177e-02 1.35798728e+00
9.50579524e-01 1.03333044e+00 6.58813166e-04 -9.01084840e-01
1.15111768e+00 -1.09559214e+00 -9.36544895e-01 -6.02652580e-02
-1.60330340e-01 -9.16626394e-01 6.65145397e-01 3.61336827e-01
-1.53433681e+00 -9.22600746e-01 -1.03777373e+00 -2.85534918e-01
4.10936326e-01 -9.53244641e-02 1.49931595e-01 1.46154329e-01
-1.13190258e+00 5.34208775e-01 -8.56285214e-01 -7.62717575e-02
2.24430755e-01 1.42556012e-01 -3.36926073e-01 -3.18753302e-01
-9.55579102e-01 4.50287789e-01 4.97362390e-02 2.96881974e-01
-8.32954824e-01 -6.70769453e-01 -7.12522149e-01 2.92713374e-01
3.74316782e-01 -7.29674280e-01 9.36042786e-01 -1.00187564e+00
-1.44097793e+00 3.59833241e-01 -6.00377202e-01 -3.03560525e-01
4.76472050e-01 -4.38159347e-01 -6.28891051e-01 1.63701102e-01
1.75426126e-01 3.24122548e-01 9.83010590e-01 -1.07529509e+00
-8.38018119e-01 -3.29746634e-01 8.78366455e-03 4.14396137e-01
-3.49137187e-01 2.32652888e-01 -1.00027788e+00 -9.89594221e-01
5.63959718e-01 -5.78247547e-01 -6.48736162e-03 -1.75322935e-01
-1.83196515e-02 7.17391074e-02 9.32082474e-01 -9.83730674e-01
1.52861965e+00 -2.27088618e+00 5.46695709e-01 -2.04547256e-01
2.87864685e-01 2.88036495e-01 -1.17435649e-01 1.24500588e-01
-2.04433769e-01 -5.27948290e-02 -7.39648379e-03 -3.57721657e-01
-4.46720779e-01 8.29681978e-02 -5.33233047e-01 4.24141318e-01
-1.23471856e-01 6.01101220e-01 -7.58615017e-01 -5.65543175e-01
2.98623472e-01 8.23114812e-01 -6.29974186e-01 2.78825313e-01
1.39537141e-01 4.42043275e-01 -4.57012594e-01 6.44444883e-01
9.16436076e-01 -2.20534861e-01 1.36145473e-01 -5.92867255e-01
-2.34686658e-01 1.51577285e-02 -1.38872623e+00 1.71613586e+00
-3.40085119e-01 4.38272208e-01 2.11744025e-01 -9.05997515e-01
8.11018109e-01 4.67395246e-01 6.69019759e-01 -7.81535923e-01
-1.84652328e-01 3.56455706e-02 -3.32073331e-01 -6.73761666e-01
6.11452579e-01 -1.08678720e-03 2.81473726e-01 2.43952364e-01
-3.45841676e-01 5.76132059e-01 1.25800982e-01 4.11682576e-02
1.09286690e+00 3.21536005e-01 2.66815782e-01 5.95968999e-02
8.61303627e-01 -5.11678874e-01 1.19174206e+00 3.23597521e-01
-1.68398112e-01 1.04610908e+00 2.08614394e-01 -5.30934691e-01
-8.36099386e-01 -1.00195992e+00 1.73081666e-01 9.34645593e-01
6.73315585e-01 -4.74086046e-01 -6.39990568e-01 -2.52432227e-01
-4.13381189e-01 2.09880576e-01 -3.43110949e-01 -7.79333785e-02
-9.46178436e-01 -6.79845989e-01 2.72405446e-02 4.84108269e-01
6.99860454e-01 -1.12175763e+00 -4.83880460e-01 4.28669572e-01
-7.22856224e-01 -1.02422822e+00 -8.77547204e-01 -5.60493112e-01
-8.52413118e-01 -9.46368635e-01 -7.84536898e-01 -9.46854830e-01
4.59737778e-01 1.08334720e+00 9.06896830e-01 3.81068200e-01
1.95440754e-01 -9.79271065e-03 -5.33711493e-01 5.11583030e-01
-4.87787724e-02 -3.46135795e-01 1.09368674e-01 3.54321212e-01
2.54844166e-02 -8.91128719e-01 -8.59189630e-01 4.51124430e-01
-1.17690170e+00 2.78999895e-01 4.78400230e-01 8.97496998e-01
1.02273571e+00 1.99530885e-01 5.88759243e-01 -3.41092408e-01
4.62738335e-01 -5.09266257e-01 -3.46213609e-01 3.23112875e-01
-3.05735350e-01 -3.70808691e-01 8.22347164e-01 -5.43531477e-01
-1.27954721e+00 -1.42527610e-01 -2.38700453e-02 -9.24725115e-01
1.15200967e-01 4.88393217e-01 -4.63981450e-01 6.91004023e-02
1.54971536e-02 6.38950765e-01 -1.29569843e-01 -7.04766750e-01
6.31759837e-02 4.76337314e-01 8.74696374e-01 -4.10364628e-01
6.07795358e-01 4.14508820e-01 -3.65085959e-01 -5.02036393e-01
-5.80554307e-01 -2.32750922e-01 -2.81079978e-01 -2.64677495e-01
8.00162375e-01 -1.22498703e+00 -6.57626808e-01 6.73599601e-01
-1.11325753e+00 9.60595682e-02 -9.83133838e-02 4.49475914e-01
-4.00433749e-01 7.07613051e-01 -8.50044847e-01 -4.34117019e-01
-4.99034107e-01 -1.45563102e+00 1.01717746e+00 5.19286513e-01
2.09346652e-01 -4.82588828e-01 -1.69868246e-01 4.43133563e-01
4.51733679e-01 -3.04111454e-04 6.36603355e-01 -5.30100940e-03
-8.86243463e-01 -2.29859562e-03 -4.66276467e-01 3.66914362e-01
2.75051326e-01 -1.15735911e-01 -6.42532825e-01 -4.82922226e-01
2.33146563e-01 1.35221809e-01 8.56864870e-01 6.17610574e-01
1.25257516e+00 -4.12754148e-01 -1.19340815e-01 6.74447834e-01
1.51282179e+00 2.40084812e-01 9.86470103e-01 3.78904998e-01
8.59643400e-01 1.35675475e-01 8.46249521e-01 4.99411881e-01
5.37662148e-01 9.02683914e-01 4.05663103e-01 1.11193687e-01
-2.52751529e-01 -1.58737786e-02 6.88445926e-01 1.10916853e+00
-3.32786739e-01 -2.36276880e-01 -4.13302779e-01 4.85278189e-01
-2.19574976e+00 -1.38533461e+00 8.57952908e-02 2.25843334e+00
6.69726491e-01 -1.06580257e-01 -5.65177463e-02 6.64622858e-02
1.02018929e+00 5.11440217e-01 -5.67061365e-01 1.09076448e-01
-3.14012766e-01 -1.58678472e-01 1.61151797e-01 5.04234731e-01
-8.86338055e-01 6.22856736e-01 5.39243746e+00 1.01303184e+00
-1.09005487e+00 3.57369810e-01 7.02261746e-01 -4.34883267e-01
-1.24071099e-01 1.14579558e-01 -6.13898814e-01 8.54511917e-01
5.54187775e-01 -2.64868259e-01 6.80393159e-01 4.13061351e-01
5.91361165e-01 -7.26555437e-02 -7.29833603e-01 1.19720209e+00
2.41443306e-01 -1.51811552e+00 2.44522139e-01 -1.24894686e-01
7.43110120e-01 -6.01438694e-02 -1.48040429e-01 3.39361541e-02
-2.13058710e-01 -7.36802578e-01 8.83746564e-01 8.26971471e-01
5.94361365e-01 -8.53924930e-01 7.21240878e-01 1.52669802e-01
-1.63790941e+00 -2.07380548e-01 -4.10675138e-01 -4.01012711e-02
5.90746343e-01 5.23047984e-01 3.70219558e-01 8.68510783e-01
1.18698502e+00 1.08780289e+00 -3.07982326e-01 1.06088877e+00
2.17967689e-01 2.72234499e-01 -4.36558239e-02 7.89462507e-01
-1.89955786e-01 -4.25691605e-01 7.04414666e-01 8.88318777e-01
6.85953796e-01 4.26349789e-01 5.15579462e-01 4.39506203e-01
9.07654837e-02 -9.49474052e-02 6.52303398e-02 5.89801252e-01
6.58368707e-01 1.17832410e+00 -4.57712948e-01 -4.90318209e-01
-7.49607086e-01 1.04247952e+00 1.57269701e-01 4.11958903e-01
-1.10142493e+00 -1.73714072e-01 8.22983801e-01 1.50681555e-01
5.75170457e-01 6.39481321e-02 -5.65668903e-02 -1.47558117e+00
5.20150363e-01 -9.72839177e-01 3.29860657e-01 -1.07182801e+00
-1.20805240e+00 7.63983011e-01 -2.54967034e-01 -1.60364354e+00
-4.42626514e-02 1.75862074e-01 -5.12118399e-01 8.04722786e-01
-1.53108037e+00 -1.02187991e+00 -7.16498971e-01 8.01130831e-01
8.62859249e-01 9.43444520e-02 3.95436734e-01 6.17601216e-01
-9.59657013e-01 4.10935968e-01 1.24800138e-01 -7.97197521e-02
8.36622894e-01 -6.53328955e-01 2.51583099e-01 1.18925524e+00
-3.64878356e-01 4.99688566e-01 6.06508315e-01 -6.88133299e-01
-1.46960533e+00 -1.11063755e+00 6.43844008e-01 2.01149583e-01
4.90181565e-01 2.69257933e-01 -1.35693026e+00 5.69823205e-01
7.12346137e-02 3.00521553e-01 1.97739840e-01 -3.90056819e-01
-3.40904355e-01 -4.49086666e-01 -1.01241279e+00 6.49662971e-01
1.13147700e+00 -3.63238871e-01 -4.86103415e-01 7.09466785e-02
8.69983196e-01 -5.81921697e-01 -1.11795580e+00 3.98643076e-01
5.05271077e-01 -1.38013864e+00 1.14592314e+00 -2.98018642e-02
6.79480135e-01 -7.31608570e-01 -3.41958940e-01 -8.98724258e-01
-6.18879318e-01 -7.52943277e-01 -3.61580253e-01 1.67483556e+00
-2.26852268e-01 -5.91118395e-01 3.80260855e-01 5.94556093e-01
-2.32889429e-01 -7.86965668e-01 -1.07446826e+00 -5.45740068e-01
-4.25282151e-01 -1.38274789e-01 7.44433284e-01 7.54035175e-01
-2.77287096e-01 1.99483886e-01 -7.56830871e-01 3.61213654e-01
5.68961442e-01 3.76947701e-01 5.82308829e-01 -8.70344579e-01
-3.19109917e-01 -2.41092384e-01 -2.86136866e-01 -1.40889502e+00
8.33383650e-02 -3.35849017e-01 5.18785678e-02 -1.49470329e+00
5.74318230e-01 -2.61921167e-01 -4.67326462e-01 3.88751864e-01
-4.91100013e-01 3.56692225e-01 3.81117225e-01 6.89037621e-01
-7.76283503e-01 6.24332130e-01 1.17184579e+00 -9.82266851e-03
-3.23110104e-01 -3.28569919e-01 -7.06812441e-01 7.26518750e-01
3.98773074e-01 -2.17659414e-01 -2.72627592e-01 -8.07262242e-01
-1.54923052e-01 6.18407190e-01 4.84262109e-01 -1.10657609e+00
4.13244635e-01 -2.04165787e-01 4.60302889e-01 -5.39803982e-01
4.99324769e-01 -7.56248474e-01 5.25821149e-01 6.16375320e-02
-1.19860312e-02 3.40218097e-01 -4.58572619e-02 7.97244310e-01
-5.34410298e-01 1.94785222e-01 7.68590212e-01 -5.83126210e-02
-9.01434064e-01 5.69672763e-01 -1.61480486e-01 -1.24243475e-01
8.31083000e-01 -4.31653619e-01 -3.93682897e-01 -4.22986656e-01
-5.93312979e-01 1.95142224e-01 8.19967389e-01 7.55947471e-01
8.03149164e-01 -1.48370671e+00 -8.83357465e-01 1.49625257e-01
-3.32866997e-01 -7.95983430e-03 1.01721382e+00 1.01786005e+00
-3.03375036e-01 -1.55269615e-02 -1.49069175e-01 -6.60928845e-01
-1.46062803e+00 7.09210098e-01 1.82680845e-01 -2.50107914e-01
-1.09289908e+00 4.06518519e-01 4.15938020e-01 2.47246802e-01
4.88931425e-02 7.16521740e-02 -3.47673267e-01 2.70020477e-02
1.07239854e+00 4.65716720e-01 -3.23739052e-02 -1.03406930e+00
-2.42842063e-01 9.52411771e-01 -2.32468098e-01 2.28117511e-01
1.39447689e+00 -7.60741591e-01 -2.85989791e-01 -3.36758383e-02
1.03871465e+00 -3.43191251e-02 -1.63095784e+00 -4.23188657e-01
-5.61901808e-01 -9.74977255e-01 3.26889127e-01 -3.78324866e-01
-1.64744902e+00 2.81859308e-01 6.26126528e-01 4.41108346e-02
1.78410006e+00 -2.64150113e-01 1.21283042e+00 -2.66896307e-01
3.40390742e-01 -7.16021419e-01 -8.30935910e-02 2.51528859e-01
9.28340018e-01 -9.54609632e-01 2.46771649e-01 -4.69386697e-01
-5.20327866e-01 9.73591268e-01 6.49895489e-01 -2.57981479e-01
3.11453342e-01 1.87224552e-01 -1.29875675e-01 1.70174450e-01
-7.99081206e-01 6.80809766e-02 1.08521469e-01 3.88336122e-01
3.67160320e-01 -2.30862543e-01 -3.36645812e-01 6.79156840e-01
2.03457251e-01 2.64430881e-01 4.88429666e-01 6.52537763e-01
-4.96920735e-01 -8.03523064e-01 -5.40285468e-01 1.80677608e-01
-5.62009215e-01 -1.79461688e-01 3.56228262e-01 3.32540065e-01
1.10049345e-01 1.15146589e+00 7.82402754e-02 -4.33214217e-01
3.19796622e-01 -4.03584689e-01 2.27201596e-01 -1.10277708e-03
-4.80900317e-01 5.41035295e-01 -3.12597811e-01 -8.78739595e-01
-7.03194082e-01 -7.63147235e-01 -1.09771287e+00 -5.81812203e-01
-1.42502666e-01 4.21744026e-02 3.36563378e-03 9.72449005e-01
6.94550395e-01 7.30929315e-01 7.65946329e-01 -1.26365006e+00
-2.54142433e-01 -7.65091896e-01 -3.96551996e-01 3.92318368e-01
5.49105167e-01 -4.22993362e-01 -2.81396061e-01 3.76589954e-01] | [11.108933448791504, -1.9059652090072632] |
ca68d985-cc1c-4c18-bf8e-62f36029ec1e | we-learn-better-road-pothole-detection-from | 2008.06840 | null | https://arxiv.org/abs/2008.06840v2 | https://arxiv.org/pdf/2008.06840v2.pdf | We Learn Better Road Pothole Detection: from Attention Aggregation to Adversarial Domain Adaptation | Manual visual inspection performed by certified inspectors is still the main form of road pothole detection. This process is, however, not only tedious, time-consuming and costly, but also dangerous for the inspectors. Furthermore, the road pothole detection results are always subjective, because they depend entirely on the individual experience. Our recently introduced disparity (or inverse depth) transformation algorithm allows better discrimination between damaged and undamaged road areas, and it can be easily deployed to any semantic segmentation network for better road pothole detection results. To boost the performance, we propose a novel attention aggregation (AA) framework, which takes the advantages of different types of attention modules. In addition, we develop an effective training set augmentation technique based on adversarial domain adaptation, where the synthetic road RGB images and transformed road disparity (or inverse depth) images are generated to enhance the training of semantic segmentation networks. The experimental results demonstrate that, firstly, the transformed disparity (or inverse depth) images become more informative; secondly, AA-UNet and AA-RTFNet, our best performing implementations, respectively outperform all other state-of-the-art single-modal and data-fusion networks for road pothole detection; and finally, the training set augmentation technique based on adversarial domain adaptation not only improves the accuracy of the state-of-the-art semantic segmentation networks, but also accelerates their convergence. | ['Hengli Wang', 'Rui Fan', 'Mohammud J. Bocus', 'Ming Liu'] | 2020-08-16 | null | null | null | null | ['thermal-image-segmentation'] | ['computer-vision'] | [ 3.61093283e-01 1.54172080e-02 2.78778553e-01 -1.54689461e-01
-6.79842353e-01 -2.56794423e-01 1.67652428e-01 -2.28678748e-01
-5.00723183e-01 4.52032983e-01 -1.72931746e-01 -4.76373166e-01
6.65946901e-02 -1.33342409e+00 -6.89176083e-01 -8.13382387e-01
4.63806897e-01 2.20004156e-01 6.69577658e-01 -3.98712397e-01
2.07969472e-01 3.63059282e-01 -1.58231938e+00 -1.92899197e-01
1.52628446e+00 1.28532577e+00 3.35601062e-01 3.26537818e-01
-2.30811730e-01 5.66025257e-01 -5.46531677e-01 -3.74771893e-01
3.40609550e-01 -1.80286691e-01 -6.49595916e-01 8.21932629e-02
7.96166658e-02 -4.03278410e-01 -3.85330111e-01 1.28795767e+00
4.89026457e-01 2.86604553e-01 6.56720281e-01 -1.15455496e+00
-1.01674902e+00 2.20154330e-01 -9.52131152e-01 2.20690042e-01
-4.01548855e-02 3.20475876e-01 5.26881874e-01 -8.01351309e-01
8.74269083e-02 1.36739838e+00 6.55518949e-01 3.16003233e-01
-6.06776059e-01 -9.44660127e-01 1.83524847e-01 2.99101830e-01
-1.27012026e+00 -1.25269592e-01 1.02937150e+00 -2.31546104e-01
4.80767697e-01 -9.93420109e-02 6.82818890e-01 7.36114264e-01
5.65683190e-03 7.34972298e-01 1.15486062e+00 -1.23900212e-01
3.11467219e-02 1.08831421e-01 -1.96823865e-01 8.13522816e-01
3.07166129e-01 1.20447256e-01 1.07742831e-01 4.12310153e-01
1.05179775e+00 5.60602546e-02 -4.68621790e-01 -1.80853575e-01
-7.85255730e-01 8.92867804e-01 1.09552526e+00 2.41273902e-02
-3.69801879e-01 1.33613482e-01 4.33398664e-01 -5.76379485e-02
5.03565490e-01 1.78645879e-01 -2.25565299e-01 1.26255304e-01
-5.16028047e-01 -1.71145454e-01 1.42490387e-01 4.39029187e-01
1.20166898e+00 1.32006854e-01 1.45778090e-01 9.47520494e-01
4.99960244e-01 7.16704607e-01 3.92661780e-01 -8.60314369e-01
7.50249743e-01 7.60618865e-01 8.73551890e-02 -1.18298590e+00
-4.51110035e-01 -2.17593774e-01 -9.38866436e-01 4.30067033e-01
2.70318359e-01 -2.39431843e-01 -1.37608147e+00 1.40986037e+00
4.04426634e-01 2.62554020e-01 2.39987195e-01 8.55459213e-01
1.01000655e+00 7.13256180e-01 3.05463612e-01 9.43477526e-02
1.31238246e+00 -1.09164798e+00 -7.17200816e-01 -6.98889494e-01
2.24867046e-01 -5.48786581e-01 1.43922484e+00 2.07171798e-01
-7.19854593e-01 -7.17491269e-01 -1.13889384e+00 -2.03607321e-01
-6.85380280e-01 1.53958276e-01 7.13278890e-01 7.56517887e-01
-6.29770279e-01 3.59928906e-01 -6.95807457e-01 -1.07433293e-02
9.87116873e-01 1.07666135e-01 -3.58746916e-01 -2.43132249e-01
-1.44106710e+00 9.56088841e-01 4.42963570e-01 5.62234223e-01
-7.75021374e-01 -4.38678682e-01 -1.09761691e+00 -5.04225530e-02
4.49144989e-01 -4.52625811e-01 9.46182251e-01 -5.84157705e-01
-1.34429646e+00 7.22697377e-01 1.48532391e-01 -2.43015990e-01
6.04738295e-01 -2.01172099e-01 -4.58846182e-01 3.97943258e-01
3.02896351e-01 7.20402956e-01 9.14540887e-01 -1.44077992e+00
-8.13308001e-01 -4.45501238e-01 6.07437156e-02 2.99634099e-01
-1.47780240e-01 -3.17318290e-01 -4.26815510e-01 -4.82454866e-01
1.99613735e-01 -4.69955325e-01 -3.73499364e-01 1.72623008e-01
-5.10839522e-01 -2.00662076e-01 1.26600492e+00 -8.45009089e-01
9.05191958e-01 -2.26033187e+00 -2.81672448e-01 2.66977370e-01
5.89522645e-02 6.32777274e-01 -5.61197884e-02 -1.66458905e-01
-2.99008805e-02 3.52866888e-01 -8.95810843e-01 -1.02831997e-01
-2.64052004e-01 2.97142595e-01 -1.74258471e-01 3.36268902e-01
5.14093220e-01 8.74311864e-01 -9.40320492e-01 -6.68772042e-01
7.42128372e-01 4.94501889e-01 -1.13059543e-01 1.29424542e-01
1.52702957e-01 6.09551430e-01 -6.59695268e-01 7.23274767e-01
1.05529547e+00 9.92844626e-03 -6.63388848e-01 -2.59577930e-01
1.36717572e-03 -7.33663961e-02 -1.03985560e+00 1.56748271e+00
-7.88355052e-01 5.16970217e-01 -2.36735493e-01 -9.29259419e-01
1.29635966e+00 -1.49386264e-02 8.91731903e-02 -9.24548507e-01
3.19933325e-01 3.00213337e-01 -2.61355698e-01 -6.97676659e-01
4.09037471e-01 -1.54428095e-01 -1.72823071e-01 8.53434652e-02
-3.90995234e-01 -5.06413996e-01 -6.20507002e-02 -1.24696247e-01
6.07912302e-01 3.44974212e-02 -1.19349442e-01 2.38680050e-01
7.42302954e-01 -3.58795039e-02 6.66962445e-01 4.71427113e-01
-2.89533854e-01 7.98803687e-01 3.93451273e-01 -2.49895364e-01
-8.73652577e-01 -1.14521396e+00 -2.48724759e-01 5.84487855e-01
8.64207923e-01 5.61232865e-01 -9.49933469e-01 -7.37872064e-01
-6.87211379e-02 5.53906083e-01 -7.57317483e-01 -5.16189277e-01
-5.16084671e-01 -8.38159800e-01 7.25089192e-01 9.49297488e-01
1.47659135e+00 -1.22303283e+00 -5.49578786e-01 1.91694647e-01
-4.08566743e-01 -1.18815482e+00 -2.38150343e-01 -1.21350989e-01
-6.99691534e-01 -1.31866193e+00 -7.94666827e-01 -7.69334316e-01
6.82638705e-01 4.25597936e-01 5.92277110e-01 1.08459055e-01
-2.73802161e-01 -5.74689172e-03 -4.34427112e-01 -5.76206565e-01
-2.85396457e-01 -1.42084599e-01 -5.36276937e-01 1.77738145e-01
1.73453480e-01 -4.94287997e-01 -8.07858348e-01 4.26801652e-01
-9.06351328e-01 -5.33227995e-02 8.54211032e-01 8.56654227e-01
5.28411567e-01 4.38946694e-01 7.86721110e-01 -9.90217388e-01
5.27113497e-01 -2.59066015e-01 -4.80061918e-01 8.63967538e-02
-4.69952583e-01 -2.95267344e-01 5.73727667e-01 -1.47551462e-01
-1.45245278e+00 -6.35707453e-02 -4.76598024e-01 -5.21734297e-01
-2.24246368e-01 2.06348643e-01 -3.98222268e-01 -3.28940541e-01
5.52760482e-01 2.11811543e-01 8.30595791e-02 -3.50516468e-01
4.17226762e-01 8.34826946e-01 1.00471580e+00 -1.56504482e-01
1.22875512e+00 6.69692993e-01 -1.09834932e-01 -5.46273589e-01
-7.05431283e-01 -4.47622776e-01 -3.86493117e-01 -1.91665247e-01
1.10169673e+00 -9.68978047e-01 -5.72219968e-01 1.14495623e+00
-1.10868144e+00 -2.52167374e-01 -1.85120687e-01 1.85268193e-01
-2.33596802e-01 6.15620971e-01 -3.88630033e-01 -7.99092412e-01
-4.15827155e-01 -1.16081047e+00 1.06864262e+00 7.54301608e-01
5.81559420e-01 -8.90646696e-01 -4.77269888e-01 6.53906286e-01
3.76684755e-01 5.46327114e-01 9.00592625e-01 -1.79487363e-01
-6.01932645e-01 -3.05224031e-01 -8.96938920e-01 7.70457804e-01
1.63675070e-01 4.07583192e-02 -1.08551133e+00 2.52562732e-01
-2.57320523e-01 -2.05653727e-01 1.06851196e+00 3.69131505e-01
1.33803499e+00 7.19205439e-02 -2.84083635e-01 4.73805696e-01
1.35509086e+00 2.49577671e-01 1.11906815e+00 4.27515864e-01
1.11024618e+00 7.00305045e-01 1.09490550e+00 3.12238242e-02
5.71939290e-01 2.26789713e-01 8.83027494e-01 -7.58412421e-01
-2.52465069e-01 -3.88146877e-01 1.02785110e-01 3.17172348e-01
-2.39197314e-01 -2.97299832e-01 -8.81445169e-01 8.29469800e-01
-1.60370147e+00 -6.93121791e-01 -3.66416872e-01 2.08002949e+00
5.09917498e-01 2.21235886e-01 -2.65205875e-02 4.59668815e-01
9.51038659e-01 2.31942356e-01 -7.46531725e-01 -3.09904605e-01
-5.40216491e-02 4.16044712e-01 7.77904689e-01 2.28468210e-01
-1.30423450e+00 9.71200883e-01 5.01121950e+00 1.00981581e+00
-1.07589924e+00 2.82068521e-01 6.56482995e-01 7.25706935e-01
-3.41249555e-01 -2.33853191e-01 -3.10495883e-01 6.46604419e-01
3.20025086e-01 1.91028461e-01 1.80830687e-01 8.74161243e-01
1.55780092e-01 -2.67484158e-01 -4.09086078e-01 7.86013186e-01
-2.01917365e-01 -9.94105160e-01 -9.63417143e-02 -1.71772808e-01
6.81676865e-01 1.00461520e-01 2.31529340e-01 3.67689848e-01
3.28328699e-01 -9.19607460e-01 4.75396484e-01 1.30005315e-01
7.93819427e-01 -9.73945498e-01 1.12565613e+00 1.65605575e-01
-1.43401122e+00 -4.46231723e-01 -4.58986312e-01 2.50202119e-01
3.68940979e-01 4.98713851e-01 -4.17658120e-01 7.10753918e-01
9.89116848e-01 8.02189827e-01 -6.33355260e-01 1.14428031e+00
-8.24462771e-01 2.81925559e-01 -2.69626558e-01 2.54874051e-01
5.67638934e-01 -2.75748849e-01 2.13992685e-01 8.74477684e-01
2.63699859e-01 1.18631544e-02 9.50674620e-03 9.69912231e-01
1.29437111e-02 -2.49469951e-02 -6.33820295e-01 4.16364491e-01
5.64707994e-01 1.15118527e+00 -7.60374427e-01 -1.22432359e-01
-3.47735554e-01 7.92341590e-01 -1.06466979e-01 4.62388575e-01
-1.07560039e+00 -9.14699912e-01 5.54121077e-01 -1.08229801e-01
3.01420063e-01 2.63161093e-01 -4.06241655e-01 -8.88101995e-01
1.42933363e-02 -4.58734334e-01 1.56554639e-01 -9.57317531e-01
-1.14503825e+00 5.34286082e-01 -2.61607885e-01 -1.35680962e+00
2.67909855e-01 -6.04340792e-01 -8.59398127e-01 8.04443657e-01
-2.09747481e+00 -1.14284587e+00 -9.88191605e-01 6.30132020e-01
4.28374141e-01 2.10706577e-01 3.80770177e-01 6.41682386e-01
-8.93400609e-01 5.02542496e-01 -1.70306578e-01 4.68052059e-01
3.12719196e-01 -9.99692380e-01 4.52755451e-01 1.09594560e+00
-4.04265016e-01 4.15148884e-02 3.07020515e-01 -3.03954393e-01
-7.17352748e-01 -1.52759731e+00 7.16080293e-02 -1.12852745e-01
4.19561207e-01 8.12484398e-02 -1.17664683e+00 4.36135858e-01
-2.36337960e-01 8.82963762e-02 8.62186477e-02 -3.92751902e-01
-1.49012089e-01 -3.68849695e-01 -1.44093513e+00 4.46781367e-01
8.38525653e-01 -5.81763268e-01 -6.80247784e-01 2.57009655e-01
1.00657034e+00 -5.74654460e-01 -6.98414207e-01 6.81445897e-01
2.18520567e-01 -1.11175370e+00 1.20870149e+00 1.39507890e-01
5.78816831e-01 -6.95264459e-01 1.78911194e-01 -1.18310761e+00
7.32464865e-02 -7.16510043e-02 4.36628252e-01 1.34695733e+00
2.45257854e-01 -1.06264436e+00 7.96200156e-01 3.28723788e-01
-5.98068058e-01 -6.41044497e-01 -1.05634987e+00 -5.88798642e-01
6.62010759e-02 -4.87015635e-01 7.50084579e-01 7.96112061e-01
-6.49602652e-01 7.95989931e-02 -2.04152852e-01 5.25156796e-01
5.34183741e-01 -2.32014023e-02 6.23186588e-01 -1.25589466e+00
8.10195655e-02 -4.25307542e-01 -4.45876867e-01 -1.06828892e+00
1.40400846e-02 -4.33674574e-01 2.97942370e-01 -1.79784358e+00
-3.04782718e-01 -6.77806795e-01 -3.40850949e-01 7.05371916e-01
-4.73691374e-01 5.52380502e-01 -3.77583317e-02 -9.10385698e-02
-2.18946084e-01 7.86597490e-01 1.79812896e+00 -3.33662748e-01
-2.00168401e-01 1.21481158e-01 -8.33396614e-01 9.18160856e-01
8.10872912e-01 -3.09500188e-01 -5.54390013e-01 -5.02920151e-01
-3.97978649e-02 -1.75913144e-02 7.47263372e-01 -1.15202320e+00
9.61615294e-02 7.78110623e-02 2.38002613e-01 -7.13871241e-01
3.09167206e-01 -8.74638855e-01 -2.97666579e-01 6.15350544e-01
3.59325349e-01 -3.45366210e-01 3.41183901e-01 8.04044783e-01
-3.78164709e-01 -2.05550075e-01 1.11446822e+00 -2.08297536e-01
-1.13905728e+00 3.52518469e-01 -1.83597893e-01 1.15949780e-01
1.05369735e+00 -6.79275095e-01 -5.58549583e-01 -1.31470338e-01
-3.03806305e-01 4.70858186e-01 3.29056710e-01 4.06553298e-01
8.72392118e-01 -1.09118819e+00 -5.09589076e-01 2.41653636e-01
1.42484546e-01 8.19346607e-01 6.75827980e-01 7.68710732e-01
-8.18692684e-01 -1.79117009e-01 -3.13998252e-01 -5.89329422e-01
-7.17902541e-01 3.91779065e-01 3.30695063e-01 -1.28558651e-01
-5.87584436e-01 9.11778331e-01 3.60073835e-01 -5.59673488e-01
-1.17188491e-01 -3.55738759e-01 -4.25501019e-01 2.93582957e-03
3.47193182e-01 5.51086843e-01 -3.24885733e-02 -4.99336034e-01
-2.31055617e-01 1.16336656e+00 2.15259101e-02 2.92978972e-01
1.07793367e+00 -2.93410450e-01 9.74133462e-02 2.09392495e-02
9.37015593e-01 -3.98329854e-01 -1.40864718e+00 -7.49319270e-02
-6.44390941e-01 -6.30280495e-01 2.73386896e-01 -7.50088632e-01
-1.57358563e+00 1.33867335e+00 8.14493179e-01 1.95960179e-01
1.49044931e+00 -1.63960814e-01 1.19035745e+00 1.45660967e-01
4.21949804e-01 -9.99188364e-01 1.33983642e-01 1.30998224e-01
5.55831432e-01 -1.51704252e+00 -1.63849905e-01 -8.14314604e-01
-7.24414468e-01 8.25924814e-01 8.24468970e-01 -2.19406798e-01
3.39256614e-01 -9.93074775e-02 3.80957484e-01 -1.88772604e-01
7.42372051e-02 -4.02395070e-01 1.34682953e-01 1.05217361e+00
-2.49646813e-01 -8.95114988e-02 9.75099653e-02 4.36109394e-01
7.04124048e-02 -2.12194443e-01 5.87354541e-01 4.68946308e-01
-6.66850209e-01 -8.23160827e-01 -4.52227741e-01 3.93091470e-01
-2.46090040e-01 -7.02760518e-02 9.54373088e-03 1.12413347e+00
5.00572324e-01 1.05118811e+00 1.75999045e-01 -3.98373932e-01
6.16207421e-01 -4.75742549e-01 5.72569817e-02 -5.20935953e-01
-2.94957876e-01 -2.62069255e-01 -1.27666950e-01 -2.81909138e-01
-4.18704897e-01 -3.82288069e-01 -1.65135610e+00 -2.70702690e-01
-6.33708358e-01 2.49812435e-02 5.29864013e-01 9.81751800e-01
8.31386372e-02 7.49892652e-01 9.67092156e-01 -8.16543937e-01
-7.24135712e-02 -9.49664772e-01 -6.34942114e-01 4.16116446e-01
3.13590586e-01 -1.09017265e+00 -5.37701488e-01 -2.01157913e-01] | [9.571032524108887, 0.06574205309152603] |
70b140e0-40e5-4ce3-b5c3-930f3d520c42 | improving-continual-relation-extraction | 2210.04513 | null | https://arxiv.org/abs/2210.04513v1 | https://arxiv.org/pdf/2210.04513v1.pdf | Improving Continual Relation Extraction through Prototypical Contrastive Learning | Continual relation extraction (CRE) aims to extract relations towards the continuous and iterative arrival of new data, of which the major challenge is the catastrophic forgetting of old tasks. In order to alleviate this critical problem for enhanced CRE performance, we propose a novel Continual Relation Extraction framework with Contrastive Learning, namely CRECL, which is built with a classification network and a prototypical contrastive network to achieve the incremental-class learning of CRE. Specifically, in the contrastive network a given instance is contrasted with the prototype of each candidate relations stored in the memory module. Such contrastive learning scheme ensures the data distributions of all tasks more distinguishable, so as to alleviate the catastrophic forgetting further. Our experiment results not only demonstrate our CRECL's advantage over the state-of-the-art baselines on two public datasets, but also verify the effectiveness of CRECL's contrastive learning on improving CRE performance. | ['Yanghua Xiao', 'Zhen Chen', 'Haoliang Jin', 'Deqing Yang', 'Chengwei Hu'] | 2022-10-10 | null | https://aclanthology.org/2022.coling-1.163 | https://aclanthology.org/2022.coling-1.163.pdf | coling-2022-10 | ['continual-relation-extraction'] | ['natural-language-processing'] | [ 1.97128087e-01 3.55680555e-01 -2.54883409e-01 -1.86148092e-01
-3.81326973e-01 -2.13174313e-01 7.23963141e-01 3.42599899e-01
-4.87249076e-01 8.74542058e-01 -3.14283781e-02 -4.07003105e-01
-3.89027089e-01 -8.32105279e-01 -5.88212132e-01 -4.35419232e-01
-9.31137651e-02 7.78399765e-01 5.65873682e-01 -3.59674156e-01
-1.78586960e-01 4.03733432e-01 -1.47237003e+00 5.47230244e-01
1.07605660e+00 1.29932439e+00 -1.62291015e-03 2.02861905e-01
6.92725107e-02 1.18821275e+00 -5.31428695e-01 -8.04091573e-01
3.67390290e-02 -1.41276434e-01 -1.02843034e+00 -2.72571027e-01
1.56286821e-01 1.65232509e-01 -6.42339528e-01 7.33790636e-01
3.51196200e-01 2.82421529e-01 6.33939624e-01 -1.47604275e+00
-8.45649421e-01 1.05233848e+00 -5.76466322e-01 5.13628721e-01
1.73889846e-01 8.41025263e-02 1.12889886e+00 -1.33869195e+00
5.69505751e-01 1.13126361e+00 7.87851930e-01 3.38342726e-01
-1.25707364e+00 -1.00224781e+00 5.17024755e-01 3.77117962e-01
-1.56485689e+00 -8.51425588e-01 6.73481047e-01 -2.69284487e-01
1.26361346e+00 4.07749832e-01 4.41426873e-01 1.14175451e+00
1.47491246e-01 9.19344127e-01 8.37978423e-01 -3.53864968e-01
8.87238085e-02 6.46253005e-02 6.61770165e-01 6.14683390e-01
4.27081823e-01 1.15011800e-02 -8.92646372e-01 -2.09678616e-03
3.13560307e-01 -7.46109616e-03 -3.37477028e-01 -3.61119449e-01
-9.58519876e-01 1.80790275e-01 5.28773367e-01 3.16882998e-01
-1.63304985e-01 -4.15595084e-01 3.53228539e-01 5.37710786e-01
6.11339808e-01 8.25057328e-01 -8.49724948e-01 1.67199925e-01
-6.08517885e-01 2.04814628e-01 6.46753669e-01 1.13813376e+00
6.11907303e-01 -5.33850849e-01 -5.40352046e-01 7.32897520e-01
-2.13470608e-01 -5.05304784e-02 6.26506269e-01 -3.87327224e-01
8.46089482e-01 1.17627454e+00 -2.65451223e-01 -9.50632036e-01
-6.38401985e-01 -9.03277755e-01 -1.20856762e+00 -2.06294358e-01
2.58261144e-01 -1.07591609e-02 -9.15188432e-01 1.70208681e+00
2.44758204e-01 2.02414945e-01 2.36124203e-01 3.66233587e-01
1.03276110e+00 2.48273239e-01 3.04672092e-01 -6.98048830e-01
1.24391496e+00 -1.06828260e+00 -7.58522987e-01 -6.21642411e-01
6.33173525e-01 -3.82176191e-01 1.09915066e+00 3.74127597e-01
-8.76617193e-01 -7.71292329e-01 -1.24263203e+00 -2.61598676e-01
-5.35880208e-01 -1.68367419e-02 9.96101201e-01 2.82500863e-01
-4.89106387e-01 6.14568949e-01 -7.43657529e-01 -6.81547390e-05
7.11251259e-01 2.52418071e-01 -5.38170695e-01 4.12086509e-02
-1.52635133e+00 9.85829234e-01 1.02494586e+00 1.36584237e-01
-4.70839560e-01 -8.90379906e-01 -8.34881902e-01 3.08063835e-01
1.10100281e+00 -5.65167785e-01 1.15039706e+00 -3.09457630e-01
-8.74610424e-01 7.49798477e-01 -2.00239986e-01 -6.01396918e-01
5.43066084e-01 -6.26885593e-01 -6.18840098e-01 -4.47746038e-01
8.80377740e-03 2.36733600e-01 7.25014687e-01 -1.36552417e+00
-8.64604890e-01 -2.72585839e-01 -9.44063738e-02 1.90467268e-01
-5.16027808e-01 -4.36535776e-01 -5.52252948e-01 -8.09367895e-01
3.49500179e-01 -6.36039734e-01 1.85768977e-01 -4.56635505e-01
-5.37841082e-01 -8.56271446e-01 8.47772181e-01 -4.44430411e-01
1.75922179e+00 -2.16898990e+00 2.14168593e-01 -7.97754154e-02
8.62765372e-01 4.54916447e-01 -1.27285570e-01 2.29238365e-02
-3.76356661e-01 -2.62366720e-02 -2.33808026e-01 -5.15718699e-01
-4.31376189e-01 5.77353910e-02 -3.18684310e-01 -6.96494207e-02
5.33163428e-01 1.23438382e+00 -1.20200813e+00 -5.95864117e-01
-4.13454175e-01 -3.24533172e-02 -1.05830982e-01 3.64344329e-01
-2.29091078e-01 5.48607036e-02 -5.19990176e-02 6.12698019e-01
7.18743742e-01 -4.07748401e-01 3.78405035e-01 -3.25712830e-01
2.60221928e-01 4.84858036e-01 -1.08063495e+00 1.46176112e+00
-1.62417516e-01 1.86732352e-01 -5.81443727e-01 -8.43693435e-01
1.00709391e+00 1.76567748e-01 2.39380464e-01 -9.32146788e-01
-9.03905928e-02 2.50858843e-01 2.73597747e-01 -4.44303960e-01
5.23748934e-01 3.19519751e-02 -1.47301983e-02 3.15765709e-01
1.98824584e-01 2.38555670e-01 3.47334236e-01 5.70802391e-01
1.28974819e+00 1.71705931e-01 4.68328685e-01 -1.28066957e-01
4.69155580e-01 -2.43852898e-01 1.11870885e+00 8.96161854e-01
-1.40735626e-01 2.03038678e-01 5.70729136e-01 -6.17049873e-01
-4.57075417e-01 -9.84300196e-01 -8.92418716e-03 1.16850042e+00
2.83980012e-01 -6.85126185e-01 -1.25525638e-01 -1.42393994e+00
1.85286686e-01 7.33279765e-01 -6.27482057e-01 -7.36079097e-01
-8.05988729e-01 -9.85538721e-01 1.99815139e-01 6.01428151e-01
8.88881922e-01 -1.12629676e+00 -2.24517524e-01 2.77607769e-01
-3.56630236e-01 -9.82388914e-01 -4.14996058e-01 6.75000370e-01
-6.61188185e-01 -1.29293668e+00 -6.75575361e-02 -9.42226171e-01
5.00063241e-01 4.57132459e-01 1.37461436e+00 4.95159179e-02
2.41080187e-02 -3.65658909e-01 -5.01917303e-01 -3.62539500e-01
-1.98498607e-01 9.55323637e-01 1.99475840e-01 -1.76520005e-01
6.13511443e-01 -9.34196770e-01 -3.93248141e-01 1.83347583e-01
-6.43214345e-01 2.00755358e-01 8.08657765e-01 1.08263385e+00
4.02987093e-01 5.85465848e-01 9.32002306e-01 -1.60614777e+00
8.13388884e-01 -6.57684922e-01 -1.83711782e-01 6.48216188e-01
-1.30842233e+00 2.08436117e-01 5.88864267e-01 -6.07071996e-01
-1.29859710e+00 -8.33755136e-02 2.42147997e-01 -6.86035082e-02
2.74386317e-01 5.42429030e-01 -4.34548646e-01 3.51503164e-01
9.76114988e-01 1.37456670e-01 -3.03672343e-01 -4.55604970e-01
3.41449946e-01 5.53205311e-01 1.01433218e+00 -5.95399082e-01
9.22126174e-01 -1.04745030e-01 -4.95453998e-02 -4.78447564e-02
-1.44883144e+00 -3.08979869e-01 -9.11706150e-01 1.22893073e-01
1.94302946e-01 -9.04176593e-01 -7.11231470e-01 7.00393736e-01
-1.24303186e+00 -3.31580281e-01 -4.37848032e-01 -1.94782000e-02
-4.21087332e-02 1.70017052e-02 -7.05788255e-01 -8.12629879e-01
-4.52064663e-01 -6.98963106e-01 6.30249262e-01 3.69730353e-01
-2.84760505e-01 -6.82187915e-01 -1.03938557e-01 2.18561009e-01
4.65245135e-02 1.25057315e-02 1.29174399e+00 -8.37402403e-01
-4.52275127e-01 -3.13508883e-02 -3.59548360e-01 1.37402162e-01
3.79461288e-01 -4.67689931e-01 -1.19756782e+00 -3.93741310e-01
-7.14769736e-02 -5.34405231e-01 1.28683579e+00 -4.24794465e-01
1.07880080e+00 -3.15169930e-01 -8.50014210e-01 5.74930787e-01
9.46620286e-01 5.26025474e-01 7.04363227e-01 3.17060590e-01
8.57226610e-01 5.47957540e-01 9.57043409e-01 1.13130836e-02
6.28419936e-01 4.50706720e-01 9.33885053e-02 -8.95351022e-02
-3.55041593e-01 -4.52278167e-01 -1.46878973e-01 9.53759134e-01
-6.92136660e-02 -2.44605437e-01 -8.97732496e-01 4.68993813e-01
-2.11949897e+00 -7.59603322e-01 1.87958986e-01 2.18673468e+00
1.42309153e+00 6.82226956e-01 -1.10835977e-01 5.00759721e-01
5.85490227e-01 1.00070849e-01 -6.43808186e-01 1.11939646e-01
-3.19667131e-01 3.76386434e-01 -4.15968383e-03 2.46396527e-01
-1.44080997e+00 1.16743171e+00 5.48722172e+00 9.54250276e-01
-7.13399708e-01 1.86940297e-01 8.42202067e-01 1.40186131e-01
-8.25032294e-02 9.79797095e-02 -9.65852439e-01 2.06128523e-01
5.61578572e-01 -1.48359522e-01 1.61214471e-01 7.03981102e-01
-6.37746990e-01 -2.51481622e-01 -1.52017629e+00 1.00880575e+00
-2.99416799e-02 -1.22335541e+00 1.88868418e-02 -3.11680257e-01
5.92276573e-01 -4.74936754e-01 9.44407657e-02 9.44759667e-01
3.72961134e-01 -9.11274195e-01 5.22944212e-01 6.33195817e-01
8.19680214e-01 -7.78050184e-01 8.39495301e-01 3.59134883e-01
-1.41292119e+00 -2.00845346e-01 -1.81283519e-01 -2.60913640e-01
-1.71847299e-01 9.27949369e-01 -7.52845109e-01 8.65623951e-01
6.62782907e-01 7.45512307e-01 -1.13204551e+00 7.34363854e-01
-3.56901705e-01 3.57594192e-01 8.12887028e-02 2.88214594e-01
-3.47235650e-01 3.59827489e-01 3.48507166e-01 1.15403664e+00
-9.53836739e-02 2.10788444e-01 1.40220121e-01 6.01630986e-01
-4.10208732e-01 2.27535539e-03 -4.14653093e-01 3.14357318e-02
1.10343790e+00 1.24390388e+00 -5.42285919e-01 -3.27853799e-01
-1.49613276e-01 8.48523915e-01 1.14291525e+00 1.32515192e-01
-5.43989599e-01 -6.86222315e-01 3.84239525e-01 3.26329917e-02
4.89494987e-02 -1.67872813e-02 -4.75308269e-01 -1.16599023e+00
3.25132519e-01 -9.55830336e-01 8.26049387e-01 -3.56864512e-01
-1.48572099e+00 9.30335999e-01 -1.20589510e-01 -8.21855962e-01
7.02497289e-02 -2.49661267e-01 -3.70761216e-01 7.52011418e-01
-1.34778166e+00 -1.32132256e+00 -4.31648254e-01 3.80464643e-01
4.80718970e-01 -4.15483356e-01 6.90306664e-01 4.49456722e-01
-9.95096743e-01 1.20715642e+00 -4.53993201e-01 1.34137809e-01
8.78290296e-01 -1.28699386e+00 6.81334376e-01 8.94320071e-01
2.32762173e-01 7.83660054e-01 3.24234813e-01 -1.07654417e+00
-9.55501974e-01 -1.26589346e+00 1.23156095e+00 -6.83666825e-01
5.50730824e-01 -5.94870806e-01 -1.19504285e+00 8.96196544e-01
-1.34660736e-01 1.68405876e-01 3.33224446e-01 8.22098374e-01
-8.07937026e-01 -5.97260535e-01 -8.37396502e-01 6.36225820e-01
1.68345654e+00 -4.08616036e-01 -8.96756411e-01 2.41824299e-01
1.13589942e+00 -4.40652370e-01 -8.20079386e-01 8.92261803e-01
4.53981102e-01 -7.24277914e-01 9.13841426e-01 -7.02522993e-01
3.69290739e-01 -1.75504684e-01 4.28254187e-01 -1.27640617e+00
-6.80782735e-01 -7.75844634e-01 -8.58540177e-01 1.74284172e+00
6.66516423e-01 -7.44755030e-01 5.77828288e-01 4.60175127e-01
-1.58765782e-02 -9.51131105e-01 -7.94388533e-01 -9.76823270e-01
-6.92296401e-02 -2.97291219e-01 8.25202763e-01 1.09750640e+00
1.38050482e-01 1.13756824e+00 -4.22812611e-01 8.75314772e-02
3.86937737e-01 9.21999514e-02 6.72034860e-01 -1.63121355e+00
-3.92342150e-01 -2.02520549e-01 -1.56155527e-01 -8.80507588e-01
1.65479943e-01 -1.08383954e+00 6.29193783e-02 -1.10874534e+00
6.04812205e-01 -1.08269536e+00 -7.61625588e-01 8.03790212e-01
-9.41260219e-01 -2.74760816e-02 -2.76400764e-02 4.68523234e-01
-9.30715382e-01 6.04698479e-01 1.25760913e+00 -2.94354051e-01
-5.81939697e-01 5.32980375e-02 -1.12774634e+00 3.55961889e-01
4.83737856e-01 -6.52786553e-01 -7.74756014e-01 -1.80716082e-01
4.58330691e-01 -2.87202477e-01 -1.34852529e-02 -1.02310824e+00
6.08321190e-01 -6.10853396e-02 4.27284837e-01 -6.27224147e-01
1.58596579e-02 -4.37092811e-01 8.53012502e-02 3.03768307e-01
-3.53811681e-01 1.16565432e-02 1.18584387e-01 5.80274701e-01
-8.21505040e-02 1.07914694e-01 5.63449740e-01 3.09267901e-02
-6.97244525e-01 3.70313793e-01 3.14608157e-01 3.80498976e-01
9.77518678e-01 2.78931081e-01 -9.55406487e-01 9.61113721e-02
-8.25787485e-01 5.80473661e-01 -1.59433693e-01 6.03733659e-01
6.60935581e-01 -1.42394984e+00 -6.57407165e-01 2.07154036e-01
4.20440227e-01 6.72830045e-01 7.20464140e-02 7.62116313e-01
1.21585079e-01 2.03497991e-01 1.73753574e-01 -1.39010891e-01
-1.10957098e+00 1.01718104e+00 2.39804506e-01 -8.47702861e-01
-7.00740099e-01 1.18427145e+00 1.92513898e-01 -2.38139868e-01
3.77922207e-01 -1.35724753e-01 -3.27984899e-01 2.08258182e-01
5.87405562e-01 2.79226363e-01 4.73069906e-01 2.26979516e-02
-4.24254596e-01 -1.51655778e-01 -8.92730713e-01 1.82654008e-01
1.34390092e+00 -2.48225946e-02 -3.24800342e-01 6.33093476e-01
5.62047660e-01 -1.20421946e-01 -1.16915703e+00 -9.09846544e-01
5.96453905e-01 -1.09571092e-01 -2.92612493e-01 -1.01686406e+00
-1.03693271e+00 3.02958697e-01 3.01037312e-01 1.69071719e-01
1.14724326e+00 1.37322351e-01 8.95292699e-01 5.76785207e-01
5.62515080e-01 -9.40450132e-01 2.40187600e-01 7.20366061e-01
8.36445093e-01 -1.25951374e+00 2.40229994e-01 -7.85371721e-01
-4.39650357e-01 7.72568166e-01 1.19137216e+00 2.79412180e-01
5.99274278e-01 4.81942594e-01 -1.66341588e-01 -2.83654600e-01
-1.08792257e+00 -3.06651920e-01 4.06167537e-01 6.98162019e-01
3.16533625e-01 -1.00307159e-01 -3.62931103e-01 1.02936673e+00
-1.65224046e-01 -1.14763826e-01 -6.06837031e-03 1.12560534e+00
-9.23033357e-02 -1.09182775e+00 1.80827290e-01 6.73760295e-01
4.46368679e-02 -4.28084344e-01 -5.32964170e-01 7.47035027e-01
6.15836620e-01 9.90708888e-01 2.64306599e-03 -8.65251839e-01
6.04363799e-01 4.47026908e-01 3.60332042e-01 -8.49041879e-01
-6.46948040e-01 -3.69267613e-01 2.92323321e-01 -2.56061971e-01
2.63379179e-02 -5.02266645e-01 -1.08318853e+00 -1.30856812e-01
-4.95890349e-01 2.77807564e-01 -2.92465955e-01 1.09096158e+00
4.36566442e-01 9.71349537e-01 5.38201213e-01 -8.00650865e-02
-3.94307226e-01 -1.24251604e+00 -4.05049294e-01 5.39878666e-01
2.08862528e-01 -1.08781767e+00 -1.17348425e-01 -1.09865882e-01] | [9.19079303741455, 8.536351203918457] |
9b2f6c9d-88ee-4458-8b82-534ba1e82b5a | audio-guided-attention-network-for-weakly | null | null | https://ieeexplore.ieee.org/document/9712793 | https://ieeexplore.ieee.org/document/9712793 | Audio-Guided Attention Network for Weakly Supervised Violence Detection | Detecting violence in video is a challenging task due to its complex scenarios and great intra-class variability. Most previous works specialize in the analysis of appearance or motion information, ignoring the co-occurrence of some audio and visual events. Physical conflicts such as abuse and fighting are usually accompanied by screaming, while crowd violence such as riots and wars are generally related to gunshots and explosions. Therefore, we propose a novel audio-guided multimodal violence detection framework. First, deep neural networks are used to extract appearance and audio features, respectively. Then, a Cross-Modal Awareness Local-Arousal (CMA-LA) network is proposed for cross-modal interaction, which implements audio-to-visual feature enhancement over temporal dimension. The enhanced features are then fed into a multilayer perceptron (MLP) to capture high-level semantics, followed by a temporal convolution layer to obtain high-confidence violence scores. To validate the proposed method, we conduct experiments on a large violent video dataset, XD Violence. Comprehensive experiments demonstrate the robust performance of our approach, which also achieves a new state-of-the-art AP result. | ['Xiaoyu Wu', 'Yujiang Pu'] | 2022-02-21 | null | null | null | conference-2022-2 | ['anomaly-detection-in-surveillance-videos', 'anomaly-detection-in-surveillance-videos'] | ['computer-vision', 'methodology'] | [ 1.94117710e-01 -6.19155705e-01 2.02973172e-01 -2.55827785e-01
-8.20164979e-01 -1.97291076e-01 6.04959428e-01 2.71364927e-01
-5.72774589e-01 4.41182405e-01 4.62068945e-01 2.26175934e-01
-1.00466385e-01 -6.71161473e-01 -4.95168507e-01 -7.54609466e-01
-3.36463094e-01 -3.65651131e-01 2.10109726e-01 -1.72381744e-01
1.80251569e-01 4.30326372e-01 -1.46029246e+00 5.42613804e-01
3.20915520e-01 1.23066521e+00 -9.73895267e-02 7.48140156e-01
3.40820216e-02 1.03815758e+00 -6.50616467e-01 -5.84629059e-01
-2.39153311e-01 -3.60973030e-01 -5.29523969e-01 -6.96161538e-02
1.09218448e-01 -5.78504026e-01 -7.18068004e-01 1.01003695e+00
7.34095693e-01 4.02239531e-01 5.89186907e-01 -1.40696049e+00
-1.53502747e-01 5.20581305e-01 -8.80547285e-01 5.84971070e-01
7.66697168e-01 2.62310207e-01 6.93020344e-01 -9.27852571e-01
1.70485750e-01 1.44117200e+00 6.08532131e-01 2.45564356e-01
-8.27706933e-01 -9.74708676e-01 1.68769434e-01 7.07861304e-01
-1.49201119e+00 -4.58820879e-01 1.20049703e+00 -4.39702153e-01
7.74187088e-01 2.12059751e-01 6.79877758e-01 1.83310723e+00
1.28346309e-02 7.18872070e-01 5.38462341e-01 6.10118173e-02
-3.98209654e-02 -3.12873513e-01 -2.18490288e-01 4.59555626e-01
-2.53162712e-01 -3.18719563e-03 -6.78217053e-01 -2.14763582e-01
4.22433466e-01 2.57449687e-01 -2.13726357e-01 2.93694049e-01
-1.08579934e+00 7.85891116e-01 3.75028163e-01 3.41838777e-01
-6.94338679e-01 2.27498218e-01 9.88473713e-01 -3.11537981e-02
1.81025088e-01 -4.60949354e-02 2.93106567e-02 -5.80480993e-01
-8.47707152e-01 2.37010434e-01 1.02884620e-01 1.98970452e-01
1.79534584e-01 1.34142876e-01 -3.62466216e-01 1.04564202e+00
2.33179271e-01 3.63298118e-01 2.91611642e-01 -8.07671309e-01
6.97268426e-01 1.41121149e-01 -3.08439136e-01 -1.80751181e+00
-7.41567969e-01 -2.39032283e-01 -1.04221833e+00 9.24902037e-02
-7.15458915e-02 -2.60251135e-01 -3.40039879e-01 1.87073076e+00
4.21420753e-01 6.50161147e-01 7.02272356e-03 1.13122416e+00
1.12208378e+00 7.81291246e-01 4.76709634e-01 -6.32416189e-01
1.40301359e+00 -5.45959651e-01 -7.86929309e-01 6.11972250e-02
1.03132151e-01 -5.25818110e-01 7.73755312e-01 5.27617395e-01
-9.10763144e-01 -5.31489730e-01 -6.71087980e-01 1.84490085e-01
3.17118801e-02 -1.92399368e-01 2.74446160e-01 2.75595933e-01
-3.23989123e-01 4.44540471e-01 -7.50236809e-01 -2.80485094e-01
5.57649851e-01 -2.09777150e-02 -5.16225994e-01 1.96038380e-01
-1.47869635e+00 5.22884309e-01 5.55796981e-01 3.97434175e-01
-9.94234562e-01 -2.54917443e-01 -9.61185753e-01 -4.59608175e-02
3.29634547e-01 -2.06859007e-01 8.55487525e-01 -8.76573563e-01
-1.22417843e+00 4.70399350e-01 9.88894477e-02 -8.13397244e-02
4.76251423e-01 -2.69674689e-01 -8.64590704e-01 6.19093060e-01
-5.99865317e-02 3.74746263e-01 8.92663836e-01 -1.05816889e+00
-7.12294817e-01 -3.47538114e-01 2.13410795e-01 3.53044659e-01
-8.63376558e-01 6.07061386e-01 -5.09900630e-01 -8.73281360e-01
-5.08038290e-02 -4.00949597e-01 1.68180078e-01 -1.59054488e-01
-4.21675891e-01 -7.31703639e-02 8.26917350e-01 -9.51679051e-01
1.58141208e+00 -2.47114325e+00 2.65789777e-01 1.40280738e-01
1.89758435e-01 3.46526690e-02 -8.37643743e-02 3.78676295e-01
-1.89677626e-01 -2.25604311e-01 -1.68086663e-01 -2.08442092e-01
-2.18238413e-01 -4.42637391e-02 -3.00047874e-01 5.15130043e-01
3.28085631e-01 3.75071883e-01 -9.12212968e-01 -8.99188995e-01
2.80834079e-01 6.44329488e-01 -4.15783107e-01 2.31381774e-01
2.92245209e-01 5.85738242e-01 -3.61554831e-01 7.10020840e-01
6.44453883e-01 3.04419696e-01 -2.15948686e-01 -2.72838563e-01
-2.35156845e-02 -1.48863211e-01 -1.16735828e+00 1.51297927e+00
-2.31514707e-01 5.67160070e-01 1.86094239e-01 -1.08086371e+00
7.59428203e-01 6.27585053e-01 5.28735876e-01 -8.50132585e-01
4.83529210e-01 -1.16961867e-01 -2.23401397e-01 -1.21080720e+00
2.52197623e-01 -9.86349955e-02 -2.73430109e-01 3.97640876e-02
-7.53669441e-02 3.98829192e-01 2.41686720e-02 9.00403336e-02
1.33730829e+00 -1.97109178e-01 6.34502321e-02 4.42933142e-01
7.20501900e-01 -4.88949120e-01 6.82966352e-01 3.82815391e-01
-4.64263052e-01 6.52093530e-01 5.44302821e-01 -2.94957578e-01
-5.88517249e-01 -1.19866991e+00 -1.90717110e-03 1.05873513e+00
3.76801878e-01 -4.83074039e-01 -6.03615046e-01 -5.47592819e-01
-4.15170789e-01 5.09734929e-01 -6.43212318e-01 -5.10102570e-01
-5.51379502e-01 -6.55927241e-01 8.32541108e-01 5.89951277e-01
6.47407770e-01 -1.42912698e+00 -6.50972009e-01 2.42746115e-01
-7.00962901e-01 -1.39598465e+00 -2.16923639e-01 -3.42949271e-01
-3.46224934e-01 -9.28510427e-01 -5.73328555e-01 -5.07019222e-01
1.34944379e-01 1.04283936e-01 4.15105104e-01 -6.31022155e-02
-4.02890295e-01 3.11973482e-01 -4.89086837e-01 -1.70897335e-01
4.06248309e-02 -2.85601318e-01 2.83804417e-01 4.07147467e-01
3.41698378e-01 -9.27899897e-01 -5.95177650e-01 -9.25281793e-02
-8.71803761e-01 -1.48743331e-01 3.88462722e-01 7.20696449e-01
2.83318609e-01 2.62181073e-01 2.77730346e-01 -5.54124117e-02
8.78407955e-01 -8.26045215e-01 -1.86835453e-02 -1.38150260e-01
3.78164321e-01 -6.40299618e-01 6.80028796e-01 -8.40162158e-01
-1.07739973e+00 -6.99229985e-02 -3.81269366e-01 -8.79691720e-01
-4.15800035e-01 6.68903947e-01 -2.58360475e-01 3.37265044e-01
5.08331180e-01 2.90547699e-01 -4.44953293e-01 -2.70707339e-01
-3.06440126e-02 8.09898138e-01 1.05271101e+00 -4.63347912e-01
7.57095754e-01 3.96036804e-01 3.38967182e-02 -1.02455759e+00
-6.33393466e-01 -2.69681484e-01 -3.50800306e-01 -7.58972943e-01
1.18928683e+00 -9.17477250e-01 -1.03187585e+00 7.43596911e-01
-1.39444041e+00 3.03448141e-01 2.66891688e-01 8.43151271e-01
-3.36081237e-01 6.31377697e-01 -8.15181196e-01 -1.19828892e+00
-2.91467994e-01 -9.72747564e-01 9.30701137e-01 3.11549753e-01
-1.97955862e-01 -4.15110499e-01 2.12086529e-01 4.01653916e-01
7.69944638e-02 6.76714242e-01 5.38088620e-01 -4.79452878e-01
2.45178416e-01 -1.44361407e-01 -2.79206783e-01 2.17718169e-01
-9.50480103e-02 1.77575171e-01 -1.03617597e+00 3.09562049e-04
-1.37545317e-01 -5.42773962e-01 7.61579514e-01 2.28306472e-01
1.53503644e+00 -2.16232553e-01 -9.89110097e-02 5.63389182e-01
9.34290707e-01 2.81351805e-01 5.49374461e-01 1.66684076e-01
8.53605509e-01 6.31094158e-01 7.49807119e-01 1.03840435e+00
2.95780540e-01 7.77172387e-01 8.14310670e-01 -7.11177960e-02
2.21432894e-01 -4.77794558e-03 4.69248831e-01 8.05011570e-01
-3.39764863e-01 -8.68445709e-02 -8.06827903e-01 6.10307157e-01
-1.84373379e+00 -1.58429682e+00 -2.15534061e-01 1.79030108e+00
6.89930856e-01 1.45174459e-01 3.58353883e-01 4.18950230e-01
9.13490057e-01 3.56945395e-01 -2.56203622e-01 -4.92249936e-01
-1.09127246e-01 5.04612066e-02 -1.90756679e-01 1.53375387e-01
-1.56498897e+00 6.31510854e-01 5.47181177e+00 1.22183275e+00
-1.26258123e+00 2.36392230e-01 3.55979353e-01 -5.42680204e-01
7.20688105e-02 -5.48162460e-01 2.85926759e-02 8.16601932e-01
6.88230872e-01 3.45149249e-01 4.72824454e-01 5.41403949e-01
4.04062331e-01 1.44592881e-01 -6.38373613e-01 1.50297868e+00
2.10941300e-01 -7.90443242e-01 -1.10980056e-01 -3.37330490e-01
1.51937194e-02 -3.17395806e-01 -8.34239740e-03 3.46441835e-01
-2.48415679e-01 -1.07817459e+00 8.14073563e-01 7.33721197e-01
7.05714583e-01 -1.29506516e+00 6.49323881e-01 2.02252984e-01
-1.40923750e+00 -3.92147362e-01 -4.08779681e-02 -4.54688482e-02
2.80439138e-01 3.62153471e-01 -6.81329742e-02 5.17515421e-01
1.14904439e+00 5.88391781e-01 -2.73776770e-01 8.93346131e-01
-2.54886895e-01 5.95899880e-01 -2.97763377e-01 1.70223653e-01
2.05270857e-01 1.34403616e-01 7.78147936e-01 1.59627807e+00
2.56156981e-01 3.86552870e-01 1.59799069e-01 4.45393920e-01
5.21468297e-02 2.32743949e-01 -4.93573964e-01 4.87651564e-02
4.39865440e-01 1.44987011e+00 -4.16855693e-01 -8.55507851e-02
-5.04458070e-01 1.10756600e+00 2.75416255e-01 2.37497970e-01
-1.42821038e+00 -7.45935977e-01 8.18826914e-01 -2.69408822e-01
1.96832985e-01 -6.49100691e-02 1.69983536e-01 -1.10718036e+00
1.67339176e-01 -8.12875569e-01 6.45614445e-01 -7.01431632e-01
-1.26910484e+00 6.22025132e-01 5.68525679e-02 -1.39496458e+00
-1.10357568e-01 -1.60879776e-01 -9.89046693e-01 4.30776209e-01
-1.03751457e+00 -1.02033865e+00 -6.10408723e-01 9.83335555e-01
4.31398660e-01 -1.06731847e-01 6.04954660e-01 7.59962201e-01
-9.63236034e-01 6.95546150e-01 -3.21458787e-01 4.82410878e-01
6.20870769e-01 -4.55198556e-01 -2.71525025e-01 9.50951815e-01
-1.26746252e-01 9.94542316e-02 7.76318789e-01 -5.35253406e-01
-1.19386721e+00 -1.17486489e+00 5.40137589e-01 1.04508586e-02
7.13936627e-01 -2.20054492e-01 -8.59194100e-01 3.07962567e-01
5.57142356e-03 1.81581095e-01 6.14552438e-01 -1.29775852e-01
-4.43687737e-01 -6.24492913e-02 -1.05007052e+00 6.44506872e-01
1.07915342e+00 -7.21219420e-01 -7.38513410e-01 7.31775016e-02
5.86764276e-01 -2.40197793e-01 -7.20390499e-01 5.45683205e-01
7.20280647e-01 -9.83358204e-01 1.03161240e+00 -7.19199240e-01
9.18239415e-01 -1.19795009e-01 -3.79874647e-01 -9.93935704e-01
-3.96481246e-01 -2.65995055e-01 -1.45736888e-01 1.54475772e+00
-1.33827001e-01 -1.36800885e-01 2.60691375e-01 7.70281479e-02
2.11514886e-02 -5.00473082e-01 -1.18466747e+00 -5.65768600e-01
-3.85942996e-01 -9.65174258e-01 3.40949237e-01 1.27599967e+00
3.67301494e-01 3.23311299e-01 -8.12861562e-01 4.61176723e-01
4.86193836e-01 -1.56418473e-01 4.83600795e-01 -9.77939129e-01
-3.88200015e-01 -5.69125533e-01 -7.82114744e-01 -4.02383596e-01
1.47975340e-01 -5.23285449e-01 4.12113778e-02 -1.10838604e+00
4.86041367e-01 2.78246343e-01 -6.58518136e-01 3.06788623e-01
-2.19073936e-01 6.39591932e-01 1.26808330e-01 -1.99011460e-01
-9.03223038e-01 8.66955459e-01 8.59424114e-01 -3.41860563e-01
-1.19119085e-01 -2.49686629e-01 -3.35548311e-01 1.10897303e+00
5.94259202e-01 -4.34628993e-01 -1.07230805e-01 -3.89217317e-01
1.67172197e-02 5.36041677e-01 4.83146369e-01 -1.28813291e+00
8.28585252e-02 -3.99818063e-01 5.76852500e-01 -5.65477788e-01
8.83959889e-01 -8.20342302e-01 8.58671963e-03 4.42834049e-01
-2.52414614e-01 -8.43170434e-02 2.34968796e-01 5.79523623e-01
-5.68371356e-01 5.99128306e-02 6.55857146e-01 3.23694229e-01
-5.27209103e-01 5.25194645e-01 -6.36295795e-01 6.57111332e-02
1.22064435e+00 -1.28934488e-01 -1.40271947e-01 -6.06389761e-01
-8.43168080e-01 2.71208346e-01 -1.43332571e-01 6.30203307e-01
9.71504450e-01 -1.68264127e+00 -9.79594827e-01 -7.74511620e-02
2.70904601e-01 -4.28655446e-01 8.63383114e-01 1.13785315e+00
-2.75690228e-01 -1.42814919e-01 -3.79842639e-01 -5.11477590e-01
-1.65824544e+00 5.66952050e-01 1.28933460e-01 2.30376855e-01
-6.18518531e-01 9.01687980e-01 1.76121473e-01 -3.00316792e-02
5.37027776e-01 2.22149849e-01 -7.11111784e-01 3.76548529e-01
8.28092217e-01 3.96472543e-01 -3.14058840e-01 -1.19590783e+00
-7.95039535e-01 5.66637576e-01 2.67170042e-01 -1.92706272e-01
1.34251249e+00 2.73381057e-03 7.35215726e-04 4.50151473e-01
1.27164137e+00 -3.85396630e-02 -9.93640661e-01 -2.35312149e-01
-4.42663342e-01 -5.46297193e-01 6.63829073e-02 -3.67315680e-01
-1.31611943e+00 1.30966949e+00 5.63028634e-01 6.96848631e-02
1.45683932e+00 -6.49455711e-02 1.02717650e+00 2.98225433e-01
8.38900506e-02 -1.17334175e+00 3.49465787e-01 4.54964429e-01
1.11014843e+00 -9.87433970e-01 -1.93316907e-01 -9.03450400e-02
-8.39337885e-01 1.18098199e+00 6.69267654e-01 8.53945967e-03
6.12138391e-01 2.30287671e-01 -9.86811742e-02 -9.52935889e-02
-5.88336885e-01 -1.68825045e-01 1.25635387e-02 5.19690096e-01
6.58076704e-02 -5.42417243e-02 -3.57250243e-01 1.14446580e+00
-5.59477396e-02 -2.65080273e-01 1.95913091e-01 7.44586289e-01
-3.13597709e-01 -3.71133924e-01 -6.77745342e-01 2.67085284e-01
-7.64411330e-01 7.32398406e-02 -4.11518782e-01 4.63012695e-01
4.01285201e-01 1.20781457e+00 2.64856219e-01 -9.63849843e-01
3.14137608e-01 -2.00510338e-01 3.85811955e-01 -6.82852715e-02
-7.50015736e-01 1.89087048e-01 2.16504544e-01 -8.22342396e-01
-2.74010181e-01 -7.57707894e-01 -1.43287230e+00 -3.95787150e-01
2.78396942e-02 -1.56824365e-01 4.37151372e-01 9.80032027e-01
-8.30237567e-02 7.37743437e-01 8.90736401e-01 -9.81522799e-01
-1.03605971e-01 -1.02085233e+00 -5.55632532e-01 7.99394786e-01
4.03548926e-01 -8.45915437e-01 -3.18068355e-01 -9.09756124e-02] | [13.439847946166992, 4.842841148376465] |
894c19b0-8009-4ff5-a8ca-ebe3eb7e9cd5 | deception-detection-in-russian-texts | null | null | https://aclanthology.org/E17-4005 | https://aclanthology.org/E17-4005.pdf | Deception detection in Russian texts | Humans are known to detect deception in speech randomly and it is therefore important to develop tools to enable them to detect deception. The problem of deception detection has been studied for a significant amount of time, however the last 10-15 years have seen methods of computational linguistics being employed. Texts are processed using different NLP tools and then classified as deceptive/truthful using machine learning methods. While most research has been performed for English, Slavic languages have never been a focus of detection deception studies. The paper deals with deception detection in Russian narratives. It employs a specially designed corpus of truthful and deceptive texts on the same topic from each respondent, N = 113. The texts were processed using Linguistic Inquiry and Word Count software that is used in most studies of text-based deception detection. The list of parameters computed using the software was expanded due to the designed users{'} dictionaries. A variety of text classification methods was employed. The accuracy of the model was found to depend on the author{'}s gender and text type (deceptive/truthful). | ['Olga Litvinova', 'Pavel Seredin', 'John Lyell', 'Tatiana Litvinova'] | 2017-04-01 | null | null | null | eacl-2017-4 | ['deception-detection'] | ['miscellaneous'] | [-7.99008533e-02 -1.76388189e-01 -5.52392080e-02 -6.27188861e-01
-6.97917104e-01 -7.20015466e-01 1.00664639e+00 2.63124734e-01
-5.95203876e-01 8.09066892e-01 3.30821365e-01 -6.89328134e-01
1.48180509e-02 -1.90734103e-01 1.93424135e-01 -4.43754494e-01
4.19784784e-01 2.46753827e-01 -6.27112612e-02 -4.05035585e-01
1.31655252e+00 7.99535632e-01 -1.47347629e+00 2.78494507e-01
8.66029382e-01 6.39645875e-01 -2.14875966e-01 1.03506207e+00
1.97508652e-02 1.22832596e+00 -1.56492209e+00 -9.62147713e-01
-1.45722538e-01 -8.65697980e-01 -1.00112748e+00 1.77189827e-01
4.39258873e-01 -3.30053866e-01 -4.65217441e-01 1.05288005e+00
5.33298552e-01 1.86014205e-01 8.83096814e-01 -1.08996832e+00
-8.84903908e-01 3.51247787e-01 -1.26619507e-02 8.56075287e-01
7.18126833e-01 -1.13659099e-01 4.68615323e-01 -8.42481434e-01
3.17619801e-01 1.50492752e+00 4.36277688e-01 7.43066490e-01
-6.57711983e-01 -8.42972815e-01 -7.04057693e-01 6.00262940e-01
-1.30218577e+00 -9.18956637e-01 9.13745224e-01 -4.97806281e-01
1.06654227e+00 5.12333393e-01 5.60340405e-01 1.40970552e+00
5.17122626e-01 5.42048573e-01 1.48882675e+00 -6.99368775e-01
6.11024983e-02 6.93632424e-01 4.09348339e-01 7.15816915e-01
3.60135317e-01 8.68722200e-02 -6.56064868e-01 -4.39729065e-01
1.15144454e-01 -7.04207420e-01 -2.27994695e-01 4.66903120e-01
-5.72799087e-01 1.24267745e+00 -3.00667614e-01 8.71856809e-01
-3.07830540e-03 -3.69242102e-01 1.02139151e+00 6.17761672e-01
8.47989082e-01 5.64105690e-01 4.74874489e-02 -7.74952471e-01
-1.02223480e+00 3.18160206e-01 1.23710382e+00 5.79164803e-01
-3.46012861e-01 4.25243229e-01 2.92476118e-01 1.09799516e+00
2.55011320e-01 4.68848169e-01 9.02124286e-01 -4.99312341e-01
4.65130001e-01 4.98933464e-01 2.30723098e-01 -1.83108544e+00
-2.28169948e-01 4.34156992e-02 -3.79093647e-01 3.91640991e-01
6.95973277e-01 -5.68610728e-02 -6.74905121e-01 8.44056010e-01
-2.18546726e-02 -9.45202947e-01 5.34334257e-02 9.32339132e-01
8.30300450e-01 7.82442570e-01 4.91357129e-03 -4.95879024e-01
1.44635904e+00 -3.84526670e-01 -1.06516373e+00 4.08268208e-03
7.62009621e-01 -1.05994534e+00 9.07355249e-01 9.00878370e-01
-1.09267116e+00 9.80790611e-03 -1.20550370e+00 -2.93194771e-01
-5.70686996e-01 -6.44315407e-02 2.79652327e-01 1.68124068e+00
-5.09683013e-01 4.49856132e-01 -3.28392714e-01 -3.21951002e-01
4.87937421e-01 -6.10544868e-02 -3.09700519e-01 5.30293919e-02
-1.16837800e+00 1.78587425e+00 3.30653369e-01 1.18094139e-01
-6.14744067e-01 -3.04023232e-02 -1.00577378e+00 -3.42468292e-01
-1.32512078e-02 -1.78105280e-01 1.08369553e+00 -1.40289557e+00
-1.40176380e+00 1.45774448e+00 -1.03913590e-01 -1.96480826e-01
6.84233010e-01 1.17443211e-01 -1.10237706e+00 4.43543226e-01
-1.87789742e-02 -4.11874354e-01 1.41644883e+00 -1.07942724e+00
-4.97726977e-01 -7.49732435e-01 -2.55061567e-01 2.58854590e-02
-1.79610997e-01 1.06859040e+00 8.35679471e-01 -6.75889373e-01
2.38556974e-02 -3.33655417e-01 7.76566088e-01 -4.23239648e-01
-2.65118539e-01 -5.38008749e-01 1.00783074e+00 -1.16716492e+00
1.55137622e+00 -1.88975704e+00 -3.26356947e-01 -5.09920414e-04
3.26645076e-01 5.61259389e-01 6.60781026e-01 8.89470041e-01
-7.81064481e-02 5.00521541e-01 -2.27818802e-01 -4.44551259e-01
2.10468739e-01 4.78715897e-01 -1.28089949e-01 1.09465814e+00
-4.57763612e-01 4.02277768e-01 -7.03557968e-01 -8.22843611e-01
1.53276235e-01 2.97769874e-01 2.58900255e-01 -3.49969342e-02
6.33566558e-01 -3.49480778e-01 -2.70804703e-01 7.98257709e-01
3.70370299e-01 5.20610213e-01 -2.20685586e-01 2.86834747e-01
-3.60795438e-01 6.04821742e-01 -5.05396008e-01 8.11114728e-01
-2.58488595e-01 1.24695039e+00 2.52444774e-01 -9.52699125e-01
1.08119667e+00 6.07786834e-01 -6.26822531e-01 -3.41988891e-01
6.82209611e-01 6.81361258e-01 3.23408693e-01 -9.81654167e-01
9.03662741e-01 -8.42503846e-01 -2.82724708e-01 4.67383802e-01
-1.65143877e-01 -5.88518202e-01 5.67078628e-02 2.40332022e-01
6.85136199e-01 -5.12146413e-01 8.42380345e-01 -4.63185221e-01
8.78721118e-01 3.73441964e-01 1.00012228e-01 5.90712667e-01
-6.53494775e-01 2.13380292e-01 5.95595896e-01 -5.32846093e-01
-1.07150626e+00 -3.52192819e-01 -4.48536009e-01 9.61152494e-01
-3.06406438e-01 1.03771083e-01 -7.93418705e-01 -7.18448699e-01
-1.77419648e-01 1.65104473e+00 -5.27795732e-01 -4.00777757e-01
-3.96589488e-01 -6.63550079e-01 1.17936218e+00 -2.58605659e-01
4.13981587e-01 -1.17918658e+00 -9.60513234e-01 -6.30818009e-02
3.46807279e-02 -7.66022146e-01 -2.72943646e-01 -1.84439525e-01
-6.42021596e-01 -8.92361641e-01 -1.88722089e-01 -6.12830281e-01
5.09564519e-01 1.65296689e-01 7.14663506e-01 5.13340116e-01
-2.28978619e-01 3.59873354e-01 -8.02850187e-01 -6.78393781e-01
-1.35177553e+00 -2.18540654e-01 7.97969028e-02 -1.27078682e-01
5.85076034e-01 -3.60467970e-01 -2.23149266e-02 7.58499727e-02
-9.90313828e-01 -5.87593377e-01 1.09866038e-01 9.83886600e-01
-6.92980647e-01 4.24978063e-02 8.10136437e-01 -6.97513700e-01
1.43304765e+00 -3.95708233e-01 -1.04076959e-01 -2.73934186e-01
-5.36462367e-01 -4.51340616e-01 7.74766505e-01 -1.98235348e-01
-9.90409315e-01 -7.76103437e-01 -3.29875290e-01 1.99943677e-01
-6.75022960e-01 4.04407233e-01 -1.87605008e-01 -3.77019554e-01
8.61412525e-01 3.77341539e-01 2.33289495e-01 -2.03123584e-01
-2.66405344e-01 1.45562458e+00 4.22554970e-01 -1.64414644e-02
5.05902946e-01 -4.51870896e-02 -3.87327999e-01 -1.33484972e+00
-7.34232366e-01 -4.27999169e-01 -5.89531779e-01 -4.55614746e-01
3.89161736e-01 -7.16020912e-02 -7.89786637e-01 6.31474316e-01
-1.31332088e+00 2.98320711e-01 3.00668985e-01 5.28137863e-01
-2.49433070e-01 8.47891688e-01 -5.50738156e-01 -1.51040781e+00
-6.00049853e-01 -6.01770103e-01 4.41021860e-01 -1.56869039e-01
-5.79771638e-01 -1.28680599e+00 -1.23259224e-01 9.62594092e-01
7.90769905e-02 1.77631095e-01 1.04865956e+00 -1.32846892e+00
5.30400932e-01 -8.35202098e-01 4.80536632e-02 7.54431725e-01
2.28364468e-02 4.24955226e-02 -8.34562361e-01 7.92122558e-02
7.03424037e-01 -5.03281593e-01 3.44856054e-01 -1.39784902e-01
5.71948349e-01 -8.54237735e-01 1.73354577e-02 -1.76342413e-01
9.79258478e-01 5.22372067e-01 8.32419813e-01 2.51490891e-01
4.25664574e-01 6.99167550e-01 5.14094710e-01 4.07816797e-01
1.17418170e-01 3.58320653e-01 3.74773026e-01 7.87884295e-01
3.55349630e-01 1.58812441e-02 5.41550577e-01 8.42225373e-01
-2.36546453e-02 -6.75064683e-01 -9.31884110e-01 7.23572433e-01
-1.18002427e+00 -1.54196477e+00 -7.26444125e-01 1.86654031e+00
7.05242455e-01 2.59146929e-01 2.07893878e-01 9.44092631e-01
5.09431541e-01 2.97336996e-01 4.76760492e-02 -1.51336277e+00
-7.34303147e-02 -1.26517490e-01 8.99517164e-02 8.90708923e-01
-6.33971274e-01 7.30383039e-01 6.09581852e+00 1.03409648e+00
-8.26511264e-01 7.53934830e-02 4.66280997e-01 3.76111642e-02
-2.61964984e-02 -4.23907280e-01 -4.00395185e-01 7.35706031e-01
1.20099294e+00 -4.17175800e-01 4.96418700e-02 6.40670896e-01
6.53669238e-01 -6.72993481e-01 -6.68593764e-01 1.11823237e+00
1.12955463e+00 -6.62921309e-01 -3.22276764e-02 -1.14257470e-01
8.68823752e-02 -8.55389893e-01 -1.59675196e-01 2.00638905e-01
-1.82175532e-01 -1.37117362e+00 1.02512753e+00 3.97372514e-01
3.38204592e-01 -1.12575734e+00 1.14692247e+00 9.49287534e-01
1.82412282e-01 -5.25621958e-02 -3.97994339e-01 -4.87266243e-01
2.50202268e-01 4.77911919e-01 -1.05867696e+00 2.68073559e-01
2.86165446e-01 1.93652317e-01 -5.83364606e-01 5.91490328e-01
-2.30706558e-01 1.04628170e+00 1.23014465e-01 -8.29089463e-01
3.22668761e-01 -2.41287678e-01 1.01146555e+00 1.44735074e+00
1.15263484e-01 3.92696530e-01 -5.14931321e-01 6.27425551e-01
2.16489390e-01 3.22290212e-01 -6.23462379e-01 -3.73788983e-01
4.05339241e-01 1.17593980e+00 -5.25111794e-01 -3.84184062e-01
-6.97961077e-02 1.18771923e+00 -4.53804396e-02 -2.57896483e-01
-4.79348451e-01 -7.66715050e-01 3.37885439e-01 2.50598371e-01
-3.06271613e-01 -2.84440011e-01 -6.79032326e-01 -7.24509895e-01
-9.99767631e-02 -1.14370787e+00 2.96227872e-01 -7.84969926e-01
-1.44593143e+00 5.98005116e-01 2.40323737e-01 -4.16564792e-01
-4.06568110e-01 -8.28148663e-01 -2.73963422e-01 1.02657294e+00
-5.19708633e-01 -9.79157031e-01 1.26061020e-02 3.84016693e-01
7.29557455e-01 -5.31028807e-01 8.46027672e-01 -3.72307701e-03
-4.95259851e-01 5.49042523e-01 -1.89724818e-01 2.58109540e-01
5.21546423e-01 -1.06910861e+00 -1.24796137e-01 8.27881455e-01
-8.26216415e-02 5.29331088e-01 1.18956220e+00 -7.17104375e-01
-9.44922626e-01 -7.84011036e-02 1.81170392e+00 -7.58633256e-01
9.99709785e-01 -4.11673546e-01 -9.10353005e-01 2.53468126e-01
4.14217263e-01 -3.91667843e-01 7.17192233e-01 -4.03907359e-01
-3.94070148e-01 2.65093803e-01 -1.62364578e+00 4.13129807e-01
7.40016460e-01 -6.72522366e-01 -1.32151532e+00 3.20630968e-01
-1.12993121e-01 -2.29445696e-01 -5.14656246e-01 -4.81233299e-01
6.01981044e-01 -1.14037883e+00 4.54349846e-01 -8.32175076e-01
5.53588212e-01 1.84092253e-01 1.52049035e-01 -1.39355576e+00
-3.08781397e-02 -6.23058915e-01 3.24324936e-01 1.15038598e+00
3.41476172e-01 -1.00551701e+00 3.75546753e-01 6.19585335e-01
-4.58205730e-01 -2.22201928e-01 -1.33929753e+00 -6.22504115e-01
2.66915023e-01 -3.69349003e-01 -4.76729013e-02 1.26783395e+00
7.40391195e-01 4.34865892e-01 -5.09342551e-01 -1.61976308e-01
3.58763993e-01 -2.09817320e-01 3.09439600e-01 -1.10585439e+00
4.33790356e-01 -6.35343134e-01 -7.70517766e-01 -5.31839073e-01
3.65428597e-01 -5.87886989e-01 -2.45140530e-02 -1.07566440e+00
5.50590083e-02 2.16043442e-01 7.66775250e-01 2.28441447e-01
1.40046543e-02 2.60070026e-01 -8.69206051e-05 1.61073714e-01
9.92724076e-02 3.33752006e-01 9.55715716e-01 1.66463763e-01
1.03559740e-01 1.43340155e-01 -7.85786808e-01 8.02373469e-01
1.17857420e+00 -8.05191636e-01 -2.15993911e-01 1.13045111e-01
2.80961990e-01 1.37064770e-01 7.14359701e-01 -5.20404696e-01
1.41279191e-01 -2.07703754e-01 3.24079067e-01 -3.72439116e-01
3.37021410e-01 -7.35125482e-01 -1.66760951e-01 3.25300694e-01
-2.51399934e-01 3.96045685e-01 -2.60088313e-02 2.81353682e-01
-2.07236528e-01 -1.18134987e+00 1.01312459e+00 -2.70142257e-01
-3.01232755e-01 -4.82528210e-01 -1.26908815e+00 1.63958911e-02
9.85810578e-01 -7.04522014e-01 -2.73449361e-01 -8.65338981e-01
-4.05819267e-01 -5.72987854e-01 4.46231514e-01 2.11450219e-01
7.64021993e-01 -8.86385560e-01 -9.22107041e-01 -2.60243744e-01
1.46049010e-02 -1.00362742e+00 -8.85277241e-02 9.02448833e-01
-9.60954785e-01 5.25509953e-01 -3.21804255e-01 2.45836765e-01
-1.82701087e+00 5.67398012e-01 4.17243898e-01 3.61131668e-01
-3.80599439e-01 6.26131058e-01 -7.49462128e-01 8.00189003e-02
-1.69876337e-01 2.49444798e-01 -3.77312243e-01 3.01650465e-01
7.85343051e-01 1.04660678e+00 1.87314495e-01 -1.42396605e+00
-3.06955069e-01 -3.82298529e-01 -2.61950880e-01 -3.49834919e-01
9.70382571e-01 -3.18567425e-01 -3.38820040e-01 7.94518530e-01
1.28364313e+00 4.02452499e-01 -1.95901841e-01 5.12890339e-01
1.47949666e-01 -9.22060907e-01 1.58547863e-01 -1.12624371e+00
-3.49914700e-01 7.59410203e-01 -1.81798622e-01 8.52210581e-01
7.46151865e-01 -2.76438862e-01 3.83621693e-01 -5.98846748e-03
2.42613330e-01 -1.29238296e+00 -2.00942636e-01 4.77385581e-01
1.41597939e+00 -1.09363842e+00 3.42720360e-01 -3.74944836e-01
-7.53911674e-01 1.31968987e+00 -6.17670491e-02 -6.47462234e-02
2.65306324e-01 -1.01975188e-01 1.85355917e-01 -5.04238427e-01
-2.22051308e-01 4.20171350e-01 2.12662101e-01 6.63392246e-01
5.77948689e-01 7.20580742e-02 -1.32266593e+00 6.45792842e-01
-8.55512917e-01 -4.47222024e-01 1.06923652e+00 8.95299613e-01
-5.89127004e-01 -8.65511954e-01 -8.31686676e-01 5.48066258e-01
-1.05164921e+00 -1.10661469e-01 -1.29030430e+00 1.25892258e+00
-2.14913934e-01 1.67448211e+00 -2.98205703e-01 -2.03142017e-01
1.27026185e-01 3.65463465e-01 5.60611784e-01 -3.71671468e-01
-1.19891191e+00 -6.42211795e-01 1.01884806e+00 5.61431348e-02
-3.10731262e-01 -1.01695561e+00 -8.07583869e-01 -1.06989944e+00
-4.11587298e-01 3.75641674e-01 7.99754918e-01 1.26611984e+00
-2.72105545e-01 -1.89516574e-01 3.38735908e-01 -4.91030782e-01
-8.16770852e-01 -1.50712645e+00 -1.00940645e+00 2.63763875e-01
2.47531936e-01 -5.03741086e-01 -1.05475903e+00 -1.15561210e-01] | [8.244256973266602, 10.443354606628418] |
67e5c6a6-8633-4d41-9f37-ce4558234e44 | a-multi-task-framework-for-skin-lesion | 1808.01676 | null | http://arxiv.org/abs/1808.01676v1 | http://arxiv.org/pdf/1808.01676v1.pdf | A Multi-task Framework for Skin Lesion Detection and Segmentation | Early detection and segmentation of skin lesions is crucial for timely
diagnosis and treatment, necessary to improve the survival rate of patients.
However, manual delineation is time consuming and subject to intra- and
inter-observer variations among dermatologists. This underlines the need for an
accurate and automatic approach to skin lesion segmentation. To tackle this
issue, we propose a multi-task convolutional neural network (CNN) based, joint
detection and segmentation framework, designed to initially localize the lesion
and subsequently, segment it. A `Faster region-based convolutional neural
network' (Faster-RCNN) which comprises a region proposal network (RPN), is used
to generate bounding boxes/region proposals, for lesion localization in each
image. The proposed regions are subsequently refined using a softmax classifier
and a bounding-box regressor. The refined bounding boxes are finally cropped
and segmented using `SkinNet', a modified version of U-Net. We trained and
evaluated the performance of our network, using the ISBI 2017 challenge and the
PH2 datasets, and compared it with the state-of-the-art, using the official
test data released as part of the challenge for the former. Our approach
outperformed others in terms of Dice coefficients ($>0.93$), Jaccard index
($>0.88$), accuracy ($>0.96$) and sensitivity ($>0.95$), across five-fold cross
validation experiments. | ['Shreyas Malakarjun Patil', 'Sulaiman Vesal', 'Nishant Ravikumar', 'Andreas Maier'] | 2018-08-05 | null | null | null | null | ['skin-lesion-segmentation'] | ['medical'] | [ 4.74367887e-01 4.72948840e-03 -1.24890395e-01 -2.52933085e-01
-9.99816716e-01 -6.32764578e-01 3.80445123e-01 5.04666507e-01
-7.75723457e-01 6.81147158e-01 -2.48866007e-01 -1.05161637e-01
-9.67085660e-02 -6.73850477e-01 -4.31292236e-01 -9.42045808e-01
-1.90492541e-01 2.64077157e-01 5.71713030e-01 1.95937514e-01
1.72560751e-01 6.57732248e-01 -1.07871366e+00 4.34914261e-01
9.34377789e-01 1.32861185e+00 -3.86303924e-02 8.01586270e-01
1.78175862e-03 4.61012214e-01 -4.25132841e-01 -4.68605638e-01
1.67460993e-01 -2.10059866e-01 -1.02036512e+00 2.54531771e-01
2.92231619e-01 -1.42912969e-01 -1.84318870e-02 7.04681814e-01
5.89916945e-01 4.45201509e-02 7.86247492e-01 -5.48384726e-01
-4.44294810e-01 3.36748213e-01 -8.38774025e-01 7.13501126e-02
-9.07057617e-03 1.92570806e-01 6.95382595e-01 -5.11557460e-01
8.06486428e-01 6.67308569e-01 1.00706983e+00 5.95016003e-01
-1.20378125e+00 -3.85917902e-01 -1.82128623e-01 7.65011013e-02
-1.71692145e+00 2.05898122e-03 2.52728015e-01 -4.19040650e-01
7.46257782e-01 4.20722336e-01 4.85213459e-01 9.58612263e-01
1.58954576e-01 6.29487395e-01 1.33443022e+00 -3.72743309e-01
1.82296202e-01 2.03603566e-01 -1.87784359e-01 7.74172127e-01
7.43792299e-03 1.40979029e-02 2.82860864e-02 9.35210809e-02
9.71866190e-01 -1.64228126e-01 -1.28812060e-01 -9.33093727e-02
-8.76622617e-01 7.83710420e-01 8.63081932e-01 3.97453219e-01
-6.57133758e-01 1.34710642e-02 4.40375179e-01 -1.82394519e-01
6.31159127e-01 2.09685817e-01 -1.34606659e-01 2.54609823e-01
-1.08917606e+00 2.43031587e-02 5.55873215e-01 1.47189245e-01
2.22719327e-01 -5.08764505e-01 -5.18983424e-01 1.02032757e+00
1.26185209e-01 2.50961259e-02 3.06908637e-01 -5.19062877e-01
1.08417578e-01 8.00483346e-01 -1.52304098e-01 -6.53120220e-01
-7.91001737e-01 -6.19430065e-01 -1.13021338e+00 2.69596666e-01
6.83942735e-01 -2.29370400e-01 -1.30227041e+00 1.32805932e+00
4.97222811e-01 3.77390981e-02 -2.96010852e-01 9.92491066e-01
7.77116776e-01 2.93142945e-01 4.14359361e-01 3.44508812e-02
1.42568731e+00 -9.59466100e-01 -3.48625511e-01 1.53442934e-01
6.28240883e-01 -6.07924819e-01 5.25536060e-01 5.48356175e-01
-1.04985332e+00 -3.85108143e-01 -7.46951818e-01 1.74810052e-01
-4.61143613e-01 5.00984848e-01 3.16639006e-01 6.11471117e-01
-1.29900944e+00 6.18575871e-01 -8.73044193e-01 -6.65714443e-01
1.00967789e+00 4.25947219e-01 -6.02861583e-01 -1.57985002e-01
-9.05953586e-01 1.06980455e+00 5.05554557e-01 3.66757154e-01
-6.00965202e-01 -8.18364143e-01 -7.13218689e-01 -2.91273683e-01
3.03486913e-01 -5.19287288e-01 8.57814193e-01 -8.51379514e-01
-1.26839364e+00 1.18052745e+00 5.34282401e-02 -6.50726318e-01
8.80397201e-01 7.30571300e-02 -3.27580184e-01 4.52129871e-01
-1.00104176e-01 9.35352147e-01 5.55255413e-01 -1.03547251e+00
-7.88637102e-01 -3.67344975e-01 -2.31270730e-01 7.19933026e-03
2.32750610e-01 2.70491749e-01 -5.70894599e-01 -5.64356685e-01
-1.76371500e-01 -8.43825221e-01 -6.27515733e-01 3.85851800e-01
-6.91849530e-01 -3.35826874e-01 5.34945905e-01 -1.23376501e+00
1.19553828e+00 -1.90320361e+00 4.65934202e-02 6.43662751e-01
3.16973120e-01 6.57040060e-01 -2.32841726e-02 8.00851583e-02
-2.26635143e-01 2.54534334e-01 -6.93547249e-01 -2.86488980e-01
-3.47682595e-01 -2.19001696e-01 4.71456558e-01 7.85538971e-01
5.47872007e-01 8.50436926e-01 -6.32294059e-01 -5.55181265e-01
5.63089311e-01 8.43177021e-01 -1.63212702e-01 1.24104045e-01
-4.70725857e-02 4.20514166e-01 -8.25289115e-02 8.67275596e-01
5.80193579e-01 -3.64110470e-01 -2.74895895e-02 -2.30648667e-01
-1.27133965e-01 -3.62595469e-01 -1.15071714e+00 1.63330507e+00
-4.23268706e-01 6.22798264e-01 1.88725248e-01 -9.29441929e-01
9.08456147e-01 4.01384294e-01 8.10526550e-01 -5.66194594e-01
3.66040260e-01 2.63741434e-01 4.65109907e-02 -5.87779284e-01
-1.78234413e-01 -7.22826943e-02 2.51365095e-01 2.52921730e-01
-8.63191709e-02 1.08486801e-01 2.97071308e-01 -4.45792228e-02
1.13230813e+00 4.05780226e-02 4.52366084e-01 -3.46303210e-02
7.89749861e-01 5.33061288e-02 1.48575380e-01 3.49981636e-01
-5.60034156e-01 9.01931405e-01 5.24777293e-01 -4.72844064e-01
-9.80978847e-01 -1.02820051e+00 -5.72199702e-01 8.12533259e-01
-1.33889958e-01 1.49674147e-01 -1.10560739e+00 -1.03511417e+00
-2.61598956e-02 3.42797756e-01 -1.02002418e+00 2.09103942e-01
-4.22723889e-01 -9.97542679e-01 4.35963452e-01 6.13035083e-01
4.54737931e-01 -1.29034042e+00 -7.09511936e-01 3.83322358e-01
-2.03702338e-02 -9.91986752e-01 -2.45224938e-01 1.57992214e-01
-3.95402938e-01 -1.41256738e+00 -1.22388375e+00 -6.82483912e-01
9.41783369e-01 -2.46691674e-01 6.48605943e-01 1.01037204e-01
-1.03554440e+00 -1.39587149e-01 -3.41933042e-01 -1.38604254e-01
-1.78831339e-01 1.67831719e-01 -4.64515805e-01 2.47024789e-01
1.20009623e-01 -1.80498585e-01 -9.01714623e-01 1.94176942e-01
-1.14035141e+00 -8.52496698e-02 1.00228751e+00 8.19521308e-01
8.03738177e-01 -3.83109748e-02 4.78630871e-01 -9.69773829e-01
5.82092702e-01 -3.85850370e-01 -5.56224763e-01 4.27703798e-01
-3.71458977e-01 -3.80537897e-01 4.33444530e-01 -2.10103095e-01
-8.20730150e-01 3.28616172e-01 -3.82515997e-01 -1.16148092e-01
-5.29041886e-01 3.60745847e-01 3.30953121e-01 -3.33122432e-01
8.61871421e-01 -1.46579677e-02 6.79990053e-02 -3.09139848e-01
4.39374238e-01 7.74572432e-01 6.00999832e-01 -6.01916090e-02
5.50586641e-01 5.50098598e-01 1.61235005e-01 -5.17363191e-01
-7.02294528e-01 -7.75308788e-01 -1.11997271e+00 -2.79072374e-01
1.17848229e+00 -5.54258466e-01 -4.88032669e-01 5.43477476e-01
-1.01763237e+00 -4.24404234e-01 -2.63436679e-02 1.90681234e-01
-2.82833099e-01 3.91968757e-01 -4.86284345e-01 -7.62341022e-01
-8.19177747e-01 -1.08610368e+00 9.47258055e-01 6.06655657e-01
-3.83730888e-01 -1.17963362e+00 1.62024707e-01 4.82832581e-01
5.28636038e-01 9.55624580e-01 6.33296847e-01 -7.42887020e-01
-5.42699955e-02 -6.91982806e-01 -8.96040916e-01 5.57888031e-01
2.34942630e-01 3.73483986e-01 -9.89044726e-01 -2.64972210e-01
-7.12603807e-01 -2.58828312e-01 1.06473315e+00 8.06858301e-01
1.60741580e+00 -1.42017066e-01 -4.76252377e-01 6.77646518e-01
1.66725242e+00 1.82545021e-01 5.68336546e-01 1.06101170e-01
4.93203759e-01 9.20765162e-01 3.63593787e-01 3.35765630e-01
2.20242962e-01 4.77240115e-01 5.04065275e-01 -6.17180467e-01
-2.98171580e-01 2.92718530e-01 -1.67456731e-01 5.42905815e-02
-3.35875332e-01 3.94877774e-04 -1.03881156e+00 8.49592149e-01
-1.47891390e+00 -4.62787539e-01 -1.89523116e-01 2.09161115e+00
8.04745376e-01 5.19482270e-02 5.48947513e-01 1.15651757e-01
9.50838864e-01 -1.11594938e-01 -4.70999569e-01 -6.12858772e-01
-5.07495478e-02 7.67333806e-01 5.44468701e-01 3.32073450e-01
-1.64645565e+00 9.18776572e-01 5.63113880e+00 8.83334279e-01
-1.47209263e+00 2.09963433e-02 9.95006561e-01 -2.95726722e-03
3.91473323e-01 -4.82966125e-01 -5.10311186e-01 4.14430529e-01
9.02146757e-01 4.12355453e-01 2.85978764e-01 5.38536966e-01
1.35924473e-01 -3.89259964e-01 -6.79249048e-01 5.79694271e-01
2.01782405e-01 -1.32061768e+00 -4.50358272e-01 4.07996168e-03
7.95345604e-01 1.26506314e-01 8.59444961e-02 -4.92710546e-02
1.47054911e-01 -1.32449329e+00 1.29582539e-01 5.32396793e-01
1.16750157e+00 -8.85612369e-01 1.04391801e+00 -2.87457537e-02
-1.07910049e+00 1.02637038e-01 -6.21866286e-02 6.17314398e-01
8.70018676e-02 6.76704109e-01 -1.19944191e+00 4.86838132e-01
6.90744221e-01 3.46378028e-01 -7.14119792e-01 1.50474274e+00
-1.54548392e-01 3.77606571e-01 -2.87332654e-01 -2.91413534e-02
5.92984200e-01 -3.04795466e-02 2.64831662e-01 1.45764792e+00
1.65637918e-02 -8.14913958e-02 -1.29955634e-01 9.39062238e-01
7.09460303e-02 4.27540570e-01 -6.39015483e-03 2.15706095e-01
2.98747662e-02 1.83090198e+00 -1.11951077e+00 -1.69135824e-01
-7.02932477e-02 1.02401257e+00 2.57932544e-01 1.58127770e-01
-9.27670479e-01 -6.54239416e-01 1.39149562e-01 4.32099812e-02
2.84760684e-01 3.09828192e-01 -3.55799198e-01 -5.58433533e-01
-7.83965811e-02 -5.55648506e-01 6.67063534e-01 -4.87779051e-01
-1.31365335e+00 6.74219012e-01 -2.65741974e-01 -8.01348925e-01
5.99951930e-02 -7.81403124e-01 -7.62809038e-01 1.11353528e+00
-1.66892326e+00 -1.35367179e+00 -5.13459861e-01 3.28285009e-01
1.90466478e-01 2.44230896e-01 8.02613437e-01 1.43961474e-01
-9.39886272e-01 7.50323713e-01 -1.10684847e-02 4.01376635e-01
8.01762402e-01 -1.48804152e+00 1.96801707e-01 6.83785200e-01
-2.13521719e-01 1.91574305e-01 7.44098648e-02 -5.83078802e-01
-6.05959833e-01 -1.51187587e+00 7.44446397e-01 -3.17201495e-01
5.73425472e-01 -4.09948453e-02 -9.57000434e-01 1.85115889e-01
-7.05174543e-03 3.99226040e-01 1.04954052e+00 -1.64580986e-01
-2.78710663e-01 -1.36075348e-01 -1.73542643e+00 4.04228210e-01
4.79377061e-01 -2.91890174e-01 -2.27518249e-02 5.05765676e-01
3.30142826e-01 -5.01194060e-01 -1.35257173e+00 5.53164899e-01
7.89161801e-01 -8.57705235e-01 1.01645827e+00 -4.90352273e-01
5.28895974e-01 9.36507434e-02 1.34109989e-01 -1.12442601e+00
-3.31963956e-01 -2.02788100e-01 9.32470486e-02 1.00838506e+00
6.02520406e-01 -4.00030494e-01 1.03327644e+00 5.00029981e-01
-1.11805767e-01 -1.47599220e+00 -1.11265576e+00 -2.72683799e-01
2.36834317e-01 -2.44880095e-01 2.34228283e-01 7.56569266e-01
-2.45702431e-01 -3.26324344e-01 3.23027372e-02 3.15827839e-02
6.52319252e-01 -3.46935451e-01 2.43124411e-01 -1.08143890e+00
2.34839264e-02 -9.19990659e-01 -2.15396509e-01 -1.12517267e-01
-2.70208567e-01 -9.50959861e-01 -1.40224639e-02 -1.81559956e+00
1.96753547e-01 -4.73718554e-01 -6.50197089e-01 7.18257546e-01
-3.56056243e-01 8.15845072e-01 6.60836324e-03 -7.02447668e-02
-4.46181983e-01 -1.86527163e-01 1.23290336e+00 -4.08721901e-02
-2.91875929e-01 1.84744954e-01 -6.05816662e-01 5.23517132e-01
9.74041045e-01 -1.35232881e-01 2.38777339e-01 4.55423491e-03
-2.87584364e-01 -9.88084897e-02 6.03301167e-01 -1.13456476e+00
1.63973480e-01 -1.14490852e-01 6.73357069e-01 -7.38923490e-01
1.98758289e-01 -7.00427711e-01 -9.99303237e-02 8.36330771e-01
-3.65459591e-01 -4.82141107e-01 2.47175232e-01 1.92340478e-01
2.09622290e-02 -1.45322025e-01 1.30506289e+00 -1.04888849e-01
-4.70790297e-01 5.45874834e-01 -1.48932666e-01 -4.77531523e-01
1.51966834e+00 -3.82755995e-01 -2.09609225e-01 1.23394892e-01
-8.32124531e-01 1.35713637e-01 8.98773074e-02 8.94263014e-02
4.36678767e-01 -9.22507882e-01 -9.71418619e-01 4.72837463e-02
2.08688244e-01 3.15796703e-01 6.82372510e-01 1.36044300e+00
-9.70364392e-01 5.20068347e-01 -1.50505424e-01 -7.55418122e-01
-1.30544233e+00 1.86687246e-01 5.41955054e-01 -5.17058730e-01
-4.35718447e-01 1.22438335e+00 -2.42732689e-01 -3.90946537e-01
2.78939277e-01 -4.34637278e-01 -3.57701629e-01 1.12099856e-01
4.48040903e-01 5.29939950e-01 1.44957811e-01 -6.26212835e-01
-3.58611733e-01 7.07434833e-01 -2.33754411e-01 9.18431357e-02
1.22444201e+00 1.62143573e-01 -2.66463757e-01 -1.94792226e-01
1.18062973e+00 -4.86179113e-01 -1.14897966e+00 -2.33211353e-01
3.56741548e-02 -1.74985513e-01 2.48421535e-01 -1.46361816e+00
-1.16649163e+00 8.13264728e-01 9.67466116e-01 2.07361102e-01
1.24542582e+00 -7.69189149e-02 8.03671896e-01 -1.88505054e-01
-1.56267256e-01 -1.05147326e+00 -1.72271162e-01 -5.23210727e-02
8.25930059e-01 -1.37113464e+00 -2.20605191e-02 -5.36231637e-01
-6.66500688e-01 1.06084406e+00 6.05869710e-01 -1.63978919e-01
6.17859781e-01 1.65092647e-02 2.38623783e-01 -1.35075077e-02
-3.64589036e-01 -3.32340419e-01 7.53810585e-01 6.96547687e-01
4.62787032e-01 2.17749327e-01 -3.73855680e-01 4.90944028e-01
1.36232764e-01 -2.57327245e-03 -3.95129994e-03 5.51213801e-01
-4.62832153e-01 -9.83319223e-01 -3.39189649e-01 6.35493279e-01
-7.51292050e-01 1.88597277e-01 -6.35084391e-01 9.61925685e-01
6.25442803e-01 7.01903045e-01 1.19761378e-01 -1.51698127e-01
1.99014135e-02 -1.86630473e-01 3.81530136e-01 -4.83278096e-01
-9.64735746e-01 2.93376863e-01 -7.59715121e-03 -5.99176288e-01
-2.96321332e-01 -4.51960266e-01 -1.18306148e+00 -2.66729537e-02
-3.53429139e-01 -7.37599060e-02 9.59274650e-01 8.58929276e-01
-4.97689703e-03 6.75811231e-01 7.11160123e-01 -4.50695753e-01
-3.60233188e-01 -9.69110608e-01 -5.30963421e-01 3.34727049e-01
1.10303238e-01 -3.41567427e-01 -2.66879853e-02 -5.57103120e-02] | [15.625797271728516, -2.9382357597351074] |
cf650c69-a916-4059-ba4c-1983b2d0bd9a | continual-learning-for-class-and-domain | 2209.08023 | null | https://arxiv.org/abs/2209.08023v1 | https://arxiv.org/pdf/2209.08023v1.pdf | Continual Learning for Class- and Domain-Incremental Semantic Segmentation | The field of continual deep learning is an emerging field and a lot of progress has been made. However, concurrently most of the approaches are only tested on the task of image classification, which is not relevant in the field of intelligent vehicles. Only recently approaches for class-incremental semantic segmentation were proposed. However, all of those approaches are based on some form of knowledge distillation. At the moment there are no investigations on replay-based approaches that are commonly used for object recognition in a continual setting. At the same time while unsupervised domain adaption for semantic segmentation gained a lot of traction, investigations regarding domain-incremental learning in an continual setting is not well-studied. Therefore, the goal of our work is to evaluate and adapt established solutions for continual object recognition to the task of semantic segmentation and to provide baseline methods and evaluation protocols for the task of continual semantic segmentation. We firstly introduce evaluation protocols for the class- and domain-incremental segmentation and analyze selected approaches. We show that the nature of the task of semantic segmentation changes which methods are most effective in mitigating forgetting compared to image classification. Especially, in class-incremental learning knowledge distillation proves to be a vital tool, whereas in domain-incremental learning replay methods are the most effective method. | ['Jürgen Beyerer', 'Miriam Ruf', 'Masoud Roschani', 'Tobias Kalb'] | 2022-09-16 | null | null | null | null | ['class-incremental-semantic-segmentation', 'continual-semantic-segmentation'] | ['computer-vision', 'computer-vision'] | [ 3.88273358e-01 -3.34964134e-02 -3.76388878e-01 -5.15649974e-01
-2.62486964e-01 -4.18218374e-01 7.68059850e-01 4.57325578e-01
-7.97005117e-01 6.80551410e-01 -6.36514187e-01 -2.70389438e-01
-3.33862245e-01 -7.91867554e-01 -8.66711617e-01 -6.35089993e-01
2.29066104e-01 8.24628472e-01 9.01045680e-01 -3.60882282e-01
2.63326257e-01 6.52372599e-01 -2.05232692e+00 2.88104445e-01
8.29768062e-01 8.91126335e-01 3.55900377e-01 5.23879290e-01
-6.19534314e-01 4.67726529e-01 -7.58403659e-01 -1.76193237e-01
2.20431879e-01 -5.49194276e-01 -1.14098656e+00 4.71475959e-01
5.34656107e-01 2.00667568e-02 -6.73338473e-02 1.04746902e+00
4.36935455e-01 3.05605590e-01 4.07126755e-01 -1.25689673e+00
-1.37134045e-01 5.44201970e-01 -1.54444575e-01 3.64396989e-01
3.95810939e-02 1.21607594e-02 3.19078475e-01 -5.91931820e-01
6.71347678e-01 9.48230028e-01 7.03842163e-01 6.50772274e-01
-1.03956151e+00 -4.37291920e-01 2.84279704e-01 8.28718662e-01
-9.87691820e-01 -2.84141839e-01 9.69288528e-01 -2.76432037e-01
8.57617974e-01 -7.78846219e-02 7.91809678e-01 9.95684624e-01
-2.02200413e-01 1.09944284e+00 1.28002715e+00 -6.00421250e-01
6.66368484e-01 6.00916386e-01 6.06442809e-01 3.09040368e-01
2.18825385e-01 -4.77199703e-02 -5.68657577e-01 3.72150898e-01
1.83366865e-01 -1.32505178e-01 1.44108072e-01 -5.50386786e-01
-7.34649420e-01 6.94532394e-01 1.44322768e-01 7.19363570e-01
-2.36849725e-01 -1.27216309e-01 7.44340777e-01 6.32105410e-01
5.82557142e-01 2.52654374e-01 -5.90249836e-01 -3.49831581e-01
-1.17735457e+00 4.38521087e-01 7.08391905e-01 7.72577226e-01
1.06274128e+00 2.41845086e-01 8.57425407e-02 7.89096296e-01
-3.56320798e-01 3.70840311e-01 7.43252099e-01 -8.16398978e-01
-1.57046597e-03 7.82191873e-01 -4.01330441e-02 -5.88949800e-01
-5.34289062e-01 -4.61554617e-01 -5.82760990e-01 4.60747987e-01
4.80210930e-01 8.84740725e-02 -1.00940478e+00 1.43623745e+00
4.27240193e-01 2.79611439e-01 2.17992350e-01 4.67048556e-01
5.40483892e-01 3.04617018e-01 2.13959634e-01 -3.16505045e-01
1.22296965e+00 -9.17079866e-01 -6.37182832e-01 -2.40943030e-01
5.66465914e-01 -5.61083555e-01 1.01736999e+00 5.92680693e-01
-7.10108936e-01 -8.86326373e-01 -1.15544868e+00 3.02657247e-01
-8.80213916e-01 -3.72119188e-01 5.02008677e-01 1.02080989e+00
-9.54442382e-01 6.36556923e-01 -8.79098177e-01 -6.58990681e-01
4.62297052e-01 2.69046128e-01 -2.97808349e-01 -7.94014186e-02
-1.20183933e+00 1.03331029e+00 9.34635162e-01 -2.08550468e-01
-6.13350987e-01 -6.07470751e-01 -5.59869826e-01 -3.21034223e-01
6.81395113e-01 -3.62658828e-01 1.55373764e+00 -1.46167421e+00
-1.48895919e+00 1.01290333e+00 -4.81360368e-02 -1.14631927e+00
8.10166657e-01 -1.03205040e-01 -4.52453613e-01 1.93182528e-01
3.99873219e-02 7.60602653e-01 9.54056978e-01 -1.24709880e+00
-9.27704811e-01 -4.37229127e-01 1.07574008e-01 1.91662893e-01
-4.72483188e-01 -4.64691132e-01 -3.71630728e-01 -3.41103047e-01
5.99659383e-02 -8.31011474e-01 -5.83323985e-02 -1.71081841e-01
8.79143998e-02 -4.13492650e-01 1.48250699e+00 -3.93655568e-01
1.01034641e+00 -2.03276658e+00 -1.79799080e-01 -1.96362317e-01
-2.59836257e-01 8.80986154e-01 1.51391532e-02 1.24741130e-01
-1.06590763e-01 -1.84480533e-01 -6.05620086e-01 -4.03298795e-01
-2.02858582e-01 4.32363749e-01 -2.99598247e-01 1.63909942e-01
-3.85710336e-02 9.36561704e-01 -9.75662291e-01 -3.33763033e-01
5.35256922e-01 2.23068088e-01 -4.00057256e-01 -1.20722085e-01
-7.40853369e-01 4.43502039e-01 -3.25069785e-01 4.07792360e-01
6.93481147e-01 1.62273571e-01 -4.02575508e-02 -1.25704527e-01
-1.97661147e-01 -2.34033450e-01 -1.06703401e+00 1.81286681e+00
-2.25988120e-01 6.71792924e-01 -3.19844216e-01 -1.74515903e+00
8.90278995e-01 2.00438112e-01 6.88134551e-01 -1.14846230e+00
-3.95961478e-02 3.15109193e-01 -2.29913101e-01 -5.70741355e-01
8.14770460e-01 -1.90929532e-01 3.85732055e-02 2.78814256e-01
6.20203428e-02 -3.29063416e-01 5.07654488e-01 -8.95382091e-02
1.03090191e+00 2.98126101e-01 1.91500068e-01 1.87732317e-02
5.36817014e-01 5.85158169e-01 3.32105845e-01 9.10180867e-01
-4.56762046e-01 4.43854272e-01 3.16403270e-01 -5.69434702e-01
-9.93455887e-01 -7.54765987e-01 -2.51759082e-01 1.03711700e+00
4.06927764e-01 5.72165474e-02 -1.05289662e+00 -9.53570247e-01
-4.27092761e-02 9.69320238e-01 -4.00770634e-01 -3.42789501e-01
-5.56079686e-01 -9.07085061e-01 4.35100347e-01 2.51367927e-01
1.19098818e+00 -1.15113223e+00 -9.92691696e-01 4.84277546e-01
8.03587586e-03 -1.14050007e+00 2.42131025e-01 2.10187212e-01
-1.25967908e+00 -1.25963771e+00 -8.21583271e-01 -6.95805311e-01
2.79287696e-01 4.25593942e-01 1.00669110e+00 -2.07620844e-01
-2.35384792e-01 8.32852662e-01 -5.60120761e-01 -6.17317557e-01
-7.76278019e-01 3.86796743e-01 -1.37890369e-01 1.35993198e-01
5.43623149e-01 -4.54529464e-01 -5.15881300e-01 3.87248993e-01
-1.16713691e+00 -1.10970154e-01 4.32469755e-01 5.73113799e-01
5.50071239e-01 2.21314266e-01 8.36055279e-01 -1.39124703e+00
4.66450959e-01 -2.75554150e-01 -5.91049969e-01 2.80611932e-01
-9.57959592e-01 -2.12461427e-02 5.13547897e-01 -4.72818255e-01
-1.18274415e+00 1.79491773e-01 -4.88797367e-01 -3.21303785e-01
-5.89853585e-01 1.80892915e-01 3.46425064e-02 -8.90030935e-02
9.60874140e-01 4.71110076e-01 2.57043112e-02 -4.80111599e-01
4.26637560e-01 5.73385060e-01 5.15250862e-01 -4.89332050e-01
4.34837610e-01 5.91823578e-01 -1.13103539e-01 -1.13914144e+00
-7.58290231e-01 -6.77788734e-01 -7.09521294e-01 -3.27495247e-01
6.87920272e-01 -5.37459731e-01 -1.87369809e-01 9.41662908e-01
-1.14381313e+00 -7.29707778e-01 -9.06108260e-01 3.91965918e-02
-8.01921844e-01 5.34252822e-01 -2.23596320e-02 -6.13495409e-01
-6.77848160e-02 -1.17739642e+00 7.38858819e-01 2.74614334e-01
9.08800140e-02 -9.69346046e-01 -5.23040965e-02 4.89438295e-01
5.97545803e-01 -1.83532611e-02 9.27348375e-01 -9.40100491e-01
-5.60778916e-01 -2.47514009e-01 -1.45845208e-02 6.75172985e-01
-1.82188489e-02 -4.48967993e-01 -1.04684293e+00 -2.22497851e-01
9.40609500e-02 -2.08428875e-01 9.24959183e-01 1.47485137e-01
9.80849922e-01 2.31190503e-01 -3.03955257e-01 1.21186696e-01
1.32055950e+00 5.08757770e-01 7.95147002e-01 5.94823003e-01
3.31076771e-01 5.56842566e-01 8.82830381e-01 1.45085648e-01
2.04868600e-01 8.05557191e-01 4.38481420e-01 1.94942862e-01
-5.54833889e-01 1.23639263e-01 2.47493945e-02 3.59076142e-01
6.85624182e-02 -9.20723155e-02 -1.01800215e+00 6.98048353e-01
-1.88077807e+00 -9.15233493e-01 -1.15635589e-01 2.26016688e+00
7.07359910e-01 5.09973645e-01 2.22797722e-01 4.29736257e-01
7.14093149e-01 -1.83240115e-03 -7.52244890e-01 -3.72953147e-01
-2.31871977e-01 1.04752094e-01 4.16890234e-01 2.90943235e-01
-1.17678058e+00 1.13653398e+00 5.67222500e+00 1.02356684e+00
-1.31072927e+00 4.76080954e-01 4.96675104e-01 3.13880146e-01
1.23304114e-01 1.98983755e-02 -8.88620913e-01 4.83321905e-01
8.92875671e-01 -7.46166706e-02 9.50130746e-02 1.21865737e+00
5.75810485e-03 -6.36669874e-01 -9.22675431e-01 9.69369531e-01
1.42888725e-01 -1.10917854e+00 9.49584842e-02 -4.37993616e-01
9.52669799e-01 4.29210477e-02 4.55337018e-02 6.21170223e-01
-1.69421673e-01 -4.10151571e-01 6.37704015e-01 4.93337125e-01
4.34947103e-01 -6.27010643e-01 6.54156208e-01 5.34224451e-01
-8.52850199e-01 -1.78079396e-01 -2.63600767e-01 2.65677661e-01
1.86297536e-01 7.71240234e-01 -8.68884563e-01 4.96476084e-01
6.96263850e-01 5.02457738e-01 -6.83030725e-01 1.27543104e+00
-4.18817401e-02 7.59867132e-01 -2.42814258e-01 1.59048289e-01
2.63516039e-01 1.64718226e-01 6.80946171e-01 1.31702304e+00
-1.38198053e-02 -3.84434104e-01 1.61487192e-01 4.22909856e-01
1.25660107e-01 9.38562974e-02 -5.30876696e-01 8.92769396e-02
2.45701209e-01 6.86072469e-01 -1.22330403e+00 -6.05278015e-01
-2.69007385e-01 1.21777546e+00 1.28449962e-01 3.44713748e-01
-6.34979248e-01 -1.58971652e-01 4.17281240e-01 2.39739090e-01
3.16971987e-01 -4.46970880e-01 -3.90839159e-01 -6.77039444e-01
-5.43729849e-02 -7.40251720e-01 4.11830813e-01 -4.68208164e-01
-8.40216875e-01 3.68615776e-01 4.13608193e-01 -9.25064147e-01
-3.12937409e-01 -4.25440013e-01 -2.90219307e-01 2.81998038e-01
-1.69101751e+00 -9.40761149e-01 -4.09497201e-01 6.05233788e-01
1.15993047e+00 -2.79704541e-01 7.12900221e-01 3.84268492e-01
-1.99866325e-01 3.45441282e-01 2.78931439e-01 -3.15139174e-01
6.46048605e-01 -1.09159195e+00 1.67030796e-01 7.47204065e-01
1.57105565e-01 1.90000534e-01 9.28739250e-01 -6.64110422e-01
-1.18155193e+00 -1.18169606e+00 5.11749983e-01 -2.65103042e-01
3.73631090e-01 -1.10840648e-01 -1.24400032e+00 3.19937050e-01
-1.52033577e-02 -7.57104456e-02 8.62028077e-02 -7.63270482e-02
-2.11061358e-01 -5.70943177e-01 -1.16232169e+00 3.42863351e-01
1.01458406e+00 -3.55129182e-01 -6.64214134e-01 4.50264424e-01
6.71655476e-01 -2.93550670e-01 -2.69413561e-01 5.13874829e-01
1.60123274e-01 -1.20350170e+00 8.40172887e-01 -1.99074596e-01
-1.65389955e-01 -2.51027316e-01 2.08146170e-01 -1.06725764e+00
3.07496220e-01 -3.09041798e-01 -2.96525732e-02 1.05573654e+00
1.31933346e-01 -7.88906515e-01 1.23619711e+00 2.32257962e-01
-1.64537892e-01 -3.25053424e-01 -1.19905365e+00 -1.11800408e+00
5.51401302e-02 -8.16132426e-01 2.17584282e-01 7.02916145e-01
-5.39226353e-01 1.46146074e-01 -3.18609066e-02 -2.97642022e-01
5.28884828e-01 1.19790129e-01 8.65558863e-01 -1.27246702e+00
-1.62297800e-01 -5.26037455e-01 -7.54490554e-01 -8.42940092e-01
1.19746976e-01 -8.63611698e-01 -3.29587385e-02 -1.49657738e+00
-1.51009724e-01 -4.92305607e-01 -3.25879514e-01 4.73532557e-01
1.62372351e-01 2.68707752e-01 1.50799513e-01 9.66685116e-02
-9.31554079e-01 3.67369801e-01 9.37120080e-01 -3.86702150e-01
-4.56391782e-01 3.29033971e-01 -2.60660857e-01 5.67078710e-01
1.03127253e+00 -4.64397997e-01 -8.63571584e-01 -1.33040681e-01
7.22979009e-02 -2.79111445e-01 4.24792856e-01 -1.43770981e+00
5.75671136e-01 7.71999881e-02 -2.58368254e-01 -6.70413733e-01
2.08469227e-01 -8.28482330e-01 5.92242032e-02 4.41968709e-01
1.17030498e-02 -3.63198757e-01 5.37643015e-01 8.86080563e-01
-4.58655626e-01 -5.15523195e-01 1.08436739e+00 -4.27094102e-01
-1.64727390e+00 1.48489445e-01 -5.15085578e-01 9.28094536e-02
1.20132899e+00 -7.02501774e-01 1.38605852e-02 -1.36855617e-02
-9.11342204e-01 -5.31938709e-02 2.99670815e-01 5.34935653e-01
4.51694548e-01 -7.46709049e-01 -4.70561355e-01 5.94017422e-03
1.38966769e-01 1.66330278e-01 5.04043639e-01 7.02364564e-01
-3.95124882e-01 3.87391299e-01 -2.57672906e-01 -9.12449896e-01
-1.29778743e+00 7.07528770e-01 3.71104687e-01 -1.15108855e-01
-6.70539141e-01 5.88779986e-01 -1.13259614e-01 -2.54198462e-01
3.79820287e-01 -1.27164600e-02 -2.62806594e-01 4.81771976e-01
3.44998479e-01 5.17117441e-01 6.70053780e-01 -3.88112396e-01
-4.68275286e-02 4.50061619e-01 -2.70400643e-01 -4.24713418e-02
1.19382656e+00 -1.66589797e-01 6.81498870e-02 6.10591173e-01
9.17696059e-01 -6.91264212e-01 -1.34208941e+00 -3.30135375e-01
4.68870312e-01 -7.61351138e-02 -1.64622277e-01 -8.33716750e-01
-7.94324875e-01 9.75488544e-01 1.01271987e+00 3.00611645e-01
1.21831787e+00 -9.57665145e-02 6.83142900e-01 7.13210881e-01
7.18129814e-01 -1.70876026e+00 3.10356319e-01 8.24682057e-01
6.32045925e-01 -1.22356462e+00 -1.82643905e-01 -1.92565933e-01
-3.61381531e-01 1.16534889e+00 4.85211909e-01 2.05681846e-01
6.54838860e-01 1.76770195e-01 4.48127203e-02 -1.90805793e-02
-1.02797180e-01 -5.39516747e-01 -1.45125613e-01 7.97449648e-01
-6.60557300e-02 -2.06351236e-01 -5.48206270e-01 2.42595896e-01
7.46854022e-02 4.65772212e-01 2.97228128e-01 1.08980358e+00
-8.05587351e-01 -1.33799684e+00 -2.74633020e-01 1.83489740e-01
-5.89854643e-02 3.25945109e-01 -1.16261564e-01 9.46224749e-01
5.04919529e-01 8.54349911e-01 -1.77970916e-01 -1.60123825e-01
6.02659702e-01 5.73850930e-01 5.14432490e-01 -6.05996668e-01
-4.15595323e-01 -5.52032828e-01 2.14082915e-02 -2.76193887e-01
-7.33144939e-01 -7.34878480e-01 -1.25814736e+00 -1.04922891e-01
-1.25238448e-01 1.59506127e-01 1.18502450e+00 1.06844425e+00
1.15121104e-01 4.82375115e-01 2.33491868e-01 -7.39059985e-01
-3.40964139e-01 -5.40135145e-01 -3.91861320e-01 5.96596956e-01
-2.70811166e-03 -6.16428077e-01 -7.97664225e-02 2.15863764e-01] | [9.507844924926758, 1.9306806325912476] |
9b7a89b9-4417-4b50-b2ba-04a9c9b143ae | hvtsurv-hierarchical-vision-transformer-for | 2306.17373 | null | https://arxiv.org/abs/2306.17373v1 | https://arxiv.org/pdf/2306.17373v1.pdf | HVTSurv: Hierarchical Vision Transformer for Patient-Level Survival Prediction from Whole Slide Image | Survival prediction based on whole slide images (WSIs) is a challenging task for patient-level multiple instance learning (MIL). Due to the vast amount of data for a patient (one or multiple gigapixels WSIs) and the irregularly shaped property of WSI, it is difficult to fully explore spatial, contextual, and hierarchical interaction in the patient-level bag. Many studies adopt random sampling pre-processing strategy and WSI-level aggregation models, which inevitably lose critical prognostic information in the patient-level bag. In this work, we propose a hierarchical vision Transformer framework named HVTSurv, which can encode the local-level relative spatial information, strengthen WSI-level context-aware communication, and establish patient-level hierarchical interaction. Firstly, we design a feature pre-processing strategy, including feature rearrangement and random window masking. Then, we devise three layers to progressively obtain patient-level representation, including a local-level interaction layer adopting Manhattan distance, a WSI-level interaction layer employing spatial shuffle, and a patient-level interaction layer using attention pooling. Moreover, the design of hierarchical network helps the model become more computationally efficient. Finally, we validate HVTSurv with 3,104 patients and 3,752 WSIs across 6 cancer types from The Cancer Genome Atlas (TCGA). The average C-Index is 2.50-11.30% higher than all the prior weakly supervised methods over 6 TCGA datasets. Ablation study and attention visualization further verify the superiority of the proposed HVTSurv. Implementation is available at: https://github.com/szc19990412/HVTSurv. | ['Yongbing Zhang', 'Guojun Liu', 'Jian Zhang', 'Hao Bian', 'Yang Chen', 'Zhuchen Shao'] | 2023-06-30 | null | null | null | null | ['whole-slide-images', 'multiple-instance-learning'] | ['computer-vision', 'methodology'] | [ 1.55804381e-01 -4.78746220e-02 -3.22294563e-01 -2.20899194e-01
-1.28462660e+00 -1.59352779e-01 3.05921435e-01 5.36896229e-01
-3.49915594e-01 8.58759522e-01 7.43545771e-01 -3.99245590e-01
-5.71108937e-01 -7.36502647e-01 -4.84826207e-01 -1.23521376e+00
-1.85142577e-01 2.89913654e-01 1.57889366e-01 -5.43129221e-02
7.83183277e-02 4.25443351e-01 -1.05220008e+00 6.09734654e-01
8.20634246e-01 9.44142401e-01 2.67189026e-01 8.00148964e-01
-2.10059695e-02 5.85024953e-01 -4.18945014e-01 9.78168771e-02
-2.53742188e-01 -2.83908308e-01 -7.29773223e-01 -3.44053537e-01
2.28085771e-01 4.43097502e-02 -4.35990840e-01 7.56670892e-01
6.94691241e-01 -2.63183415e-01 5.92024088e-01 -1.04627514e+00
-4.21445638e-01 5.62096655e-01 -8.12183201e-01 3.99640739e-01
-9.98734087e-02 4.02581632e-01 1.16894054e+00 -8.16580653e-01
5.75510383e-01 7.92198777e-01 7.80353904e-01 3.15795928e-01
-9.77117836e-01 -6.01466596e-01 6.06167950e-02 1.75842375e-01
-1.71863222e+00 -1.96163952e-01 5.56961238e-01 -2.96863228e-01
8.77613723e-01 8.46683025e-01 8.29612970e-01 8.50380898e-01
6.24382675e-01 8.45526040e-01 1.13561678e+00 -2.24878117e-01
-1.29933566e-01 -1.19757339e-01 6.27975702e-01 6.10848010e-01
1.10295095e-01 3.81780416e-02 -4.75188583e-01 -2.67065138e-01
7.71842659e-01 5.11859536e-01 -5.89809954e-01 9.28215086e-02
-1.59176159e+00 6.81409359e-01 9.81745899e-01 4.94537145e-01
-3.60966325e-02 1.49550378e-01 4.80179459e-01 5.14999032e-03
3.70863557e-01 1.02599688e-01 -4.14432824e-01 3.07441771e-01
-9.84571099e-01 -1.11488432e-01 1.51190653e-01 6.84969723e-01
1.95218250e-01 -5.43820441e-01 -7.57727921e-01 6.32916927e-01
2.81769633e-01 1.52144328e-01 8.47281456e-01 -2.16376066e-01
2.51104385e-01 8.36697698e-01 -3.42307746e-01 -7.78378367e-01
-9.69344914e-01 -7.74364114e-01 -1.53261328e+00 -2.90603101e-01
2.83427984e-01 2.46252030e-01 -9.74030614e-01 1.58023512e+00
2.18774542e-01 4.77758139e-01 5.65387532e-02 7.66029775e-01
1.39440179e+00 3.75748873e-01 2.57485420e-01 -2.44222432e-01
1.75730550e+00 -8.61855984e-01 -5.95404208e-01 1.71658084e-01
1.08403027e+00 -2.68031985e-01 1.06577170e+00 -2.35430710e-02
-7.96313345e-01 -3.52863938e-01 -8.72394562e-01 -1.79411843e-01
-4.55445051e-01 2.74164379e-01 7.40440428e-01 1.79913655e-01
-1.17545474e+00 2.64910966e-01 -7.35991359e-01 -3.84106666e-01
8.63074183e-01 5.28784931e-01 -5.71225822e-01 -8.72126818e-02
-1.30768204e+00 5.34788251e-01 2.90698975e-01 2.43051514e-01
-8.16277802e-01 -1.09710026e+00 -7.79087424e-01 1.02443390e-01
1.16412409e-01 -9.28481519e-01 6.43157005e-01 -3.15806627e-01
-9.95464861e-01 8.60326171e-01 -4.23517734e-01 -1.22900739e-01
3.82918835e-01 3.40830445e-01 -2.27212295e-01 -2.12368853e-02
-2.97356043e-02 5.12574911e-01 2.32531726e-01 -8.69641006e-01
-6.07621193e-01 -8.17503214e-01 -4.17902410e-01 4.60263550e-01
-4.35189486e-01 -2.46555239e-01 -6.63733959e-01 -7.68925667e-01
1.79674670e-01 -6.95863962e-01 -5.13069212e-01 -1.12994440e-01
-7.92718351e-01 -2.04562813e-01 5.13292551e-01 -7.61167169e-01
1.48929560e+00 -2.28017426e+00 1.81827351e-01 3.10070157e-01
5.68301320e-01 -1.04992792e-01 -7.86140412e-02 4.79546823e-02
-3.53191495e-01 2.39890903e-01 -3.03043962e-01 -3.22028846e-01
-4.46936011e-01 -1.09371558e-01 1.35020301e-01 5.26044488e-01
-5.74215464e-02 1.41685903e+00 -6.67571485e-01 -6.15561783e-01
3.00458409e-02 3.67691994e-01 -5.50959766e-01 3.17339301e-02
3.24252605e-01 5.96030414e-01 -4.61717904e-01 9.22875226e-01
5.22688329e-01 -6.71067417e-01 -3.00150197e-02 -4.11209226e-01
1.01685725e-01 -1.45487487e-01 -5.63911498e-01 1.69607425e+00
-1.11050941e-01 4.52387184e-01 -1.34635285e-01 -9.64215577e-01
5.31029999e-01 2.69048601e-01 6.03673577e-01 -4.95430380e-01
1.13901667e-01 -1.24647327e-01 7.12455884e-02 -4.48940665e-01
2.64470875e-02 -1.86181232e-01 -3.58078063e-01 8.06844886e-03
-1.24928787e-01 1.17361180e-01 -2.00224668e-01 2.64082134e-01
1.45248020e+00 -4.29584920e-01 5.27622879e-01 -4.74799484e-01
5.38661599e-01 -9.22724009e-02 8.22693944e-01 7.25449324e-01
-3.69697779e-01 8.62701476e-01 6.75962090e-01 -3.56501162e-01
-6.07948422e-01 -1.05545366e+00 -5.36158323e-01 8.25466156e-01
2.54106242e-02 -2.89597422e-01 -4.15846229e-01 -6.91588104e-01
1.60676569e-01 3.70534897e-01 -1.15145218e+00 -2.77127802e-01
-3.95256072e-01 -1.50445306e+00 6.98449194e-01 7.67818749e-01
5.38866758e-01 -7.74311066e-01 -2.04266638e-01 1.11729890e-01
-2.45523348e-01 -6.32862568e-01 -7.17846632e-01 1.78952038e-01
-8.73622656e-01 -1.05082893e+00 -7.49004662e-01 -6.45453632e-01
8.78027916e-01 1.40678361e-01 1.01517487e+00 3.39712232e-01
-8.41666341e-01 -1.06705636e-01 -1.24944337e-01 -4.01250422e-01
9.00665969e-02 4.81565520e-02 -2.82389224e-01 -8.49837139e-02
4.36169475e-01 -3.76139194e-01 -9.82014894e-01 1.94699764e-01
-7.41585076e-01 4.42534328e-01 8.54827881e-01 1.31713700e+00
8.20845664e-01 5.52457646e-02 5.20534813e-01 -7.84475505e-01
3.21022213e-01 -5.28634489e-01 -2.76108772e-01 4.26461279e-01
-1.67946398e-01 -3.31939399e-01 4.27160054e-01 6.47456245e-03
-7.53830194e-01 -1.36308977e-02 -1.99089661e-01 -6.63674027e-02
-1.80506125e-01 6.75693512e-01 -2.77506292e-01 -3.08567826e-02
4.53568697e-01 3.73299092e-01 6.91270754e-02 3.40610035e-02
-9.16466862e-03 6.93893790e-01 1.97450206e-01 2.03063879e-02
4.15513664e-01 4.64173824e-01 -1.54794548e-02 -6.99874163e-01
-8.47489178e-01 -4.20004606e-01 -6.48388743e-01 1.13073356e-01
9.30140615e-01 -8.55027199e-01 -1.10907435e+00 6.76554859e-01
-6.31107509e-01 -4.17458653e-01 -2.49659851e-01 3.29909623e-01
-1.67883024e-01 -7.64898630e-03 -9.32056487e-01 -2.50930399e-01
-6.70601130e-01 -1.26375270e+00 1.26027989e+00 3.71673584e-01
-1.27600327e-01 -1.02902627e+00 -4.89603654e-02 5.72769403e-01
4.42411572e-01 4.05690193e-01 1.12568688e+00 -6.16584361e-01
-5.81903994e-01 -3.16116899e-01 -6.10966682e-01 -2.60243595e-01
2.50157207e-01 -8.87663737e-02 -9.54370856e-01 -4.28418547e-01
-4.53010738e-01 -8.36693421e-02 1.21112859e+00 9.59820628e-01
1.77497733e+00 -6.95897192e-02 -8.79456460e-01 9.86102402e-01
1.49098706e+00 2.93341607e-01 6.06478274e-01 1.48710996e-01
8.62758040e-01 4.24562573e-01 3.93517435e-01 3.20926279e-01
6.02461159e-01 4.72073495e-01 2.32614070e-01 -4.53839421e-01
-8.77111033e-02 -2.58383285e-02 -2.31025338e-01 7.89209247e-01
-6.01900034e-02 -2.01303244e-01 -1.12445498e+00 5.75882673e-01
-1.75629151e+00 -7.86709666e-01 -3.07574738e-02 2.06376290e+00
9.10044134e-01 3.09560332e-03 -3.66144210e-01 4.07033451e-02
3.91013086e-01 1.12763077e-01 -5.37957251e-01 1.44197270e-01
-1.80287227e-01 -8.57766047e-02 5.47021270e-01 4.38758463e-01
-1.22472453e+00 6.05027020e-01 5.29007530e+00 1.10365009e+00
-8.84343147e-01 1.37519196e-01 1.37915313e+00 -4.00195301e-01
-4.04227197e-01 -3.46235961e-01 -1.00212181e+00 5.37814438e-01
6.33888483e-01 2.08459981e-02 -1.66336015e-01 2.26070419e-01
2.35969096e-01 -1.73186705e-01 -9.78905499e-01 9.97700870e-01
-1.21234180e-02 -1.78242064e+00 -8.09402689e-02 2.49923304e-01
4.13779110e-01 -1.60258319e-02 9.43567455e-02 3.19116861e-01
2.65922278e-01 -1.41522539e+00 -9.77726281e-02 7.95593977e-01
1.12390685e+00 -6.60910547e-01 1.21140218e+00 2.34739751e-01
-1.13867724e+00 -1.19782113e-01 -6.66896701e-02 4.17670846e-01
-1.17068693e-01 6.57815933e-01 -8.36097300e-01 8.40857863e-01
8.55428159e-01 7.81630874e-01 -8.14730763e-01 1.07862020e+00
2.87964672e-01 6.03059888e-01 -4.01264429e-01 -1.08976550e-01
1.56438380e-01 1.82511434e-01 3.14597458e-01 1.26915741e+00
5.10810018e-01 5.71910024e-01 -1.05824023e-01 2.13178426e-01
7.91315511e-02 2.02822983e-02 -2.38712117e-01 2.66660690e-01
3.64141256e-01 1.37290633e+00 -7.14976788e-01 -2.55239397e-01
-3.89676452e-01 8.31706822e-01 3.13060880e-01 4.31921542e-01
-8.16266000e-01 -3.99629176e-01 5.11147976e-01 1.82998255e-01
-8.32159631e-03 3.41505527e-01 -6.37114704e-01 -1.06742096e+00
-2.95837432e-01 -6.37349010e-01 9.03271616e-01 -5.42614996e-01
-1.34107649e+00 6.10196710e-01 -1.61288008e-01 -1.26895106e+00
3.95704955e-01 -5.02131224e-01 -8.86830389e-01 7.75024176e-01
-1.38244343e+00 -1.40104222e+00 -7.62144208e-01 6.54599249e-01
3.47940356e-01 -1.31256670e-01 9.90790665e-01 1.13883622e-01
-9.33393896e-01 1.06334376e+00 1.41399086e-01 2.64194012e-01
6.53741896e-01 -1.21486866e+00 -1.37629807e-01 2.23797292e-01
-4.58904475e-01 5.77488005e-01 2.85851151e-01 -5.62607408e-01
-1.27755284e+00 -1.32009077e+00 6.51001215e-01 -4.74546939e-01
6.17021799e-01 -3.01786125e-01 -1.01112545e+00 7.46559441e-01
7.74452686e-02 5.22934139e-01 1.19093096e+00 1.37827486e-01
2.12586865e-01 -2.92090088e-01 -1.21476388e+00 7.77854443e-01
9.98752117e-01 -2.18497828e-01 -2.35586956e-01 3.13505590e-01
8.07754934e-01 -5.05624413e-01 -1.27122748e+00 7.83422232e-01
3.54510367e-01 -6.50705695e-01 1.07795334e+00 -5.25019407e-01
2.86070883e-01 -3.90314132e-01 -4.34497707e-02 -1.26016057e+00
-7.80313492e-01 -6.03102222e-02 1.84172213e-01 9.91305649e-01
6.17440224e-01 -7.88889468e-01 8.20364892e-01 4.12764609e-01
-4.17934120e-01 -1.38510454e+00 -1.24061310e+00 -1.36683673e-01
1.27219766e-01 -2.14614823e-01 6.70728147e-01 9.55306411e-01
1.37520656e-01 8.19649026e-02 -1.15941912e-01 2.92497694e-01
6.39204443e-01 6.39105737e-02 3.20441633e-01 -9.92067695e-01
-2.98035413e-01 -6.97982013e-01 -5.85044205e-01 -5.31687140e-01
-3.75252604e-01 -1.06368876e+00 -1.74074993e-01 -1.60174382e+00
7.39542544e-01 -5.45213819e-01 -9.96724904e-01 7.08360553e-01
-5.19321263e-01 3.00977796e-01 -2.97700465e-01 1.49576828e-01
-5.66566586e-01 6.14254117e-01 1.50954425e+00 -2.92647183e-01
-1.55756995e-01 -1.89973980e-01 -9.47548509e-01 6.25495970e-01
8.20864618e-01 -1.17733546e-01 -1.57110691e-01 -2.23475378e-02
-1.57703057e-01 3.33176136e-01 5.41354477e-01 -8.97941709e-01
4.48992252e-01 -2.40609854e-01 9.13978994e-01 -9.76173520e-01
2.53658265e-01 -5.47902465e-01 1.70419708e-01 5.92402995e-01
-2.96450406e-01 -2.10176528e-01 3.46373588e-01 4.81647730e-01
-3.63997400e-01 1.82425916e-01 5.68733752e-01 -5.67153376e-03
-4.24471229e-01 7.67487288e-01 -1.52251035e-01 -2.43061885e-01
1.39308822e+00 -2.67056316e-01 -5.35340250e-01 2.42356494e-01
-9.14895475e-01 6.70438111e-01 1.12973496e-01 1.03754804e-01
7.77933538e-01 -1.48993552e+00 -9.12210464e-01 4.30675000e-01
5.02973795e-01 3.43517095e-01 1.02508664e+00 1.26147032e+00
-4.24914151e-01 4.52662915e-01 3.75324190e-02 -7.14820802e-01
-1.41821015e+00 4.49905097e-01 2.74745613e-01 -8.02411497e-01
-6.71816945e-01 1.35986161e+00 7.97619164e-01 -4.15307999e-01
1.30312666e-01 -3.75574499e-01 -3.64192784e-01 1.37255892e-01
6.44865870e-01 -5.78136824e-04 6.74303323e-02 -4.31515217e-01
-4.93814588e-01 6.65202439e-01 -4.83325571e-01 3.15492302e-01
1.23516965e+00 3.10177188e-02 -3.49571258e-01 3.29167753e-01
1.32010853e+00 -2.82714218e-01 -1.01677263e+00 -3.18089366e-01
-2.10145637e-01 -2.86352158e-01 2.44181663e-01 -8.14639926e-01
-1.01774967e+00 7.63292789e-01 7.70821273e-01 -1.36560082e-01
1.44517195e+00 2.17723697e-01 7.42115021e-01 -2.33708292e-01
1.07578672e-01 -6.44283831e-01 -1.84730828e-01 2.44640842e-01
9.90593076e-01 -1.31413805e+00 -3.02045215e-02 -3.48107249e-01
-5.64134896e-01 7.53677309e-01 6.86163247e-01 1.50581256e-01
8.54055524e-01 4.99095052e-01 3.11264135e-02 -2.89423823e-01
-9.27100062e-01 7.31711015e-02 3.35793793e-01 2.43572280e-01
4.59382504e-01 4.09338713e-01 -1.37833253e-01 1.03247750e+00
-1.13531299e-01 -3.31470892e-02 1.03873491e-01 6.50739491e-01
-2.43992507e-01 -5.89785516e-01 -2.99191773e-01 9.95482326e-01
-4.04510111e-01 -4.22810107e-01 9.89228860e-02 7.89691627e-01
-9.21214931e-04 4.42347437e-01 2.16681272e-01 -3.94925296e-01
2.55506426e-01 -1.72162607e-01 6.04535043e-02 -3.62119615e-01
-5.87022841e-01 1.53271511e-01 -2.49546915e-01 -3.99093628e-01
4.70165396e-03 -7.12617755e-01 -1.43085706e+00 -4.34802957e-02
-1.56844705e-01 5.99365495e-02 8.31440836e-02 7.65776634e-01
3.71738762e-01 1.03350580e+00 4.85214621e-01 -5.02531111e-01
-6.20040745e-02 -1.03572953e+00 -5.45334220e-01 1.58209562e-01
5.71589291e-01 -4.15269881e-01 -2.98987865e-01 -2.49678597e-01] | [15.106377601623535, -2.8883988857269287] |
1332919f-714e-43e3-ac25-ce77db2cc4c4 | mbptrack-improving-3d-point-cloud-tracking | 2303.05071 | null | https://arxiv.org/abs/2303.05071v1 | https://arxiv.org/pdf/2303.05071v1.pdf | MBPTrack: Improving 3D Point Cloud Tracking with Memory Networks and Box Priors | 3D single object tracking has been a crucial problem for decades with numerous applications such as autonomous driving. Despite its wide-ranging use, this task remains challenging due to the significant appearance variation caused by occlusion and size differences among tracked targets. To address these issues, we present MBPTrack, which adopts a Memory mechanism to utilize past information and formulates localization in a coarse-to-fine scheme using Box Priors given in the first frame. Specifically, past frames with targetness masks serve as an external memory, and a transformer-based module propagates tracked target cues from the memory to the current frame. To precisely localize objects of all sizes, MBPTrack first predicts the target center via Hough voting. By leveraging box priors given in the first frame, we adaptively sample reference points around the target center that roughly cover the target of different sizes. Then, we obtain dense feature maps by aggregating point features into the reference points, where localization can be performed more effectively. Extensive experiments demonstrate that MBPTrack achieves state-of-the-art performance on KITTI, nuScenes and Waymo Open Dataset, while running at 50 FPS on a single RTX3090 GPU. | ['Song-Hai Zhang', 'Yu-Kun Lai', 'Yuan-Chen Guo', 'Tian-Xing Xu'] | 2023-03-09 | null | null | null | null | ['3d-single-object-tracking'] | ['computer-vision'] | [-1.32969141e-01 -5.52403510e-01 -3.16920221e-01 -2.34137222e-01
-5.74730277e-01 -4.98149127e-01 4.94808793e-01 3.90120894e-02
-4.59343463e-01 4.58803058e-01 -1.53771713e-01 1.08321644e-01
2.34054521e-01 -7.64717460e-01 -9.46656108e-01 -7.93886185e-01
-2.66534574e-02 3.07977855e-01 1.11181462e+00 1.11831121e-01
3.28911901e-01 7.77374506e-01 -1.72228718e+00 1.06430920e-02
5.06243885e-01 1.39785123e+00 4.85868126e-01 4.43903148e-01
-5.88751510e-02 6.52534068e-01 -5.18899679e-01 -1.24599934e-01
4.16216165e-01 3.11399311e-01 -9.47270244e-02 1.56694785e-01
8.57674778e-01 -4.43701357e-01 -7.30341852e-01 1.09780228e+00
2.83928394e-01 3.28992248e-01 1.85259998e-01 -1.27279675e+00
-2.42699236e-01 7.42934644e-02 -9.81030285e-01 5.35802186e-01
5.79248331e-02 4.25645709e-01 4.92118835e-01 -9.71227467e-01
5.16099095e-01 1.35631561e+00 6.75522387e-01 3.66696894e-01
-9.86437321e-01 -1.11837828e+00 4.94639069e-01 2.54022628e-01
-1.29310024e+00 -3.74210596e-01 5.88721931e-01 -4.45021451e-01
6.84687614e-01 5.33089004e-02 6.80013239e-01 6.30840003e-01
4.48630959e-01 5.59261143e-01 9.07246411e-01 2.88477447e-02
3.56928110e-02 -1.39864177e-01 1.97082594e-01 7.97600746e-01
4.32408005e-01 4.42439407e-01 -7.10687697e-01 -1.41282737e-01
9.27750349e-01 3.54332119e-01 -3.44335556e-01 -7.09189177e-01
-1.48995233e+00 6.24271750e-01 8.41026127e-01 -1.74255729e-01
-2.67015636e-01 3.28724921e-01 2.76432008e-01 -1.33001164e-01
4.20015365e-01 -2.60124356e-01 -4.29130286e-01 1.18341558e-02
-7.80859351e-01 2.16576591e-01 3.20703387e-01 1.24022973e+00
9.84034419e-01 -1.15719862e-01 -3.95950437e-01 3.55416059e-01
4.18202400e-01 1.10344768e+00 1.75087139e-01 -9.85063136e-01
4.80276763e-01 5.46165884e-01 4.34288353e-01 -1.09826148e+00
-2.50978321e-01 -6.06892884e-01 -4.91722465e-01 4.99525040e-01
4.46197778e-01 -2.99483426e-02 -8.62254918e-01 1.46937811e+00
1.04970896e+00 6.02585673e-01 -3.12455922e-01 1.27947736e+00
6.85964823e-01 6.97090030e-01 8.89983699e-02 -1.51885450e-01
1.53935707e+00 -1.17288733e+00 -3.94600332e-01 -4.73979384e-01
3.96118999e-01 -8.95458341e-01 4.60513979e-01 2.02290788e-01
-7.51306951e-01 -9.89200354e-01 -9.09957588e-01 -3.18103954e-02
-1.36970401e-01 3.27628642e-01 5.24988234e-01 5.53156972e-01
-6.42588019e-01 3.96929979e-01 -1.22805989e+00 -2.21285015e-01
5.54629922e-01 3.03210855e-01 -2.64884025e-01 -1.37445524e-01
-6.36845112e-01 8.41567397e-01 3.12147051e-01 1.75750062e-01
-8.71531129e-01 -9.40367281e-01 -8.64392996e-01 -2.20204726e-01
5.23484290e-01 -6.41174316e-01 1.05852044e+00 -4.43392068e-01
-1.24337602e+00 7.05166578e-01 -4.90341187e-01 -7.11218655e-01
5.21217525e-01 -5.61111271e-01 -3.72423261e-01 -1.28458530e-01
2.50763059e-01 7.43077576e-01 9.19162989e-01 -1.02330291e+00
-1.24243319e+00 -6.59709454e-01 -1.00180440e-01 2.30920106e-01
7.55722895e-02 -2.55360641e-02 -8.11147451e-01 -3.86428356e-01
3.49958897e-01 -1.00667226e+00 -1.43215254e-01 6.33390665e-01
-9.01959240e-02 -2.33661890e-01 1.30587518e+00 -1.90909654e-01
9.29595292e-01 -2.30318403e+00 -1.56529039e-01 -5.35389893e-02
2.90472984e-01 1.32115528e-01 1.91914141e-01 -2.46264815e-01
3.12079698e-01 -7.67962813e-01 2.80647188e-01 -4.20751572e-01
-1.35908380e-01 7.24096373e-02 -5.61936140e-01 8.67541373e-01
2.27038134e-02 7.64549553e-01 -8.70771468e-01 -6.53578758e-01
5.72215855e-01 6.23411834e-01 -3.39949995e-01 2.07000330e-01
-2.62769312e-01 4.46684033e-01 -7.59408176e-01 7.59388506e-01
1.03992629e+00 -1.79012835e-01 -4.90646929e-01 -4.84212250e-01
-6.93189681e-01 -4.38628644e-02 -1.21557260e+00 1.69292235e+00
-1.11854583e-01 6.58398867e-01 6.16806298e-02 -3.54180694e-01
9.52744663e-01 -3.40784132e-01 4.81710613e-01 -6.15871847e-01
3.14470768e-01 1.29054831e-02 -1.39444917e-01 -1.18053742e-01
7.47738242e-01 2.93695509e-01 2.79431846e-02 2.38871966e-02
-3.38929266e-01 1.47072196e-01 4.97147217e-02 -1.09832525e-01
1.08469176e+00 4.10491824e-01 2.64318902e-02 -4.36984263e-02
5.43632030e-01 5.77415526e-01 7.86822975e-01 7.02644944e-01
-5.41148782e-01 5.34177303e-01 -2.81159163e-01 -6.64040148e-01
-7.47156322e-01 -1.08163965e+00 -2.28495806e-01 1.08912778e+00
7.38836288e-01 -1.08503990e-01 -2.42493838e-01 -4.96046901e-01
2.31860384e-01 3.00758719e-01 -6.10923707e-01 -1.03038475e-01
-9.01770592e-01 -3.05413604e-01 1.95574053e-02 5.83875656e-01
6.70298517e-01 -7.97020495e-01 -1.25071108e+00 3.32707733e-01
1.05732799e-01 -1.21385574e+00 -7.24304855e-01 6.97811618e-02
-8.66968870e-01 -1.03876007e+00 -7.03239202e-01 -5.65250039e-01
5.27242422e-01 9.95862126e-01 9.06862915e-01 -1.46928906e-01
-3.68256241e-01 2.09030196e-01 -1.05703317e-01 -3.09538007e-01
8.99989381e-02 -3.06463003e-01 2.50317738e-03 4.07319926e-02
4.12647486e-01 -2.39056218e-02 -9.04078424e-01 6.32879317e-01
-4.00155872e-01 6.70328885e-02 6.02024853e-01 5.00580132e-01
1.04483020e+00 -5.03189787e-02 -2.10002750e-01 -3.81858319e-01
-3.06129485e-01 -3.41304809e-01 -1.27800941e+00 1.03252299e-01
1.09118201e-01 -1.39093190e-01 1.88101366e-01 -7.89251268e-01
-9.08788383e-01 4.84765351e-01 2.86143422e-01 -1.05003130e+00
-3.61239314e-02 -2.32573688e-01 -1.72480777e-01 -4.43920553e-01
4.17987078e-01 3.09713036e-01 3.02827102e-03 -2.41223752e-01
2.81447649e-01 2.47514069e-01 8.01044226e-01 -5.04559457e-01
9.47568238e-01 8.66570175e-01 7.28259012e-02 -4.81460571e-01
-1.07719278e+00 -6.14531040e-01 -5.62083423e-01 -5.48969269e-01
9.53130007e-01 -1.21422708e+00 -9.08958495e-01 4.14465874e-01
-1.21500003e+00 -1.90075070e-01 -1.79903820e-01 5.55770040e-01
-2.31553376e-01 9.66791213e-02 -2.81230301e-01 -6.14236712e-01
-2.63808429e-01 -1.21355665e+00 1.39060962e+00 6.71521008e-01
2.74038911e-01 -6.06473386e-01 2.57938951e-02 -1.67584382e-02
3.39626759e-01 2.81054914e-01 9.89654660e-02 -4.28478867e-02
-1.18088055e+00 -1.61974236e-01 -5.22724271e-01 -3.30521286e-01
2.35891119e-02 -6.65018335e-02 -7.56103754e-01 -4.50086474e-01
5.07055856e-02 1.15817577e-01 8.03903341e-01 4.89118814e-01
1.03784645e+00 3.01798254e-01 -9.83374357e-01 7.54391432e-01
1.32534575e+00 2.44295970e-01 2.26135120e-01 4.65829462e-01
6.78353190e-01 1.12271927e-01 1.26558805e+00 4.35530573e-01
4.71771926e-01 7.58435786e-01 5.57234287e-01 1.38016909e-01
-3.52298796e-01 -1.60129592e-01 3.22327316e-01 2.54279375e-01
2.31224641e-01 1.16648011e-01 -5.76041281e-01 3.36673707e-01
-1.78121054e+00 -8.75634074e-01 -2.70432472e-01 2.31503725e+00
4.48129743e-01 4.71813083e-01 5.46333380e-02 -3.65861088e-01
9.96211946e-01 2.36461520e-01 -9.55915213e-01 5.10558248e-01
4.89218310e-02 -1.72718614e-01 9.06263113e-01 2.41415903e-01
-1.22092593e+00 9.48428035e-01 5.18930674e+00 8.73148739e-01
-1.37563968e+00 1.95379153e-01 4.20969605e-01 -1.33310899e-01
4.28246349e-01 4.83210124e-02 -1.59384573e+00 6.91680670e-01
4.66156632e-01 -8.11013356e-02 -1.78126059e-02 1.05918777e+00
2.69384347e-02 -3.81584346e-01 -8.98568928e-01 1.02572429e+00
-1.89558547e-02 -1.20995247e+00 -3.12500387e-01 1.97744891e-02
5.60910881e-01 4.30594087e-01 1.03609934e-01 3.31172496e-01
1.73102006e-01 -4.48275357e-01 1.09762943e+00 4.46222633e-01
5.16686499e-01 -5.55782795e-01 3.25356275e-01 3.93863767e-01
-1.86657023e+00 8.66190158e-03 -9.35843945e-01 4.38940749e-02
1.69541791e-01 5.95063031e-01 -5.37961781e-01 3.10776442e-01
9.45158124e-01 7.54609585e-01 -5.73091269e-01 1.53345835e+00
8.10378864e-02 1.66135445e-01 -6.45114601e-01 4.78911307e-03
3.57024581e-03 1.31519035e-01 5.95297754e-01 8.62650812e-01
5.41938961e-01 1.85502872e-01 5.63199699e-01 6.77301168e-01
1.93368807e-01 -2.93689042e-01 -3.15482736e-01 6.44042432e-01
7.83823550e-01 1.31839800e+00 -8.71345878e-01 -3.98563623e-01
-5.25889993e-01 7.61636198e-01 3.71253997e-01 -3.48772854e-02
-1.33323359e+00 -2.17059299e-01 8.40345025e-01 2.61600673e-01
7.38579214e-01 -5.39449453e-01 -4.13580164e-02 -1.09009159e+00
5.84312901e-02 -3.10156852e-01 3.37477356e-01 -7.86445260e-01
-9.76177633e-01 8.24477196e-01 -1.06369726e-01 -1.72845304e+00
3.46077025e-01 -5.37823677e-01 -4.79557276e-01 7.39376843e-01
-1.67467546e+00 -1.05602729e+00 -8.58343601e-01 6.18432820e-01
6.25641465e-01 1.94557413e-01 2.37165704e-01 3.35197449e-01
-5.06833196e-01 3.33047897e-01 -8.09239075e-02 4.11135331e-02
7.95796990e-01 -7.31602967e-01 4.97780710e-01 8.51829708e-01
5.23717888e-02 3.86686385e-01 6.13816977e-01 -8.73602867e-01
-1.66300094e+00 -1.50233972e+00 3.51478934e-01 -5.74188590e-01
7.14631736e-01 -2.45391861e-01 -8.44874978e-01 6.27830565e-01
-1.60198256e-01 8.21210027e-01 3.01085087e-03 -4.91846979e-01
-2.10110337e-01 -3.14010233e-01 -9.04118299e-01 3.66186172e-01
1.20961463e+00 -6.11268282e-02 -5.03225207e-01 2.41640151e-01
8.82395804e-01 -1.21056223e+00 -6.61891103e-01 3.51444066e-01
5.03953516e-01 -8.71389747e-01 1.13647676e+00 4.22313251e-03
-1.80682778e-01 -9.28876162e-01 -1.15614049e-01 -9.27984118e-01
-4.13353801e-01 -4.35274094e-01 -4.30917352e-01 1.06047928e+00
-8.37764964e-02 -6.24589324e-01 1.01559031e+00 5.37101984e-01
-2.49959558e-01 -6.33621156e-01 -1.15354025e+00 -7.31650710e-01
-4.44498450e-01 -4.03715402e-01 7.00719535e-01 3.19488525e-01
-6.45518720e-01 -7.69943893e-02 -2.48528659e-01 6.68308556e-01
1.09511590e+00 6.45709097e-01 1.15574932e+00 -1.10732555e+00
-6.76280782e-02 -1.65418327e-01 -6.16040468e-01 -1.73353612e+00
2.03504842e-02 -4.00943816e-01 3.87598813e-01 -1.09380662e+00
8.50144848e-02 -8.00112069e-01 -1.10746369e-01 2.48871610e-01
-2.30411261e-01 5.59964061e-01 2.49588072e-01 3.18559766e-01
-8.97977531e-01 5.56962967e-01 1.30895627e+00 -1.52328625e-01
-1.46545600e-02 9.45360214e-02 -1.56440973e-01 6.91622615e-01
4.26187515e-01 -6.53308213e-01 2.48504803e-02 -7.15689182e-01
-3.98554176e-01 2.25539088e-01 5.81969440e-01 -1.67266488e+00
6.70806646e-01 -2.38907129e-01 8.09608340e-01 -1.39022064e+00
7.45619237e-01 -1.06237328e+00 2.73309559e-01 6.38534367e-01
1.94087133e-01 2.01918945e-01 4.37916040e-01 8.39394331e-01
-1.28865186e-02 1.99137852e-01 7.86760151e-01 1.68352917e-01
-1.08812785e+00 7.43131876e-01 1.58129022e-01 -6.73708394e-02
1.39478528e+00 -3.74215394e-01 -4.24649030e-01 3.22909892e-01
-4.37085301e-01 4.93523061e-01 6.89582169e-01 5.90400934e-01
6.35454774e-01 -1.40214205e+00 -4.21145529e-01 3.57835352e-01
1.68920100e-01 3.18483949e-01 4.76781160e-01 8.95941734e-01
-3.64593834e-01 3.68626148e-01 -1.61236182e-01 -1.31988275e+00
-1.36322021e+00 6.62344992e-01 3.20819169e-01 7.16705397e-02
-9.81945038e-01 8.93562138e-01 4.85031426e-01 4.16552387e-02
4.37985808e-01 -5.34841895e-01 -1.44397160e-02 -3.16852152e-01
8.43651712e-01 2.47595102e-01 -5.88602945e-02 -9.42436397e-01
-4.62805182e-01 9.86070454e-01 -1.69049084e-01 2.71068782e-01
9.39959049e-01 -3.21384341e-01 2.89889872e-01 1.56315163e-01
9.09347057e-01 -2.70649046e-02 -2.05622983e+00 -4.87129837e-01
-2.39732280e-01 -1.05530822e+00 9.93011445e-02 -2.21521944e-01
-1.27948689e+00 6.26428127e-01 9.52603221e-01 -2.65915364e-01
7.83776522e-01 8.79876167e-02 9.17654097e-01 1.66925341e-01
7.77523160e-01 -5.28520405e-01 -4.08830121e-02 6.12160206e-01
4.83898342e-01 -1.37049878e+00 9.84530896e-02 -5.78031659e-01
-2.72878379e-01 1.06415284e+00 1.00896287e+00 -3.48880887e-01
5.96484601e-01 4.60455716e-01 -6.41028583e-02 -7.98308626e-02
-6.52944326e-01 -2.91231960e-01 2.83917755e-01 5.91996074e-01
-1.02312773e-01 -1.70734331e-01 2.03444302e-01 3.12844962e-01
-9.87244770e-02 -8.78681540e-02 6.74552321e-02 1.10687613e+00
-7.74807334e-01 -6.49243951e-01 -9.38851535e-01 2.87235826e-01
-1.58322051e-01 2.91789174e-01 3.50526512e-01 7.74488211e-01
3.26034129e-01 6.86250031e-01 4.79910105e-01 -2.26422623e-01
3.72655690e-01 -5.04507780e-01 5.48378944e-01 -4.31955874e-01
-3.35131884e-01 1.62918970e-01 -3.35441262e-01 -7.44076848e-01
-4.20812756e-01 -8.03001642e-01 -1.41411996e+00 -2.27163523e-01
-4.92011338e-01 3.21259014e-02 7.11536825e-01 7.33250856e-01
5.62074661e-01 4.91081268e-01 4.64446604e-01 -1.46959805e+00
-3.54810238e-01 -7.20338225e-01 -1.48334488e-01 1.31804705e-01
4.19345766e-01 -1.22038138e+00 -1.28597304e-01 -7.09552467e-02] | [6.585860252380371, -2.254053831100464] |
4f1fc457-4f2a-4362-aee6-daa1e919834d | comparing-statistical-and-neural-models-for | null | null | https://aclanthology.org/2020.lt4hala-1.12 | https://aclanthology.org/2020.lt4hala-1.12.pdf | Comparing Statistical and Neural Models for Learning Sound Correspondences | Cognate prediction and proto-form reconstruction are key tasks in computational historical linguistics that rely on the study of sound change regularity. Solving these tasks appears to be very similar to machine translation, though methods from that field have barely been applied to historical linguistics. Therefore, in this paper, we investigate the learnability of sound correspondences between a proto-language and daughter languages for two machine-translation-inspired models, one statistical, the other neural. We first carry out our experiments on plausible artificial languages, without noise, in order to study the role of each parameter on the algorithms respective performance under almost perfect conditions. We then study real languages, namely Latin, Italian and Spanish, to see if those performances generalise well. We show that both model types manage to learn sound changes despite data scarcity, although the best performing model type depends on several parameters such as the size of the training data, the ambiguity, and the prediction direction. | ['Beno{\\^\\i}t Sagot', "Cl{\\'e}mentine Fourrier"] | 2020-05-01 | null | null | null | lrec-2020-5 | ['cognate-prediction'] | ['natural-language-processing'] | [ 2.50829816e-01 1.80260316e-01 -6.86792657e-02 -9.75061581e-02
-3.53091478e-01 -6.23453200e-01 8.27994466e-01 2.52368093e-01
-5.70125639e-01 6.37437284e-01 2.79727101e-01 -6.56881392e-01
-1.40317231e-01 -5.60524940e-01 -7.61645734e-01 -5.60795426e-01
1.83663592e-01 6.26303792e-01 4.94981110e-01 -3.20050031e-01
2.51828372e-01 7.32238889e-02 -1.48074949e+00 5.87410592e-02
7.50349700e-01 3.64040792e-01 3.29095006e-01 4.30940509e-01
-1.52893513e-01 3.32657874e-01 -1.50872573e-01 -5.62597454e-01
2.03700855e-01 -5.70414126e-01 -8.67541790e-01 -1.08205445e-01
1.72087148e-01 3.54528874e-01 1.07036479e-01 8.63889158e-01
3.31416875e-01 -8.32209438e-02 7.50467718e-01 -5.68847477e-01
-3.15369874e-01 1.05601013e+00 -8.76845494e-02 4.52481449e-01
3.48333418e-01 -3.66607942e-02 1.22785449e+00 -7.31600702e-01
8.22190225e-01 1.06129241e+00 7.76885748e-01 4.32926416e-01
-1.22212815e+00 -9.81572568e-02 -9.29419100e-02 3.63602161e-01
-9.42384422e-01 -5.54953933e-01 7.95917392e-01 -4.70402122e-01
5.05241990e-01 1.43615142e-01 6.43750846e-01 8.30766022e-01
1.23007027e-02 3.97771508e-01 1.35128200e+00 -1.09324586e+00
4.35252339e-01 4.01401430e-01 -7.49477595e-02 4.25258666e-01
1.69864625e-01 2.34829396e-01 -6.71768427e-01 1.07271560e-02
4.33222651e-01 -8.49801600e-01 -3.20140541e-01 -1.40634611e-01
-1.17439306e+00 7.21002936e-01 1.34524032e-02 1.06329799e+00
-1.04990281e-01 -7.94130638e-02 3.44691932e-01 4.59755599e-01
2.94385403e-01 5.56493402e-01 -5.79033732e-01 -1.43490061e-01
-7.53895164e-01 2.33325616e-01 8.74270380e-01 4.68504667e-01
5.11558115e-01 -2.62682494e-02 4.02986795e-01 8.51049602e-01
1.68602869e-01 1.90242872e-01 8.16021860e-01 -5.04951596e-01
3.65781516e-01 2.38127902e-01 -1.44066453e-01 -9.54637468e-01
-3.30523908e-01 -4.18413550e-01 -3.99738073e-01 -2.82878786e-01
9.72096503e-01 -6.65091425e-02 -4.08786416e-01 1.98049796e+00
2.20637724e-01 -4.60688919e-02 1.03559598e-01 6.45018578e-01
4.04053718e-01 6.32467806e-01 1.14307538e-01 -5.33715010e-01
1.17643988e+00 -4.88418102e-01 -3.89152616e-01 -3.95029962e-01
7.15298593e-01 -8.22760284e-01 1.44555438e+00 2.86582798e-01
-1.19419563e+00 -4.52381313e-01 -8.91945481e-01 1.01937450e-01
-2.22926095e-01 1.44690186e-01 5.08321047e-01 6.42401397e-01
-7.35737681e-01 9.34204280e-01 -8.63737702e-01 -5.39414823e-01
-2.45284051e-01 1.73835218e-01 -2.56259114e-01 3.74495387e-01
-1.20466995e+00 1.21350992e+00 5.62659740e-01 2.02377483e-01
-2.68386275e-01 -3.22712600e-01 -4.61761802e-01 -6.35200441e-02
8.25438797e-02 -5.32639802e-01 1.22559822e+00 -1.29426932e+00
-1.42918158e+00 1.19929409e+00 -9.44677740e-02 -5.33508956e-01
5.56577027e-01 7.73802325e-02 -6.21471219e-02 -2.29674518e-01
-1.18931122e-01 3.08924824e-01 4.90492433e-01 -1.10973346e+00
-6.12547278e-01 -5.57055295e-01 -2.09085509e-01 5.02556898e-02
-2.05900431e-01 1.24308847e-01 3.80022191e-02 -5.32763004e-01
3.78474504e-01 -9.92188752e-01 -1.03234418e-01 -2.52790272e-01
1.39389308e-02 -3.61688226e-01 -5.12022851e-03 -6.76891744e-01
9.71819043e-01 -1.94232118e+00 3.70738119e-01 2.37434000e-01
-5.07949412e-01 3.16760950e-02 6.26136363e-02 4.03266281e-01
9.44399312e-02 -7.70559832e-02 -4.47424829e-01 -2.13011637e-01
5.51844969e-05 4.03925568e-01 -2.94456780e-01 4.50150669e-01
9.66543928e-02 6.79156661e-01 -5.51724911e-01 -3.26329976e-01
-1.09011948e-01 3.32392275e-01 -6.85364723e-01 -1.33255184e-01
-4.35012400e-01 6.34890854e-01 -2.75848866e-01 1.39729470e-01
6.85390607e-02 3.10853451e-01 6.82832301e-01 1.98382050e-01
-4.46744531e-01 8.56589019e-01 -1.12857044e+00 1.49181604e+00
-6.46452129e-01 5.60174704e-01 -1.26151755e-01 -1.13139272e+00
9.55167830e-01 3.54445189e-01 4.46447084e-04 -7.70175219e-01
2.69045591e-01 7.87299275e-01 5.02897143e-01 -4.26939249e-01
4.79064137e-01 -7.11130500e-01 5.19588478e-02 4.46388990e-01
-5.39555661e-02 -1.58192679e-01 4.18506801e-01 -3.88671130e-01
6.83940530e-01 2.77034611e-01 4.06373501e-01 -6.00712955e-01
5.00203192e-01 -5.77195212e-02 6.28414035e-01 4.76384997e-01
1.77765116e-01 4.97648388e-01 5.22397161e-01 -5.45027375e-01
-1.04649425e+00 -7.54052222e-01 -4.31456059e-01 1.16105378e+00
-1.16008129e-02 -2.51011044e-01 -8.31750453e-01 -2.22914860e-01
-4.30436969e-01 8.90182912e-01 -3.60277146e-01 -7.82104507e-02
-1.02591991e+00 -9.52179372e-01 4.53091949e-01 6.94346726e-02
-1.20785721e-01 -1.54557598e+00 -8.94411862e-01 4.17277753e-01
-1.23078190e-01 -9.58849907e-01 5.82643636e-02 1.91368639e-01
-9.28235948e-01 -8.65235507e-01 -4.66153473e-01 -9.13435817e-01
1.15225233e-01 -4.29602802e-01 1.22789013e+00 6.47226050e-02
7.79443458e-02 1.16339900e-01 -3.15125078e-01 -4.93241161e-01
-1.02919281e+00 4.85010564e-01 1.98055983e-01 -1.85563996e-01
2.32467979e-01 -1.07789755e+00 7.06982464e-02 5.76674528e-02
-9.72901940e-01 2.71682795e-02 4.06122148e-01 6.20032668e-01
4.19177294e-01 -1.49952322e-01 3.58870655e-01 -1.09856045e+00
4.27302301e-01 -3.79820377e-01 -4.87278402e-01 3.50152254e-01
-5.37452817e-01 3.92946005e-01 7.02062666e-01 -4.29284394e-01
-8.13796103e-01 1.09653801e-01 -3.21661800e-01 3.65779847e-01
-1.93188593e-01 6.73958600e-01 -2.85247058e-01 3.22788119e-01
7.82718837e-01 2.85590351e-01 -3.12928736e-01 -6.39392853e-01
4.07342762e-01 4.07719910e-01 6.66520178e-01 -8.14705908e-01
7.04278708e-01 5.18093929e-02 -3.37234810e-02 -9.55782235e-01
-5.30316532e-01 -2.09087487e-02 -8.87482524e-01 4.07838225e-02
6.15788043e-01 -4.43489820e-01 -2.58214146e-01 1.59687862e-01
-1.39372218e+00 -2.76870400e-01 -4.12663430e-01 5.09237051e-01
-7.60053515e-01 2.23772839e-01 -5.44255316e-01 -5.99475801e-01
-1.07833743e-01 -8.40670764e-01 6.17798865e-01 -1.70785077e-02
-2.31416956e-01 -1.19283223e+00 5.65607429e-01 -3.78918052e-02
3.56064498e-01 2.18970552e-02 1.55623889e+00 -7.11648524e-01
-3.13336164e-01 2.43636236e-01 3.86298388e-01 1.26382321e-01
-1.93589414e-03 -7.24171102e-02 -8.19777191e-01 6.54308945e-02
3.24908704e-01 -2.37918030e-02 7.61278510e-01 2.08888739e-01
3.30640286e-01 -3.15437078e-01 1.02469102e-02 3.67594928e-01
1.38543510e+00 6.81325095e-03 5.06824136e-01 4.44216520e-01
2.55588830e-01 1.06541348e+00 1.97876021e-01 5.19867130e-02
3.43500972e-01 7.54013658e-01 -1.52064837e-03 2.47929633e-01
-7.55142272e-02 -3.73587370e-01 4.00582582e-01 1.27936530e+00
-2.01557651e-01 4.13473472e-02 -1.20710027e+00 6.70005500e-01
-1.72544432e+00 -7.76962340e-01 -1.76676661e-01 2.18994308e+00
1.09594452e+00 4.03657317e-01 3.33608985e-01 5.09902000e-01
5.86757243e-01 7.40595609e-02 9.19670761e-02 -4.65151608e-01
-3.12555254e-01 4.21877235e-01 3.17947231e-02 6.34087801e-01
-5.73817968e-01 1.05941677e+00 5.98464918e+00 6.39100492e-01
-1.20958102e+00 1.60356894e-01 4.25460905e-01 2.96485513e-01
-6.72061741e-01 4.29183990e-01 -4.95801330e-01 4.69150156e-01
1.20196795e+00 9.84226689e-02 5.05046964e-01 5.60973823e-01
2.02477291e-01 -1.25962332e-01 -1.31110644e+00 5.81161559e-01
5.10628335e-03 -1.01671696e+00 2.21248623e-03 -2.36559778e-01
4.65506166e-01 -4.51777652e-02 -2.70005792e-01 -9.61104706e-02
-1.77796304e-01 -9.03402090e-01 1.14538276e+00 4.29173261e-01
3.92707348e-01 -4.91194695e-01 5.84243298e-01 6.31924450e-01
-7.10664332e-01 8.95702317e-02 -4.14080739e-01 -2.76849151e-01
2.46276900e-01 4.93714273e-01 -5.17696261e-01 2.67476767e-01
2.08439827e-01 2.94651002e-01 -6.41351283e-01 8.08827996e-01
-3.62090409e-01 9.71888185e-01 -5.20013154e-01 -2.34831944e-01
-4.97766137e-02 -2.86989152e-01 6.91865981e-01 1.08796680e+00
3.26392055e-01 -6.88865185e-02 -1.68130338e-01 8.62289011e-01
3.46131921e-01 5.95936000e-01 -4.31435972e-01 -1.67920366e-02
4.09035265e-01 8.08961570e-01 -7.90888190e-01 1.99058533e-01
-2.18803480e-01 5.96012473e-01 4.42494065e-01 -1.11076586e-01
-4.58410919e-01 1.04458518e-01 8.88895541e-02 5.20378292e-01
1.18060187e-01 -4.31728870e-01 -3.67911160e-01 -9.61948633e-01
3.54433775e-01 -8.02138686e-01 6.10083938e-02 -2.82498986e-01
-1.13431215e+00 6.34067535e-01 -5.25184423e-02 -7.46100903e-01
-6.09652460e-01 -6.19052231e-01 -6.73826516e-01 6.25103891e-01
-1.23715496e+00 -1.18034816e+00 4.92145300e-01 1.39170870e-01
4.60256308e-01 -1.69334099e-01 6.77233636e-01 5.24273217e-02
-2.88517863e-01 3.36341232e-01 1.84340164e-01 -9.66559649e-02
4.06520426e-01 -1.08633220e+00 3.64769012e-01 8.78264546e-01
7.71929741e-01 4.02581602e-01 1.01301563e+00 -3.37853283e-01
-1.08171844e+00 -4.63640302e-01 1.59326398e+00 -4.22346950e-01
7.03014851e-01 -3.69070858e-01 -8.90852630e-01 5.00814438e-01
1.26762554e-01 -1.37810215e-01 3.32760930e-01 3.55873197e-01
-2.25175500e-01 -5.54701837e-04 -5.62687099e-01 5.94375551e-01
1.13421881e+00 -4.87040967e-01 -9.28347528e-01 1.72650397e-01
4.85178828e-01 -1.05577558e-01 -4.50980693e-01 2.61085421e-01
5.34518957e-01 -1.14833748e+00 4.81277704e-01 -6.27876401e-01
4.98661786e-01 -8.91465172e-02 -1.33622527e-01 -1.19753003e+00
-1.41236007e-01 -5.42908013e-01 4.45304751e-01 1.40807807e+00
6.60492003e-01 -5.96823156e-01 5.06729305e-01 2.25598216e-01
6.83599338e-02 -6.07245803e-01 -1.15744936e+00 -6.36012018e-01
5.55476367e-01 -5.50243497e-01 4.79645997e-01 8.13538074e-01
-2.00189333e-02 8.27816427e-01 -8.26774985e-02 -1.77991211e-01
5.56763494e-03 1.81738362e-01 4.86973286e-01 -1.41915059e+00
-7.82385767e-01 -4.93763685e-01 -2.55559295e-01 -5.25747359e-01
4.35649782e-01 -9.94531810e-01 1.65832758e-01 -7.52172112e-01
-1.55338317e-01 -7.97948658e-01 2.17211284e-02 2.56220818e-01
-2.03708857e-01 1.50173292e-01 2.73884624e-01 4.28156167e-01
-9.63848531e-02 4.54994440e-01 6.29036844e-01 2.91367054e-01
-5.35589099e-01 2.41166756e-01 -3.43520463e-01 8.98528397e-01
8.35498512e-01 -6.25580311e-01 -1.58011764e-01 -5.71968436e-01
6.54383183e-01 -1.63339060e-02 1.71303019e-01 -9.48959768e-01
1.60720542e-01 -9.98205319e-02 -9.59092006e-02 1.11970462e-01
-1.31978452e-01 -6.19001150e-01 1.69482976e-01 4.66107935e-01
-4.60512221e-01 3.12392861e-01 1.04383543e-01 2.44191647e-01
-1.63008168e-01 -8.04409027e-01 7.46101081e-01 -3.16222042e-01
-3.73675257e-01 -1.86642319e-01 -4.00020868e-01 3.30442548e-01
4.55452830e-01 -5.16238995e-02 1.74735919e-01 -1.09652720e-01
-7.12025046e-01 -4.39672291e-01 4.39108044e-01 3.71462941e-01
-3.31521481e-02 -1.00423288e+00 -8.31471920e-01 1.54223561e-01
-8.13358352e-02 -2.49944508e-01 -1.76615223e-01 7.43983805e-01
-5.99527180e-01 3.79856259e-01 -9.33617949e-02 -4.79440957e-01
-9.41947341e-01 4.89358008e-01 3.65836233e-01 -2.36316502e-01
-3.34713042e-01 4.50321108e-01 -5.60198762e-02 -6.18751109e-01
1.28902018e-01 -2.86078215e-01 -2.51732498e-01 6.55089468e-02
2.08673570e-02 1.80077374e-01 1.92806646e-01 -7.95369864e-01
-2.09815159e-01 7.01680958e-01 3.06188852e-01 -4.38778549e-01
1.36378527e+00 -1.76523358e-01 -3.00038964e-01 9.11238790e-01
6.57162666e-01 3.29296082e-01 -6.27249658e-01 -4.05161798e-01
6.80569172e-01 -7.59083629e-02 -2.85412014e-01 -4.30235028e-01
-5.22540092e-01 9.56952691e-01 3.07709277e-01 2.51868576e-01
9.53070998e-01 1.58422515e-01 6.41006231e-01 1.56411886e-01
3.69513214e-01 -1.05977547e+00 -4.29154754e-01 6.10181093e-01
7.22024679e-01 -9.16524172e-01 -2.37345874e-01 -3.09572637e-01
-2.59025335e-01 1.17496622e+00 2.09418148e-01 -1.51539236e-01
4.87412930e-01 1.45909097e-02 -1.85683779e-02 6.16848171e-02
-8.22043657e-01 -2.99302846e-01 3.30592781e-01 3.20881397e-01
7.03200221e-01 1.02906108e-01 -1.08461845e+00 5.26833177e-01
-8.90032828e-01 -2.23614857e-01 3.04802418e-01 5.39762914e-01
-4.97796506e-01 -1.45120680e+00 -2.98891008e-01 -9.99674648e-02
-5.91839552e-01 -2.43440926e-01 -5.75481713e-01 8.49938035e-01
5.09992480e-01 6.57493412e-01 1.11307949e-01 -1.63603067e-01
1.96335986e-01 3.53492886e-01 7.36172259e-01 -5.99635601e-01
-6.44049644e-01 -3.88512909e-02 2.04917669e-01 5.23299873e-02
-5.70234835e-01 -8.41815531e-01 -1.03844500e+00 -2.21207231e-01
-3.92822236e-01 2.71891624e-01 7.55301058e-01 1.32308865e+00
-2.09178612e-01 -8.04797560e-02 2.90309191e-01 -6.65235221e-01
-6.50934517e-01 -9.58549619e-01 -5.79427540e-01 2.69121885e-01
1.35904634e-02 -2.14230046e-01 -4.87032533e-01 1.14105061e-01] | [10.740528106689453, 9.767122268676758] |
d7ac5b50-afd1-4cd4-ac13-810c1c7691bf | fmfcc-a-a-challenging-mandarin-dataset-for | 2110.09441 | null | https://arxiv.org/abs/2110.09441v1 | https://arxiv.org/pdf/2110.09441v1.pdf | FMFCC-A: A Challenging Mandarin Dataset for Synthetic Speech Detection | As increasing development of text-to-speech (TTS) and voice conversion (VC) technologies, the detection of synthetic speech has been suffered dramatically. In order to promote the development of synthetic speech detection model against Mandarin TTS and VC technologies, we have constructed a challenging Mandarin dataset and organized the accompanying audio track of the first fake media forensic challenge of China Society of Image and Graphics (FMFCC-A). The FMFCC-A dataset is by far the largest publicly-available Mandarin dataset for synthetic speech detection, which contains 40,000 synthesized Mandarin utterances that generated by 11 Mandarin TTS systems and two Mandarin VC systems, and 10,000 genuine Mandarin utterances collected from 58 speakers. The FMFCC-A dataset is divided into the training, development and evaluation sets, which are used for the research of detection of synthesized Mandarin speech under various previously unknown speech synthesis systems or audio post-processing operations. In addition to describing the construction of the FMFCC-A dataset, we provide a detailed analysis of two baseline methods and the top-performing submissions from the FMFCC-A, which illustrates the usefulness and challenge of FMFCC-A dataset. We hope that the FMFCC-A dataset can fill the gap of lack of Mandarin datasets for synthetic speech detection. | ['Xianfeng Zhao', 'Xiaowei Yi', 'Yewei Gu', 'Zhenyu Zhang'] | 2021-10-18 | null | null | null | null | ['synthetic-speech-detection'] | ['audio'] | [ 3.32499027e-01 -2.78143555e-01 3.89569163e-01 -2.11751796e-02
-1.52785420e+00 -5.79474211e-01 8.07194650e-01 -4.60490435e-01
-1.44490927e-01 2.58066118e-01 2.24700600e-01 -5.88014305e-01
5.97188115e-01 -5.27888089e-02 -8.76663208e-01 -5.80951214e-01
1.30260155e-01 2.00946704e-01 4.58397746e-01 2.07084939e-02
2.97500551e-01 3.36936086e-01 -1.44601619e+00 7.61294544e-01
3.94098133e-01 8.03982854e-01 3.75332952e-01 1.09268677e+00
2.66917437e-01 5.65105915e-01 -1.34470963e+00 -5.22091150e-01
2.20893547e-01 -6.82066858e-01 -5.89483678e-01 1.43064424e-01
5.44881344e-01 -5.02217591e-01 -4.87827063e-01 1.30447173e+00
8.01185608e-01 -3.74156356e-01 4.94912863e-01 -1.23432779e+00
-8.77871573e-01 8.83307576e-01 -3.72463554e-01 4.82383341e-01
2.40556672e-01 6.09784186e-01 5.04756510e-01 -1.29924357e+00
6.70951366e-01 1.70958483e+00 4.23398763e-01 6.98208570e-01
-6.98079050e-01 -9.63059247e-01 -4.76911664e-01 2.93782860e-01
-1.52997327e+00 -8.72696996e-01 7.70255983e-01 -5.77954292e-01
6.85874939e-01 4.07700121e-01 1.85552493e-01 1.68944001e+00
-3.40186626e-01 1.22928488e+00 9.76976931e-01 -5.85011780e-01
7.20171928e-02 2.95273304e-01 1.10893346e-01 3.69958073e-01
-1.20819040e-01 4.85213727e-01 -6.48975790e-01 -3.12360615e-01
3.20915014e-01 -9.36314166e-01 -4.27139431e-01 8.13675702e-01
-1.33083427e+00 8.45077932e-01 -3.34642738e-01 2.43734658e-01
1.90995350e-01 -1.70192912e-01 7.11124718e-01 6.30434334e-01
4.05987471e-01 1.90062106e-01 -7.04664388e-04 -3.04769486e-01
-1.13416553e+00 2.71213740e-01 4.26882446e-01 1.32686913e+00
-7.54196346e-02 7.37808168e-01 -1.24031352e-02 9.80111420e-01
2.64528811e-01 9.00150895e-01 5.36156118e-01 -9.04273689e-01
8.44660938e-01 -3.22943270e-01 -6.76038489e-02 -7.06497073e-01
4.21522617e-01 -3.53633426e-02 -4.49755549e-01 7.45255947e-02
2.72292316e-01 -2.25703306e-02 -6.76368237e-01 1.12771356e+00
2.19656304e-01 2.38955036e-01 1.24952532e-02 8.51414084e-01
7.70032227e-01 6.90493405e-01 -5.22309601e-01 -3.32651794e-01
1.49951231e+00 -8.81249666e-01 -9.32584822e-01 1.59389712e-02
4.55397666e-01 -1.62167311e+00 1.34007025e+00 7.09852457e-01
-7.97901511e-01 -7.05526590e-01 -1.27270794e+00 3.23646754e-01
-2.41834372e-01 2.30490059e-01 -1.73793837e-01 1.34896266e+00
-6.35249019e-01 -8.96004681e-03 -3.49651635e-01 8.35263580e-02
5.11259198e-01 -3.27958316e-01 -2.28244498e-01 6.18122937e-03
-1.18738496e+00 6.60160720e-01 6.53780773e-02 -4.77157114e-03
-1.51835287e+00 -3.94027710e-01 -6.39813602e-01 -3.66122067e-01
3.20832402e-01 3.26976657e-01 1.47464418e+00 -5.85011601e-01
-1.23218524e+00 9.02180672e-01 1.31602362e-01 -7.24466980e-01
7.69022048e-01 -1.94528028e-01 -1.33409154e+00 3.75843674e-01
3.84618223e-01 3.99488330e-01 1.66873097e+00 -1.13571084e+00
-3.78561080e-01 6.62102848e-02 -8.82430792e-01 -5.23962438e-01
4.03334983e-02 9.18646634e-01 -2.57959723e-01 -1.21834195e+00
-1.84641406e-01 -7.14039266e-01 4.80819583e-01 -3.70272756e-01
-8.51844907e-01 -1.12876311e-01 1.78862727e+00 -9.96748209e-01
1.13944972e+00 -2.61855030e+00 -6.99477732e-01 -2.69888639e-01
-9.70452577e-02 9.36571777e-01 -3.49009484e-01 2.27555871e-01
-1.34112418e-01 3.30181062e-01 -3.47334534e-01 -6.28327906e-01
3.07473019e-02 -2.12288857e-01 -1.05735636e+00 7.03540742e-01
4.95710790e-01 4.73091692e-01 -7.98775136e-01 -6.74681425e-01
2.00316548e-01 2.48135760e-01 -1.17670700e-01 5.59512153e-02
-2.36827180e-01 3.07022363e-01 -2.14008689e-01 1.19098520e+00
9.94466543e-01 5.82012057e-01 -2.85390258e-01 1.22341044e-01
-2.32120037e-01 4.14644450e-01 -1.06219637e+00 9.47632492e-01
3.43894362e-01 1.31964600e+00 3.76336962e-01 -5.88123381e-01
8.88275087e-01 6.76333666e-01 -1.85200199e-01 -6.43312573e-01
3.78040336e-02 5.91321528e-01 -1.15704246e-01 -5.51506460e-01
6.76414669e-01 2.69826829e-01 -1.22543685e-01 5.97148776e-01
1.01123616e-01 -5.73191524e-01 1.48828164e-01 3.36063921e-01
1.04218745e+00 -3.06347579e-01 -2.98374206e-01 -2.77738627e-02
8.01707029e-01 -1.29624546e-01 4.35298681e-01 8.15045893e-01
-7.28306592e-01 1.04899395e+00 1.98193967e-01 -9.03047845e-02
-1.18904793e+00 -1.15044916e+00 -3.27998191e-01 5.22586644e-01
-4.07524407e-01 -5.40645182e-01 -1.26969612e+00 -6.51279628e-01
-3.71825904e-01 9.71304595e-01 -7.03317523e-02 4.97370735e-02
-9.04555023e-01 -3.38655472e-01 1.86609161e+00 -7.25929141e-02
7.17116952e-01 -1.18518829e+00 1.04613617e-01 3.95093411e-02
-4.81748551e-01 -1.61392081e+00 -1.08835089e+00 -3.36559385e-01
-1.93746239e-01 -1.18476379e+00 -5.71455181e-01 -9.92018759e-01
5.24283051e-02 6.07584476e-01 6.74167454e-01 8.23776647e-02
-5.38743198e-01 5.21137901e-02 -4.93797362e-01 -7.34820962e-01
-1.57661009e+00 -5.99030674e-01 3.56602401e-01 7.30256513e-02
2.79251963e-01 -5.09474128e-02 -4.51867585e-04 6.88681245e-01
-7.87841380e-01 -3.69926184e-01 1.42407879e-01 7.17194617e-01
3.33173156e-01 8.37216079e-02 6.91681921e-01 -5.51928639e-01
6.13175452e-01 -3.75484318e-01 -9.05374050e-01 8.12330991e-02
-3.69703472e-01 -6.26068473e-01 6.11822784e-01 -6.80725753e-01
-1.04461563e+00 -1.82655975e-01 -2.82704145e-01 -7.98426926e-01
-2.07484350e-01 -2.51261126e-02 -2.63966143e-01 1.63175151e-01
8.48616183e-01 6.28922880e-01 -1.39353082e-01 -8.56484473e-01
3.54027063e-01 1.48401630e+00 1.25748217e+00 -3.72288018e-01
9.84153152e-01 9.95955616e-02 -6.16661131e-01 -1.55638802e+00
-3.31394225e-01 -2.36646160e-01 -1.83994576e-01 -3.96409065e-01
6.40193343e-01 -9.19250727e-01 -3.60175282e-01 1.15941286e+00
-1.63293707e+00 -1.94744058e-02 -7.03375861e-02 5.39801180e-01
-2.19144702e-01 1.05219531e+00 -7.92666614e-01 -1.23959017e+00
-2.90381163e-01 -1.61397088e+00 1.28327203e+00 -6.74713552e-01
-1.06875248e-01 -1.76978022e-01 2.53976453e-02 9.06771898e-01
1.92782924e-01 4.28584851e-02 5.95213234e-01 -5.88445008e-01
-5.92622280e-01 -3.21309358e-01 2.41523981e-02 1.02127564e+00
-5.75944707e-02 3.89805436e-01 -1.39390278e+00 -2.05579296e-01
5.78941762e-01 -3.45451146e-01 8.02469611e-01 -7.96960946e-03
7.45407224e-01 -4.32001561e-01 1.84007078e-01 3.12325299e-01
6.21246636e-01 4.54165548e-01 8.34248126e-01 1.24944255e-01
5.47358692e-01 4.08029139e-01 7.08022296e-01 2.13945359e-01
-1.45849660e-01 6.59753323e-01 3.23269397e-01 5.21466434e-01
-7.23763883e-01 -3.27803940e-01 1.02414262e+00 1.20233727e+00
3.60433698e-01 -2.91097999e-01 -6.96001232e-01 5.46549916e-01
-1.07977712e+00 -1.23510063e+00 -6.19539022e-01 2.07560396e+00
7.74819791e-01 2.60456115e-01 2.93573916e-01 6.22498810e-01
1.35416162e+00 2.39652097e-01 -1.77314997e-01 -3.74160320e-01
-6.22227967e-01 3.10800970e-01 1.46494523e-01 4.97691035e-01
-1.27630639e+00 1.19549835e+00 6.76855373e+00 1.32974648e+00
-1.00856090e+00 3.74056965e-01 5.02248704e-01 1.13066584e-01
1.84576720e-01 -1.93783969e-01 -9.15404558e-01 8.27285767e-01
1.39111769e+00 9.38883498e-02 4.55769151e-01 9.74765062e-01
4.88744706e-01 1.96728244e-01 -8.28201056e-01 9.72889245e-01
2.59464681e-01 -1.56350195e+00 2.49806400e-02 6.63654730e-02
5.86959302e-01 2.12952882e-01 2.41170347e-01 3.01277488e-01
1.92760248e-02 -7.67406106e-01 1.35528374e+00 -1.15005456e-01
1.08105958e+00 -5.92843056e-01 4.29554015e-01 5.80767453e-01
-9.56790805e-01 1.99841514e-01 -3.23988169e-01 4.35125917e-01
1.99959412e-01 4.75737125e-01 -1.03815138e+00 1.65723488e-01
7.61928141e-01 2.40978941e-01 -5.14235198e-01 7.49636889e-01
-1.34844586e-01 1.30663729e+00 -1.45939425e-01 1.84217885e-01
1.02962583e-01 1.27025276e-01 1.09021950e+00 1.48896861e+00
3.95039260e-01 -4.72191930e-01 -4.25232440e-01 1.09086704e+00
-2.06910282e-01 -1.85423195e-01 -5.24922609e-01 -6.94555461e-01
1.04723096e+00 7.49058843e-01 -4.31900591e-01 -3.76923382e-01
-2.66601771e-01 8.11565876e-01 -3.33482295e-01 2.67413016e-02
-1.00571549e+00 -3.77573729e-01 5.51187217e-01 -2.51059849e-02
3.56760323e-01 -2.43390664e-01 -4.34443280e-02 -9.78367627e-01
1.12006456e-01 -1.52421403e+00 1.01835780e-01 -7.25772023e-01
-1.25560427e+00 8.30448985e-01 -2.79653460e-01 -1.42459393e+00
-3.38160247e-01 -5.52143931e-01 -8.08704019e-01 7.14541316e-01
-1.14660585e+00 -9.96812582e-01 4.41654697e-02 6.94882810e-01
1.11227703e+00 -1.02817559e+00 6.82012498e-01 5.18325269e-01
-5.71315050e-01 7.96481192e-01 1.61568329e-01 5.07639349e-01
8.76374543e-01 -7.32962191e-01 1.29850757e+00 1.26097250e+00
3.37041587e-01 3.87155890e-01 6.88715160e-01 -7.24375248e-01
-1.26126707e+00 -1.29091728e+00 8.66545498e-01 -5.42709351e-01
6.91971242e-01 -7.65900552e-01 -7.91022420e-01 4.58032936e-01
7.20625147e-02 -1.85393095e-01 3.77452433e-01 -1.00892651e+00
-5.77811897e-01 2.12697729e-01 -1.30695093e+00 3.49482208e-01
6.09665096e-01 -1.11803353e+00 -7.02503204e-01 5.78412414e-01
1.18448281e+00 -1.22593276e-01 -4.92352068e-01 -7.50903636e-02
2.00966761e-01 -9.30011809e-01 1.04181242e+00 -1.38312444e-01
8.85450542e-02 -5.55926383e-01 -6.24364257e-01 -7.66541779e-01
3.84965748e-01 -1.30954933e+00 -1.38004959e-01 1.71484017e+00
2.84264237e-01 -3.69964153e-01 5.08713663e-01 -2.14344159e-01
-5.72828710e-01 9.02028829e-02 -1.25924623e+00 -1.25288141e+00
3.38867567e-02 -1.11418283e+00 7.85635829e-01 1.09215212e+00
-4.78200585e-01 1.79314002e-01 -8.22833776e-01 3.49788368e-01
9.09525156e-01 -3.21230590e-01 1.03864062e+00 -4.60080385e-01
-4.76093054e-01 -2.00217351e-01 -3.03246886e-01 -1.08084381e+00
-1.73178345e-01 -6.55656576e-01 3.17530632e-02 -4.40179884e-01
-3.43827307e-01 1.30184755e-01 3.78306091e-01 -4.53742407e-02
1.63121112e-02 3.82151574e-01 2.93569028e-01 4.68178511e-01
-3.61881591e-02 4.55347538e-01 1.08483636e+00 -4.04266566e-01
1.99203536e-01 2.59744018e-01 -1.18775524e-01 7.04924405e-01
5.11660039e-01 -7.74079621e-01 -1.16627350e-01 -3.07856530e-01
-7.17347562e-01 1.37108460e-01 4.42049235e-01 -9.68614578e-01
1.50286257e-01 1.35485411e-01 1.77240279e-02 -1.19458008e+00
5.26095510e-01 -2.97940373e-01 -3.10199130e-02 4.61901546e-01
2.46295836e-02 -1.82877272e-01 -2.02220663e-01 4.87106651e-01
-5.15650630e-01 -3.23075205e-01 1.00974202e+00 -1.78532541e-01
-5.06886780e-01 -1.61459625e-01 -9.07500327e-01 1.48705229e-01
7.17115283e-01 -2.50879914e-01 -8.00915241e-01 -4.06293899e-01
-3.15082729e-01 -3.41817379e-01 -1.21484529e-02 5.29510677e-01
9.67782319e-01 -1.22435963e+00 -1.04742837e+00 4.50091332e-01
9.35919657e-02 -3.96225035e-01 7.54333436e-02 6.13245904e-01
-7.63781488e-01 4.78667587e-01 2.04430282e-01 -6.54156864e-01
-1.52345955e+00 7.42140472e-01 6.01644181e-02 3.54303628e-01
-4.52443600e-01 8.21996033e-01 3.07209557e-03 -3.19220901e-01
4.40178305e-01 -5.34736067e-02 4.32113588e-01 -3.28566849e-01
8.27858984e-01 6.95256233e-01 1.33573845e-01 -1.18580103e+00
-4.03399825e-01 -2.40202710e-01 -8.25754628e-02 -3.58474910e-01
1.09601986e+00 9.14780796e-03 8.99275616e-02 2.20977247e-01
1.23891139e+00 4.51103628e-01 -8.57077003e-01 -1.14363320e-01
7.90169984e-02 -7.30783105e-01 -3.79712731e-02 -6.12211287e-01
-7.87794888e-01 9.95268226e-01 5.59386194e-01 3.15973461e-01
7.73964643e-01 -3.88501920e-02 1.32006037e+00 1.21846303e-01
3.01435381e-01 -1.18328214e+00 4.28319126e-01 6.64546967e-01
1.09492958e+00 -1.07070327e+00 -3.60828131e-01 -5.40828943e-01
-6.69204295e-01 1.16577244e+00 3.11559200e-01 1.72672465e-01
3.88398290e-01 4.52714831e-01 3.76471251e-01 5.75737990e-02
-3.64504009e-01 2.24065736e-01 -7.69873559e-02 8.03613722e-01
1.24406636e-01 1.66479036e-01 1.45961329e-01 7.01051712e-01
-7.92905033e-01 -5.16274095e-01 5.42538285e-01 7.98932791e-01
-5.29013097e-01 -1.24740565e+00 -9.40284431e-01 -6.41724914e-02
-6.12217665e-01 -3.22012752e-01 -8.23065817e-01 7.20543623e-01
8.54898691e-02 1.57514358e+00 -2.77746707e-01 -6.59605503e-01
2.12257966e-01 6.00148849e-02 6.86774179e-02 -3.61517370e-01
-6.99768126e-01 5.35355747e-01 3.91515464e-01 -2.90557683e-01
8.88594314e-02 -8.05053890e-01 -8.33206594e-01 -6.41102672e-01
-7.15666831e-01 7.33319148e-02 8.16912532e-01 7.12905645e-01
1.37956098e-01 -4.89835814e-02 9.06364501e-01 -5.11672556e-01
-7.45732009e-01 -1.20254040e+00 -6.67046428e-01 2.04250440e-01
6.36712849e-01 -1.50749803e-01 -7.45340168e-01 2.06416667e-01] | [14.170060157775879, 5.814658164978027] |
6c7041a6-d0bb-4caa-84e7-79857d45c30d | opinion-mining-with-deep-recurrent-neural | null | null | https://aclanthology.org/D14-1080 | https://aclanthology.org/D14-1080.pdf | Opinion Mining with Deep Recurrent Neural Networks | null | ['Ozan {\\.I}rsoy', 'Claire Cardie'] | 2014-10-01 | null | null | null | emnlp-2014-10 | ['fine-grained-opinion-analysis'] | ['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] | [-7.253962516784668, 3.5042145252227783] |
bdbb7163-46bf-4243-a008-a497eff45363 | end-to-end-offline-goal-oriented-dialog | 1712.02838 | null | http://arxiv.org/abs/1712.02838v1 | http://arxiv.org/pdf/1712.02838v1.pdf | End-to-End Offline Goal-Oriented Dialog Policy Learning via Policy Gradient | Learning a goal-oriented dialog policy is generally performed offline with
supervised learning algorithms or online with reinforcement learning (RL).
Additionally, as companies accumulate massive quantities of dialog transcripts
between customers and trained human agents, encoder-decoder methods have gained
popularity as agent utterances can be directly treated as supervision without
the need for utterance-level annotations. However, one potential drawback of
such approaches is that they myopically generate the next agent utterance
without regard for dialog-level considerations. To resolve this concern, this
paper describes an offline RL method for learning from unannotated corpora that
can optimize a goal-oriented policy at both the utterance and dialog level. We
introduce a novel reward function and use both on-policy and off-policy policy
gradient to learn a policy offline without requiring online user interaction or
an explicit state space definition. | ['Li Zhou', 'Kevin Small', 'Charles Elkan', 'Oleg Rokhlenko'] | 2017-12-07 | null | null | null | null | ['goal-oriented-dialog'] | ['natural-language-processing'] | [ 9.12942514e-02 8.12600553e-01 -2.64221698e-01 -7.83902287e-01
-1.02721786e+00 -1.00906467e+00 8.81986678e-01 3.26617628e-01
-6.27372265e-01 1.15001464e+00 2.25848421e-01 -5.50621927e-01
4.03128654e-01 -4.84262228e-01 -3.99186462e-01 -3.55165273e-01
1.81477681e-01 9.62507486e-01 -7.21701980e-02 -4.12304074e-01
1.76760092e-01 3.10699254e-01 -9.57204819e-01 2.45902110e-02
8.05568933e-01 7.77846992e-01 2.64126182e-01 1.05324507e+00
-4.62240458e-01 1.20711219e+00 -8.32597017e-01 -3.95392448e-01
3.68137211e-01 -7.08085299e-01 -9.46336865e-01 4.13060784e-01
-2.04925999e-01 -6.57748759e-01 -1.70049388e-02 8.78134668e-01
3.39828551e-01 2.94700712e-01 3.08671981e-01 -9.82409716e-01
-1.71556428e-01 6.97653770e-01 9.88585129e-02 -1.18081383e-01
4.33825850e-01 5.73585689e-01 1.19189358e+00 -3.72649133e-01
5.99710405e-01 1.43597317e+00 1.99433975e-03 7.36970901e-01
-1.24408877e+00 -7.70684704e-02 3.86492163e-01 -9.03146788e-02
-6.30986989e-01 -5.01518607e-01 7.22829223e-01 -3.42182755e-01
1.14467061e+00 -6.60714433e-02 4.53485966e-01 9.00324285e-01
5.72822280e-02 8.99834991e-01 1.06880867e+00 -5.93677580e-01
5.38807988e-01 5.39672196e-01 -1.22317530e-01 7.91367114e-01
-5.52782834e-01 1.47661969e-01 -2.84619957e-01 -2.16754705e-01
7.86448121e-01 -4.09275919e-01 8.30318481e-02 -3.36291611e-01
-6.65547192e-01 1.25072575e+00 -2.49669980e-02 8.74255896e-02
-5.27969837e-01 2.85046250e-02 5.99875748e-01 6.76589429e-01
4.00226682e-01 7.38884509e-01 -6.20045185e-01 -6.81853533e-01
-4.15582955e-01 2.91485339e-01 1.26405406e+00 8.79830241e-01
8.26767147e-01 2.87597358e-01 -6.64551184e-02 9.31554377e-01
3.91254574e-01 2.54584819e-01 4.35061991e-01 -1.19810426e+00
5.48169076e-01 5.64864457e-01 5.92875659e-01 -2.67444670e-01
-2.70052403e-01 1.59001365e-01 -8.09875652e-02 3.29282999e-01
7.36196339e-01 -7.78071940e-01 -5.11962295e-01 1.69130147e+00
4.11392391e-01 -3.42466325e-01 4.93424654e-01 8.95116806e-01
3.12811226e-01 8.55401814e-01 1.49580508e-01 -6.49220288e-01
1.13286150e+00 -1.10858178e+00 -9.01730061e-01 -4.52347755e-01
7.16216564e-01 -5.13771713e-01 1.12560761e+00 4.11798686e-01
-1.34230828e+00 -1.92962423e-01 -7.97688127e-01 2.85102036e-02
-1.28685772e-01 -4.10461538e-02 5.46851754e-01 5.70427895e-01
-1.06557655e+00 3.85411203e-01 -7.98108041e-01 -1.75120637e-01
-2.63306648e-01 6.33885324e-01 1.27202123e-01 3.82975250e-01
-1.21129978e+00 1.22462940e+00 3.69478375e-01 -3.99216823e-03
-9.72680151e-01 -1.64354160e-01 -9.61613655e-01 1.18248351e-01
7.19616055e-01 -1.88044786e-01 2.21101332e+00 -1.38148034e+00
-2.73131561e+00 4.28554296e-01 -1.33459410e-02 -7.88712680e-01
5.88528514e-01 -1.60945863e-01 -1.92337036e-02 9.98854116e-02
-3.11192185e-01 6.56234205e-01 8.24745655e-01 -1.07584190e+00
-7.37576425e-01 -8.86283815e-02 6.90412343e-01 7.05884516e-01
-7.21106958e-03 -2.16014739e-02 -1.33845448e-01 -6.79374021e-03
-4.28821474e-01 -9.75210726e-01 -5.02717197e-01 -4.38637286e-01
-1.42684773e-01 -5.84700882e-01 6.12290502e-01 -6.01536036e-01
8.24331284e-01 -1.87140834e+00 1.00548685e-01 -9.31280032e-02
-1.56520337e-01 2.88102090e-01 -1.66916743e-01 7.41206884e-01
3.21294308e-01 -2.03164667e-01 -4.23582830e-03 -3.99612248e-01
1.35881424e-01 3.27069372e-01 -3.43377769e-01 1.87115967e-01
2.40359977e-01 7.74856448e-01 -1.14183497e+00 -4.30011690e-01
5.03875077e-01 -6.41688108e-02 -7.19106376e-01 8.67364526e-01
-9.59136605e-01 8.82666826e-01 -5.70477247e-01 1.86618388e-01
-7.26817697e-02 -1.12309217e-01 8.39589775e-01 4.02943015e-01
-2.86240220e-01 8.82328987e-01 -7.58297563e-01 1.50797451e+00
-9.21807349e-01 4.49076921e-01 5.69464087e-01 -1.00549376e+00
1.05541325e+00 6.04417026e-01 4.22835022e-01 -5.06369531e-01
3.34050119e-01 2.92551685e-02 1.44630730e-01 -3.44235241e-01
5.51632941e-01 -4.15790975e-01 -3.19350123e-01 6.80633664e-01
2.42161170e-01 -3.33755553e-01 1.41507939e-01 1.83751881e-01
8.47082675e-01 4.75804955e-02 4.86532748e-01 -4.69735079e-02
5.96059382e-01 2.81878412e-01 3.98263186e-01 6.15384936e-01
-2.12789834e-01 -5.45318425e-03 7.23067045e-01 -2.75358140e-01
-9.65932667e-01 -6.63232982e-01 4.06633765e-01 1.42115963e+00
-1.30552009e-01 -1.47517279e-01 -8.42107654e-01 -8.68364930e-01
-1.53134122e-01 1.05940294e+00 -4.84800413e-02 7.35902563e-02
-7.50593066e-01 -1.01062879e-01 1.96163103e-01 2.20007241e-01
3.42677146e-01 -1.27514505e+00 -6.66700244e-01 6.36594713e-01
-3.51054110e-02 -1.21010995e+00 -5.50059140e-01 3.53071362e-01
-7.67685771e-01 -6.64976478e-01 -4.19612825e-01 -5.07257581e-01
6.89789534e-01 -2.74356753e-01 9.40498650e-01 -2.09273696e-01
1.53412297e-01 5.88430226e-01 -1.95671499e-01 -3.35528910e-01
-1.01865160e+00 1.72188595e-01 8.15884918e-02 -8.53472874e-02
2.62195855e-01 -3.09429735e-01 -2.68722057e-01 1.50282666e-01
-4.70869362e-01 5.88028803e-02 4.70805943e-01 9.59215283e-01
7.54416338e-04 -3.63689929e-01 1.08805203e+00 -9.53232467e-01
1.18490064e+00 -2.06422836e-01 -1.20436394e+00 1.67228475e-01
-6.19142473e-01 5.34214973e-01 9.88221049e-01 -2.76571214e-01
-1.38596046e+00 4.63961273e-01 -1.25767395e-01 -2.68755704e-01
-1.84918240e-01 4.01699007e-01 -1.42701879e-01 3.76430184e-01
4.85898674e-01 1.36715218e-01 3.59757066e-01 -2.84975588e-01
6.32808983e-01 8.52560759e-01 3.13246161e-01 -8.08170915e-01
4.42088753e-01 -1.17183536e-01 -5.41886210e-01 -7.03301668e-01
-7.53615320e-01 -3.87214392e-01 -4.98602539e-01 -4.28867400e-01
9.24954355e-01 -7.01848388e-01 -1.29936624e+00 -1.56658009e-01
-1.27265263e+00 -9.29617941e-01 -4.37637061e-01 4.12659436e-01
-1.00043190e+00 1.09919488e-01 -7.41347730e-01 -1.35504103e+00
-2.08317056e-01 -1.27682674e+00 7.71060705e-01 3.89256269e-01
-4.76539433e-01 -1.13489711e+00 -7.91030824e-02 4.48309749e-01
1.86635628e-01 -1.81876004e-01 7.93876350e-01 -1.05553854e+00
-6.17223978e-01 -2.44415835e-01 1.29620120e-01 2.89728820e-01
2.73955882e-01 -3.32345515e-01 -7.91677475e-01 -3.20197701e-01
7.01679066e-02 -9.74970460e-01 -1.39860988e-01 1.86397895e-01
5.21004081e-01 -9.05810416e-01 1.43206269e-01 -1.83656648e-01
1.04333758e+00 7.44363666e-01 6.35176226e-02 1.10904604e-01
1.43410951e-01 9.55990136e-01 7.62321889e-01 6.41397476e-01
4.01497066e-01 7.67373681e-01 2.35899016e-01 3.20825845e-01
3.56451601e-01 -5.78397751e-01 6.64012551e-01 6.70077980e-01
1.60574034e-01 -3.37682813e-01 -6.76825941e-01 4.18468148e-01
-2.11260056e+00 -7.13438272e-01 4.60301220e-01 2.17593145e+00
1.24548030e+00 2.74186909e-01 4.16067094e-01 -4.44922209e-01
5.05309582e-01 1.08549796e-01 -9.05197322e-01 -9.51868832e-01
4.47051436e-01 -3.76724191e-02 5.73207915e-01 1.12780666e+00
-7.33819425e-01 1.31168556e+00 6.54281855e+00 3.02954525e-01
-1.15426135e+00 1.00213267e-01 6.12306595e-01 -1.15157656e-01
-3.81148532e-02 1.07106671e-01 -8.37469637e-01 2.08701536e-01
1.30031168e+00 -2.41263613e-01 7.07037985e-01 1.12439322e+00
6.17127478e-01 -2.06473529e-01 -1.36882639e+00 7.41064310e-01
-4.07707036e-01 -1.15243971e+00 -5.02342999e-01 1.53022543e-01
4.50244814e-01 -2.84500480e-01 -2.44193941e-01 7.51949787e-01
1.08245182e+00 -9.29030716e-01 4.07958835e-01 7.79689401e-02
4.81261462e-01 -6.91321850e-01 4.87392873e-01 7.66031742e-01
-7.60269165e-01 -1.87184677e-01 -3.03203374e-01 -2.71262765e-01
6.16774321e-01 2.71030795e-03 -1.67811346e+00 1.79133952e-01
-2.59200007e-01 1.07356705e-01 1.87630296e-01 2.95142502e-01
-4.10794348e-01 6.37306511e-01 -2.50053406e-01 -4.74977642e-01
7.45658398e-01 -5.46678543e-01 4.75199789e-01 9.74568903e-01
-1.51110291e-01 3.85008454e-01 8.24407816e-01 6.18192732e-01
1.40812220e-02 1.27886802e-01 -4.98382747e-01 -4.09980327e-01
3.91890019e-01 1.16446805e+00 -5.59081674e-01 -4.01191354e-01
-3.43047559e-01 9.15886045e-01 4.12026554e-01 3.23795259e-01
-5.98249614e-01 5.22044152e-02 5.93797684e-01 -1.49768189e-01
1.41949937e-01 -4.29423600e-01 -4.53106202e-02 -9.12499607e-01
-2.72704571e-01 -1.12578046e+00 7.79539421e-02 -3.01701456e-01
-9.31415439e-01 4.78549451e-01 -3.45198028e-02 -8.42366755e-01
-1.27869070e+00 -4.28324312e-01 -3.71248394e-01 7.60194898e-01
-1.18301344e+00 -5.99060178e-01 3.85357320e-01 3.15533340e-01
1.12213838e+00 -3.23350012e-01 8.35033715e-01 -1.34946287e-01
-3.92727464e-01 3.97972554e-01 -5.51341102e-03 9.07545835e-02
5.22124112e-01 -1.48255408e+00 5.73709272e-02 2.09277645e-01
-8.31538886e-02 2.75068253e-01 1.04612112e+00 -5.36957800e-01
-1.43379974e+00 -7.14407921e-01 8.16956758e-01 -2.11170658e-01
7.09089637e-01 -5.05144000e-01 -5.49287677e-01 8.35078776e-01
6.54823780e-01 -5.64305902e-01 3.75840396e-01 1.62469029e-01
1.03042491e-01 2.09884178e-02 -1.15933228e+00 7.28283048e-01
3.87584329e-01 -5.56867421e-01 -5.15023351e-01 5.80532491e-01
7.88308740e-01 -6.66745484e-01 -8.06810200e-01 -7.25415349e-02
1.61891595e-01 -6.98511899e-01 4.30823863e-01 -6.37230575e-01
1.81390896e-01 1.12017773e-01 8.30482766e-02 -1.40428650e+00
2.31844291e-01 -1.15629435e+00 -1.03566937e-01 1.05773520e+00
6.65635288e-01 -5.96276999e-01 9.29299951e-01 1.31113827e+00
-1.16281189e-01 -6.66320503e-01 -7.29636431e-01 -7.85419762e-01
-1.75986381e-03 -4.72105630e-02 2.64137030e-01 6.44247472e-01
6.88841879e-01 9.63276982e-01 -6.16901040e-01 -1.76993012e-02
1.20641664e-01 1.26094505e-01 8.89459431e-01 -8.40507388e-01
-5.43674529e-01 -2.79589325e-01 1.23089753e-01 -1.67154920e+00
4.45782453e-01 -5.36777198e-01 6.61983728e-01 -1.26167464e+00
-4.69373137e-01 -3.21751237e-01 2.37775639e-01 4.25668836e-01
1.06559753e-01 -7.31393993e-01 3.68391871e-01 1.12606242e-01
-6.03049040e-01 7.19704807e-01 1.19787180e+00 1.38873840e-02
-6.29662216e-01 2.76556760e-01 -3.79833162e-01 5.61738253e-01
1.03223598e+00 -2.32446045e-01 -8.08234155e-01 -1.41145483e-01
-1.17605947e-01 7.61511862e-01 -1.89645186e-01 -2.77785420e-01
2.36364454e-01 -5.01119494e-01 -3.32718402e-01 -1.46660432e-01
6.87566280e-01 -5.90740442e-01 -3.25495303e-01 3.88670594e-01
-9.50331926e-01 9.90284383e-02 4.06210689e-04 5.93859673e-01
-2.63721317e-01 -6.24004960e-01 8.46410334e-01 -4.13169950e-01
-5.08420587e-01 -1.26970068e-01 -9.59456503e-01 -2.46573351e-02
1.07482314e+00 1.37765780e-01 6.28381297e-02 -1.18811619e+00
-8.03769886e-01 5.06090283e-01 2.24460378e-01 3.16318214e-01
1.95063025e-01 -7.57863104e-01 -3.94268721e-01 -2.64297307e-01
-3.38730901e-01 6.49489239e-02 -1.96082145e-01 3.46643597e-01
-2.60604262e-01 8.18647206e-01 9.35815200e-02 -4.05151278e-01
-9.83339250e-01 4.59401548e-01 3.72064441e-01 -7.91578889e-01
-4.25315171e-01 6.66826308e-01 9.30033550e-02 -7.19505787e-01
4.42421526e-01 -1.42678961e-01 -6.95441067e-02 -3.07992846e-02
3.34722430e-01 -1.34847254e-01 -1.15149736e-01 -2.59646684e-01
8.86134952e-02 -4.07754511e-01 -2.73286104e-01 -8.16042423e-01
9.60543156e-01 -1.82628885e-01 3.40156108e-01 7.02784896e-01
8.97331119e-01 -2.93277681e-01 -1.73402345e+00 -2.51223266e-01
4.43588257e-01 -2.21673042e-01 -1.67009652e-01 -9.20015156e-01
-5.51616490e-01 6.28111780e-01 2.78332174e-01 6.58002734e-01
6.00968063e-01 -1.04792610e-01 7.22916961e-01 7.73329854e-01
5.10378063e-01 -1.49477470e+00 3.68414372e-01 8.05054605e-01
7.42896080e-01 -1.40248513e+00 -3.74765962e-01 -2.13980138e-01
-1.20263696e+00 1.10531175e+00 8.06220889e-01 -6.74583912e-02
2.07705557e-01 3.20053160e-01 5.11467814e-01 1.39832556e-01
-1.17431080e+00 -2.37532049e-01 -2.67607868e-01 4.70258743e-01
6.97920620e-01 -2.94297282e-03 -4.86566395e-01 1.19922660e-01
-2.55633682e-01 -1.77570552e-01 5.69377482e-01 1.02657402e+00
-6.50623441e-01 -1.63362467e+00 -1.35597095e-01 2.34901249e-01
-3.31347317e-01 3.62472124e-02 -5.86827397e-01 5.63704789e-01
-7.11502016e-01 1.22772992e+00 1.80565957e-02 1.77030102e-04
1.67651311e-01 5.10136783e-01 3.33848417e-01 -1.00903177e+00
-9.86708045e-01 2.37131968e-01 6.83514416e-01 -3.07541847e-01
-1.80360153e-01 -6.09839022e-01 -1.55071890e+00 1.14113195e-02
-4.34961081e-01 6.48310900e-01 7.15771794e-01 1.20867848e+00
2.03138649e-01 2.92283058e-01 9.63192880e-01 -7.73233235e-01
-1.11965251e+00 -1.04689848e+00 -2.09958956e-01 1.59476385e-01
2.92269677e-01 -2.65981823e-01 -3.36438641e-02 1.01182371e-01] | [13.016447067260742, 8.068645477294922] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.